MellyForex RSS feed of MellyForex Blog Articles en-us Copyright (C) 2019 MellyForex Forex GTO Forward Test Forex GTO
Robot Name:
Forex GTO

Type of Robot:
Multi-currency grid trader. To counter drawdowns, the EA has been set up on a $50k demo account trading microlot sizes, and profits/losses are being divided by 10 so that the EA can be compared against the other EAs in test at MellyForex.

30 minute

Currency Symbol(s):

Forward Test Broker:
Alpari (UK) Ltd.

View a detailed analysis of all Forex GTO trades

Forex GTO can be purchased for $195.00 in the MellyForex Store
Forex GTO is available in the MellyForex Store
Visit the Forex GTO website]]> Tue, 18 Feb 2014 00:00:00 +0000
iProfit HFT Forward Test iProfit HFT
Robot Name:
iProfit HFT - version 1.02

Type of Robot:
Based on trading currency symbols with high correlation simultaneously with symbols which have low/medium/negative correlation.

1 hour

Currency Symbol(s):

Forward Test Broker:
FX Central Clearing Ltd.

View a detailed analysis of all iProfit HFT trades

iProfit HFT can be purchased at the reduced price of $48.00 with a 20% discount in the MellyForex Store
iProfit HFT is available in the MellyForex Store
Visit the iProfit HFT website]]> Mon, 25 Nov 2013 00:00:00 +0000
Forex Earth Robot Forward Test Forex Earth Robot
Robot Name:
Forex Earth Robot - version 1.0

Type of Robot:
Unknown strategy

5 minute

Currency Symbol(s):

Forward Test Broker:
FX Central Clearing Ltd.

View a detailed analysis of all Forex Earth Robot trades

Forex Earth Robot can be purchased at the reduced price of $79.20 with a 20% discount in the MellyForex Store
Forex Earth Robot is available in the MellyForex Store
Visit the Forex Earth Robot website]]> Wed, 20 Nov 2013 00:00:00 +0000
FX SEER Forward Test FX SEER
Robot Name:
FX SEER - version 1.0.2

Type of Robot:
Multi-strategy according to trading session

1 hour

Currency Symbol(s):

Forward Test Broker:
FX Central Clearing Ltd.

View a detailed analysis of all FX SEER trades

FX SEER can be purchased at the reduced price of $213.75 with a 5% discount in the MellyForex Store
FX SEER is available in the MellyForex Store
Visit the FX SEER website]]> Tue, 05 Nov 2013 00:00:00 +0000
Forex Spectre Forward Test Forex Spectre
Robot Name:
Forex Spectre - version 1.0.3

Type of Robot:
Multiple strategies

Any chart timeframe

Currency Symbol(s):

Forward Test Broker:
FX Central Clearing Ltd.

View a detailed analysis of all Forex Spectre trades

Forex Spectre can be purchased at the reduced price of $185.25 with a 5% discount in the MellyForex Store
Forex Spectre is available in the MellyForex Store
Visit the Forex Spectre website]]> Thu, 27 Jun 2013 00:00:00 +0100
Forex Megabot Forward Test Forex Megabot
Robot Name:
Forex Megabot - version 1.1b

Type of Robot:
Pre-Asian Scalper which uses fractals to determine the direction of the trade. The EA uses a Stop-Loss of 50 and a Take-Profit of 8.

15 minute

Currency Symbol(s):

Forward Test Broker:
FX Central Clearing Ltd.

View a detailed analysis of all Forex Megabot trades

Forex Megabot can be purchased for $99.00 in the MellyForex Store
Forex Megabot is available in the MellyForex Store
Visit the Forex Megabot website]]> Mon, 10 Jun 2013 00:00:00 +0100
Forex Scalpino Forward Test Forex Scalpino
Robot Name:
Forex Scalpino

Type of Robot:
Asian Scalper

15 minute

Currency Symbol(s):

Forward Test Broker:
Alpari (UK) Ltd.

View a detailed analysis of all Forex Scalpino trades

Forex Scalpino can be purchased at the reduced price of $158.40 with a 20% discount in the MellyForex Store
Forex Scalpino is available in the MellyForex Store
Visit the Forex Scalpino website]]> Thu, 30 May 2013 00:00:00 +0100
Joker EA Forward Test Joker EA
Robot Name:
Joker EA - version 3.6

Type of Robot:
Trend watching scalper, trading on a pullback in trend direction.

5 minute

Currency Symbol(s):

Forward Test Broker:
Alpari (UK) Ltd.

View a detailed analysis of all Joker EA trades

Joker EA can be purchased at the reduced price of $263.12 with a 12% discount in the MellyForex Store
Joker EA is available in the MellyForex Store
Visit the Joker EA website]]> Tue, 21 May 2013 00:00:00 +0100
Ray Scalper Forward Test Ray Scalper
Robot Name:
Ray Scalper - version 2.03

Type of Robot:
Ray tracing

15 minute

Currency Symbol(s):

Forward Test Broker:
Alpari (UK) Ltd.

View a detailed analysis of all Ray Scalper trades

Ray Scalper can be purchased at the reduced price of $280.00 with a 20% discount in the MellyForex Store
Ray Scalper is available in the MellyForex Store
Visit the Ray Scalper website]]> Mon, 13 May 2013 00:00:00 +0100
CabEX Forward Test CabEX
Robot Name:
CabEX - version 1.01

Type of Robot:
Follows price action near moving averages to identify trends and reversals

1 hour

Currency Symbol(s):

Forward Test Broker:
Alpari (UK) Ltd.

View a detailed analysis of all CabEX trades

CabEX can be purchased at the reduced price of $280.00 with a 20% discount in the MellyForex Store
CabEX is available in the MellyForex Store
Visit the CabEX website]]> Mon, 13 May 2013 00:00:00 +0100
Omega Trend EA Forward Test Omega Trend EA
Robot Name:
Omega Trend EA - version 7.0

Type of Robot:
Trend following


Currency Symbol(s):

Forward Test Broker:
Alpari (UK) Ltd.

View a detailed analysis of all Omega Trend EA trades

Omega Trend EA can be purchased at the reduced price of $137.90 with a 30% discount in the MellyForex Store
Omega Trend EA is available in the MellyForex Store
Visit the Omega Trend EA website]]> Fri, 10 May 2013 00:00:00 +0100
Forex Pips Bag Forward Test Forex Pips Bag
Robot Name:
Forex Pips Bag - version 2.0

Type of Robot:
Trend Follower

1 hour

Currency Symbol(s):

Forward Test Broker:
FX Central Clearing Ltd.

View a detailed analysis of all Forex Pips Bag trades

Forex Pips Bag can be purchased at the reduced price of $83.85 with a 35% discount in the MellyForex Store
Forex Pips Bag is available in the MellyForex Store
Visit the Forex Pips Bag website]]> Wed, 19 Sep 2012 00:00:00 +0100
ZenX EA Forward Test ZenX EA
Robot Name:
ZenX EA - version 1.1

Type of Robot:
Hedging Martingale Grid Trader

1 hour

Currency Symbol(s):

Forward Test Broker:

View a detailed analysis of all ZenX EA trades

ZenX EA can be purchased for $119.50 in the MellyForex Store
ZenX EA is available in the MellyForex Store
Visit the ZenX EA website]]> Tue, 03 Jul 2012 00:00:00 +0100
How To Test Your Broker's Latency A rise in recent months of the number of EAs which trade so-called impulse spikes (such as Forex Thor) has meant that a lot of questions are now being asked about latency.

Latency is, broadly speaking, a measure of the time taken to send the packets of data from your PC to your broker's server. A poor connection to the broker's server could result in orders failing to execute at the desired price and trading opportunities being missed. What might seem like a very good EA to some users, may easily turn into an absolute nightmare for others!
I say 'broadly speaking' because, to the best of my knowledge, there's absolutely no way of measuring how long it takes to make the one-way connection between the two machines. Instead, we can only measure something that is called round-trip latency.

Round-trip latency is actually a three-part process because it includes not only the two each-way trips from one machine to the other and back again, but it also includes the element of time which is taken by the broker's machine to process the instructions that have been sent and compile a suitable response. For instance, if each trip takes 200 milliseconds, and the broker's server takes 100 milliseconds to execute any instructions and compile its response, then the round-trip latency will be 500 milliseconds; the one-way latency will only be 200 milliseconds and all the important bits and bobs will have been sorted out on your broker's server after 300 milliseconds.


So how can we measure latency to get some sort of an idea whether our connection is any good or not?Well, firstly, the only MetaQuotes Language functions which involve two-way communication between the user and his broker are the order sending functions. As part of this process, the MT4 client terminal sends an instruction to the broker's MT4 server to either buy or sell a market. The MT4 server executes that instruction and then sends a response back to the MT4 client terminal which either says "Bingo" or "Houston, we got a problem".

So, we simply make a note of the time immediately before sending our order to the broker, and then make another note of the time as soon as we get a response back from the broker to say the order has been executed. Subtract one time from the other and, hey presto, we've got our round-trip latency. It's easy innit? :roll:

Well no, it's actually a bit more complicated than that, because we don't want to be waiting for an EA to be ready to trade in a live environment before we can find out whether our connection is any good or not. To get over this problem, we need to be sending a series of dummy limit or stop orders at a price way away from the current price, so there's no risk whatsoever of the test order actually being filled. Having sent a dummy limit order, we can then delete it straight away. And if we send the dummy orders at regular intervals, we can easily find the maximum, minimum and average latency over a sample of 100's or 1,000's of dummy test orders. Cool, huh?

The good news is that I've written a simple EA which does precisely that, and I've put it onto MellyForex as a FREE download for my readers.

There are no strings attached. You simply attach the Latency EA to the chart window of ANY currency symbol, running on ANY timeframe. Set the frequency in minutes that you want to send the dummy orders, and the EA will do the rest.

The Latency EA will display a comment in the top left of the chart window showing the minimum, maximum and average round-trip latency.

The limit orders themselves are sent at the absolute minimum price and for the minimum contract size allowed for that market so, like I say, there's no real chance of any of the dummy test orders being filled. The EA automatically detects 4- or 5-digit brokers and, for the technically minded, it uses the somewhat obscure magic number 4156434123, so there's unlikely to be any conflict between its trades and those of any other EAs operating on the account. In any event, I've made the EA open source, so you can change or modify the code if you wish.

Please bear in mind that it's an EA, rather than an indicator or a script, so its successful operation is dependent upon a flow of incoming price ticks for it to work. If you notice that it doesn't close an open order immediately, it will simply be because it's waiting for the next incoming tick.

As I've said, you are at liberty to modify the EA if you wish and, throwing a couple of ideas for improvement into the melting pot, you could save the results as a global variable so they could be added to whenever the EA is reloaded or you could write the details of each trade to a CSV file for posterity. You could even run some deviation algorithms on the results if you wish. But that's down to you - please don't ask me to make modifications to the EA for you, as I really don't have the time.

If you like the EA, then all I ask is that you tell your friends about MellyForex and send them to this page to download the EA. Please don't link directly to the download file in case I decide at some future date to either move the file's location or update it with a newer version. And if you do decide to modify the EA, please be sure to leave the original copyright notice intact.

I suppose I should also include a disclaimer to the effect that anybody who uses the Latency EA does so entirely at their own risk. Otherwise, enjoy!

Download the FREE MellyForex Latency EA here

Installation Instructions

There's every chance that the above download link will open the EA as a readable text file in your browser window, rather than presenting it as a file for download. If that's the case, when it opens in your browser window, you just need to hit File > Save As and choose where you want to save it on your PC. You should save the LatencyEA.mq4 file in your MetaTrader Installation Folder > experts folder. The MetaTrader Installation Folder will normally be located at C: > Program Files, although that won't always be the case.

When you restart your MetaTrader 4 client terminal, the MQ4 will automatically compile, and you'll be able to view it as an Expert Advisor in the Navigator panel. From there, you can drag it into a suitable chart window and set the frequency at which you wish the test limit orders to be sent (default is 5 minute intervals). You also need to ensure that Live Trading is enabled, otherwise no orders will be sent. :)

]]> Mon, 30 Apr 2012 12:07:44 +0100
Cut Your Losses And Let Your Profits Run
As part of the Advanced Statistics of each EA in test, there is a tab called 'Duration' that readers can click on. Clicking this tab displays a scatter graph comparing the profit or loss in pips of each trade taken by the system against the duration of each trade.

As I was browsing through the different pages for each EA, checking the data and charts for accuracy as part of the testing process, I noticed one EA which had taken quite a lot of trades whose trade duration scatter graph looked like this:

This graph shows that the EA holds on to its losing trades for a longer period of time than it holds its winners

What struck me the most about this graph was that it was a classic example which displayed everything that I didn't want to see, despite the fact that the EA in question has demonstrated both reasonable profitability and safety during its time in the MellyForex forward test.

So what's wrong with that graph?Well, the first thing I should make clear is that I'm not targetting any specific EA in my example above. I've chosen the chart I did simply because this particular EA has made sufficient trades in its forward test for its results to be considered meaningful. There's no reason other than that.

Now hands up in class if you've ever heard the following saying ...

Cut your losses and let your profits run

I'm sure that most traders have heard the above quote, and I'm equally sure that most people know and understand what it means. In case you don't, the concept is simply that a trader should exit a trade as soon as it starts to go wrong. If he doesn't do this, and chooses instead to hold on to the trade in the hope that the market will turn, then there is every probability that the situation will worsen, and the loss will get even bigger still before he finally gets round to pulling the plug.

Before any wise soul tells me that my above statement is nothing more than plain old-fashioned common sense, let me also say that an awful lot of traders act like rabbits staring at a car's headlights when faced with a losing scenario, and a great number will hang on to their losing trades far longer than they should in the blind hope that things will change.

But there is also a double meaning to the quote, which I'll attempt to explain by asking if anybody has heard of another old saying ...

Nobody ever went skint by banking a profit

The above saying suggests that it's worth banking a profit regardless of how little the amount is. Hey, a profit's a profit, and if you keep banking 3 or 4 pips at a time, those pips soon mount up. Until, of course, you hang onto one of those losers longer than you should and end up taking an unnecessary 100 pip hit which wipes out the gains of the last 25 or 30 profitable trades. And how many times have you banked a measly profit, only to watch the market then soar over the next half hour and put on 200 pips or more, leaving you feeling pig-sick that you'd banked prematurely? So, whilst you might not go skint from banking a small profit, wouldn't it be better to resist temptation and at least give those barely profitable trades a bit longer to see if they can turn into something worthwhile?

It's OK, we've all been there, seen it, done it and most of us have even got a T-shirt to prove it. The ability to successfully run profits is a skill which requires great discipline, and it's not really surprising when natural human emotion sometimes gets in the way of a trader's decision making.

Going back now to the scatter graph example showing the EA's trade duration, all of the little green profitable dots are clearly visible on the left of the graph, whilst all of the red dots which show the losing trades are on the right of the chart. The graph is clearly telling us that the longer this particular EA holds on to it's trades, the larger its losses tend to be. It certainly isn't following the old proverb about cutting losses and running profits. Or, put another way, this particular EA is a robot which portrays human emotions. :)

As I said previously, this observation is generalised and isn't targetting any single EA specifically, as there are a lot of EAs which exhibit similar tendencies. Also, my example shows a profit/loss range of between -60 and +20 pips with durations of typically less than 12 hours, but it wouldn't matter if I was talking about scalpers or long term trend followers where trades last days on end, as it's the general concept of running profits and cutting losses that I'm trying to put across here.

In an ideal world, my scatter graph would have a look and feel something like the attempt I've concocted below using PhotoShop. I've had to use PhotoShop to prepare the image, because I haven't yet seen an EA produce a chart looking like how I would like it to look. Hey ho, maybe one day. :(

This graph shows how an ideal system would cut its losses and run its profitable trades

So, if any EA developers are reading this article, one thing that they might like to consider is the impact of closing out of any losing trade after a certain period of time. In my initial example above, would overall performance improve and would the losses be less if the EA closed everything regardless after, say, 4 hours? I don't know the answer to that question, but I certainly think it's something worthy of investigation.

In the meantime, seeing as we've now got robots that can mimic humans, I'd just like to say that I've got some dishes that need washing and pile of clothes that need ironing. ;)
]]> Wed, 25 Apr 2012 09:55:11 +0100
Yen Drive Forward Test Yen Drive
Robot Name:
Yen Drive - version 201

Type of Robot:
Multi-currency basket grid trader

30 minute

Currency Symbol(s):

Forward Test Broker:

View a detailed analysis of all Yen Drive trades

Yen Drive can be purchased at the reduced price of $169.15 with a 15% discount in the MellyForex Store
Yen Drive is available in the MellyForex Store
Visit the Yen Drive website]]> Mon, 09 Apr 2012 00:00:00 +0100
Forex Thor EA - Expert Advisor Review Forex Thor 2One of the first steps in preparing a Forex EA review for MellyForex is to run some strategy tests on the EA which can then be analysed to provide the content for the review. The hope is then that the EA will perform in the same manner going forwards as the backtests suggest.

In the case of Forex Thor, however, which has been sitting proudly at the top of the MellyForex leaderboard for several weeks, I have to be honest at the outset and say that I don't think that those backtests are going to be of much use at all.

To explain my reasoning behind this, I feel firstly that I need to discuss something which isn't specifically related to Forex Thor and which some readers will already be fully aware of. I can only apologise in advance, therefore, if you feel that what I am about to say covers some old ground.
The problem comes about because of the exceptionally short trade duration of all of Forex Thor's trades. A quick peek at the stats page shows that the average trade duration of Forex Thor's trades in its forward test is less than 1 minute, with the longest lasting trade being all over in less time than it takes to boil an egg!

So what difference does that make exactly?Melly's Rant

Well, the biggest limitation is that the history data used in a MetaTrader strategy test consists of just the open, high, low and close prices on the 1-minute timeframe. MetaTrader doesn't store data on a 'per-tick' basis, and this means that we have no way of knowing the correct sequence of price movements in the time between a 1-minute bar opening and the bar closing. Sure, we know what the high and low prices were in between the open and close, but did the market make its 1-minute high price before it made its low price? Or vice versa?

MetaTrader attempts to overcome this fundamental problem by using a method called Fractal Interpolation to 'guess' the sequence of ticks within each and every 1-minute price bar. But how do we know if MetaTrader has 'guessed' the tick sequence correctly? The simple fact is that we don't know, and this matters a great deal when we are trying to evaluate an EA such as Forex Thor whose typical trade duration is less than a minute, because we've no way of knowing if any of the trades taken in the tester would have actually been taken a real environment. In turn, this means that the backtest results aren't worth the paper they're printed on. :(

To make matters even more complicated, whenever you restart the MT4 client terminal the strategy tester will 'guess' a totally different sequence of ticks because the Fractal Interpolation algorithm employs a random process. This means that you could run two strategy tests on the same EA with the same broker using the same history data and the same price spread, and get two totally different sets of results. Confusing, or what?

I suppose I should say that, for 90% plus of EAs, this MT4 limitation won't matter in the slightest. Expert Advisors with longer trade durations and with SLs and TP of 10 pips or more are, generally speaking, unaffected by the lack of 1-minute tick data.

Now, before I get 101 emails from readers, I should also point out that I am well aware of the excellent work that Birt has done in preparing his tick data scripts and modules which are aimed at improving the accuracy of backtest results. Birt has put a lot of hard work and effort into providing something which wouldn't otherwise be available to retail traders, and I honestly think the guy deserves a medal for his dedication to the cause. The limitation of any 99% modelling quality data, however, is that the tick data should really come from the same broker as the EA is going to run live with. There is no guarantee whatsoever that the Dukascopy tick data is going to be the same as that of any other broker and, for all of Birt's good work, we still can't say for sure precisely how accurate and indicative a Dukascopy 99% modelling quality tick data backtest really is when it comes to evaluating EAs like Forex Thor which trade inside a 1-minute price bar. Most people have witnessed first hand the extent to which some EAs are broker dependent and, for such an EA, it wouldn't matter whether a strategy test used tick data or not because the tester's results could never be relied upon. Unfortunately, until a majority of brokers start releasing their own tick history data (and I should add that there is absolutely no incentive whatsoever for them to do so!), I fear the average retail punter will have to continue playing this game on an uneven surface.

About Forex Thor

Now that my rant is over, I should make it clear that the above mentioned limitations are not the fault of the Forex Thor team in any way, nor are they the fault of any other EA developer for that matter - this is, purely and simply, an MT4 problem which serves to make life difficult for EA users and developers alike!

Despite the reservations that I've attempted to explain above, I'm still going to conduct an analysis of the Forex Thor developer's own 99% Dukascopy strategy tests going back to 2007 (tick data is unavailable prior to then) for the purposes of this review, although there are some caveats which I'll try to highlight as I proceed.

Before looking at those backtests, I suppose that I should firstly say a little bit about Forex Thor and explain what little I can about how it works. The EA runs without any time filtering, 24/5 on the 30-minute chart timeframe of the EURUSD pair, and comprises of three independent strategies which it calls 'cores'. Users can use either just one core, or they can use a combination of cores, but each core needs to run in a separate chart window. Despite the fact that the EA runs on 30-minute charts, that doesn't in itself mean that it uses 30-minute trade signals. As a part of its inner workings, the EA copies the rates of the 1-minute timeframe across to its DLL which is where it conducts all its hocus pocus and spell-casting behind a mysterious veil of secrecy.

The fact that the EA does its work inside its DLL means that I can only hazard a guess as to Forex Thor's exact modus operandi, and I can't be 100% sure of what it's actually doing. That said, there are several EAs which work in a similar manner to Forex Thor. Their strategy generally involves waiting until the price hits either the top or bottom of its range and then detecting the early signs of a price reversal which the EA attempts to latch onto. Very often, the reversal signals manifest in the form of a two-way battle between the bulls and the bears which can be detected by an increase in the frequency at which incoming price ticks flow. Having taken its signal and entered a trade, the Forex Thor then runs a very tight trailing stop-loss and the vast majority of its trades result in a mere 2 or 3 pip profit or loss.

Occasionally, however, Forex Thor manages to latch onto the start of decent price move and will end up bagging itself a profit of perhaps 20 or even 30 pips.

The fact that Forex Thor runs a very tight SL also means that it can afford to take larger than norm lot sizes and still only risk a small percentage of the account. Some users might initially be a little perturbed by the trade sizes being larger than they're used to, but they needn't be alarmed.

The larger lot sizes do mean, however, that when the EA does latch onto one of its 20 or 30 pippers, the outcome will be a very healthy upwards spike in the account balance as can be seen from my own forward test of the EA on the left!
It's also worth saying that Forex Thor's trade filtering method means that it can often go a week or two without trading at all before suddenly bursting into like, and taking several trades within a few short minutes before falling back into snooze-mode for another week.

Bearing in mind that the vast majority of Forex Thor's trades involve a profit of just a couple of pips, broker selection is going to be very important otherwise the performance of these 'bread and butter' trades will suffer. The best performance will invariably be obtained by using a true ECN broker who charges a commission on each trade, rather than one who pads their commission within the spread.

When selecting a broker, users will also need to ensure that the broker's stops level is very tight. Forex Thor enters its trades using stop orders and, if you choose a broker with a 5 pip stops level, the EA will never manage to get any trades away. You really do need to be using a broker whose stops level is 0.01 pips! It's also my personal suspicion that the EA will work better with 5-digit brokers, rather than with older style 4-digit ones and, finally, the developer also suggests that good latency may improve the EA's performance in terms of trade execution speeds.

Strategy Tests

Now that we've learned a little of how Forex Thor works, it's probably time to look at the developer's backtests. The tests have been conducted using the latest version Forex Thor 2, which is only different from the original version of the EA in so far as it includes a new 'Aggression' parameter. Following a number of user requests for a modification which made the EA trade more frequently, the developer has introduced an updated version containing this new parameter. Set the aggression value to FALSE, and the EA will trade identically to the original version; Set it to TRUE, it will take the extra trades that some users were asking for, and everybody's happy. Easy, innit? ;)

There are separate strategy tests for each of Forex Thor's cores, and they can be viewed in detail by clicking each of the balance curves below. Be aware though that the file size for the Core 1 strategy test is quite large (nearly 14MB), so the report may take a while to load which could cause a problem if you're on a slow connection.

Forex Thor Core 1 Strategy Test

Forex Thor Core 1 Strategy Test

Forex Thor Core 2 Strategy Test

Forex Thor Core 2 Strategy Test

Forex Thor Core 3 Strategy Test

Forex Thor Core 3 Strategy Test

A close inspection of the test settings shows that the aggression parameter has been set to false, meaning that these tests represent the original version of Forex Thor. To analyse these results, my first step was to merge all three cores into a single set of results, and I then standardised the trades to the same 0.1 lot size. This enabled me to conduct some risk simulation tests, and I have summarised the important bits of the analysis below.

Forex Thor Risk Simulation

It seems that Core 1 is responsible for around 59% of all of Forex Thor's trades, although all 3 cores exhibit very similar levels of performance with similar win rate and trade expectancies. Very often an analysis like the one above will show that best performance can be obtained by disabling either certain symbols or strategies, but there's nothing to suggest that is the case with Forex Thor.

Having bored most readers to death with my earlier rant about the failings of MetaTrader 4, I also thought it appropriate to parse the trades taken in the test from the date my MellyForex forward test started, and compare the tester's trades with the forward test's trades. This would give me some sort of idea of just how reliable the strategy test results were likely to be. My original Forex Thor forward test started on 14th December 2011 and the developer's strategy tests ended on 9th March 2012, so there is a period of around 3 months during which the forward tests can be compared against the 99% modelling quality strategy tests.

I've adjusted the start deposit to $1,000 and standardised the lot sizes to 0.1 lots to make the comparison as straightforward as possible. For anybody who wants to compare the physical trades, this can be done by clicking the relevant summary image below.

Forex Thor Developer's Strategy Test

Forex Thor MellyForex Forward Test

I don't intend to go into too much detail about the differences between my forward test trades and the Dukascopy strategy test trades for the corresponding period, other than to say that I'm sure everybody is able to spot that the strategy tester is showing around 16% more trades than have been taken in my forward test during the corresponding period, and this clearly results in better all-round performance. This simply reinforces my earlier assertion that there's no effective way of reliably predicting the future performance of an EA of Forex Thor's nature from its backtests.

Despite the obvious statistical differences, however, both equity curves do go in pretty much the same direction and the basic concept of a system which exhibits short trade durations, a low trade expectancy and a win rate of around 75% - 80% is apparent in both the strategy and forward test. In addition, it's difficult to be sure precisely how representative the comparison actually is when the sample is less than 40 trades taken over a relatively short period of time.

The other point to note is that the combined strategy tests for all 3 cores suggest a total of 4,165 trades in 1,799 days of test - in other words, over 16 trades on average each week. In addition, the longer term win rate seems to be around 55%. So far, after around 4 months in forward test, my copy of Forex Thor has been averaging just 2 trades each week and winning over 75% of its trades. Again, it's quite difficult draw any firm conclusion from this. The initial suggestion is certainly that the tester is likely to overstate the number of trades that are seen in reality, but it's also quite possible that the EA has been able to successfully reject the worst trade entries in recent weeks and only select the good ones. That may explain the anomaly.

Forward Tests

The original version of Forex Thor has been in forward test at MellyForex for quite a while now. It has been running on a an FXCC true-ECN demo account and has already built up a reasonable amount of trade history which is currently being analysed in detail at MellyForex on this link - Forex Thor Performance Analysis and Charts.

As I stated earlier, the only difference between Forex Thor 2 and the original version is the new 'Aggression' parameter. Given the unreliability of the strategy tests, the Forex Thor developer has kindly agreed to allow me to publish two sets of the results to MellyForex.

The original version will continue to publish with its aggression parameter set to FALSE. The version that I am referring to as Forex Thor 2 has its aggression parameter set to TRUE. This will enable readers to monitor and compare the effect of this new parameter. Otherwise, the settings I'm using are identical to the original version's settings.
Both copies of the EA appear on the MellyForex leaderboard, and a detailed analysis of Forex Thor 2 is available on MellyForex.


It's very difficult to draw any firm conclusions about this EA given the unreliability of the strategy tests, and I should repeat that this is a fundamental MetaQuotes issue rather than a fault of the EA's developer. Despite those problems, I do believe that Forex Thor offers tremendous potential if used in the right environment, and I find it encouraging that my forward test is so far heading in a broadly similar direction to the strategy tests, albeit there are clear differences in the route being followed.

We do know from both the strategy and forward tests that the EA's trade expectancy is low, and that the choice of broker is very important. Users should be prepared to see the EA possibly take different trades with different brokers. I'm sure that some users will become frustrated by this aspect of the EA but, over the fullness of time (and I mean a year or more, not just a few weeks!), I would hope to see those difference even themselves out to produce a similar end result.

The other caveat is that there are no results available to call upon which might suggest how beneficial the new aggression parameter is likely to be. This means that any user who opts to set the aggression parameter to TRUE is likely to be flying blind. We do know that setting the parameter to TRUE will result in the EA taking more trades, but I suspect that its extra signals will be of a poorer quality and users could see the win rate fall and the drawdowns increase if they opt to use aggression. That's where I hope MellyForex readers will benefit from being able to compare copies of the EA running both settings in identical environments.

For anybody who is interested, the developer has recently started offering a free trial of Forex Thor 2. There are some minor limitations and, rather than go into detail about the trial here, readers can find out more about the offer by visiting the Forex Thor website.
]]> Sat, 07 Apr 2012 11:46:23 +0100
TradeVader Forward Test TradeVader
Robot Name:
TradeVader - version 2.7

Type of Robot:
Indicator driven

15 minute

Currency Symbol(s):

Forward Test Broker:
FX Central Clearing Ltd.

View a detailed analysis of all TradeVader trades

TradeVader can be purchased for $179.00 in the MellyForex Store
TradeVader is available in the MellyForex Store
Visit the TradeVader website]]> Thu, 05 Apr 2012 00:00:00 +0100
Volatility Factor EA - Expert Advisor in Forward Test The Volatility Factor EA is in forward test at MellyForexThe Volatility Factor EA is the latest Expert Advisor to enter its forward test at MellyForex, and it comes from the developers of the popular and successful Wall Street Forex Robot which is also in forward test and has been reviewed elsewhere on the site.

The EA itself works on two currency pairs - EURUSD and GBPUSD - and runs on the 15-minute chart timeframes. It can only be described as a grid trader, which means that, at times, it's going to have multiple concurrent open positions which are in drawdown.

I'm not going to offer any opinion upon this particular trading style, as it's an approach which undeniably popular amongst many traders. I should also say that, in fairness to the developers, they have dedicated no fewer than six pages of their manual to explaining the EA's exit strategy and how it copes with drawdowns so that users know at the outset what to expect.

Forewarned is forearmed!
I've installed my copy of the Volatility Factor EA on a $5k demo FXCC ECN account which offers razor-sharp spreads.

Its trades are being published to MellyForex, and readers can analyse the EA's performance extensively in the Expert Analysis section of the site by clicking on this page.

I'm using the EA's default settings with the exception that I have set the risk value to 3.0% which is described as being 'moderate' within the user manual. Obviously, if I find myself in the situation where I have multiple open losing trades, I'll be risking considerably more than 3.0% of my account!]]> Tue, 27 Mar 2012 09:23:20 +0100
New analysis features at MellyForex It's not very often that I get a chance to blow my own trumpet, but I am rather pleased with the latest feature to go live at MellyForex so, hey, I guess it's OK to bring out the bunting and have a bit of a toot. :)

Since MellyForex started, regular readers will know that I've been publishing the results of all the EAs in forward test to MyFxBook. MyFxBook is a great website in my opinion; it's stylish and some of their analysis tools are very useful, however, I have had problems with some accounts not verifying and I recently noticed that a couple of EAs hadn't updated at all for 4 or 5 months.

With that in mind, I decided a while ago to build a new section on MellyForex where the EAs could publish their results and their performance could be analysed.

I'm pleased to announce that work on the new analysis section is now complete, and all results publishing is now being done in-house. Readers can access the statistics for any EA in forward test by visiting the leaderboard here and then clicking on the name of the EA whose performance they wish to view.

Each review/article at MellyForex has also been updated so that the old links to MyFxBook pages now point to the relevant page on MellyForex, and the old MyFxBook pages have been deleted.

I'm always open to suggestions for improvement of the analysis layout and features, so don't be shy with your suggestions.

In the meantime, please enjoy!]]> Wed, 21 Mar 2012 12:09:47 +0000
Fast Forex Millions - Forward Test Halted I'm convinced that, if Albert Einstein were alive today, he'd quickly be able to crunch a few numbers and work out the inversely proportional relationship that seems to exist between hype and garbage.

Hey, if old Albert could have had all of those theories about negativity and cope with that funky E=mc2 stuff, it should have only taken him a couple of seconds to figure that the more an EA is hyped, the sooner it is likely to end up in the dustbin.

Anyway, my good nature prevents me from suggesting that the Fast Forex Millions EA is an over-hyped piece of trash, but I have decided to stop its forward test at MellyForex nonetheless.

To explain, ever since starting the test I've been having nothing but problems with some symbols failing to login and validate on the vendor's authentication server, and now the whole thing has stopped working completely. Given that the vendor's support is pretty weak (Oh alright then, more like bordering on non-existent!), I really can't be bothered to waste any more of my time with this EA.

There is a current open trade which will close either when its TP or SL is hit, and I'll just leave that trade to run. Ordinarily, the EA would have closed the trade long ago, but if it won't login to its authentication server, it certainly won't trade and the trade will have to stay open pending either of the orders being hit.

In any event, its performance hasn't exactly been stellar since starting the test. The EA has drawn down substantially, and I suspect a lot of punters will be giving this particular vendor a very wide berth from now on.

As the late Freddie Mercury used to sing, another one bites the dust!]]> Tue, 13 Mar 2012 18:54:59 +0000
Black Belt EA - Expert Advisor in Forward Test Black Belt EAI'm happy to admit that I don't know the first thing about either Judo or Karate, but I get the impression that the Black Belt EA developer might be a martial arts expert.

I've therefore removed my witches hat to write this short article, and I'm going to try to say some nice things about the EA, through fear that I might get a visit from Bruce Lee or Jackie Chan to 'sort me out'. :roll:

Seriously though, this brand new Expert Advisor seems to have already captured a lot of interest, and the developer has kindly provided a copy to put into forward test at MellyForex so that our readers can monitor its performance.

The Black Belt website states that the EA is a non-grid, non-Martingale high frequency trader that works on the EURUSD symbol on the 30-minute chart timeframe. It claims to use price action rather than indicators for its trade entry signals, it's supposedly ECN and NFA compliant and, although it is meant to work best with brokers whose EURUSD spread is under 2 pips, spread is not meant to be that big an issue.

Now that the standard marketing spiel has been dealt with, what else can I tell you about the EA?Well, the first thing I noticed from the backtests on the Black Belt website was that those backtests were performed using the 'Open prices only' model. The suggestion here is that the EA would only ever open and close its trades on the first tick of a new price bar, and this is something that the developer confirmed to me. In addition, the EA doesn't declare its stop-losses or take-profits which means that, if the EA gets a signal to close a trade shortly after a price bar has opened, the user will have to wait patiently at the mercy of the markets while the current price bar completes its formation and a new bar opens before the exits the trade. Obviously the market could go either way during that time, and there's an obvious risk that an extra 100 pips or more could be lost while the EA is waiting to close its trade.

My next observation is that the EA uses a form of 'pyramiding' as part of its strategy. Pyramiding is the process of using the free margin generated by an open, profitable trade to open additional new trades. In its purest form, the lot sizes of the continuation pyramid trades are calculated to use the free margin which has been generated by the other trade(s). This approach frequently means that the continuation trades will be of a smaller size than the initial trade, although Black Belt's additional positions are most likely to all be of the same size as the initial trade. As I scrolled down to view one of the backtest statements, I didn't get very far before I noticed that the EA had opened no fewer than 9 concurrent trades. When I enquired, the developer advised me that the default was for the EA to open as many as 16 open concurrent trades. This might seem excessive, but remember that the additional positions are only opened on the basis that a continuation entry signal has been received and the previous trades are in profit. It most certainly shouldn't be confused with a typical grid system which adds to losers.

My biggest concern here is that, if you've got umpteen open pyramid trades, and the market turns against you, and you've then got to wait for a new bar to start before the EA closes out, you could potentially find yourself losing umpteen hundred pips while you're hiding behind the sofa overdosing on Valium. Pyramid trading can be both high risk and high gain, so widows and orphans should take note!

Although I've put the Black Belt EA straight into forward test without running any strategy tests of my own, I did spend a bit of time analysing the developer's own backtests. In particular, I examined a statement from 2000 through to the end of 2011 and can confirm that the win rate is approximately 53% with an average winning trade size of 51.8 pips and an average loss trade size of 38.4 pips.

The total number of pips won in those 12 years was 75,286 which, in 4,379 days of test, equates to 523 pips won each month, achieved from an average of around 12 trades each week. Given that this pyramid style of trading is similar to that employed by Forex Growth Bot, which I also have in test at MellyForex, obvious comparisons are going to be drawn between the two EAs.

Also of importance from the developer's backtests, the trade expectancy is 10.0 pips per trade which should seemingly support the developer's assertions that the EA isn't broker dependent.

Sounds great so far? Unfortunately, when things look like they're too good to be true, they very often are.

In the case of of Black Belt EA, there was something unusual that I could see in the 12 year fixed lot equity curve which was sticking out like a sore thumb. When an EA trades variable lot sizes, which increase as the account balance also increases, the equity curve tends to have an exponential appearance and often rises steeply in the latter part of the curve. With a fixed lot equity curve, however, the line should be nearly straight and rise at an even rate.

A quick look at the Black Belt fixed lot curve clearly shows a steep jump in the curve in the middle section as I've highlighted below.

Black Box EA Equity Curve

The rate of incline is reasonably constant from 2000 through 2007, but the balance curve rises steeply in 2008 and 2009, before resuming at a slope angle very similar to its previous incline from 2010 through to the end of the backtest.

It's common knowledge that the global financial crisis affected markets dramatically during both 2008 and 2009. Fiber quickly fell from its perch at $1.60 to a low of $1.41 in less than 4 months during 2008, before subsequently rallying back up to $1.51 before the end of 2009. Fiber's longer term average monthly range is around 500 pips, and it very quickly jumped to three times that amount back in 2008. Those were clearly exceptional circumstances, and it's therefore not surprising that a pyramid trend-following EA like Black Belt would have been able to latch onto some of the trends and profit much more than it would have been able to whilst operating during normal, more sedate market conditions.

When I looked at both 2008 and 2009 as individual years, I found that Black Box would have made over 25,000 pips in 2008 and nearly 13,000 pips in 2009. Removing these two years from the 12-year backtest results leaves us with 37,218 pips generated from 6,046 trades over 3,648 days. In other words, over half of all the pips gained in the 12-year backtest would have been won in just two of those twelve years. The win rate was unaffected by removing those two years and stays close to 53%, but the average winning trade size reduces to 44.7 pips whilst the average loss size only changes marginally to 36.9 pips.

We're now trading in 2012, markets have returned to a much more orderly fashion with average monthly ranges having fallen back to 600 or 700 pips, and the crazy events of 2008 and 2009 have become little more than a distant memory. I don't really feel, therefore, that it's fair to include Black Box's extraordinary performance of those two years within a current analysis of the EA.

Despite this, the results for the other 10 years of the backtests are still quite impressive, although we're now talking about an EA which looks to be gaining 310 pips a month instead of the 523 pips suggested previously, and whose trade expectancy is only 6.2 pips per trade instead of 10.0 pips. The reduced expectancy means that the EA is getting into the realms of possibly being broker dependent, as an extra pip on the spread, or a pip or two of slippage here and there, is likely to impair results noticeably. Unfortunately, I don't know what spread was used in the developer's backtests.

Remember also that the developer's backtests appear to be using MetaQuotes history data which has some notorious gaps of a month or more where history data is missing completely. I've not looked closely enough to say whether or not the MetaQuotes gaps are impacting 2010 and 2011 backtest performance for better or worse.

The worst case drawdown when trading 0.1 lots in the backtests was $2,294. Ordinarily, a risk simulation on the test results would give an idea of the minimum balance needed to run an EA but, in the case of Black Box, I'm uncertain as to precisely how much correlation there is between the pyramid trades. Risk simulations involve replaying a set of trades in a random order to get an idea of the worst thing that could happen, but I'm not so sure that it's appropriate to replay Black Box's trades in a random order because the continuation trades are only ever opened on the basis that the initial open trade is in profit and the baskets of open trades should remain in order.

Nonetheless, I did conduct a risk simulation to get an idea of the minimum balance required to run Black Box, and the simulation suggested a minimum deposit of around $4,300 for an account trading 0.1 lots. This roughly concurs with the developer's assertion on his website which suggests that $500 would suffice for a microlot account. My only reservation here is that using such a small account small could potentially spook users into turning the EA off once they witness it first hand opening multiple pyramid trades.

Bruce LeeReaders should also note that the largest single winning trade in the backtests was over 600 pips and the largest losing trade was over 500 pips. Admittedly, these worst case scenarios both happened during 2008/2009, but the biggest loss during the other 10 years was still as high as 473 pips which might leave some users running scared.

Despite the various caveats above, the performance is still likely to be better than that of most EAs, and I certainly hold the belief that this EA could prove a winner in the weeks and months ahead. In all seriousness, if Black Box can achieve an average return of over 300 pips each month, it will quickly find itself sitting very close to the top of my leaderboard.

Besides which, I really don't want to receive a visit from Bruce Lee! :)

For now, I've put the Black Belt EA into forward test on a $5,000 FXCC demo account. FXCC have recently changed their account structure and are now offering micro lot sizes.

The EA does not contain much by way of user settings to adjust, which is the way I think things should be. There is an option to either trade fixed or variable lots, and I've chosen the variable lot option for my forward test and have set the 'balance_percent' parameter to 2.0.

Readers should note that this EA handles risk in the opposite way to most commercial EAs, and users need to INCREASE this parameter in order to REDUCE the risk. I'm expecting to see the EA take trades of 0.05 lot size until the account balance has increased sufficiently for the EA to risk larger lots.

As with all my EAs in forward test, I'm publishing the Black Box EA's trades to MellyForex, and its performance can be viewed in detail by clicking here.]]> Mon, 27 Feb 2012 11:36:57 +0000
PhiBase PRO EA - Expert Advisor in Forward Test PhiBase PROThe latest EA to enter a forward test at MellyForex is called PhiBase PRO, and it claims to use advanced chart pattern recognition techniques to determine its trade entries.

In fact, there's quite a bit of marketing blurb on the developer's website which explains some of the thinking of modern-day patternmeisters like Gann and Elliot, and subjects such as Phi and Fibonacci numbers are also discussed.

Elliot waves have actually been around for some time, but I've got to be honest and say that I've never been able to spot them when it matters the most, which is during their formation. I've only ever been able to pick them out on historical charts, and it's invariably a bit too late by that stage to make any money from them. :(

Despite all the fancy gobbledegook about Ermanometry and logarithmic spirals, PhiBase PRO appears to use much more basic chart patterns as part of its modus operandi. The website refers to trade entries being based upon things that I can understand - like double and triple tops and bottoms, support and resistance breaks and wedges - none of which has anything whatsoever, in my opinion, to do with any of the fancy names I referred to above. That said, I'm not sure that it matters precisely how the EA works, just so long as it does actually work and it's profitable.

Let's try and find out if that's the case with this EA ...First and foremost, PhiBase PRO runs on the EURUSD symbol on the 1-hour chart timeframe and there are some strategy tests on the vendor's website which extend back nearly 5 years using some 99% modelling quality tick data. Those backtests certainly suggest that the PhiBase PRO EA is likely to be profitable, and that the EA is likely to return a win rate slightly in excess of 70%.

Although I haven't carried out any backtests of my own, I did conduct an analysis of the tick data backtests which show an average winning trade size of 46.3 pips against an average losing trade size of 78.7 pips. Whilst I don't know what spread was used in the backtests, the results show that the trade expectancy averages a whopping 11.1 pips, so the suggestion is that there is a bit of leeway when it comes to using PhiBase PRO with brokers with wider spreads or who like to slip a pip or two to help pay for their Christmas parties.

Those backtests also showed a total of 788 trades in 1,705 days of test, so this EA is only going to trade around 3 times each week. The backtests show its largest recorded stop-loss was in the region of 250 pips, so users are going to need to keep a stash of Valium to hand if they are of a nervous disposition. Similarly, the largest winning trade in the backtests was in excess of 700 pips, so it's important that traders resist any temptation to take manual control and close prematurely, otherwise they may find that the EA's overall long term performance is impaired. The EA sets a stop-loss, which it also trails, but I noticed that it also closes trades of its own accord, generally at the start of a new 1-hour price bar.

My only other observations at this stage are that the EA seems to be ECN-style broker compliant and it is also NFA hedging/FIFO compliant.

I've put PhiBase PRO straight into a forward test and it's my aim to analyse and review it in closer detail if and when it performs profitably in that forward test. My test is running on a $5k FXCC demo account which charges a 1.0 pip round turn commission per trade.

I'm assuming that the developer optimised his EA properly before release and that the default settings are optimum, so I've left the default settings as they are, with the exception that I've lowered the default variable lot size setting of 10.0 to a value of 3.0. The backtests at the default setting experienced far too high a drawdown for my personal comfort!

There do not appear to be any settings which limit either the maximum spread or slippage, although the EA does display a message on screen whilst in use which suggests that it only trades within the confines of a maximum allowed spread.

FXCC are a raw spread ECN broker, so slippage shouldn't even be an issue within my test.

As with all my EAs in forward test, I'm publishing the PhiBase PRO trades to MellyForex and its performance can be viewed in detail by clicking here.]]> Thu, 16 Feb 2012 11:01:58 +0000
Fast Forex Millions EA - Expert Advisor in Forward Test Fast Forex MillionsSome EA vendors take a softly, softly approach when releasing a new Expert Advisor onto the market, and product launches can often be relatively modest and quiet affairs, limited to a solitary email blast to prospective users.

This was most certainly not so in the case of the Fast Forex Millions EA which has been hyped like crazy since I first announced its release in an earlier article. I guess the fact that the developer has been behind some other popular EAs has simply served to heighten expectations.

Anyway, the Fast Forex Millions EA has now been released and I'm pleased to report that the vendor has provided MellyForex with a review copy which I've put into forward test so that readers can monitor its performance.

At this stage, I have to be perfectly honest and say that I still don't know too much about the EA other than that it works on five of the major currency pairs where liquidity shouldn't be an issue, and that it's a non-scalper (the manual suggests that trades can typically last up to 17 hours) which doesn't use Martingale.

So what else can I tell you about the EA?Well firstly, it's my guess that, despite being a black box, the Fast Forex Millions EA is indicator driven. I say that because the EA's installer unpacked a template which included ATR, Fractals and Force Index indicators displaying within a chart window. In addition, the user manual suggests that the EA's stop-loss and trailing profit are dynamically adjusted and relative to the ATR. Step aside Sherlock Holmes, MellyForex is after your job. :)

Secondly, it looks as though the EA will work with both NFA brokers and other brokers, because there is an external parameter called "DualMode" which either enables or disables hedging. In addition, the EA looks as though it will work with ECN/STP type brokers where the stop-loss can only be set once a trade has been opened.

Finally, the EA contains a money management system which allows users to either trade fixed lot sizes or to risk a percentage of the account balance on each trade.

To begin with, I've put Fast Forex Millions straight into a forward test and it's my aim to analyse and review it in closer detail if and when it performs profitably in that forward test. My forward test is running all five suitable pairs (EURUSD, GBPUSD, AUDUSD, USDCHF and USDCAD) on a $5k FXCC demo account which charges a 1.0 pip round turn commission per trade.

For the time being, I've followed the developer's recommendations and set Fast Forex Millions up using the default risk settings, with the exception that I've lowered the default risk setting of 10.0 such that I'm only risking 3.0% of the account balance per trade in my forward test.

There do not appear to be any settings which limit the maximum spread, but there is a parameter which controls maximum slippage, and I've set this to zero. FXCC are a raw spread ECN broker, so slippage shouldn't even be an issue within my test.

As with all my EAs in forward test, I'm publishing the Fast Forex Millions trades to MellyForex and its performance can be viewed in detail by clicking here.]]> Fri, 20 Jan 2012 11:57:06 +0000
An Update Regarding Plimus Many of you reading this will already be aware that, as of a couple of days ago, Plimus decided to suspend accepting new orders for Forex related products.

Without prior warning, on the 10th January, Forex vendors selling their wares through the Plimus network received an email similar to the one below:

The credit card associations (including Visa and MasterCard) are increasing their scrutiny of merchants selling products within a number of online segments. Affected market segments include, but are not limited to, the following:

- Online auctions
- Foreign Exchange (FOREX)-related products
- Media
- Poker chips and other virtual currency

As you offer products that fit within one or more of these segments, Plimus must now undertake a detailed and exhaustive review of your account, and its compatibility with the Plimus platform. This review will take approximately 30 days; at its conclusion, we hope to be able to advise you that we can continue to provide e-commerce services to your business.

During the period of this review, we must undertake two steps. First, the payout scheduled for the 15th of this month will be suspended for a period not to exceed 30 days, pending our confirmation that all transactions will be honored by the card associations. Second, your account will be suspended as of Thursday, January 12th, which will mean that no further transactions can take place until further notice.

We recognize that this represents a significant business challenge, and apologize for the inconvenience. If you believe your account should not be considered to be within one of the stated market segments, or if you have any questions of concerns, please contact Vendor Support directly."

So, what does all this mean?Well, first and foremost, the chargeback rate of Forex EAs is significantly higher than the industry average. Chargebacks occur when somebody buys something and pays using their credit card. Let's suppose the buyer subsequently asks for a refund, and gets an unsatisfactory response from the vendor. He then goes directly to his credit card company, tells them that the product is faulty and requests that the credit card company refund him directly themselves. The credit card companies automatically make the refund and then pursue the matter with Plimus. Given that the credit card companies pay in arrears, they are able to call the tune and simply withhold the disputed money from Plimus. Plimus then have to adjust their own accounts and claw the money back from the vendor which, often, they will have already paid out themselves.

It can all get a bit messy, to say the least.

Obviously a small level of chargeback is only to be expected. I don't know if it's true, but I have heard that between 0.25% and 0.5% of all sales is the typical industry chargeback level for software downloads, and that chargebacks within the Forex niche are many times higher than these levels.

The chargeback system is one which was designed to protect innocent victims who have had their credit card stolen and then used fraudulently. The system is also designed to protect purchasers who don't receive their goods. It's increasingly being used, however, as a means of obtaining a refund by buyers for no reason other than that they simply don't like what they've bought. Clearly, the dissatisfaction level amongst purchasers of Forex EAs is pretty damn high, and it's fair to say that the credit card companies have had enough!

Now I could write a 10,000 word essay about Plimus's actions and also about my personal opinions of this murky little industry in general but, fortunately, I won't bore you all with that.

In a nutshell, Plimus have been pressured to do something about the high level of chargebacks, by suspending the vendor accounts of Forex related products as per their circular which I pasted above. On MellyForex, around 70% or 80% of the EAs I have in forward test are sold by vendors using the Plimus network. As of today, my original affiliate links which take readers to a vendor's website via Plimus no longer work because Plimus has suspended all of those products.

I am in the process of editing all of my links so that the links take readers directly to the vendor's website via a non-affiliated route. The problem is that, when you get onto the vendor's website, you probably won't be able to buy anything because Plimus won't processs the sale.


In the meantime, the EAs in forward test at MellyForex will continue to function and I hope that the vendors are able to implement their necessary changes as quickly as possible. It certainly strikes me that not all Forex EA vendors are rogues, and there are a lot of good guys being unfairly penalised here. I have to say that it would have been nice of Plimus to have given a bit more notice than they did for vendors to find alternative payment processing networks!]]> Fri, 13 Jan 2012 09:43:03 +0000
Fast Forex Millions - New EA Coming Soon Fast Forex MillionsI hope to have a copy of the new Fast Forex Millions Expert Advisor in forward test at MellyForex very soon.

Although the EA hasn't been released yet (official release is an as yet undisclosed date sometime this month), it certainly looks very interesting and suggests great potential.

Fast Forex Millions claims to work on no fewer than five currency symbols, it's a non-scalper and the suggestion is that it will work with any broker. The backtest summaries on the website's homepage suggest a win rate of around 60% and an average winning trade size which is greater than the average losing size. I'll certainly be looking to put it through its paces and reporting back to you with my findings and views.

Watch this space!]]> Fri, 06 Jan 2012 10:07:30 +0000
Forex Thor EA - Expert Advisor in Forward Test Forex ThorIn Viking Norse mythology, Thor was a hammer wielding thunder god, and it's my best guess that this old mythology has played a big part in the naming of the Forex Thor EA which I've just put into forward test at MellyForex. Was it really that difficult to make my deduction by looking at the streaks of lightning and the hammer which are clearly displayed on the Forex Thor box? :roll:

Anyway, now we've established that Inspector Clouseau needn't worry about the safety of his employment, please allow me to tell you what little snippets of information I know about Forex Thor.

Firstly, the Forex Thor EA is a scalper which, at the time of writing this article, is brand new to the market. It runs on the 30-minute chart timeframe of the EURUSD symbol, and I have to be be perfectly honest and say that I'm not entirely sure what logic it uses to get its mojo. The video on the developer's website talks about money flow, so I can only hazard a guess that trade decisions are volume related.

The literature suggests that the EA employs a combination of three different strategies, and each strategy can either be used independently of the other two, or all three strategies can be used together concurrently. From the developer's own backtests, it looks at first glance that employing all three strategies together may render the EA as NFA non-compliant in terms of FIFO. The backtests certainly showed Forex Thor as having more than one trade open at a time, and the stop-loss handling didn't look to me as though it complied with FIFO. It's my initial suspicion, therefore, that US-based traders who want to run Forex Thor with all three strategies will have to run the EA on three separate broker's accounts, with only one strategy per broker. Obviously users in the rest of the world won't have any problem trading all three strategies together.An examination of the developer's own 2011 strategy tests (it's a 7.90MB file download if you're on a slow connection) suggests to me that Forex Thor is a reasonably frequent trader, as it looks to take between 15 and 20 trades each week across all three of its strategies. Those strategy test results also suggest that the EA doesn't employ any time filtering, as it seems to place its orders 24/5. I did notice, however, that Forex Thor may take several trades within very quick succession on some days yet, at other times, it may go several days on end without taking any trades whatsoever.

Forex Thor appears to work by placing individual stop orders and waiting for a fill. The stop orders are very close to the market price, and the EA frequently adjusts those orders very quickly before they are filled as it attempts to get a better entry level. The EA also recognises if the trade entry criteria have changed, in which case it will have no hesitation in deleting a pending order ahead of a fill. Having entered a trade, Forex Thor trails a tight stop-loss and also dynamically adjusts its take-profit. The outcome is that trade durations are generally only a minute or two with an average winning trade size of around 7 pips and an average loss of about 3.5 pips. The expected win rate is around 75%, and the combination of low risk/reward ratio and high win rate means that Forex Thor has the potential to be very profitable.

This style of trading does have a drawback, however, and it is most likely that the EA is best suited to ECN brokers with the tightest of spreads who permit scalping and who have stops levels of only 0.1 pip so that the EA can work efficiently so very close to the market. With that in mind, I've set my copy of Forex Thor up on a $5k FXCC demo raw spread ECN account, in the hope that I can exploit the EA's true potential. FXCC have no issues with scalping whatsoever and, because they don't need to worry about the NFA rules, I'm able to run all three of Forex Thor's strategies concurrently.

For the time being, I've followed the developer's recommendations and set Forex Thor up using the default risk setting. FXCC do charge a 1.0 pip round turn commission per trade, and I do hope that this won't impact upon profits. Ordinarily, I would adjust an EA's TP and SL settings to reflect this commission, but Forex Thor doesn't provide this option.

I'm permitting a maximum spread of 2.0 pips and have set the slippage level to 1 (I assume this means 0.1 pip as FXCC are a 5-digit broker), as this is the minimum value allowed by the EA.

As with my other robots in test, you can monitor Forex Thor's performance at MellyForex by clicking here. If Forex Thor performs well over the next month or so, that will be the point in time that I may conduct a more detailed review.]]> Wed, 14 Dec 2011 14:40:40 +0000
Orange Forex Robot - Expert Advisor in Forward Test Orange Forex RobotHere in the UK, we have a mobile telecoms provider called Orange whose advertising slogan used to be - "The future's bright, the future's Orange". I certainly hope that the future will be bright for the Orange Forex Robot, which is the latest EA to go into forward test at MellyForex.

Is it just me, or are the names of these EAs becoming increasingly obscure? :-P

Anyway, I'm not the kind of person to allow a name to cloud my judgment, so I'll quickly tell you a few important snippets of information about the Orange Forex Robot.

First and foremost, it's a scalper which works 24/5 on the 5-minute chart timeframe of the EURUSD currency pair. I already have several EURUSD scalpers in forward test at MellyForex, but they generally apply extensive filtering which means they may typically trade only once a day.

What's different about the Orange Forex Robot, however, is that the backtests on the developer's website suggest that this EA is a reasonably high frequency trader. The strategy test from January 2008 to date shows around 12,000 trades in under 4 years, which equates to 10 or 11 trades each and every trading day.The high frequency aspect of the Orange Forex Robot certainly offers the potential to increase account balances exponentially, and the aforementioned strategy test shows a phenomenal 100,000% return in under 4 years, but there are also trade-offs which are necessary to achieve these kind of results.

The first trade-off is that the EA seems to maintain its very high win rate (around 94%) by banking only very small profits. The average win size from the strategy tests is only 3.35 pips, so it's pretty clear that this EA is only designed for ECN-type trade execution. The likelihood of Orange Forex ever working with a fixed spread broker who also acts as a market maker is, therefore, pretty remote. What's more, from my discussion with the developer, I'm not entirely sure that the EA will work at all with brokers who charge a commission on their trades. The developer suggested to me that the EA be set up on a ThinkForex demo account, simply because they don't charge commissions.

When I looked closer at the subject of using the Orange Forex Robot with a 'commission broker', I noticed that the EA's expectancy per trade is a mere 0.7 pips. If I were to set the EA up on a raw spread FXCC ECN account, as I would normally use for my MellyForex forward tests, I would be subject to a 1.0 pip commission per trade. Depending upon how the EA has been coded, it's possible that the commission charges could result in a negative expectancy, conceivably as low as -0.3 pip per trade, in which case the EA would be loss-making overall. Regular readers will know by now that I generally counteract any broker commission charges by increasing the take-profit amounts by 1.0 pip and reducing the stop-losses by the same amount. This approach allows me to benefit from great execution with razor-sharp raw spreads without profits being affected by commissions but, unfortunately, there is no way of making these adjustments within the Orange Forex Robot.

The other sacrifice made by this EA is a rather high risk/reward ratio of over 12, so a few loss trades in quick succession are invariably going to impact heavily upon profits.

For the time being, I've followed the developer's recommendations and set the Orange Forex Robot up on a ThinkForex $5k demo account. Whilst claiming to be ECN, I should reiterate that this is not a raw spread account, and the ThinkForex spread appears to be padded by a pip or so to disguise the broker's commission.

The EA only contains one parameter relating to risk, and I have set its value to 0.35. Unfortunately, this value is arbitrary and doesn't really relate to anything as such.

As with my other robots in test, you can monitor the Orange Forex Robot's performance at MellyForex by clicking here. If the Orange Forex Robot performs well over the next month or so, that will be the point in time that I may conduct a more detailed review.]]> Wed, 30 Nov 2011 11:01:57 +0000
Forex Cleaner EA - Expert Advisor Review Forex Cleaner EA has shown much early promise since I put it into forward test at MellyForex.

Some of that is down to good fortune in timing, of course, as I also have EAs in forward test which started badly and are only now showing their true potential after several months of sub-standard performance. Likewise, it's only fair to say that I also have EAs which have started well before going into terminal decline. In the case of Forex Cleaner, it's been in forward test for little more than a month at the time of writing this review and, in that time, it has zoomed up towards the top of my leaderboard where it currently sits proudly in third place.

Forex Cleaner

In terms of its modus operandi, I suppose Forex Cleaner could be described as a 'part-time swinger'.

Before my lusty male readers jump to false impressions that this article is all about wife-swap parties, it's probably best if a describe my term of 'swinger' in greater detail. For a start, swing trading is the process of identifying short-term tops and bottoms in the market, and trading the move between those tops and bottoms. As soon as a trade in one direction is exited, the trader generally hops back on board in the opposite direction in the hope that he can profit from the price reversal. Using this approach, trades generally follow a buy-sell-buy-sell pattern.

Forex Cleaner attempts to identify market tops and bottoms for its trade exits, but also applies some additional filtering which means that it won't take every entry unless those additional criteria are met. In other words, Forex Cleaner only swings part-time. What did you think I meant? ;)About the EA

Forex Cleaner trades without time restriction 24/5 on the 30-minute charts of the EURUSD pair, and determines its trade entries and exits in a manner such that it only trades at the very start of a new 30-minute price bar.

Initially, I think there was a bit of confusion over this behaviour because a forward test on the developer's own website clearly showed trades being closed mid-bar, which I referred to within my earlier Forex Cleaner introductory article on MellyForex here. Having watched Forex Cleaner trade in the MellyForex forward test for over a month now, and having conducted my own strategy tests for this review using an 'every tick' model, I'm now satisfied that my copy of Forex Cleaner is behaving as intended and I'm going to put the anomalies of the developer's own forward test down as being one of life's great unsolved mysteries, ranking alongside that other great mystery in life of why it is that wearing red makes my bottom look big. :?

As might be expected of any EA trading on the 30-minute chart timeframe, the moves between the swing highs and lows can be reasonably substantial, and my strategy tests will show that Forex Cleaner looks to win around 70% of all its trades with an average winning trade size of just over 40 pips and an average losing trade size of about 70 pips. Trade expectancy is around 7 pips per trade, which is markedly higher than the majority of EAs that I have in test at the moment.

Finally, Forex Cleaner only ever has one trade open at a time and is fully NFA hedging and FIFO compliant.

Setting Forex Cleaner Up

In common with many EAs of this day and age, Forex Cleaner is delivered wrapped in its own installer which extracts the relevant files to the necessary MetaTrader folders. The developer tells me that each Forex Cleaner licence permits use on up to three demo and two live MetaTrader accounts.

Regular MellyForex readers will know by now that I like the user interface of my EAs to be simple and easy to use, with only the minimal number of external parameters to adjust. In the case of Forex Cleaner, I'm happy to report that the EA only contains eight external parameters, the main ones being to do with the EA's trade position sizing.

There is also an option within the external parameters which enables users to decide whether to set a stop-loss on each trade. By default, Forex Cleaner sets neither a stop-loss nor a take-profit on its trades and, instead, it uses some dynamic internal hocus-pocus to exit its trades. Obviously some users will be unhappy about the idea of trading without a stop-loss being declared, so the EA also contains an option to set a fixed 135 pip SL on its trades. Use of the fixed stop-loss does not affect the dynamic closing of trades in any way and, if users choose to use the stop-loss function, those stop-losses aren't entered until after an order has been opened, making the EA fully ECN/STP broker compliant.

Strategy Tests

Because Forex Cleaner is supposed to only open and close its trades at the start of a new 30-minute price bar, it should be possible to use the 'Open Prices Only' model for strategy tests, and these should produce identical results to if the 'Every Tick' model were used. Having noticed anomalies in the developer's own forward test, with some trades being closed mid-bar as explained above, I decided to use the 'Every Tick' model for my strategy tests. I wanted to examine the results in detail to ensure that trades were only being opened and closed on the new price bar.

In truth, however, the Forex Cleaner strategy tests run very quickly and it really doesn't matter which model is used.

Ordinarily, I would conduct a first round of strategy tests using Tadawul who are an older style 4-digit broker with a fixed 2.0 pip EURUSD spread. As I've explained previously, these results will never be the best, but I do think they are fairly representative of what the average user should be able to achieve. When I started to run some Tadawul tests, however, the first thing I noticed was that countless order send price and volume errors were being reported to the logs, which resulted in only around 6% or 7% of all trades getting away as intended. Anyway, to cut a long story short, the developer is aware of the bug, it only relates to 4-digit brokers, it's very easy to rectify, and it will have hopefully already been fixed with an update issued by the time you read this article. In the meantime, if your Forex broker uses older style 4-digit pricing, you're going to have a problem running Forex Cleaner and will need to switch to a modern 5-digit broker.

Unable to use Tadawul for my strategy tests, I opted to use Pepperstone instead, and I fixed the spread at 1.5 pips which is a fairly typical EURUSD spread nowadays.

For my first strategy test, I used fixed 0.1 lots without a stop-loss, instead allowing the EA to control its stop-losses internally.

Pepperstone - EURUSD - 1.5 pip spread - Fixed 0.1 lots without stop-loss

Forex Cleaner Strategy Test

For my second test, I used a variable risk setting of 3.0, again without a stop-loss.

Pepperstone - EURUSD - 1.5 pip spread - Variable risk, setting = 3.0 without stop-loss

Forex Cleaner Strategy Test

Although I haven't mentioned it previously within this article, using Forex Cleaner without a stop-loss means that the risk setting doesn't actual relate to anything. Ordinarily, I would expect it to mean that I was risking a fixed percentage of my account on each trade, be it a percentage of account balance, equity or margin. However, Forex Cleaner doesn't close its trades once the stop-loss value is reached and, instead, holds onto the trade until the start of the next price bar. The market could obviously move much further against the trade while the EA is waiting for a new bar to start!

An examination of my two strategy test results above show that, although the stealth stop-loss is 135 pips, there were 37 losses in a total of 1,576 trades which were between 150 and 200 pips, and 5 losses of 200 pips or more, with the largest of those losses being 250 pips. Obviously, this doesn't tell the full story, as there were also numerous trades which were successfully closed for considerably less than 135 pips. So, for those traders with a nervous disposition, I also ran the strategy tests allowing the EA to set a fixed 135 pip stop-loss so that I could compare results.

Pepperstone - EURUSD - 1.5 pip spread - Variable risk, setting = 3.0 with 135 pip stop-loss

Forex Cleaner Strategy Test

Having carried out these tests, I did look closer at those trades which had run to a full 135 pip stop-loss, and noticed that, with the risk setting that I had used of 3.0, the losses seemed to vary randomly between 3.6% and 8.1% of the account balance. This suggests to me that Forex Cleaner calculates its lot sizes based upon something other than the distance away of the stop-loss.

Finally, and mainly because the strategy tests do run very fast, I carried out some optimisation of the risk parameter setting. Obviously the net profit increases with greater risk but, of equal (if not greater) importance, the drawdown also increases with that extra risk. Readers may use my optimisation results to select a risk level to suit their personal needs.

Forex Cleaner Risk Optimisation

Walk Forward Tests

One of the benefits of reviewing an EA once it has been in forward test for a month or more, is that I can include a minimal amount of walk forward testing within my review. To explain a little bit more, longer term 10-year backtests within my reviews are carried out using data from the MetaQuotes History Centre. I personally use the data for my reviews because it's easy to obtain, but EA developers also generally use it when they optimise the settings of their EAs. In reality, however, different brokers use the price feeds of different liquidity providers and these feeds will not only differ from the MetaQuotes feed but, because there's no such thing as a central regulated Forex exchange, the price feeds will also vary from one broker's liquidity provider to the next. This raises obvious questions as to the reliability of strategy test results and, as a result, we can never be sure if an EA is going to perform live in the manner that the strategy tests suggest.

By having some live results from the MellyForex forward test over the last few weeks, I can compare the trades taken in my forward test against some backtests for the corresponding period to see if the same set of trades would have been taken. Firstly, this comparison gives me some sort of an idea of just how reliable the MetaQuotes backtest results actually are and whether or not an EA has been curve fitted towards the MetaQuotes data. Secondly, by downloading the small amount of recent history data which is available directly from different individual brokers' own servers for the period in question, it's also possible to compare performance with those different brokers to get an idea of how broker dependent an EA might be.

In the case of Forex Cleaner, I compared the MellyForex live forward test with FXCC against a strategy test using MetaQuotes History Centre data for the identical period. I also compared those results against backtests for the same period using broker's own data obtained from both Pepperstone's and GoMarkets' own servers.

It's quite evident from the equity curves below that there are differences in the number of Forex Cleaner trades taken between different brokers during the same six week period between 12th October and 23rd November 2011, and users should therefore be prepared to accept that they are likely to witness different levels in performance with Forex Cleaner from one broker to the next. From my understanding of how the EA works, I can see there is an important reason for these trade differences, but I don't feel that it would be appropriate to explain the inner workings of the EA within this article.

Forex Cleaner Walk Forward tests

I should also say that, just because Forex Cleaner's performance from my FXCC forward test is clearly better than the suggested performance with other brokers, there's no evidence to suggest this trend will continue into the future. I'm an IB for FXCC, so I'd obviously like you to click on my link and open an account, but please don't do it just because one EA has performed well with them for a few weeks. FXCC are, in my opinion, a good honest Forex broker who don't trade against their clients, but I do need to be fair about these things.


In keeping with my normal practice, my final piece of analysis involves some risk simulation tests on the 10-year strategy test results. I felt it best to analyse the results both without and without the fixed 135 pip stop-loss being applied.

Forex Cleaner Risk Simulation

Not surprisingly, use of the fixed SL results in a marginally lower risk/reward ratio together with a lower drawdown. More important than this, in my opinion, is the Monte Carlo simulation which suggests that only around 20% of users are likely to not see their account balance drop below their initial deposit. Conversely, around 80% of users will be out of the money at some stage, which is a relatively high figure when compared to most profitable EAs. Similarly, the likely minimum number of trades necessary before the Forex Cleaner's return exceed its drawdown which has occurred in the meantime is also relatively high. Given that the EA typically trades around 3 times each week (albeit my walk forward tests suggest that figure is likely to vary between brokers), it may take around 11 weeks to achieve this if the fixed SL is being applied, and around 14 weeks if it's not being used.

Notwithstanding its level of broker dependency, Forex Cleaner certainly looks like it will be profitable over a sustained period of time. In keeping with most profitable EAs, it is likely to endure periods of weeks or months where it merely breaks even, but users should accept these periods for what they are as an occupational hazard of trading.

Forward Test

Forex Cleaner is currently in forward test running on a $5k demo FXCC raw spread ECN account.

My initial settings were such that I thought that I was risking a maximum of 3.5% of the account on each trade and I had been using the EA without a stop-loss.

Having carried out my analysis of the EA, however, I'm now of the opinion that my initial risk setting was almost certainly too high.

I have therefore lowered the risk setting to 2.5, and also set the 'stealth mode' parameter such that the EA will now start setting 135 pip stop-losses on its trades.

Forex Cleaner's performance can be monitored on MellyForex by clicking here, and there is a thread on the MellyForex forum here where readers are welcome to ask questions and discuss their experiences with Forex Cleaner.]]> Thu, 24 Nov 2011 10:48:02 +0000
Forex OverDrive EA - Expert Advisor in Forward Test Forex OverDrive EA is another Expert Advisor that I've recently put into forward test at MellyForex. If it shows some promise once it's been in test for a month or so, then I'll consider a full review of the EA.

Forex OverDriveThe EA itself is a non-martingale scalper which runs 24/5 on the 5-minute charts of two currency pairs - EURUSD and GBPUSD, and it appears to get its trade signals when about 500 different indicators are all standing up like soldiers and facing the right direction. Hey, 500 might be a slight exaggeration of the number of indicators used by Forex OverDrive, but I'm sure that readers will understand the concept that this EA is being driven by a shedload of different indicators. :)

Forex OverDrive sets a fixed stop-loss (default setting is 328 pips) and take-profit (default is 25 pips) once it's entered a trade, and it then trails a stop-loss to exit if or when the trade has moved into profit.

A consequence of the EA's high risk/reward ratio is that the EA should enjoy a very high rate of winning trades, and some forward tests on the developer's website suggest that the EA actually wins around 95% of all its trades. The downside of this high risk/reward ratio, however, is that a full stop-loss being hit will wipe out the entire profits of the last umpteen trades. I ran a quick analysis of the developer's forward test, and it suggested that Forex OverDrive's average winning trade is around 14 or 15 pips and its average loss trade is around 260 pips, giving a risk/reward ratio of approximately 18. That same analysis also suggested that the EA should trade around 7 or 8 times each week on both its currency symbols, so it's relatively easy to work out that it's going to take around two or three weeks to recover from one of the big losses.Interestingly, a deeper analysis of the same forward test suggested that whilst Forex OverDrive seems to have been profitable over the last 18 months or so on the EURUSD pair, where it has been winning around 97% of its trades, it has been loss making on GBPUSD where it has only won around 93% of its trades during the same period. I suspect this is down to the fact that both pairs seem to use the same stop-loss setting expressed as pips, and that the EURGBP cross valuation means that Forex OverDrive's EURUSD stop-loss is, in effect, looser than its GBPUSD stop, leading to a fewer number of full stop-losses being hit.

For my forward test of Forex OverDrive on MellyForex, I'm running the EA on a $5k demo FXCC true ECN account. This account offers raw spreads which are as tight as 0.1 pip, but it charges a 1.0 pip round turn commission on each trade, so I've increased the default take-profit value from 25 pips to 26 pips on both suitable currency symbols to compensate for the commission charges. Ordinarily, I would also reduce the stop-loss setting by 1.0 pip to reflect the commission charges but, seeing as the default SL is 328 pips as mention previously, I decided that an odd pip here would only make a negligible difference to overall performance.

Aside from the take-profit settings, I have set the EA up to risk no more than 4.0% of the account on any single trade.

Forex OverDrive also contains an option on whether or not to trade on Fridays. I've set this to true (ie. It will trade), as I'm not convinced from the developer's forward test that Fridays are a historically worse day than any other day.

Finally, I'm not allowing any slippage whatsoever on the trades.

As with my other robots in test, you can monitor Forex OverDrive's performance at MellyForex by clicking here.]]> Tue, 15 Nov 2011 07:20:22 +0000
MellyForex October Update Despite recent unsettling effects on the markets caused by global economic woes, and the unending advice suggesting that automated trading systems should be turned OFF completely until after the dust has settled, the number of profitable EAs in forward test at MellyForex has increased over the last month, from seven at the end of September to no fewer than twelve by the time everybody was out 'Trick or Treating' over Halloween.

Admittedly, I've got more robots in forward test now than I had a month ago, but I find it particularly refreshing that there are a couple of EAs which I've reviewed previously that backtested well, but hadn't been performing in forward test. This month they've made it into the top part of my leaderboard!

The point here is, I believe, that it's very difficult when you look at a 10-year/3,000 trade backtest curve which has been condensed into a relatively small 820 pixel wide GIF image, to fully appreciate that the backtest might show the EA having endured periods lasting 6 months or even a year during which it may not have been profitable. It's understandable, therefore, that the best EAs aren't always going to return a profit immediately they come out of the box. This is highlighted by the Monte Carlo simulations that I conduct within my reviews which suggest most EAs only offer a 30% or 40% probability of an account balance never dropping below its initial deposit. In other words, as many as 60% or 70% of EA users are going to have to suffer a bit of pain before they get the gain!

For what it's worth, I sometimes get the impression that a lot of people just look at the start and end balances of a backtest equity curve, but don't look closely enough at the shape of the curve in between. This leads to systems being dismissed unfairly after two or three bad months, simply because the average EA user doesn't fully appreciate either the risk involved or quite how much patience is needed to trade successfully.On the subject of patience, readers might like to wind the clock back a couple of years and take note of this news article about Man Group, the world's largest hedge fund which manages assets worth in excess of $40bn. At the time, their AHL division which specialises in automated trading had just posted a 16.9% trading loss.

It just goes to show that these things happen to the best of us. :)

Of the EAs which have made it into the top part of my leaderboard during October, I'm particularly pleased to see the Steady Winner EA and Forex Combo amongst them. I've previously reviewed both of those EAs in a favourable light and I hope to see them both stay high up in the leaderboard. I'm also pleased to see the Forex Real Profit EA strengthen its position, as it also endured a difficult time at the start of its forward test, but it moved into profitable territory a month or so back and has continued to perform well since.

Anyway, by way of a reference for posterity, here is the leaderboard as it stood at the end of October.

MellyForex October Leaderboard

The euphoria of the winning EAs aside, things are not looking so hot at the moment for one EA, namely Forex Gale, which I put into forward test towards the end of September, as it barely lasted a month before blowing the entire account. For anybody who is unaware, Forex Gale is a grid trading EA and it opened too many positions which went the wrong way forcing a stop-out. This is always going to be an occupational hazard of this type of EA, but I am worried that I may be partially to blame for the problem, as I started my forward test in the knowledge that I may have been trading the EA at excessive risk.

To explain a little more, the minimum lot size of raw spread, true ECN brokers such as FXCC (with whom the Forex Gale forward test was being conducted) is 0.1 lots and, putting it bluntly, my $5k account deposit was too small to trade Forex Gale at this lot size. I'm all for using forward tests to expose a system's strengths and weaknesses, but I don't really feel that anything productive has been learned from my little escapade. The only thing I've discovered is that I was over-trading (which I already knew anyway), so I've decided to try Forex Gale again and have restarted the test on a significantly larger FXCC demo account of $50k. The EA will still trade fixed 0.1 lot sizes, and I'm hoping it will survive this test in the hope that readers can get a clearer picture of the maximum drawdown, and find out what minimum account size is needed to run the EA.

I should say at this stage that I expect Forex Gale to behave in a completely different manner in this new test to how it performed previously. It banks profits when a basket of trades reaches a profit of 0.35% of the account balance. On my earlier test, it started banking profits once the profit reached $17.50 - now it won't bank its profits until the basket reaches $175.00, representing a far greater price movement within the various markets and leading to a completely different basket trade structure. As mentioned previously, Forex Gale backtests are totally worthless, and I've got no idea whatsoever how the forward test will go.

The one thing I would suggest to anybody who decides to use this EA on a live account is that they regularly withdraw any profits until they get to the point that it is using 'free' money. For now though, I'm going to leave the original 'blown' test in place on my leaderboard as a warning so that readers can witness the dangers of this type of EA, and I'm publishing the new forward test result to MyFxBook on this link and hope that it fares better than its predecessor.

At the time of writing this article, everything seems to be kicking off in Greece. What will be the result of their referendum? What question will be asked of the people within that referendum? Do they have enough money to survive until the referendum? And if the government doesn't survive its pending no-confidence vote, will there even be a referendum? It's impossible to predict what effects the constantly changing scenarios will have on the markets so, on that note, I'll say αντίο which (Babelfish tells me) is Greek for goodbye.]]> Thu, 03 Nov 2011 11:01:36 +0000
Forex Combo EA - Expert Advisor Review
Forex Combo EAThat can be loosely translated to read that I'll probably only review something if it's either profitable or if I like it a lot. :P

By way of an example, I've had the Forex Combo EA in forward test for over three months now, but it's only recently that it's moved into profitability and I feel it's become worthy of a full review.

The Forex Combo system gets its name because it is a combination of three separate EAs which have been merged into one. In fact, that's not entirely true, because users get one EA which is designed to work on cable (GBPUSD currency symbol) and a second EA which works on fiber (EURUSD symbol). The two EAs are essentially the same, but their parameters use different optimum values for each of the two symbols.
Each of the three strategies acts independently of the other two strategies, and users have the option of running either all or any of the strategies they wish. None of the strategies are particularly prolific traders in their own right although, with all strategies combined, the EA should trade on average once a day on each of its two currency symbols, and I guess that the strategies have been merged into one as a way of providing better value for money from the product.
About the EA

Forex Combo runs on two currency symbols as mentioned above on the 5-minute chart timeframe. The first thing to be said about the Forex Combo concept is that using multiple strategies within the same EA is invariably going to mean that the EA will, at times, attempt to take trades in contravention of NFA hedging and FIFO rules, thereby rendering it unsuitable for the vast army of US-based users who unfortunately have to abide by the NFA rulings. Fortunately the EA contains a couple of external parameters which disable both its hedging and FIFO aspects to satisfy those US users. I should say at the outset, however, that I haven't tested these particular features of the EA in any way and can't offer any guidance as to their effect on the EA's performance.

Secondly, and whilst on the subject of compliance, I suppose I should quickly mention that the EA sets its SLs and TPs only once an order has been opened, and is therefore fully ECN/STP broker compliant.

Regarding the three different strategies which are employed by Forex Combo, I can best describe them as follows:

Strategy 1 is a scalping strategy. Users set a parameter which evaluates the strength of a trend. If the trend possesses sufficient strength AND a different oscillator reaches a certain level, Forex Combo will enter a trade. The EA then sets a fixed pip stop-loss (default values are 100 pips fiber, 160 pips cable) and take-profit (22 pips fiber, 27 pips cable) which remain fixed throughout the duration of the trade. If the oscillator reaches a different level, Forex Combo closes the trade. The outcome is that trades either hit the fixed SL or TP, or they are closed using the internal witchcraft because the oscillator has hit the right number. Unsurprisingly, and in common with most scalping EAs, the combination of low take profit values and high stop-losses results in a relatively high win rate of between 75% and 80%, and Forex Combo should take around 3 trades a week on each of its two symbols using the scalping strategy, which represents over half of the EA's total trades.

Strategy 2 is described as a trending/breakout strategy. Forex Combo casts a spell to enter a trade, and then sets a relatively low SL (33 pips fiber, 60 pips cable) with a high TP (default 300 pips fiber, 300 pips cable). It then trails a stop-loss behind the price action if or when the trade moves into profit. The high TPs are very rarely hit and trades tend to be closed either by some internal magic or because the price retraces causing the trailing stop to be hit. My strategy tests will demonstrate that the effectiveness of the trending strategy varies considerably, and win rates are less than 50%. This strategy isn't employed very often, and only represents 2 or 3 trades each week on average split between both currency symbols.

Finally, Strategy 3 is a counter-trend/reversal strategy. This strategy is time filtered and it only seems to enter its trades during pre-Asian hours when the market is often at its quietest, only taking on average around 1 trade per week per symbol. Again, having entered a trade, the EA sets a fixed stop-loss (default 70 pips fiber, 260 pips cable) and take-profit (160 pips fiber, 100 pips cable). It also trails a stop-loss and, like the other two strategies, trades can be closed using some internal hocus-pocus. Although trades are entered using a time filter, they can stay open for several days which struck me as being rather unusual for a system with a time filter. Time-filtered trades are typically closed within several hours and normally before the next time session opens. In addition, my backtests suggest that the win rate of this strategy varies considerably between 67% for fiber and 79% for cable. I suspect those win rate inconsistencies are probably due to the different default risk/reward ratios being used on each symbol, although I do get the impression that this particular strategy just happens to be something that the developer has stumbled across which works, rather than it having any proven technical foundation.

Setting Forex Combo up

The fact that there are separate EAs for each symbol makes for a relatively easy setup, as the EA can run 'straight from the box' if wanted without a need to adjust any settings. That said, there is an extensive list of external parameters that the user can adjust if he's that way inclined. I am assuming that Forex Combo's indicator and TP/SL values have already been optimised prior to the EA's release so, from my perspective, I am only really interested in the obvious external parameters such as risk, slippage, etc.

On the subject of risk, Forex Combo can trade either fixed lots or, as stated within the manual, it can set its trade sizes as a percentage of the account free margin. On this point, I have to admit that I initially misinterpreted that paragraph of the manual, and I will explain more about this problem in a moment. As explained previously, the EA also contains parameters to turn each of its three strategies on or off, and there is another ominous external parameter called "LossFactor" which is Swahili for Martingale. The merest mention of the word Martingale brings me out in a cold sweat, and it certainly won't feature in any of my tests as part of this review.

Each Forex Combo licence can be used on one live account and an unlimited number of demo accounts. Users need to visit the members area of the Forex Combo website to enter and activate the account number(s) on which they wish to use the EA.

The Strategy Tests

Normally, I would run my strategy tests through with 2 or 3 different brokers to get an idea of how sensitive the EA is to different spreads, broker digits and broker lot sizing. In the case of Forex Combo, however, I wanted to test each strategy individually on each of its two currency symbols and it was pretty clear from the outset that my review would go on forever and the web-page would take an eternity to load if it were littered with countless images of balance curves from different brokers. I therefore opted to only test the EA using MetaQuotes History Centre data on a Tadawul MT4 platform, as Tadawul's performance should, in my opinion, be reasonably indicative of what the average user can expect to achieve. Obviously if you prefer to use a true ECN broker with razor-sharp spreads, then it should go without saying that the EA's performance is likely to show an improvement on the Tadawul performance.

For my first tests, I ran the EA through on both symbols with all three strategies set to run concurrently. For this test, I set the risk parameter to 3.0 for all three strategies, and the results are shown below:

Tadawul - EURUSD - 2.0 pip spread - Trading All Strategies - Risk = 3.0
Forex Combo Strategy Test

Tadawul - GBPUSD - 3.0 pip spread - Trading All Strategies - Risk = 3.0
Forex Combo Strategy Test

Having run these tests, I immediately noticed something that I wasn't comfortable with. I had wrongly assumed from the instruction manual that a risk setting of 3.0 meant that I would be risking 3.0% of my account free margin on any single trade. WRONG!! When I read the manual once more, I realised that the risk value is little more than an arbitrary number which has nothing to do with how much you stand to lose if a trade runs straight to its stop-loss.

To explain a little more, I come from a school where I was taught to never risk more than a small percentage of my account balance on any one trade. For relatively small accounts of $10k or $20k, that percentage may typically be around 4%, although the percentage risked would decrease as the account got larger. At school, I learned how to calculate my position sizes according to the percentage I was risking and the distance away of the stop-loss. For instance, if the stop-loss was 100 pips away, I was taught to take twice the size position than if the stop-loss was 200 pips away. If either of those example trades went wrong, my loss would still be the same even though one trade was twice the size of the other.

In the case of Forex Combo, it was pretty evident that the EA doesn't use the stop-loss distance in its lot sizing calculations. Combo's stop-losses vary by default between 33 pips and 260 pips, and its lot sizing should vary from trade to trade so that it suitably reflects those variations.

As mentioned previously, I've had Forex Combo in forward test for a while, and it had been set up to risk what I'd understood to be 3.0% of the account. Its trades are being published to MyFxBook on this link, and when I looked at them in detail I noticed that a couple of the individual losses so far have been as high as 7%. Oh dear. :(

Forex Combo Risk

Having identified what I considered to be a problem with how I had set up Forex Combo, I then attempted to adjust the risk settings of the individual strategies to balance the lot sizes such that they were risking closer to the 3.0% of the account balance I was looking for. I have to say that my attempts at doing this are very much a case of 'trial and error' as it's something that would be far better done through an internal code modification than through altering an external parameter setting 'on the fly'.

Having adjusted the various risk settings to risk approximately 3% of the account per trade, I then repeated my tests. The risk settings I decided upon were:

EURUSD - Strategy 1 - risk = 2.5
EURUSD - Strategy 2 - risk = 6.0
EURUSD - Strategy 3 - risk = 3.0
GBPUSD - Strategy 1 - risk = 2.5
GBPUSD - Strategy 2 - risk = 3.0
GBPUSD - Strategy 3 - risk = 1.3

Tadawul - EURUSD - 2.0 pip spread - Trading All Strategies - Risk = variable as described above
Forex Combo Strategy Test

Tadawul - GBPUSD - 3.0 pip spread - Trading All Strategies - Risk = variable as described above
Forex Combo Strategy Test

I must admit that I found it interesting that my risk adjustments resulted in reduced relative drawdowns on the GBPUSD symbol, but increased drawdowns on EURUSD, even though individual losses were now being limited closer to 3% of the account balance as intended. I guess it just goes to demonstrate that trading isn't an exact science and, if it were, everybody would be doing it. ;)

Unperturbed, my final set of tests involved running each strategy on its own for each symbol. I've reduced the image sizes to help the page load a bit quicker. In the image links below, strategy 1 is displayed on the left, strategy 2 is in the centre, and strategy 3 is on the right.

Tadawul - EURUSD - 2.0 pip spread - Individual Strategies 1, 2 and 3 - Risk = variable as described above
Forex Combo Strategy TestForex Combo Strategy TestForex Combo Strategy Test

Tadawul - GBPUSD - 3.0 pip spread - Individual Strategies 1, 2 and 3 - Risk = variable as described above
Forex Combo Strategy TestForex Combo Strategy TestForex Combo Strategy Test


My main frustration with Forex Combo relates to the issues with setting suitable risk levels as I described above. It would be nice to see a coding modification to address this problem in a later release. As things stand, prospective users do need to do a bit of homework before running Forex Combo to satisfy themselves that they will not be over-trading. Those frustrations aside, clearly each individual strategy tests differently to either of the other strategies as can be seen from the obvious differences within the set of six balance curves above. Overall, however, the three strategies do appear able to work together with reasonable harmony, although the combined balance curves do clearly show several periods during the course of the strategy tests where the EA would be susceptible to prolonged drawdown periods lasting several months on end.

In keeping with previous practice, I have carried out a risk simulation of my strategy test results to get a better idea of the EA's weaknesses. I have performed a risk simulation on each individual strategy as well as on the combined analyses.

Forex Combo Risk Simulation

What is interesting about Forex Combo's scalping strategy is that, at around 17 or 18 pips, the average pips won by the strategy is considerably higher than the typical scalping EA which often banks profits of under 10 pips. The scalping strategy also demonstrates a reasonably attractive risk/reward ratio and should require only a relatively small account deposit to get started.

Although the trending and reversal strategies both enhance the Combo system's overall performance, their use does also result in increased drawdowns and they require a considerably higher deposit in order to run the EA safely. The drawdowns would obviously be controlled through the individual risk parameters as I described in detail above, but I can see that anybody starting to run Forex Combo with only a small initial account deposit might wish to turn strategies 2 and 3 off completely until such time as their account balance had grown and they could run the other two strategies with a greater degree of safety. This would be quite understandable.

What did concern me from the risk simulation was the fact that the maximum drawdown when running both currency symbols was nearly twice the amount of each of the individual symbol's drawdowns. This suggested to me that the EA might be susceptible to drawing down on both symbols concurrently. If this were to happen, it would be best to either only run it on one symbol, or two run it at half the normal risk on both symbols so as to limit the drawdown.

To find out if this were indeed the case, and to err on the side of caution, I plotted fixed lot balance curves for each symbol to get an idea of where the drawdowns were happening exactly. The worst case drawdowns are highlighted by the red section of each plot below.

Drawdown Corellation EURUSD

Drawdown Corellation GBPUSD

Despite the fact that the maximum drawdown on both currency symbols occurs very close to the start of each test, it looks to me that this was just a minor blip as the plots then continued along their own separate paths and any correlation between the two symbols appears to be only minimal. It's my considered opinion, therefore, that there's insufficient correlation between the two plots to warrant either removing one of the symbols or lowering the risk, and I would be looking for the two plots to match each other more closely if I were to take those steps.

Forward Test

Having carried out the various tests and prepared my review, I've decided to alter some of the initial risk settings in the hope that it will result in a smoothing of the equity curve. The forward test is on a $5k raw spread true ECN FXCC demo account, and it was originally set up with all strategies trading at risk 3.0 at a time when I mistakenly believed that this level represented a percentage of the account balance. I've subsequently changed those original risk settings as follows:

EURUSD - Strategy 1 - risk = 3.5
EURUSD - Strategy 2 - risk = 4.5
EURUSD - Strategy 3 - risk = 3.0
GBPUSD - Strategy 1 - risk = 2.5
GBPUSD - Strategy 2 - risk = 3.0
GBPUSD - Strategy 3 - risk = 1.3

Hopefully, those new risk settings will result in improved performance onwards of this review date!

FXCC do not need to comply with NFA hedging and FIFO rules, so these features of the EA are disabled within my forward test.

The EA's other settings have generally been left at default, although I have set the allowed slippage to zero and am limiting Forex Combo's maximum allowed spread to 2.0 pips on EURUSD and 3.0 pips on GBPUSD.

As with my other robots in test, you can monitor Forex Combo's performance at MellyForex by clicking here and there is an existing thread on the MellyForex Forum here.]]> Fri, 28 Oct 2011 17:17:43 +0100
Forexbody Money Printer EA - Expert Advisor in Forward Test
Forexbody Money Printer The reasons behind my beliefs are two-fold. Firstly, a limited number of options means that there is less margin for user error. I can't think of anything worse than setting a parameter incorrectly and losing money, when others around you have used a different setting and are making money.

Secondly, my assumption is that EA developers spend considerable time optimising their products to obtain the best settings before releasing them on the market. If, having carried out the optimisation and discovered the best settings, where is the logic in giving users the option of changing those settings? Why not simply hard-code the best settings internally so that they can't be changed?

That said, I guess that variety is the spice of life, and there is no shortage of people in the world who do actually like the ability to change settings as this gives them a greater sense of being in control. Those people are certainly going to love the latest EA to go into forward test at MellyForex. It's called Forexbody Money Printer, and it contains approximately 700 different external parameters that can be adjusted by its users. Yes, you did read that figure correctly, and I use the word 'approximately' because I gave up counting them halfway through! The good news for those readers who, like me, want to keep things simple, is that Money Printer is delivered with an accompanying template so you don't need to worry too much about altering anything.

The good news doesn't stop there, however, and the best bit of news is that Forexbody Money Printer's developer is offering his EA on a FREE trial for two months, and trial users are at liberty to use the EA on both live and demo accounts during that time before deciding whether or not to buy!Money Printer works on the EURUSD symbol, and its default is to use the 1-minute timeframe, although it can be set up to work on any timeframe. From what I can gather, it uses a combination of strategies to take trades based upon numerous different chart candlestick patterns such as Dojis, Engulfing candlesticks and Star patterns. In addition, Forexbody Money Printer also features a scalping strategy which works as a compliment to the chart pattern strategy. The EA can have multiple trades open concurrently, although there is a setting to limit the number of open trades to prevent over-exposure and a margin drain. Finally, Money Printer also includes a hedging ability which is used to offset losses when the EA gets on the wrong side of the market, although setting the EA to hedge in this way will obviously render it unsuitable for those US-based users who are unfortunately restricted by NFA hedging and FIFO rules. Fortunately, the option can be turned off.

Although it does declare stop-losses with the broker, Money Printer doesn't declare a stop-loss on its trades immediately after it has been opened in the manner that other EAs do. Instead, it uses some internal hocus-pocus to wait for the trade to mature before setting its stop-losses which it then trails to exit. Some trades may even be closed by the EA without a SL having been set. Similarly, some trades may be closed quite quickly and others may take days or weeks to close.

The fact that the EA doesn't immediately set its SLs means that the risk settings don't relate to the account balance in the conventional way, and users will find themselves risking a greater percentage of their account balance on some trades than on others.

In my case, I'm running Money Printer on a $5,000 FXCC raw spread ECN demo account where the minimum lot size is 0.1 lots and I expect that trades will be taken at this minimum size for the foreseeable future.

Given the fact that the maximum number of open trades is being limited to three in either direction for each of the EA's two strategies, I'm hoping to not see as excessive margin use as frequently happens with some EAs which can open multiple concurrent trades when they get the wrong side of a major trending move. I guess we'll just have to see how it performs.

Users can compare Forexbody Money Printer's performance against the performance of other EAs on my live leaderboard, and also its results are being published to the detailed analysis pages at MellyForex which can be accessed by clicking here.]]> Thu, 20 Oct 2011 12:14:18 +0100
Forex Cleaner EA - Expert Advisor in Forward Test Forex CleanerOne of the first things I always do when I start to write a new article about any Forex product is to visit the developer's website looking for some suitable graphics material to include within my header. That's normally in the form of a boxshot, but not exclusively the way things work out. If I can't find anything suitable, I'll search the web to find something else which is appropriate and has some sort of link to the product's name.

In the case of the latest EA to go into forward test at MellyForex, I couldn't see any boxshots or anything else suitable on the developer's website. Undeterred, I put some old clothes on, picked up a duster and got busy in front of the camera.

The things I have to do sometimes to please MellyForex readers. :)

That's right. The latest Forex EA to grace the shores of MellyForexLand is called Forex Cleaner. Strewth, what's a cleaner got to do with Forex? Where do some of these names come from?

All joking aside, The Forex Cleaner developer emailed me and invited me to put a copy of the EA into forward test at MellyForex. As I'd like to increase the number of Expert Advisors that I've got in test, I was more than happy to oblige.The EA is described as a non-scalper and runs on the EURUSD symbol on the 30-minute timeframe, only opening trades at the very start of a new 30-minute price bar. It supposedly obtains its market entries using a combination of several standard indicators including something described by the developer as a "price analysis algorithm". Stop-loss and take profits are calculated dynamically and are managed internally. The default method is to use the stealth approach of hiding their values from the broker and using the EA to close trades as and when the time is right.

The risk settings are seemingly calculated using a maximum SL of 135 pips, and the Forex Cleaner website states that no stop-loss will be greater than this. However, the developer's own backtests use the 'Open Prices Only' model. Use of this model means that trades will not only be opened in the tester on the start of a new 30-minute price bar, but they will also only be closed by the tester at the start of a new 30-minute price bar.

This is acceptable if that is how the EA is coded to behave, but I'm concerned that the Forex Cleaner forward tests I've seen so far suggest that this isn't actually the case and, in the developer's own forward test which I've highlighted below, trades are clearly being closed mid-bar which is contrary to how they would be closed in an 'Open Prices Only' strategy test. I can only conclude, therefore, that the backtests on the developer's website are not to be relied upon.

Forex Cleaner Trade Times

In any event, the worst case loss on one of the backtests I looked at was 250 pips - a far cry from the 135 pip suggested maximum. I am assuming that this particular trade would have been closed some time during the prior price bar for a significantly lower loss, but the test also shows a lot of trades being closed for losses greater than 135 pips and it is quite likely that a number of the winning trades are also being overstated through the use of the 'Open Prices Only' model in the developer's strategy tests.

Anyway, I've put Forex Cleaner straight into forward test without carrying out any strategy tests of my own, and I therefore don't know what to expect of it going forwards. I want to see first hand whether it closes its trades with the start of a new price bar or if it closes trades mid-bar as has been happening in the developer's forward test. If Forex Cleaner is performing well after a month or so, that's the point in time that I'll conduct some proper tests of my own and analyse the results in greater detail.

Although I feel that the developer's backtests are unreliable in terms of how the open trades are managed and when they are closed, the strategy tests do suggest that Forex Cleaner is likely to trade just over three times each week and I am expecting to see a similar level of trading activity going forwards.

I've put Forex Cleaner into test on a $5k demo FXCC raw spread ECN account.

The EA is very easy to set up with only minimal parameters to adjust, and I'm using default settings which include using the non-declared stealth stop-loss, although I've set the allowed slippage to zero and I'm risking a maximum of 3.5% of the account on each trade.

It'll be interesting to see if the EA is capable of adhering to this value using its stealth approach.

Forex Cleaner's performance can be monitored at MellyForex by clicking here, and there is also a thread on the forum here where readers are welcome to discuss Forex Cleaner.]]> Wed, 12 Oct 2011 09:50:49 +0100
Forex Avatar EA Version 2 - Expert Advisor in Forward Test Forex Avatar Expert AdvisorI've put the latest version 2 of the Forex Avatar EA into forward test at MellyForex.

Forex Avatar is described by its developer as a non-Martingale, non-grid scalper which trades through 24 hours. It runs on the 5-minute timeframe charts of both the EURUSD and GBPUSD currency symbols.

Setup is very easy, as users only really need to decide whether to allow the EA to trade on Friday (default is true, meaning it will trade), and set the EA's maximum spread and risk level.

I've put the EA straight into forward test without carrying out any backtests of my own, and I haven't yet analysed the EA in any detail to understand how it works exactly. If it does well in forward test, then that will be the point in time that I will probably decide whether to look at Forex Avatar in closer detail.

From the developer's website, the backtests suggest that it's likely to take around 5 trades each week on the GBPUSD symbol and less than 4 trades each week on the EURUSD symbol. These aren't exactly a prolific number of trades for a scalper, so I'm guessing there must be some pretty comprehensive trade filtering going on inside the EA so that it only selects those trades with the highest win probability. If the developer's backtests are to be relied upon, I'm expecting to see a win rate of just under 90% and a risk/reward ratio of around 3.I've set Forex Avatar up in forward test on a $5k FXCC raw spread demo ECN account. The fact that the FXCC account is using raw spreads will, hopefully, ensure that Forex Avatar is able to exit its trades quickly and efficiently. However, a 1.0 pip round turn commission is being charged to each trade, and I would normally increase the TP values by 1.0 pip and reduce the SL values by the same amount to compensate for those commissions. Because Forex Avatar is a black box system and it's impossible for users to manually adjust either the TP or SL values, I'm hoping that the commission charges won't impair the EA's performance. I raised the matter with the developer and he advised me that it would be OK.

Regarding maximum spreads, I've set the EURUSD maximum allowed spread to 2.0 pips and the GBPUSD maximum spread to 3.0 pips. Seeing as FXCC generally quote prices much lower than these values (and often as tight as 0.1 pips), I'm not expecting this to be an issue.

My only other observation is that Forex Avatar appears to use no fewer than 9 different magic numbers in its operation.

The multitude of different magic numbers most probably means that Forex Avatar will have several trades open at a time.

Although I've set the EA up to risk no more than 3.0% of the account balance on any single trade, I'll almost certainly be risking much more than this if the EA decides to open several trades together.

Users can compare Forex Avatar's performance against the performance of other EAs on my live leaderboard, and also its results are being published to MellyForex which can be accessed by clicking here.]]> Mon, 10 Oct 2011 10:16:45 +0100
FX Speed Trader and ParaSwing EA - Expert Advisors in Forward Test
FX Speed Trader is an EA which falls into the category of pre-Asian scalper, as it works during the few hours after New York closes and before Asia opens. Most Europeans tend to be in their wincyette jimjams at that time of day, and markets tend to get range-bound in thin trade as a result.

Unsurprisingly, robots that trade at these times have very low take profit values, and FX Speed Trader is no exception with a default TP of 4 pips and a default SL of 35 pips.

Whenever TPs are this low, broker stops levels and spreads are always going to be an issue and, for that reason, this type of EA is most likely to work best with a raw spread ECN broker.

To get the best out of the EA, I have set FX Speed Trader up on an FXCC raw spread ECN account which charges a 1.0 pip round turn commission per trade. To compensate for the commission charges, I have increased FX Speed Trader's take profit value to 5 pips and reduced its SL to 34 pips. FX Speed Trader works on four different currency symbols, namely EURCHF, EURGBP, GBPUSD and USDCHF, and is delivered in the form of a separate EA for each different symbol. I presume this is because the optimised settings for one symbol are different to the optimised settings for each of the other symbols, but I'm only guessing here as FX Speed Trader's precise Modus Operandi is hidden away inside a black box.ParaSwing Expert AdvisorFX Speed Trader supposedly works identically on all chart timeframes, but I've set it up on 5-minute charts because that's the timeframe the developer uses himself.

The EA can either trade using fixed lot sizes or it can risk a percentage of the account's free margin on each trade. I've elected to risk 4.0% of the free margin for my forward test and readers are welcome to monitor FX Speed Trader's progress at MellyForex by clicking here.

The ParaSwing EA is supposedly a swing trading Expert Advisor which looks to track trends on the 15-minute timeframe of the EURUSD symbol.

It casts some magic spells on an assortment of indicators to decide when to enter a trade, and then trails a stop-loss behind the price action if or when that trade becomes profitable. Aside from the trailing stop-loss, ParaSwing also uses some internal magic to decide when the time is right to exit trades.

In terms of simplicity of use, ParaSwing is pretty much idiot-proof, and users really only need to validate their copy and decide whether they wish to trade fixed lot sizes or whether they want to risk a percentage of their account on each trade. Regular readers will know by now that's the manner in which I believe commercial EAs should be designed.

I've set ParaSwing up on an FXCC raw spread demo ECN account to risk 4.0% of the account balance, and readers can monitor the EA's performance on MellyForex by clicking here.

As things stand at the moment, I don't really have the time necessary to prepare a comprehensive review of every EA that I've got in forward test, and it's therefore impossible to say yet whether I'll be carrying out a full review of either FX Speed Trader or ParaSwing.

If either Expert Advisor starts to appear in the top part of my Live Leaderboard after being in forward test 30 days or more, then I guess that will be the best time to examine them in further detail.]]> Thu, 06 Oct 2011 12:11:17 +0100
Normal Service is Partially Restored - MellyForex September Update Having witnessed several EAs struggling in drawdown over the last couple of months as I discussed within my August Update, I'm now pleased to report a modest upturn in fortunes during September. Although positions at the top of my Live Robot Leaderboard have rotated slightly from a month ago, the top slots are still occupied by the same half dozen Expert Advisors as have been there for a reasonably long time, and I expect that trend is likely to continue for the foreseeable future.

The difference this month, however, is that the top six EAs have now been joined in their ranks by a couple of additions to the Profitability Club, namely FxPowerCube and Forex Combo.

Although I've had Forex Combo in forward test now for well over two months, this is an EA that I haven't yet got round to analysing in any detail.

Despite my honourable early intentions, there is actually a very simple reason why I haven't yet got round to reviewing Forex Combo. To explain more, the EA started off very well in forward test but, after a couple of weeks (and in line with so many other EAs at the same time), it then went into drawdown which has so far peaked at nearly 23%.

As my current 'To Do List' occupies nearly a whole ream of A4 paper (I really should try harder to save trees :-)), the drawdown meant that my plans to review Forex Combo got pushed further and further down the list.

Forex Combo's recent move into the profitable part of my leaderboard, however, means that the time is probably right to revisit the idea of a review!As for FxPowerCube, this is an EA which I had previously dismissed as I had been unable to carry out any long term profitable backtests as mentioned in other articles previously. To be honest, I'm not surprised to see it in the top half of my leaderboard now. The backtests I carried out certainly suggested that its performance was likely to be erratic and, similarly, I wouldn't be surprised to see it back in drawdown again at some stage in the not too distant future. Despite this negativity, I do hope that my assessment of the EA is wrong.

I took a screenshot of my leaderboard at the end of September so that readers could compare current performance with performance from previous months. I don't see any point in discussing the table further as, aside from the two EAs I referred to above, it doesn't look too different to last month's table.

September MellyForex EA Leaderboard

I also seem to have spent a bit of time recently trying to get my head around Forex Gale which I recently put into forward test. As I explained in this article, Forex Gale is not an EA that I'm ever going to be able to review, because all of its backtests are completely meaningless and aren't indicative in any way of the EA's live performance. As mentioned previously, this is due to the fact that the EA aggregates trades from other currency symbols in deciding when to simultaneously close a group of trades, and it's impossible to replicate this type of behaviour within the MT4 Strategy Tester.

It's probably best, therefore, to add a few more comments about the EA here and now, and to then just leave it to run in forward test to see how it performs over the coming days, weeks and months.

Firstly, despite taking around 90 trades within its first few hours in forward test, it seems to have now settled at approximately 350 trades per week. I'm not currently testing any other robots which trade as prolifically as Forex Gale trades!!!

Secondly, I'm concerned that the results as they appear on my leaderboard might be leaving users confused, because they undeniably look a little strange. For instance, Forex Gale tops the leaderboard in terms of the number of pips won each month (4,700 pips won), yet its monthly return appears as a negative value (-65%).

Forex Gale Performance

The reason for these seemingly strange statistics is that Forex Gale seems to be continually closing trades and banking profits from one of its three strategies yet, at the same time, there are plenty of other trades taken by other strategies of the EA which it isn't ready to close and which are in drawdown. The result is that the balance curve has been steadily rising, yet the equity curve has gone in the other direction. Forex Gale's MyFxBook screenshot shows that the equity is currently around 17% lower than the initial deposit.

If the EA gets to the point where it closes the majority of its trades out, and has either no or very few remaining open trades, then the equity curve will catch back up with balance curve and Forex Gale will most probably soar towards the top of my leaderboard on virtually all counts. That is, until the next time it starts to carry a large open drawdown!

Forex Gale MyFxBook

At the time of writing this article, the EA has no fewer than 51 trades open across all three of its strategies dating back various times over the last 8 trading days which are responsible for the high open drawdown. These trades are consuming just over 40% of the account's available margin although, to be fair, the trades are all of 0.1 lot size and margin usage would be considerably less if I had opted to use a microlot account for my forward test.

The two obvious questions are, therefore, will Forex Gale be able to successfully exit those open trades for a net profit? And, if it can exit those trades, is it likely to happen any time soon? I certainly hope that it can make a graceful exit, but I must admit that I do have my concerns! The developer has suggested to me that the EA can sometimes have as many as 60 open concurrent trades. If that proves to be the case, my demo $5k FXCC forward test account should be quite capable of withstanding the margin pressures, even if the resultant drawdown is higher than I would like to see. If, however, the EA gets up to as many as 80 open concurrent trades, I fear the account may start to come under margin pressure. We can only wait and see what happens.

Without any way of backtesting Forex Gale, however, I can only really categorise this EA as a 'Good Fun EA' which should only be used for enjoyment purposes as a gamble with money that users are fully prepared to lose. I think it could possibly give a great deal of pleasure under those circumstances, but users should be aware that its strategy is totally unproven. My recommendation is that readers regularly monitor Forex Gale's performance on my FXCC account to get an idea of likely drawdowns and margin requirements before using this EA themselves. Remember that FXCC's minimum lot size is 0.1 lots, so drawdown and margin usage would be considerably less on a microlot account.

Finally, it's my aim to get as many EAs as possible into forward test even if I'm not able to fully review each and every EA in test. My hope is that MellyForex can ultimately become a definitive resource for EA testing with its verified results and unbiased reviews. I've got the spare server capacity, and I may as well use it, so please say on the forum if there's a particular EA that you'd like to see in test at MellyForex. I'm always happy to ask developers for copies of EAs to test, even if some of my requests aren't extended the courtesy a reply!

On the subject of the MellyForex forum, can I ask you guys once again to not be shy and to please start using it in greater numbers? Contrary to popular opinion, the MellyForex business model does not revolve around IB and affiliate commissions. If that were the case, some of my reviews would be biased to help me earn more commission! However, I'd rather not go there, and I'll always try to remain impartial and keep my reviews unbiased, even if I don't always say what some people want to hear.

Instead, MellyForex's business model revolves around paid advertisers. To attract those advertisers, I really do need you guys to start talking to each other on the forum and help boost traffic on the site. It becomes a kind of self-fulfilling prophecy because, the more you guys talk to each other, the more likely it is that someone will reply and the site will grow.

So, if you are looking for an answer to anything Forex or trading related, don't hesitate to start a new topic and fire away with your questions on the forum. PLEASSSSSSE!!!]]> Wed, 05 Oct 2011 15:30:40 +0100
Forex Gale EA - Expert Advisor in Forward Test
In case anybody is thinking that's an unusual way to begin a Forex EA article, you're dead right! :)

Forex GaleThe truth is, though, that I really don't have a clue at this stage how I'm going to explain my thoughts on the Forex Gale Expert Advisor which I've put into forward test this week. I'm just going to scribble everything down and hope that it makes some kind of sense by the time readers get to the end of this article.

Anyway, I've had my eye on Forex Gale for a few weeks and, a couple of days ago, I asked the developer for a review copy. His initial response was to send me a copy of the full version of the EA and he also suggested at the time that I open an Alpari demo account for the purposes of testing his EA.

When I replied that I was currently testing EAs on FX Central Clearing raw spread ECN accounts, his response was that he had never tested Forex Gale on an FXCC account, so "didn't know how it would work".

Although this remark didn't alarm me as such, it did give me some clues as to what to expect, and I'll explain more about that shortly.

The Forex Gale EA is delivered in the form of six different Expert Advisors. The EA employs three different strategies, each strategy being coded within a separate EA, and different copies of each EA are provided for use with either 2/4 digit brokers or 3/5 digit brokers. I don't understand why the broker digits couldn't have been hard-coded into the EA to make the product more user-friendly. Similarly, I'm not sure why all three strategies couldn't have been combined into one, as I happen to believe that commercial products should be made as 'idiot-proof' as possible.

Those personal niggles aside, Forex Gale users can choose which strategies they wish to run, be that just one strategy on its own or all three strategies combined. In addition, each strategy runs on six different currency symbols, meaning that users will need no fewer than 18 chart windows open in order to run all three strategies on every suitable symbol.To quickly summarise Forex Gale's strategy, it is a grid trader which uses limit and stop orders to enter the market with a normal 3 hour expiry time before an unfilled order is deleted. Having entered a trade, the EA doesn't use either a stop-loss or a take profit value and, instead, uses internal logic to exit trades once the combined profit of all the open trades reaches a certain amount.

All grid traders invariably fail when they get on the wrong side of a major trend, and add to the losing positions to such an extent that the trader gets margined out of the market and is forced into closing everything for a massive loss. By using multiple symbols and strategies, Forex Gale looks to hedge its open losing trades from one strategy with profitable trades taken in the opposite direction by one of its sister EAs running the strategy on a different currency symbol. Once the basket of trades taken on the group of six chart windows reaches the desired profit level, then every open trade on every symbol which has been taken by that strategy will be closed simultaneously. In addition, any open unfilled orders will also be deleted ahead of their expiration, and the EA will start its cycle over again.

The EA's hedging approach does mean that that any US-based trader who is subject to the NFA's non-hedging rules will only be able to use one of Forex Gale's strategies at a time, although they would still be able to use multiple symbols to group the trades together.

Before putting Forex Gale into forward test, I firstly ran the EA through the strategy tester on a couple of symbols using Tadawul who are an older-style 4-digit fixed spread retail broker. I generally use these backtests to get some sort of idea of the sort of performance typical Forex clients can expect. There's no point in discussing these tests in any detail at this stage, but the tests did suggest that Forex Gale would be likely to take between 50 and 100 trades per year for each strategy/symbol. I thought from those initial tests that I was looking at an EA which would probably take around 30 trades a week if used on all 18 chart instances.

I then looked once more at the developer's website and noticed a link to an Alpari account running all strategies on all symbols and saw that, at the time of writing, it has taken 491 trades in 3 weeks. This is around 5 or 6 times as many trades as my Tadawul tests had suggested.

Having then put Forex Gale into my FXCC forward test, it started opening orders left, right and centre. In its first 7 hours, it opened and closed no fewer than 90 trades and also had a further 19 open trades. If Forex Gale continues at that rate, it could potentially take over 1,500 trades a week!!

If that does indeed prove to be the case, the forecasters will have got this EA wrong and it will require reclassification as a full Category Five Hurricane instead of just a gale. :roll:

In short, therefore, I really don't know what to expect from Forex Gale going forwards. I can only monitor its performance in the days, weeks and months ahead, and I hope readers can now understand why I began this article in the manner I did.

Having noticed the lack of trade correlation between the strategy tests and forward tests, and having given the matter some thought, there is an important issue here which I can see will render Forex Gale's strategy test results as worthless. Because it's only possible to test one EA at a time in the MT4 Strategy Tester, it's impossible to tell precisely when trades would have been closed as a result of the hedging performed by trades taken on other symbols within the group. The trades in my backtests would invariably have been closed far sooner than the backtests suggest because of this. Likewise, fresh trades would have been opened in their place which won't show at all in my backtest results because the previous trades in the tester hadn't yet been closed.

This certainly explains why my initial backtest results suggested Forex Gale would take far fewer trades than is being seen going forwards.

Regular readers will know by now that strategy test results normally form a significant part of my reviews. Those strategy tests not only give an insight as to how and when an EA is likely to trade, but they also give indications of suitable safe risk levels to use when setting the EA up in forward test. Being unable to properly backtest this EA is going to make it very difficult to draw any early conclusions of its capabilities, and the only way to proceed looks to be through forward testing alone. The only likely conclusion is that it will almost certainly work best with a 3/5 digit raw spread ECN broker outside of the US who doesn't have to comply with NFA regulations. It will then be able to hedge losing trades and will be able to exit its basket of trades quickest.

US-based users who are restricted to running just one of the three strategies, or users who run Forex Gale on just one or two symbols, are invariably going to find that average trade durations are longer and drawdowns will probably be higher because the EA won't be able to profitably exit its trade basket so easily.

For now, I've set Forex Gale up on a raw spread FXCC ECN account. It is trading fixed 0.1 lots and is aiming to bank profits of 0.35% of the account balance for every symbol which has an open trade. So, if it has trades open on 3 symbols, it would bank profits once the total open profit was 1.05% regardless of whether there were 3 trades or 30 trades open within the group at the time.

Hey, at least I think that's how it's meant to work. :?

[edit - 27 Sept 2011: The developer has confirmed that this assertion is actually incorrect and the target profit is only 0.35% (default setting) across the whole group of trades. My first few trades seemed to be closing for considerably more than 0.35% profit which led to me to deduce that the target profit was being multiplied by the number of symbols with open trades. I suppose it's possible that I was witnessing several groups closing at the same time and mistakenly thinking that the trades were all from the same group]

Being a raw spread broker, FXCC charge a commission on each trade which will erode the profits. If Forex Gale needs to open 30 trades to bank its profit, it means that the commissions will erode profits more than if the EA can bank its profit from just 3 trades. For some reason, Forex Gale seems to include overnight swaps within its open profit calculations, but it doesn't include commissions. It would be very easy for the developer to make the necessary modification so that Forex Gale's open profit calculations included the commissions as well.

Despite the commissions, I still believe that raw spread ECN is the way to go with Forex Gale as profit targets are going to be reached that much quicker. I also noticed that the EA allows slippage which is equal to the spread at the time it trades. This means that using Forex Gale on a conventional fixed spread retail broker could work out very expensive if the broker decides to help himself to some slippage at the trader's expense!

The EA is running using all three strategies on all 6 suitable symbols (EURUSD, GBPUSD, AUDUSD, USDJPY, USDCHF and USDCAD). Although I'm using fixed lot sizes at the moment, the EA does seem capable of increasing lot sizes proportionally to the account balance. Because I don't know what level of risk I'm taking and what sort of drawdowns I should expect at this current stage, I want to see first how it performs using fixed 0.1 lots before investigating this particular feature of the EA further.

You can monitor Forex Gale's performance at MellyForex by clicking here and there is also a thread on the MellyForex forum here for users to discuss the EA.]]> Sun, 25 Sep 2011 13:19:00 +0100
Forex Trend Hunter EA - Expert Advisor in Forward Test Forex Trend Hunter EA into forward test on MellyForex, as I intend to carry out a full review of this Expert Advisor the very moment that one of my readers invents a 12 day week. :P

Forex Trend Hunter Aside from the fact that things are getting increasingly busy in MellyForexLand, there were a couple of things in particular that attracted me to this EA.

Firstly, unlike the vast majority of Expert Advisors, this EA isn't a scalper. Instead, if the strategy tests on the developer's website are anything to go by, it looks to win, on average around 180 pips profit on each winning trade, with a low Risk/Reward ratio of approximately 0.5.

Secondly, aside from working on the EURUSD pair, it also runs on a couple of Yen crosses, namely the GBPJPY and EURJPY symbols. I'd need to check thoroughly to make sure, but I don't believe that I have any profitable EAs working on either of those two crosses at the moment, so my initial hope is that Forex Trend Hunter might redress the balance.

The larger than normal profits do have a trade off, however, and the downside is that Forex Trend Hunter is expected to win less than 45% of all its trades. Hey, you can't win 'em all. :roll:

All joking aside, the lower than normal win rate doesn't concern me in the slightest, but I'm not sure that this type of system will be too popular with everybody. I've never been one to judge a book by its cover, however, and I'm pretty sure that the tests I plan to carry out will soon determine whether or not Forex Trend Hunter is any good and if it's capable of mixing it with 'The Big Boys'.For the time being I've put Forex Trend Hunter into forward test on a $5,000 FXCC demo account. As I've stated previously in other articles, FXCC are a raw spread ECN broker and the performance of the EA in forward test should match near identically to the performance users can expect on a live FXCC account.

I've set Trend Hunter up to run on all three of its recommended symbols that I mentioned above on the 1-hour timeframe.

Because FXCC charge a 1.0 pip round turn commission on each trade, I've increased Trend Hunter's recommended take profit values for each symbol by 1.0 pip and decreased the recommended stop-loss values by the same amount to compensate for the commission cost.

Otherwise, I'm generally using Trend Hunter's default settings although I have limited the maximum allowed spread and I'm not allowing any trade slippage whatsoever.

I'm hoping to risk no more than 3.5% of my account balance on any single trade, although I've noticed already that this might not be possible, and I plan to discuss this matter in detail when I publish my full review. Anybody interested in Forex Trend Hunter's performance can monitor it at MellyForex by clicking here.

Before I close this post, I'd just like to put in a shameless plug for an old friend of mine who has recently set up a new business to provide a VPS service for automated traders. I've known Gordon now for close on 10 years from when we first worked on a trading project together. Both Gordon and his partner, Norman, have got a lot of experience of automated trading software at institutional level and their Beeks FX VPS has some interesting introductory offers which might appeal to some of you.

I can promise that I'm not making a penny out of Beeks FX and I'm literally giving Gordon this plug to help raise his profile as a favour for an old friend. I know how difficult it can be to get yourself heard when you first set up a business, and I'm always happy to help anyone who's brave enough to give it a go.

Unfortunately, I can't vouch for the quality of the product as I don't have any need for a VPS service myself (MellyForex has its own servers), but I'm sure that you're all capable of clicking a link and deciding for yourself if Beeks FX might be of benefit to you.

There's also a Beeks FX thread on the MellyForex forum, and both Gordon and Norman are happy to visit the thread and answer anybody's questions.]]> Fri, 16 Sep 2011 09:32:02 +0100
Steady Winner EA v5 - Expert Advisor Review
Steady Winner EANot only is the Steady Winner website informative, containing test results together with lots of useful information about the EA, but Henry and Ming, the EA's developers, are very helpful and open in their approach, and you get the instant impression that they do genuinely care about their users and understand what it will take for them to succeed.

There's even a free copy of the EA available for download which only works on small accounts so that prospective users can try Steady Winner before they buy.

What's most impressive about Steady Winner, however, is the fact that the developers' ethos is not about how much money you stand to win through using their EA but, instead, it focuses on how much you stand to potentially LOSE if and when things don't go quite to plan.

That's the same approach as is used by professional traders and serious investors across the globe. Those guys are able to accept losses in their stride because they always ensure that their losses are never large enough to spook them out of the game, and they have the confidence and ability to move on to better and greater things.

Regular readers of my reviews will know by now that it's an understatement to suggest I'm not the world's greatest fan of the Martingale gambling technique. What's unique about the Steady Winner EA is that it features a money management approach which I believe can only really be described as a type of anti-Martingale. In its strictest form, an anti-Martingale system would actually increase the next lot size after a winning trade, but safety is all important with Steady Winner and there's absolutely no increasing of lot sizes whatsoever within its workings.

So how does that work exactly?Well, with a Martingale-based system, an EA will increase the trade size following a losing trade in an attempt to recover from the loss quicker. The Steady Winner EA does the exact opposite and REDUCES the lot size of the trade following a losing trade. I'll explain a little more about how they go about implementing this in a moment but, for now, it's fair to say that their anti-Martingale approach can be used to reduce Steady Winner's drawdowns considerably.

About the EA

The Steady Winner EA works on the EURUSD symbol and applies its jiggery pokery to a combination of several moving averages to identify a trend in the market and open a trade in the same direction as that trend. Having entered the market, it sets an initial stop-loss of 110 pips and a fixed take profit of 55 pips, although it also uses logic to lock in a profit and trail its stop-loss once a trade is more than a certain amount in profit. This results in the vast majority of trades being exited ahead of either their full SL or TP being hit. I suppose I should also mention at this stage that Steady Winner works with ECN and STP-style brokers and, because it only has one open trade at a time, there are no NFA issues with hedging or FIFO.

Having identified a trend, the EA is then reliant upon this trend gaining some momentum and continuing for a while so that it can continue taking trades in that direction. Personally, I'm 100% in favour of trading in the same direction as the trend, as I wrote in an earlier article entitled Let the trend be your friend till the bend at the end. As each trend develops, the Steady Winner EA will continue to trade its signals and my strategy test results will suggest that it should win, on average, nine trades consecutively.

When the bend arrives at the end of each trend, however, the Steady Winner EA is nearly always going to incur a losing trade. At that point in time, there can often be much to-ing and fro-ing in the market as the bulls and bears wrestle for control, and it's not uncommon for traders to mistakenly think that one side has won the battle and enter a trade prematurely, only for the other side to regain control again and the trader incurs a loss. In other words, these are the times that Ming and Henry try to keep out of the market as they have identified them as being the most dangerous to trade.

Whilst Ming and Henry would ideally like to avoid trading completely during those difficult times, they figured that this approach would be very difficult to implement within an MT4 Expert Advisor, so they sought an alternative solution to the problem. They decided that the best thing to do after their EA incurred a loss at the end of a trend would be to take the next trade at the smallest lot size possible. If that trade subsequently won, it could be fairly assumed that a new trend had commenced and that it would be safe to start trading with normal lot sizes once again. If that trade lost, the Steady Winner EA would continue trading its smallest lot sizes until such time as the EA did start to win again.

Setting Steady Winner up

The Steady Winner EA comes with its own 48 page PDF manual and is very simple to set up. There are only ten external parameters which can be adjusted. Of these parameters, there are two which I feel are more important than the others. The first one relates to the percentage of free margin that you want to risk on each trade which happens to be capped at 6.0%, and the second important parameter is used to control the EA's anti-Martingale feature which happens to be called "TestLotScale".

Steady Winner EA To explain a little more about this parameter, its setting can range anywhere between 0.0 and 1.0. Setting it at 0.0 means that the next trade after a losing trade will be at your broker's minimum lot size. Setting it at 1.0 will mean that the next trade will be at its normal size and test lot scaling won't be applied. The parameter can also be set anywhere between the two extremes to control the size of the next trade on a kind of sliding scale.

The EA's other parameters control things such as the maximum allowed spread and whether the EA is to trade on a Friday afternoon and in late December. If you do want the EA to trade during those periods, it's worth noting that test lot scaling will be applied to those trades, so there's no need to worry about risking your shirt while everyone around you is tucking into their Christmas turkey. :)

The only thing to note about trading on a Friday afternoon is that users will need to calculate their broker's GMT offset and apply the trading cut-off time manually, as the EA doesn't calculate it automatically.

Finally, I noticed that the EA allows 3 pips of slippage which isn't adjusted in any way to discriminate between brokers with 4 and 5 digit pricing. Personally, I never allow any slippage whatsoever so I set this at zero. Users should nonetheless adjust this value to suit their own needs.

The Strategy Tests

My initial strategy test was aimed at finding out how good the EA is normally, without reducing the lot sizes after a losing trade. I figured this approach would enable me to ascertain firstly if the EA had underlying profitability in pips and, secondly, it would provide a benchmark so that I would be able to judge just how beneficial the anti-Martingale test lot scaling strategy is.

As with previous reviews, I once again chose to use Tadawul for these first tests. It's not that I have a preference towards Tadawul in any way, far from it, but their combination of fixed 2.0 pip EURUSD spread and older style 4-digit pricing can often help to portray an EA in a manner which highlights its weak spots. Basically, if an EA can survive a Tadawul backtest, I would hope that the vast majority of retail Forex clients would be able to achieve similar results.

Tadawul - EURUSD - 2.0 pip spread - Trading All Times - Test Lot Scaling 1.0
Steady Winner Strategy Test<

This test certainly suggested that the EA is profitable of its own accord, with a healthy Profit Factor of 2.19 and a relatively modest drawdown of 17.61%. My next test, therefore, was to run Steady Winner with the test lot scaling applied so that I could compare the results. The white 'gaps' along the length of the green lot size plot below indicate those points where the lot sizes were being reduced to their minimum following a loss trade. These are the points in time where the drawdowns should hopefully be reduced and you should see a smoothing of the equity curve.

Tadawul - EURUSD - 2.0 pip spread - Trading All Times - Test Lot Scaling 0.0
Steady Winner Strategy Test

Although the net profit is reduced by around 20%, there are obvious benefits for applying the test lot scaling, as the Profit Factor increases to a very impressive 4.02 and the relative drawdown is reduced by around 2/3 to 5.92%.

Because the Steady Winner EA doesn't use a DLL which can slow things down considerably in the Strategy Tester, its backtests were running quite fast which afforded me a bit of time to look at a couple of other parameters within the EA to see whether or not they were of any overall benefit.

The EA includes options to trade late on a Friday afternoon and to trade throughout the Christmas period. If a user does choose to trade through either of those periods, the trades at those times will all be at the test lot sizes. I decided firstly to test the EA so that it didn't trade through those two periods at all.

Tadawul - EURUSD - 2.0 pip spread - No December Trading & No Friday Trading - Test Lot Scaling 0.0
Steady Winner Strategy Test

Although the EA avoids approximately 10 or 11 trades each year by keeping out of the market completely during both of those periods, it seemed to me that any advantage gained by not trading was only relatively minor. I also tested each of the two periods individually to ensure that the effect of not trading one period wasn't offsetting the effect of trading during the other one, but that doesn't look to be the case. I guess the fact that you'd only be trading minimum lot sizes during those periods helps to keep the difference in performance low.

To reduce pageload times by avoiding an excessive number of images in this review, I've provided text links to each of those backtests in case anybody reading this article wants to see the results for themselves.

Tadawul - EURUSD - 2.0 pip spread - With December Trading - with full Test Lot Scaling
Tadawul - EURUSD - 2.0 pip spread - With Friday Trading - with full Test Lot Scaling

Finally, having decided in my own mind that it was probably better to trade through Fridays and through December with full test lot scaling, I decided to see what impact different brokers were likely to have on Steady Winner's performance.

One of the features of the Tadawul account that I've been using for the Steady Winner tests so far is that both Tadawul's minimum lot size and lot increment is 0.5 lot. I wanted to see if running some tests with a microlot (i.e. 0.01 lot) broker would affect performance in any way.

For my microlot test, I opted to use Alpari and I fixed the spread at 0.8 pip which is obviously tighter than Tadawul's 2.0 pip that I had been using.

Alpari - EURUSD - 0.8 pip spread - Trading All Times - Test Lot Scaling 0.0
Steady Winner Strategy Test

It's evident that there is no significant improvement in drawdown within the Alpari test, and I guess this is probably because the Tadawul drawdown was already as low as it is every likely to get. Let's face it, if the risk is 3.5% on any single trade and the worst case drawdown is only minimally higher at around 6%, it's going to be nigh on impossible to improve things any further except by not trading completely!!

At 89.28%, the Alpari win rate was over 1% better than Tadawul's, however, and this is almost certainly attributable to the tighter spread being used in the Alpari backtest. In turn, the higher win rate probably resulted in fewer test lot trades being taken overall which had the effect of increasing both net profit and Profit Factor in the report.

I didn't really see the point in carrying out any further strategy tests at this stage, as I felt that I'd already covered both the best and worst case scenarios that users could expect.


Clearly the Steady Winner EA looks to be profitable over a sustained period of time. Although it is not an exceptionally prolific trader (Steady Winner only takes around 10 trades each month), it wins in terms of the number of pips without being reliant upon any money management, and it comfortably survives an 11-year backtest with only relatively modest drawdowns when test lot scaling isn't being applied.

Theoretically, the use of test lot scaling will improve results further, although there is a significant caveat which I believe is likely to affect performance significantly from one user to the next.

I'll try to explain this caveat by taking two sample users who are both using the same broker. It could be any broker. One user has a $1,000 account and the other user has a $50,000 account. Both users start using Steady Winner at the same point in time, applying the same level of risk and they share exactly the same trades. The $1k user will be taking trades of, for example, 0.02 lots and the $50k user might be taking trades of 1.0 lots. When Steady Winner takes its first loss, both users will take the same size next trade of 0.01 lot. The $50k user will see a 99% reduction in his lot size, whereas the $1k user will only see a 50% reduction. The $50k user is clearly better placed to take advantage of Steady Winner's test lot scaling feature than the $1k user. If this test lot size trade loses, then the smaller user is going to see a higher relative drawdown than the larger user whose drawdown will be barely noticeable.

Similarly, you could have two users, both with 1k accounts, but one has an account with a microlot broker and the other user has his account with a minilot broker. The minilot user is likely to suffer higher drawdowns than the microlot user because Steady Winner will find it more difficult to apply the test lot scaling to the minilot account and that scaling is less likely to have the same beneficial effect.

A further consequence of this is that every user's drawdowns are likely to be at their worst when they first start using Steady Winner because that's the point in time that their account balances are likely to be at their lowest. Assuming users leave their money invested and continue to run Steady Winner for a few years, however, their account balance should increase markedly, test lot scaling will be able to work more efficiently and the relative drawdowns should be lower as a direct result.

My only conclusion from this is that the majority of users will probably need to exercise considerable patience and run Steady Winner for a number of years to allow their account balance to increase sufficiently to allow the test lot scaling feature to work with maximum impact. Unless, of course, you're already a BSD (that's City slang for a Big Swinging Dick trader) with a spare $100k to play with. ;)

To assist my determining whether an EA is any good or not, I would normally carry out a risk simulation at this stage which involves replaying all of the strategy tester trades in a multitude of different random sequences to try to ascertain what might happen to an EA in a worst case scenario. In the case of Steady Winner, however, I believe it will be impossible to carry out a proper risk simulation with test lot scaling being applied, as there is an essential correlation between certain trades in so far as the small lot size trades are ONLY applied immediately after a losing trade. Therefore, replaying the trades in a random sequence would completely destroy that trade correlation and the results wouldn't be indicative of what would happen in reality.

What I can do, however, is to carry out a risk simulation of my very first strategy test results where test lot scaling isn't being applied. If users then run the EA live and opt to apply test lot scaling, it should then be a fair assumption that actual performance is likely to show an improvement on this risk simulation by an indeterminable amount.

Steady Winner Risk Simulation

The risk simulation suggests that a reasonably high number of users should be satisfied with the EA, and that a relatively low account deposit is needed in order to run the EA.

I can also tell you that the average winning trade size expressed in pips is 10.5 pips and that the average loss is 37.3 pips leading to a Risk/Reward ratio of 3.55. When test lot scaling is applied, the Risk/Reward ratio based upon my first Tadawul strategy test results reduces to 1.94 which highlights the fact that using test lot scaling is definitely beneficial to Steady Winner's performance.

I also decided to look at the risk value being applied to each trade, to see how increasing it affected the level of drawdown. In my Tadawul tests, with test lot scaling being applied, I had been risking 3.5% on each trade and relative drawdowns were only around 6%. It's not for me to suggest users may want to risk any more than this amount, but I thought it might be useful to see the possible effects of different risk levels anyway.

Steady Winner Risk Optimisation

Having carried out my tests, I have to say that I have a reasonably high expectation level of Steady Winner, although I'm pretty sure that I will need patience in order to see it really start to pay. Aside from the prospect that the test lot scaling feature is really only likely to be of benefit once the EA has been running for two or three years, a close inspection of the backtests suggests that, like many other EAs, Steady Winner can go several months and only break even during that time. Personally, that's not something that bothers me in the slightest, but I can understand how some users might have a higher level of expectation and not be quite so patient.

Forward Tests

As part of my testing, I've set the Steady Winner EA up on a $5k FXCC demo account trading the EURUSD symbol on the 1 hour timeframe. It's possible that I may be doing Steady Winner a disservice by using FXCC, as their minimum lot size is 0.1 lot so, as I discussed above, the test lot scaling feature may not start to work properly until the balance has increased after a couple of years or even longer. I'm trading at 3.5% risk per trade, and on the assumption that the test lot scaling feature will be ineffective initially, I'm prepared for drawdowns up to around 15% or even 20% over the first few months of forward test in accordance with the backtests I've performed together with my risk simulation.

Whilst I could easily put the EA onto a different broker's account, and hopefully witness a lower drawdown as a result, I don't believe it would be appropriate to do that. If nothing else, this forward test will hopefully serve to prove that my reasoning above has merit.

As further evidence of how ineffective test lot scaling could be on a minilot account with a relatively low balance, I've taken a screenshot of some trades taken by Steady Winner since it has been in its FXCC forward test.

Steady Winner Account History

If Steady Winner had been on a different account which allowed microlots, the two trades taken at test lot size which followed the $8.85 losing trade would have both been taken at 1/10 the size they were. The first of those two trades would have only lost $10.90 and the second trade would have only won $0.62. Taking into account the original $8.85 loss, the combined loss of the three trade sequence would have only been $19.13 instead of $111.65.

In setting Steady Winner up, I also had to consider that FXCC are a raw spread ECN broker who charge a 1.0 pip round trip commission per trade.

To compensate for those commission charges, I have increased Steady Winner's take profit amounts and reduced the stop-losses by 1.0 pip.

I've also set the EA up to allow a 2.0 pip maximum spread with no slippage on trades whatsoever.
Steady Winner will also trade on Friday afternoons and during late December, as I'm not convinced that disabling trading during those periods would be that beneficial.

You can monitor Steady Winner's performance at MellyForex by clicking here and there is also a thread on the MellyForex forum here for users to discuss the EA.]]> Thu, 15 Sep 2011 06:23:17 +0100
Primeval and Primus Alpha EAs - Expert Advisor Updates Primeval EAWhen I first put the Primeval EA into forward test, I had already decided that it wasn't an Expert Advisor that I was going to spend time reviewing.

To explain, the EA had already been around for a while and had been well sold. The chances of my review actually generating sufficient fresh sales to cover the time it would take me to prepare a review, therefore, were pretty slim. I had also read stories of the developer being aggressive and abusive to anyone who bad mouthed his product on a public forum and, to be honest, I didn't need the hassle.

Despite not intending to conduct a review, I was nonetheless aware that the Primeval EA was supposed to work profitably. Against that backdrop, I thought it would be a good idea to put a copy of the EA into forward test and to use it as some sort of a benchmark to compare the performance of the latest EAs against this older, more established workhorse. It was to satisfy my own curiosity that I put Primeval into forward test more than anything else.

Anyway, I've had the Primeval EA running now for around four months. During that time it has drifted in and out of profitability but, over the last couple of weeks, it has stormed to the top of my Live Leaderboard in terms of safety, where it has replaced some of the earlier leaders which have struggled throughout July and, in particular, August.

This has meant that I've started to get a few emails enquiring about Primeval because a search of the MellyForex website hasn't provided any real information.In addition, and because it has been performing well in forward test with low drawdowns, I'm wondering if it might be suitable for use on any of my personal live accounts. Before doing that though, I would need to carry out proper strategy tests and analyse their results in the same way as I do with any other EA that I decide to use live. The EA parameters list includes Martingale options, so I need to ensure that Martingale isn't being applied to any trades, and that any risks being taken by the EA aren't excessive.

Just because it's had a good few weeks on a demo account, regular readers will know that I consider good long term strategy test reports to be very important when deciding whether to run an EA live. If an EA can't withstand a long term backtest, I believe it'll only be a matter of time before it self destructs in a forward test.

I've got a couple of other reviews that I'm currently working on, but I'll aim to get a Primeval EA review onto MellyForex over the coming weeks.

For now though, I can confirm that Primeval version 2.1 is running on the 15-minute charts of 5 symbols, the EURUSD, GBPUSD, USDJPY, USDCAD and USDCHF using the EA's default settings on a $5k Alpari demo account.

Primeval's performance can be monitored at MellyForex by clicking here.

My second update concerns the Primus Alpha EA which, as I mentioned in this article, I put into forward test several weeks ago with the intention of conducting a review.

The EA is protected against piracy by an elaborate registration system. When he supplied a review copy, the Primus Alpha developer also provided a serial number and password for me to use. Primus Alpha worked fine for a short while, but the EA failed to login to validate the licence a couple of weeks back. I haven't heard back from the developer since with instructions on how to overcome the problem, so the EA hasn't traded in that time.

Primus Alpha EA Unfortunately therefore, I've had no option other than to remove the EA from its forward test. In addition, my server on which Primus Alpha was running started to experience "Blue Screen Traps" causing it to freeze and reboot unexpectedly two or three times each week while Primus Alpha was installed and running. The server didn't experience these issues before installing Primus Alpha and, since removing the EA, the traps have stopped, so I can't help but wonder if there is an issue with the EA which was causing this problem.

Anyway, if the developer is able to resolve the problems, I'll be happy to try again. I can't do more than that.]]> Fri, 09 Sep 2011 12:07:07 +0100
FxPromax - Expert Advisor Review FxPromaxBelieve it or not, I've spent far longer trying to evaluate the FxPromax EA than this review is ever likely to suggest.

The reason that I've spent so long on the review is that, despite the fact FxPromax appears to be a long term profitable Expert Advisor, all of my analysis points towards it being a somewhat unappealing EA. That doesn't mean that it's a bad robot in any way, on the contrary, but I do believe that it will only suit a relatively small minority of Expert Advisor users.

The only thing I can do is to try and explain more about my reasoning later within this review.

First things first though, and the developer's website consists of several pages, and includes information on recommended brokers and VPS services, although there isn't much information on FxPromax itself by way of an explanation about the type of EA it is or how it works. Let's see if I can fill in some of the blank spaces.About the EA

FxPromax is a scalping EA which works round the clock primarily on the EURUSD symbol on the 1-minute timeframe. I say primarily, because there is a link on the developer's website to a MyFxBook page which includes several trades on a couple of other symbols, but more about that later.

The EA relies upon a combination of several different indicators to work their mojo and deliver trade entries. Having decided that it would like to trade, FxPromax then places a limit order with a 20 minute expiry to enter the market only if the the price retraces to the level of the limit order.

The result is that just fewer than 50% of the limit orders get a fill and the majority of them end up being deleted after 20 minutes.

Having entered a trade with a filled order, FxPromax then sets a fixed stop-loss and take profit (the default values are 9 pips SL and 120 pips TP). Because it doesn't set the TP and SL until after the trade is open, the FxPromax EA works fine with ECN and STP type brokers. There is no stealth system within the EA to close trades ahead of these values, so all trades either run to a TP or to a full SL.

FxPromax will only allow one trade at a time and, as such, it is fully NFA hedging and FIFO compliant.

The fact that FxPromax only uses the fixed TP of 9 pips and a SL of 120 pips makes understanding the performance of the EA a straightforward exercise in mathematics. It's very easy to calculate that the percentage of winning trades which is required for the EA to breakeven is 93.02%. If FxPromax wins more than 93.02% of its trades, it will be profitable. If it wins less than 93.02%, it will lose money.

Easy innit! Well no, this scenario is really only applicable if constant lot sizes are being employed, but it's made a little messier because FxPromax includes an option to apply a form of Martingale to attempt a quick recovery after a losing trade by increasing the trade sizes of the following batch of trades.

I'll discuss the Martingale aspect of FxPromax in greater detail a bit later on. For the moment though, the developer's website claims that FxPromax enjoys a 95% rate of winning trades. The expectation, therefore, is for it to win 255 pips on average for every 100 trades.

Setting FxPromax up

The EA is downloaded from the developer's website after purchase in the form of a zipped folder which contains the EX4 and DLL. FxPromax isn't wrapped in its own installer, so it is necessary to copy the files across to the relevant MetaTrader installation folder. There is also an 8 page PDF manual together with a number of preset files available for download, with the preset names ranging from "conservative" to "very aggressive".

Following installation, users are invited to submit a form to the FxPromax developer containing the account numbers on which they intend to use the EA, so that the EA can be activated for use. Each licensed copy can be used on one live account and up to four demo accounts.

FxPromaxThe EA contains 15 parameters which users can adjust as part of its operation. These parameters control whether the EA is to trade fixed lot sizes or if it is to risk a fixed percentage of the account balance. In addition, users can decide whether or not they want to trade on a Monday and/or Friday, and the SL and TP distance can also be adjusted, as can the length of time before limit orders expire. There are also parameters to control if or how Martingale is to be used to effect a loss recovery, plus a mysteriously sounding strategy parameter called "DeltaPips".

The DeltaPips parameter is dimissed within the instructions which simply say to leave it at its default value which is 5, yet all the preset files had it set to 3 which only served to heighten the mystery. :)

With temptation having got the better of me, I looked in closer detail and figured that the DeltaPips are the distance away from the current price at which the limit order is placed to enter a trade. One potential issue is that the value for DeltaPips must be equal to or greater than the broker's stops level, and users are likely to see the EA take fewer trades if a broker has a relatively high stops level because the price is less likely to retrace sufficiently to fill the limit order.

Finally, there are a couple of parameters that seemed to me to serve no real purpose. Firstly, there's a slippage parameter. If limit orders are being used for market entries, I don't see the point in allowing a broker to help himself to a few pips of slippage at my expense. Secondly there's a parameter to lock in 1 pip of profit for every 10 pips of open profit that the EA makes. Seeing as the default TP of FxPromax is only 9 pips, I struggled to see how this parameter would be of any practical use.

The Strategy Tests

Readers of my previous articles will know by now that I'm no fan of Martingale so, for my first strategy test on the EURUSD symbol, I opted to use Tadawul as a broker with the Martingale element of the EA turned off. Tadawul are an older style 4-digit broker who have a fixed 2.0 pip EURUSD spread and their stops level is 5.0 pips. This meant that I had to leave the DeltaPips parameter at its default level, as setting it any lower than the stops level would have no effect whatsoever.

Readers will also know by now that I'm not one to take excessive risks, so I set the EA to risk no more than 5.0% of the account balance on any single trade.

I left the EA's other parameters at their default values, which also meant that FxPromax would take trades on Fridays but not on Mondays. I presume that the developer opted for these settings as he had carried out his own tests and found those settings to be best.

Tadawul - EURUSD - 2.0 pip spread - 5.0 pip stops level - without Loss Recovery
FxPromax Strategy Test

Anybody who looks closely at the above image should not be concerned that the equity curve only shows a 25% modelling quality. This is a standard aspect of any MetaTrader strategy tests which are being run on the 1-minute timeframe.

Before considering the above results in any detail, I decided to run the same test with the Martingale loss recovery feature turned on. To explain a little bit about how FxPromax uses Martingale, I'd like to explain first that the EA has a very high risk/reward ratio of 13.33. This value is obtained by dividing the stop-loss of 120 pips by the take profit of 9 pips.

In order, therefore, to recover a loss within one trade, the EA would need to multiply the lot size of the losing trade by a factor of 13.33 simply to claw back the loss and break even over two trades. The problem with this approach is that, assuming your first trade risked and lost 5% of the account balance, you would need to risk 66.7% of your account balance on the second trade just to get back to where you started. If the second trade went wrong, you'd be down nearly 3/4 of your account!!! :(

The approach that FxPromax uses towards Martingale is that it attempts to recover a loss over a number of trades which the user can determine himself. The default seems to be to multiply the losing lot size by a factor of 2.5 for the next 5 trades. Assuming that all five trades were to win, you'd have recovered nearly 94% of the loss over those 5 trades. Please bear in mind that this figure is dependent upon the particular lot sizes that your broker allows you to trade.

In addition, it's also important to remember that recovery is dependent upon the next five trades in the sequence all being winners and that, if any of them lose, you'll have lost a further 12.5% of your account balance on each losing trade (assuming you use the defaults).

Anyway, that's the theory. Here are the backtest results with the default Martingale being applied. Note the relative drawdown level which is nearly 50%.

Tadawul - EURUSD - 2.0 pip spread - 5.0 pip stops level - with Loss Recovery
FxPromax Strategy Test

I wouldn't want to be risking as much as 12.5% of my account balance on any one trade, and I certainly don't want to incur drawdowns of nearly 50%, so I'm going to dismiss the Martingale feature of FxPromax within this article right here and now.

Having carried out these two tests with Tadawul, there were a number of things that occurred to me leading me to suspect that a broker like Tadawul might not be the best for this EA. Firstly, Tadawul's stops level meant that I couldn't test the DeltaPips setting properly. Secondly, I wondered if Tadawul's lot step and the fact they do not use 5th digit pricing may be encumbering performance.

I therefore decided to repeat my first test (i.e. without Martingale) using an Alpari demo account featuring a 0.8 pip spread, microlot sizing and a 2.0 pip stops level.

I ran my first Alpari test with DeltaPips set to its default of 5.0 pips. Clearly, the combination of tighter spreads, microlot sizes and 5th digit pricing improves overall performance, although the drawdown remained at a similar level to the earlier Tadawul drawdown.

Alpari - EURUSD - 0.8 pip spread - 2.0 pip stops level - without Loss Recovery - DeltaPips 5.0 pips
FxPromax Strategy Test

I then ran the test once more with DeltaPips set to 3.0 which is the value used within the developers various preset files. This resulted in about 40% more trades being taken, but the quality of those extra trades looks to be poorer, as the net profit only sees a minimal increase which is accompanied by a significant increase in the relative drawdown and a slight drop in the rate of winning trades.

Alpari - EURUSD - 0.8 pip spread - 2.0 pip stops level - without Loss Recovery - DeltaPips 3.0 pips
FxPromax Strategy Test

Finally, readers may have noted at the beginning of the article that I commented on a MyFxBook account of the developer's which showed FxPromax being used on other symbols. I therefore decided to run a test on the GBPUSD symbol with Tadawul using default settings to see how it performed.

Tadawul - GBPUSD - 3.0 pip spread - 5.0 pip stops level - without Loss Recovery
FxPromax Strategy Test

The poor default performance on the GBPUSD symbol discouraged me from carrying out further tests on other symbols.


Firstly, it is clear that if FxPromax is to work with any currency symbol other than EURUSD, the vendor will need to provide a different set of TP and SL values for those symbols which will actually be profitable.

Secondly, it is my firm opinion that the EA shouldn't be used with Martingale turned on, and that FxPromax shouldn't be used with a risk level greater than 5%. I say this simply because the higher drawdowns which will be incurred from using a higher risk setting will almost certainly spook users into removing the EA from their account completely.

That said, with a long term strategy test win rate in excess of 95%, the EA certainly seems like it will be profitable over a sustained period of time and therefore is deserving of a place as part of a portfolio of robots. FxPromax also seems to be broker dependent, and it looks far better suited to ECN style brokers than to older style, fixed spread brokers.

As seen above, my best strategy test results were obtained from Alpari using the EA's default settings including a DeltaPips value of 5.0 pips, so I carried out a risk simulation on those results.

FxPromax Risk Simulation

The risk simulation in itself suggests a reasonable level of satisfaction amongst users, with around 45% of users unlikely to see their account balance drop below the initial deposit. I would, however, view the minimum number of 26 trades merely as a likely indication of outcome, and wouldn't rely upon that figure. The figure was obtained from 50,000 random passes over 10 years, but it may be that more random passes were needed with this particular system because of its combination of very high win rate and very high risk/reward ratio.

So what's the catch?

Well, the biggest problem the EA has is its very high risk/reward ratio followed, a close second, by the fact that it's an infrequent trader. After FxPromax has a losing trade, the high risk/reward ratio means that it is going to take a further 14 trades to recover the drawdown. And that's assuming those 14 trades are all winners - if one of them loses, it's going to take longer! But, even if you use an ECN broker who can replicate the Alpari demo strategy test results going forwards, this EA only trades between 90 and 100 times a year. So, best case scenario, it's going to take nearly two months to recover a drawdown.

In actual fact, if you look closely at the backtest results, you can see that the EA wouldn't have recovered from the drawdown which started in November 2007 when the backtest account balance peaked at $11,036 until October 2009 when the balance returned to its former level. Users would certainly need a high pain threshold and considerable patience!

Of course, there is also a potential upside to this. Studying the strategy test results some more shows that the longest winning streak was 116 trades. In other words, the EA could easily go for a whole year without seeing a single losing trade!

The other thing to note is that, to use the EA safely with only a 5.0% risk per trade, the average annual return is only around 11%. Against that return, there is a possibility of a 20% drawdown at some stage.

Finally, use of the EA involves a one-off payment of $149.00 plus a monthly licence payment which is currently $39.95. Assuming an average annual return of 11% at 5% risk, users would need to deposit a minimum amount of over $4,300 just to cover the annual $480 licence fee. This doesn't take into account the initial payment either.

That's what I meant in my opening paragraph when I said that FxPromax wouldn't suit everyone. That aside, the EA looks to be profitable, and I certainly believe that it should form part of a portfolio of EAs to be used by serious, long-term hardened investors who understand that markets can, at times, be irrational and are prepared to accept this EA for what it is.

Forward Tests

As part of my testing, I've set the FxPromax EA up on a $5k FXCC demo account trading the EURUSD symbol on the 1 minute timeframe.

I'm using the EA's default settings with the exception that I have set the slippage to zero and have adjusted the TP to 10 pips and the SL to 119 pips.

FXCC are a raw spread ECN broker who charge a 1.0 pip round trip commission per trade. I believe that they are well suited to being able to replicate the Alpari strategy test results, and my adjustments to the TP and SL will compensate for the commission they charge.

You can monitor FxPromax's performance at MellyForex by clicking here and there is a thread on the MellyForex forum here where people are welcome to discuss FxPromax.]]> Wed, 07 Sep 2011 08:29:12 +0100
Take Heart, The St Leger Stakes are being run on 10th September - MellyForex August Update In case anybody either knows nothing about horse racing (I've got to admit, I don't know much about it myself), or is wondering what on earth a horse race has got to do with a Forex Reviews website, I suppose I'd best explain.

Nearly everyone who has been around the London Stock Market for any period of time will have heard of the age-old saying "Sell in May, go away, don't come back until St Leger's Day". I suspect that religious types may rush to point out that St Leger is the patron saint of some lost cause or another but, in the context of this article, it refers to the four day horse racing festival which takes place at Doncaster racecourse every September culminating in the world's oldest Classic, The St Leger Stakes.

So what does this age-old saying actually mean?The inference is that there is no point trading in the summer. All the brokers and fund managers will be hanging out on the beach, or attending Wimbledon and Ascot. Volumes will be low and markets more likely to fall. Once the racing season's over, the City boys get back to their desks...and the stock market picks up.

Nowadays, a large section of the City considers this complete nonsense. There's a "harder working culture to the City" than there used to be, insists Hilary Cook of Barclays Stockbrokers. Selling in May and buying back in September is an "eccentric form of market timing", agreed David Kuo on Motleyfool recently. It is, he reckoned, "no more accurate than flipping a coin". But they're both wrong. On average, the stock market actually loses 1.8% of its value each summer. Compounded over a couple of decades, that makes for a tidy overall loss of 30%.

That's certainly been the case during the summer (apologies to Antipodean readers who will need to substitute 'summer' for 'winter') of 2011, which has been particularly bad for investors. The Dow Jones Industrial Average is currently around 12% off its May highs, while the FTSE100 Index is down over 15% from its corresponding high.

Whilst the St Leger saying was originally intended to refer to stock markets, it doesn't stop there though. A consequence of thin trade at any time of year can be excessive market volatility, and just about every market (including Forex and commodities) has seen wild and often unpredictable daily swings over recent weeks as those remaining players who elected to stay behind their desks this summer have mulled over the consequences of a fresh global debt crisis.

Unfortunately, this can have a damaging impact on nearly all automated trading systems. The vast majority of automated systems will follow a combination of indicators upon which they base their trading signals, and these indicators, by their nature, will lag the price action. The result of a thinly traded volatile market is that market prices are finishing their vicious moves before the indicators are able to react and serve up their signals to enter trades. And then, by the time the indicators have caught up with the price action and finally decide that it's right to enter a trade, the market is ready to swing back in the other direction and a lot of those automated trades end up being stopped for a loss.

To illustrate the woes that have befallen automated trading systems over recent weeks, let's go back to the end of June when my Robot Leaderboard looked something like this:

June Leaderboard

Everything was running tickety-boo at that time, and a number of robots were starting to forge a reputation for themselves in terms of both profitability and safety.

And now, just two months later on, that same leaderboard looks like this:

August Leaderboard

Positions may have changed slightly but, clearly, the same group of 5 or 6 robots continue to dominate the top of the Leaderboard. The good robots haven't gone totally bad, and are still profitable overall in my forward tests, but the profit factors of all robots seem to have fallen, win rates have declined, drawdowns have increased and, in general, none of the top robots can be considered as safe now as they were just a couple of months earlier.

I don't believe there's anything to be gained from targetting specific robots which have underperformed these last few months, simply because just about every robot has underperformed during the recent conditions. In any event, it's easy enough for anybody to see which ones they are from my Leaderboard without me adding to their humiliation. It is worthwhile noting, however, that the Real Profit EA is finally clawing its way back up the leaderboard after a lean spell lasting several months. I do believe that this could prove to be a good long term EA on the proviso that it can successfully get its trades away with minimal slippage at the time of day it trades.

Now, for what it's worth, none of the robots have drawn down any more than their expected level from the backtests that I carried out as part of my reviews. In truth also, I can't even claim to be surprised that the drawdowns have occurred (the expectation was always for them to drawdown at some stage), so I'm certainly not alarmed by it in any way. One reason I'm not concerned at this stage is simply that I make a point of limiting my risk to ensure that I'm never overtrading.

Being the eternal optimist, I also believe there are some valuable lessons that can be learned from the last couple of months. One such positive should be to never underestimate the benefit of quality long-term strategy tests when deciding whether to use an EA and setting it up to trade live.

Another lesson should be that there is no such thing as a get-rich-quick Forex robot. A robot which has been backtested over 10 or 11 years and which exhibits long term profitability and stability should never be judged on two or three bad months which are irrelevant when viewed as part of the long term outcome.

Finally, a lot of people are successful through good timing rather than through anything else, and there's certainly an argument that some form of contrarian investing could be used with Forex robots to enhance performance. By that I mean, provided you have a long term profitable robot, it could be better to turn it off completely when it's outperforming and to turn it back on again when it's underperforming.

Remember that the Warren Buffetts and George Soros's of this world are renowned for buying low and selling high, and the same methodology can be applied to any number of different instruments.

St Leger Ladies Day So how do you know when a robot is outperforming and when to turn it off?

Well, the obvious answer is from the backtests! For instance, if the backtests suggest a long term Profit Factor of 1.5 and your EA is currently returning a PF of 1.6, then it's undeniably outperforming. Similarly, if a recent drawdown has taken the PF down to 1.4, then now could be as good a time as any to start using it again.

It doesn't have to be PF though; you could easily use other statistics such as Win Rate or simply track the current drawdown to decide when best to turn a robot back on.

Hopefully, in my next update in a month or so's time, the good quality robots will have started to perform again once the City boys have been back at their desks for a few weeks. For now though, I've seen a new frock that I like and I'm dusting off my hat ready for Ladies' Day. :)]]> Mon, 29 Aug 2011 13:17:40 +0100
FxPromax and Steady Winner EAs - Expert Advisors in Forward Test FxPromaxIt was only a few weeks ago in the MellyForex July Update, that I was bemoaning the lack of EAs that had appeared on the market recently which I felt would be suitable for testing. Since then, I don't seem to have stopped following up users' various recommendations of EAs to test, and I'm now able to report that I've put a couple more Expert Advisors into forward test over the last few days.

The first of those EAs is called FxPromax and, if it weren't for a user's request, I may never have even noticed this Expert Advisor at all. In truth, FxPromax probably isn't the type of EA that would ordinarily interest me, as it's not only a scalping robot with a very high risk/reward ratio, but the developer's website states that it is "Equipped with automated loss-recovery algorithm!"

Oh dear. In MellyForexLand, that only means one thing. Martingale! :(

The obvious question, therefore, is why have I even bothered to give this EA the time of day?

Well, the answer is simply that my preliminary tests do suggest that, unlike many Martingale-style EAs, FxPromax is profitable in terms of the pips that it wins and, also, that it is profitable over an 11 year backtest. Fortunately, the Martingale element of the EA is also optional, so I decided to investigate FxPromax some more to try to find out just how close this EA sits to that very fine line which can divide success and failure in trading.To illustrate my point here, the developer's website suggests that FxPromax enjoys a winning trade success rate of over 95%. I did a few quick sums, and figured that this gives an average profit expectation of 2.55 pips per trade. If, however, that success rate were to drop by a couple of percentage points to 93%, then FxPromax would merely breakeven. Anything less than 93%, and FxPromax will lose money!

The probability of the EA attaining such a high win rate is a subject that I intend to look at in closer detail within my full review, and spread and slippage are most probably going to be important influential factors in the performance of the EA.

With performance in mind, I've once more chosen to use an FXCC ECN account for my forward tests, because of the fact that they quote raw spreads which can be as a tight as 0.1 pip.

FXCC also charge a 1.0 pip round trip commission on each trade, so I've compensated for this by increasing the TP of each trade by 1.0 pip and reducing the SL by the same amount.

I've set the Martingale loss-recovery option of the EA to false, until I've analysed the EA in greater detail for my full review.

The EA doesn't seem to feature a maximum spread parameter, although it does include a slippage parameter which I have to set to zero so as not to allow any slippage whatsoever. The EA runs on the 1-minute EURUSD chart, and I've left the other settings at their default levels, which includes a risk of 5% per trade.

FxPromax's trades are being published to MellyForex, and the EA's performance can be monitored by clicking here.

Steady Winner EA The second Expert Advisor that I've put into forward test is the Steady Winner EA which has been around for a while, although it has recently been updated to version 5 which includes a number of improvements.

Again, it was on the suggestion of a MellyForex user that I contacted the Steady Winner developers to ask for a review copy, and they were more than happy to oblige.

Steady Winner is interesting because it uses what I feel can only be described as an ANTI-Martingale system. I'm sure most readers will be familiar by now with the Martingale approach which increases the next trade size after a losing trade. The specific intention is to recover the earlier loss in the quickest way possible. Well Steady Winner does precisely the opposite to this, and it actually DECREASES the next trade size after a losing trade (to the smallest size allowed by your broker, if needs be).

To understand a little more of the logic behind this approach, perhaps I should firstly explain that Steady Winner works by following trends. If the market trends well, then Steady Winner will also perform well and pick up lots of pips as the trend develops.

Unfortunately, however, trends don't last forever, and Steady Winner is nearly always going to incur a loss trade at the end of a trend. Once the next trend develops, Steady Winner will start to make money again, but the trend may take a little while to establish as the bulls and bears fight it out to see who will take control the market. During this pitched battle, the likelihood of Steady Winner incurring another loss trade is increased, so the EA takes smaller sized trades until the direction becomes clear once more and it can then start to trade its normal lot sizes again.

The developers have obviously worked hard to identify those periods during which they believe the worst case drawdowns are likely to occur, and the aim is to restrict the drawdowns at those times which can leave traders unsettled and nervous.

Within my review, I intend to look at the anti-Martingale approach in closer detail and I'll be able to analyse its effectiveness quite easily.
For now though, I have once more opted to use an FXCC ECN account for my forward testing, and I've put a copy of the Steady Winner EA into a $5k demo account.

Steady Winner runs on a 1-hour EURUSD chart, and I'm using the EA's default settings with the exception of the risk parameter, which I've set to risk no more than 3.5% of my account balance on any one trade.

In the event of a losing trade, the following trade will be at a reduced size of 0.1 lot.

Again, you are welcome to monitor Steady Winner's performance and detailed trades at MellyForex by clicking here.]]> Thu, 11 Aug 2011 08:46:12 +0100
Primus-Alpha EA - Expert Advisor in Forward Test NexGen4x Primus-AlphaA while ago I asked NexGen4x who market the Primus-Alpha EA if it would be possible to conduct a review of their Expert Advisor for MellyForex. My request was initially declined, but they've recently contacted me again and offered a copy of the EA which they've now asked me to review.

The Primus-Alpha EA works on the 5-minute EURUSD chart and seems to employ two separate strategies within its logic.

The NexGen4x website describes the first of these two strategies as being medium/high risk.

Primus-Alpha's first strategy seems to trade quite infrequently (less than 1 trade per week), and the money management logic of the EA applies comparatively small lot sizes because of the elevated level of risk which is associated with each trade.

The second strategy is described as being low/medium risk. The trade frequency of this strategy seems to vary between 6 and 20 trades each week, and lot sizes are larger because of the lower risk associated with this strategy.The Primus-Alpha EA also seems to have an option of supporting each of its two strategies with an additional grid strategy. At this stage, I can only guess that this works by adding to a losing trade if it doesn't go quite to plan and moves out of the money by a certain amount.

Readers of my earlier articles will be aware that adding to losing trades is a massive trading no-no in my eyes, so this will be something that I look at in closer detail within my review. You've only got to look at another grid system, Forex Crescendo, which I've previously reviewed and also have in forward test to see what can go wrong with this type of approach. I certainly hope that Primus-Alpha will be different!

While I'm carrying out my tests on Primus-Alpha and preparing my review, I've set the EA up in forward test on a $5k demo Forex Central Clearing account. As I've mentioned elsewhere, this is a true ECN account quoting raw spreads which are often as tight as 0.1 pip, and the FXCC demo price feed is identical to their live feed, so it will hopefully simulate Primus-Alpha's live performance as accurately as is possible. In common with ECN accounts, a commission of 1.0 pip is being charged by FXCC per round trip trade.

The Primus-Alpha EA has a lot of parameters which can be adjusted by the user, but I've set it up using the defaults as recommended within the instruction manual. These defaults include having the back-up grid strategy that I mentioned above switched on.

One parameter that I do like to control myself (or, at least, be aware of how it is being employed) relates to slippage, but there is no slippage parameter that I can see to adjust.

Another parameters that I do generally monitor myself is the level of risk, but Primus-Alpha is different to many other EAs in so far as users need to enter an amount in lots per $1,000 of account balance or account equity. I've used the EA's default settings, but I've not calculated at this stage what the risk in percentage terms will be if a trade runs to full stop-loss.

As with my other robots in test, you can monitor Primus Alpha's performance at MellyForex by clicking here.]]> Mon, 08 Aug 2011 11:31:23 +0100
Forex Combo System - Expert Advisor in Forward Test Forex ComboI've recently placed the Forex Combo EA into forward test while I carry out some strategy tests and analyse the results for a review.

Forex Combo seems to have been performing profitably for a while now and one of our MellyForex forum members asked if it was possible for me to review it. I took up the suggestion and emailed the developer who promptly replied to my request and kindly provided a review copy.

The EA trades on two currency symbols and gets its name because it uses a combination of three different strategies within the EA. The three strategies are described within the Forex Combo user guide as being "Scalping", "Breakout" and "Reversal" which all sound very encouraging.

I'm not entirely sure what my tests are going to reveal but, seeing as the EA seems to have been performing well in both a forward test running on the developer's website and it has also started well in my own forward test, I'm certainly hoping that this EA will live up to its early promise!!!While I'm carrying out my analysis of the Forex Combo System and preparing my review, I've set the EA up in forward test on a $5k demo Forex Central Clearing account. As I've mentioned elsewhere, this is a true ECN account quoting raw spreads which are often as tight as 0.1 pip, and the FXCC demo price feed is identical to their live feed, so it will hopefully portray the Forex Combo System in as accurate a light as is possible. As is the case with ECN accounts, a commission of 1.0 pip is being charged by FXCC per round trip trade.

The Forex Combo System works on both the EURUSD and GBPUSD currency symbols on the 5-minute timeframe. Because it combines three separate strategies, there is the possibility that it could take trades in contravention of NFA hedging or FIFO rules. The EA does include a setting to make it FIFO compliant, but I've turned this off as FXCC aren't required to comply with NFA rules.

I've set Forex Combo up such that it risks 3.0% of the account free margin on each trade on both of its recommended currency symbols. In common with the other EAs that I've got in test, I'm not allowing any slippage on the trades whatsoever.

I am also limiting Forex Combo's maximum allowed spread in pips as follows:

EURUSD - 2.0
GBPUSD - 3.0

As with my other robots in test, you can monitor Forex Combo's performance at MellyForex by clicking here.]]> Thu, 28 Jul 2011 13:58:05 +0100
MellyForex July Update I had hoped to be able to report around this time that I'd put another robot or two into forward test but, unfortunately, that isn't the case. The truth is that I haven't seen anything new over the last few weeks which has captured my interest sufficiently to put into test and write a review.

C'est la vie!

All is not lost, however, as it has given me the opportunity to provide a general update on progress and to cover several minor points which wouldn't otherwise be worthy of an article of their own.

Recent movement on my Robot Leaderboard has been less dramatic and it's becoming increasingly obvious which are the good quality robots and which are the ones that will never make the grade.

PipRider has regularly occupied one of the top spots since first going into test and, despite struggling a little with market conditions over the last few weeks, it is clearly developing a reputation for being one of the safest robots around.

Wall Street Forex Robot has been going from strength to strength and it is now safely holding on to first place on the leaderboard. One of the reasons for this is the fact that it is able to trade four different currency symbols, so those extra trades mean that it is always going to win more pips than, say, PipRider which only trades one symbol. Coupled with this, the effect of using an EA with multiple symbols can have a smoothing effect on the balance curve and serve to lower drawdowns.Interestingly, the Morpheus OddBot has quickly risen up the leaderboard, and is currently sitting in second place after a relatively short space of time. Given the similarity of some of OddBot's trade entries with those of the Wall Street Forex Robot, this probably isn't too surprising, although I maintain my previous stance that the two robots act independently of each other and that it should be quite safe to run both robots together on the same account, and it should also be noted that OddBot's current average return of pips per month far exceeds that of Wall Street.

Before leaving the subject of OddBot completely, it appears that the earlier 10% discount available on OddBot is no longer available. I have therefore edited my original article accordingly to remove any mention of the offer.

Forex WindFall has been profitable, but not to the same degree as either Wall Street or OddBot but, as I alluded within my recent article which directly compared the three robots, I'm sure this will change in the future and you will then witness either Wall Street or OddBot as under-performer out of the three for a while.

Another robot which has disappointed slightly in forward test has been the Forex Real Profit EA which is still underwater after a couple of months of test, although its ranking has been improving recently. One of my initial concerns when reviewing this EA was that, because of the times of day it trades, it might not be able to replicate the good backtest performance going forwards. That said, I believe that the EA's current drawdown has nothing to do with the ability to get trades away and is simply down to some unfortunate trade timing, and I do believe that it will recover the current drawdown and progress to greater things in the not too distant future.

What does concern me about the Real Profit EA, however, is that an updated version has recently been released which contains what I believe is an untested news filter. Filters such as this are all very well, but I believe they should be backtested properly in advance of being released to determine the benefits. If they aren't tested in this way, there is absolutely no evidence whatsoever to suggest that the inclusion of a news filter will actually be of any benefit whatsoever. For that reason, and seeing as I don't have either the time or inclination to rig up a news history file to test the modification for myself, I will be leaving the news filter switched OFF in my ongoing forward test.

Despite offering a healthy monthly return, Forex Growth Bot has slipped down my leaderboard slightly because of its relatively high drawdown level. This is most likely due to the fact that my version supplied by the vendor for forward test is a basic version which uses fixed lot sizes. Had I been using one of the advanced versions which employs variable lot sizes, then it's my guess that the drawdowns would be easier to regulate.

I previously mentioned that I wasn't planning to carry out a full review of FxPowerCube because of its poor backtest results as seen in the following erratic equity curve.

FxPowerCube Strategy Test

The developer did contact me after that to suggest that there was some kind of bug within my copy of his EA and he subsequently provided an updated version which he asked me to retest and report upon, but I'm sorry to say that I am still unable to produce any noticeable improvement on my earlier tests. It's certainly not for want of trying, as I've retested FxPowerCube using a multitude of settings and experimented with different brokers using different spreads in an attempt to produce some worthwhile backtest results, but it's all been to no avail.

The FxPowerCube developer has produced his own 11 year profitable backtest which, as a last resort and because I want to be fair to the developer, I analysed in some detail. Unfortunately, however, a risk simulation of those results didn't fill me with any great confidence as it suggests that users of the EA will almost certainly end up disappointed.

FxPowerCube Risk Simulation

To explain the risk simulation results, the EA looks to be susceptible to high drawdowns, it requires a relatively high account deposit to run, and less than 1 in 5 users will probably never see their account balance drop below their initial deposit. Although FxPowerCube is currently profitable in my forward test (which I intend to leave running), this is an EA which I would never contemplate running live, as it is not only too risky, but I feel it could go into serious unexpected decline at any stage.

My most pleasant surprise comes from the Primeval EA which is really only in forward test as a benchmark. It's an older EA which I've already stated that I don't plan to review thoroughly, but it's managing to hold its own with a top 4 spot on my leaderboard against its newer rivals. Maybe there's something to be learned from that?

At the bottom of my leaderboard, I can't claim to be shocked by Forex Shocker's poor performance, nor am I surprised that Forex Crescendo has fizzled to a whimper, or that Forex Hippo has gulped over half its account balance. :(

Another EA which has managed successfully to live up to its name is Forex Cash Protector which has managed to protect capital by trading so infrequently that Alpari deleted my demo account following an extended period of inactivity by the EA.

Before concluding this article, I would like to repeat myself once more and stress the importance of using a good broker with the tightest spreads. So many people seem to think that the best ECN brokers are those which offer the lowest commission but, frequently, this isn't the case. Many ECN brokers will "pad" the spread by 0.3 or 0.4 pip each side of the trade which enables them to offer a low commission and give an illusion of being competitively priced, but retail clients often don't realise what's actually going on. A true ECN broker will offer spreads as low as 0.1 pip, whereas other brokers claiming to operate via ECN will mysteriously never offer spreads that tight. In addition, a true ECN broker will not present any restrictions on trading style (such as scalping), they won't force any requotes on their clients and they certainly won't take the other side of your trade because they really do want you to win.


At this point, I'm going to unashamedly mention FX Central Clearing once again, as they are a true ECN broker who are totally transparent and fulfill all of those criteria I've just mentioned. The only thing that you need to watch out for with any broker who charges commission is that, when using an EA such as OddBot which locks in a small profit of, say, 1.0 pip, you must increase the amount of locked-in profit by the amount of commission being charged, otherwise your locked-in profit will be wiped away by the commission.

Finally, I'd like to ask a favour of anybody who enjoys MellyForex, that they help promote the site and assist it to grow. By that, I don't necessarily mean that you've got to click on any of my affiliate links to buy the products I brazenly promote (although it is nice when you do, so please don't let me stop you from clicking those links :)).

What I mean is that you can help in other ways which don't even involve parting with any money. I simply want people to start posting on the forum, telling their Forex friends about MellyForex, and to generally help develop a nice, caring and considerate community of users who talk to and help each other through sharing their personal experiences.

To explain in a little more detail, it's alway been my aim to fund the site through paid advertisers (I promise that I'll never litter the site with countless ads, and no more than are already here at the moment). Development of MellyForex is progressing to plan and I'm happy to continue posting my reviews, but I really do need you guys to get talking on the forum to develop the community and provide the levels of web traffic that the advertisers are looking for.

Please keep that at the back of your mind and, if I can find some worthwhile products to test and review, there'll be some new Forex articles coming soon.]]> Thu, 21 Jul 2011 08:13:40 +0100
Forex WindFall, Morpheus OddBot and Wall Street Forex - Expert Advisor Comparison Morpheus OddBot and Forex WindFall Expert Advisors and putting them into forward test, I've had a number of emails in which it has been suggested that these EAs are little more than clones of the more established Wall Street Forex Robot which I also reviewed a while back. The fact that all three robots trade the same two currency pairs (GBPUSD and EURUSD) is something which has added weight to this theory, despite the fact that Wall Street Forex Robot does trade other symbols as well.

If anybody reading this articles happens to follow me on Twitter, they will know that, as well as providing up to the minute news and snippets of useful information from Forex dealers' desks, the trades of the top performing robots in my Live Leaderboard are all tweeted as they are opened and closed.

A couple of weeks or so back, I was looking at my tweets flowing thorough when I noticed that the three aforementioned robots had all opened trades in the EURUSD pair within a couple of minutes of each other.....

Twitter trades

I guess it's only fair to make clear that, despite opening at the same time, the three trades on my Twitter screen all closed at different times and, therefore, ran to totally different outcomes, and I'm confident that the three robots certainly aren't clones of each other.

Seeing the tweets appear simultaneously, however, gave me the idea to produce an article which examined the robots in greater depth to try to determine if the three robots could work in harmony with each other on the same MetaTrader account.Equity curveAny lingering notion that the three robots may be cloned can be dispelled by the fact that OddBot and WindFall were both put into forward test at around the same time, and their equity curves so far show no correlation whatsoever. Wall Street Forex EA has been in test for a longer period of time but, if I were to recode the script which produces those equity curves so that it standardised the growth from the same starting point, I'm fairly certain that the lack of correlation would be highlighted further still.

What I'm hoping to find out in this article is whether the three robots all drawdown together at the same time.

If they do, then there will be no point in running them together that I can see. If, however, they all tend to drawdown at different times, then the equity curve should hopefully be much smoother indicating that it should be safe to use the robots together in harmony.

The first thing to remember, however, is that using all three robots on the same account would result in them trying to take non-FIFO compliant trades. This wouldn't matter for anybody whose account was outside of the USA, and wasn't subject to NFA regulations, but US-based users would need to run the EAs on separate accounts if they wished to use more than one of the robots.

In determining if the three EAs can be used successfully together, the first step involves taking some of the long term strategy test reports for each of the robots on both the GBPUSD and EURUSD symbols and using some of the software at my disposal to merge each set of results into a single report with the trades all sorted into the correct chronological order. For the purposes of this article, I'm going to compare the performance using both of the suitable symbols (ie. with the GBPUSD and EURUSD results merged) with each robot rather than attempting to examine each symbol individually, as this would complicate things unnecessarily in my opinion.

When carrying out tests such as this, it's very important to compare apples against apples. The tests to be compared should all be carried out with the same broker, using identical spreads and covering the same period in time. Fortunately, in preparing out my individual reviews for each EA, I already had a set of 10 year backtest results for all three EAs from 1st Jan 2001 through 31st Dec 2010 which had been carried out using fixed spreads (EURUSD 2 pips, GBPUSD 3 pips) with the Tadawul broker, so I decided it would be easiest to merge these reports. Remember that the Tadawul tests are indicative of what every user should be able to achieve as Tadawul are a typical retail broker who uses fixed spreads. Better performance would be obtained by using a broker with a tighter spread, but choosing the best broker with tightest spreads isn't particularly relevant in this comparison test.

These reports had been prepared with the EAs using their internal money and risk management system which resulted in varying trade lot sizes. Having merged the reports, therefore, the first step was to adjust the lot sizes so that each trade would have been of a fixed 0.1 lot size.

OddBot Balance curve

WindFall Balance curve

Wall Street Balance curve

Looking at the individual fixed lot size balance curves, it's pretty clear that the balance curve of each EA does follow a slightly different path. Despite this, the thing which stuck out the most was that both OddBot and Wall Street Forex Robot look as though they would have both drawn down at the same period in time.

The next step in the process was to start merging the results of the EAs in pairs to see what effect this had on the balance curves.

OddBot & WindFall Balance curve

OddBot & Wall Street Balance curve

Wall Street & WindFall Balance curve

As expected, the merged curve for OddBot and Wall Street Robot clearly shows the nasty period (which looks to have been during the first couple of months of 2009) where the combined effect of both EAs drawing down concurrently impairs the overall performance of the system. It should be said that this nasty period was unique within the balance curve, as the two EAs certainly seemed to work well together for the rest of the time.

The final step was to merge the results of all three robots into one.

Combined Balance curve

Interestingly, the positive influence of WindFall at around the time that OddBot and Wall Street would have been experiencing their maximum drawdown looks to have negated the ill effect considerably. The maximum drawdown looks to have been shifted to the very beginning of the 10-year tests in this balance curve.

I should also point out that I don't believe this makes WindFall the best EA out of the three in any way. At some stage in the future, it is possible that two different EAs will be drawing down concurrently and it may then be either OddBot or Wall Street which comes to the rescue and relieves the drawdown.

Finally, I carried out a risk simulation on the merged results, and set everything out in tabular form below.

Risk Simulation

I've got some observations from the results which I believe are worth mentioning,

1) Morpheus OddBot is using default settings in these tests with the SecureProfit parameter set at 1. Anybody who has read my Morpheus OddBot review will be aware that I believe significantly improved performance can be obtained from the EA by increasing the setting of this parameter to about 8 or 9.

2) Using multiple robots does not noticeably affect the minimum deposit needed to run them on the same account.

3) Using all three robots should mean that nearly every user has made a sufficient profit after just 3 or 4 days to cover the drawdown incurred during that time. If each robots is used individually, it can be expected to take a week or two for the return to exceed the drawdown.

4) These tests were performed using a typical retail broker with fixed spreads. Improved performance would invariably be obtained when using an ECN broker with tighter spreads.

Those observations aside, I don't want to draw too many conclusions from the above results, as I've already written detailed reviews of each of the EAs used in the comparison tests and I would only end up repeating myself. I'm also confident that anybody reading this article is able to study my results for themselves and form their own opinion upon whether to use multiple EAs. At the risk of repeating myself, this exercise isn't about finding the best EA from a selection of three.... It's focus is simply to find out if all three EAs can operate together and enhance overall performance.

The fixed lot balance curve is clearly straightest when using all three EAs together, and the maximum drawdown when using all three robots is less than the worst case drawdown from any of the individual robots. The only conclusion that I will draw, therefore, is that all three EAs can work together in harmony.

]]> Wed, 06 Jul 2011 07:06:01 +0100
Forex WindFall - Expert Advisor Review When a windfall apple fell on Sir Isaac Newton's head all those years ago, I'm quite sure that, even with his great mathematical skills, his vision didn't extend to an age where robots ran on computers to trade financial markets. Now, over four centuries later, one of the latest Expert Advisors to gravitate into the apple box of MT4 currency traders is called Forex WindFall.

In truth, the Forex WindFall EA has got absolutely nothing whatsoever to do with either Sir Isaac or falling apples, although I make no apologies whatsoever for the cheesiness of my introduction, given the seriously limited amount of material I could find on the web relating to anything of the same name.

Now we've established that my sense of humour has taken a severe turn for the worse, it's probably best that I get to the point and move on to discuss the Forex WindFall EA in some detail. :)

The WindFall website isn't glossy by any stretch of the imagination. Instead, it is quite factual and contains backtests of the EA together with a link to a verified MyFxBook account so that people can monitor WindFall's live progress. The website also includes a members area where WindFall users can download their copy of the EA, nominate and change the three MetaTrader account numbers upon which they are allowed to use the EA, and also view a copy of the user guide.

The WindFall user manual is also well set out, containing a total of 16 pages of which around half relate to installing MT4 and the remainder are dedicated to setting up the WindFall Expert Advisor itself.About the Forex WindFall EA

The Forex WindFall user manual suggests that the EA offers "Automatic, intelligent and low risk wealth creation". This basically means that it is a non-Martingale scalping EA which trades through 24/5 and offers a high rate of winning trades.

For a long time, a feature of scalping robots was that they achieved high win rates simply through running excessively high stop-losses. One stop-loss could easily eradicate weeks of gains. Things have changed, however, and modern-day scalping EAs are generally able to exit their trades far sooner and much more efficiently than was previously the case. Forex WindFall is no exception, and my tests will show that its risk/reward ratio is very impressive for a scalping EA at less than 2.5.

The EA works on the 30-minute timeframe of two currency pairs - EURUSD and GBPUSD. Although its precise modus operandi is hidden away within protected code, it seems to identify an underlying trend within the market and then waits in the sidelines until the time is right to jump into a trade.

WindFall declares a fixed pip stop-loss and take profit value once it has entered a trade, but also employs a stealth method to decide when to exit its trades. As a result, the majority of losing trades are closed before the stop-loss is reached, something which serves to improve the EA's risk/reward ratio. For reference, the fixed SL and TP values employed by Forex WindFall are:

EURUSD - SL = 62.0 pips, TP = 13.0 pips
GBPUSD - SL = 102.5 pips, TP = 11.5 pips

I noticed that the GBPUSD SL and TP values are rounded up to 103 and 12 pips respectively when Forex WindFall is being used on an older-style 4-digit broker.

NFA compliance isn't a problem for Forex WindFall, as it only has one trade open at any time which makes the EA suitable for US-based users. Likewise, the EA works with both 4- and 5-digit brokers as well as ECN/STP type brokers.

Setting the Forex WindFall EA up

Forex WindFall is delivered within its own self-installer which extracts the relevant WindFall files to the correct folders of whichever MetaTrader installation that the user chooses. It's then just a simple task to drag the EA into a chart window and set the various parameters.

Readers of my previous reviews will be aware by now that I like my EAs to be as idiot-proof as possible, with only a few parameters needing to be set so that the risk of user error is minimised. I was pleased to see that Forex WindFall fulfills this criteria, as it only has eight external parameters to adjust, of which one parameter relates to the unique licence key. The other parameters relate to setting and controlling the trade lot sizes and level of risk which is to be used. The EA can trade using either fixed lot sizes or it can risk a predetermined percentage of the account free margin on each trade, which is based upon the initial stop-loss distance.

Users can also set a maximum spread level at which they are prepared to trade together with a parameter called "adjustTarget". From my reading of how Forex WindFall works, it seems to wait for a retracement in the price before entering a trade. If, for instance, it was looking to go long when the market fell back to 1.6105, users could enter a deviation via the adjustTarget parameter of, say, 0.0002 in which case they would be prepared to go long at 1.6107. It seems to me to be an alternative form of slippage which would be applied to every trade regardless.

Forex WindFall's final parameter enables users to manually enter their broker's GMT server offset although, in this day and age, most EAs calculate this offset internally themselves. The user manual suggests that the GMT offset is required so that the EA can suspend trading when high impact news is expected.

Although the EA does include a news reader which reports any relevant forthcoming news in the chart window, I gather from my dialogue with the developer that trading isn't actually suspended at all, and that the GMT offset parameter has only been included as part of a planned future improvement. I have to conclude that, if the parameter is non-functional at the moment, I'm not entirely sure why the it has even been included.

The Strategy Tests

As with other recent reviews, and given that spread is always going to be an important factor with any scalping EA, I decided to test Forex WindFall first using a typical broker with a 2 pip EURUSD spread and a 3 pip GBPUSD spread, and I opted again to use the Tadawul broker for my first tests because they not only satisfy the spread criteria, but they're also an older style 4-digit broker and this type of broker isn't generally suited too well to scalping robots.

The default value of the adjustTarget parameter is 0.001, so I decided to leave it at that value for my tests. Also, for backtesting, the GMT offset parameter has no consequence whatsoever, as the EA trades through 24/5 regardless of any prior news events.

Anyway, here are my 10-year Tadawul strategy test results.

Tadawul - EURUSD - 2.0 pip spread
Strategy Test

Tadawul - GBPUSD - 3.0 pip spread
Strategy Test

Having ensured that Forex WindFall could survive using a typical fixed-spread retail broker, I wanted to see how it could perform with some tighter spreads. Recently I've been fixing the FX Primus spread for these tests, but I had some difficulties achieving that on this occasion which I won't go into now, so I decided to carry out my next set of tests using a FX Central Clearing demo account. FXCC are a true ECN broker with very tight spreads and offer their clients a choice of either quoting raw interbank spreads and charging a commission on each trade or of padding the spread to include the commission within the spread.

The raw FXCC spreads for many symbols are often as tight as 0.1 pip, so there is obviously a significantly higher chance of banking a profit using this approach than if the spread is padded to include the commission. This is normally the most cost efficient way of trading, but robots which only bank a single pip or two of profit at a time can see profits totally eradicated by the broker commissions, so care needs to be exercised over which type of robot to run on a true ECN account.

I was pretty sure that Forex WindFall would work with a true ECN broker, so I felt this would be a very good way of testing the EA. FXCC add a 1.0 pip commission to each round turn trade.

FXCC - EURUSD - 0.3 pip spread plus 1.0 pip round trip commission
Strategy Test

FXCC - GBPUSD - 0.6 pip spread plus 1.0 pip round trip commission
Strategy Test

The only observation I had from the backtests were that there a number of OrderSend errors being reported in the tester. This is only a minor irritation and certainly isn't something that would affect Forex WindFall's performance in any way, and it should be quite straightforward for the developer to rectify the problem in a future update.


My first observation about Forex WindFall is that, for a scalping EA trading through 24/5, it isn't exactly what I would describe as a prolific trader. It averages 215 trades a year on the GBPUSD pair and only 152 a year on the EURUSD. I guess, therefore, that there is some pretty extensive trade filtering going on inside the EA to get it to select the seven highest win probability trades that it takes each week across both suitable symbols.

In line with my recent reviews, I also carried out a risk simulation on Forex WindFall. Because commission was being charged on each trade, it was necessary to run the simulation based upon a 0.1 lot standardised monetary average win and loss amount, rather than use the average win/loss size expressed in pips as I would normally do.

Risk simulation based upon the FXCC strategy test results
Risk Simulation

The other thing which I noticed from the backtests was that, although the net profit and profit factor significantly improved with the tighter spreads of an ECN broker, the relative drawdown didn't improve by a corresponding amount as might be expected. I can't put my finger on precisely why this is, and I'm not entirely sure how relevant it actually is, given that the relative drawdown isn't particularly large.

More important than this, however, is going to be the significance of a broker's stops level. Because Forex WindFall sets take profit values which are quite low (13.0 pips EURUSD, 11.5 pips GBPUSD), it might not always be possible to exit a trade via the stealth approach which is used by the EA if the price is closer to the TP than allowed by the broker's stops level. FXCC set their stops level at 0.1 pip, so Forex WindFall was able to close every trade that it wanted within my strategy tests ahead of the TP being reached. This wasn't the same case in the Tadawul tests, however, where the stops levels were significantly higher and Forex WindFall was unable to close out of some trades as intended.

Users therefore need to give some thought to their choice of broker, and the EA is most certainly going to be best suited to either an ECN or STP type broker.

Otherwise, Forex WindFall looks to be a safe EA which can be run from a relatively low initial deposit and users should experience a reasonable level of satisfaction with around half of all users being unlikely to see their account balance drop below their initial deposit.

Forward Tests

For these tests, I've chosen to use a $5k demo Forex Central Clearing account. As described above, this is an ECN account and the demo price feed is identical to the live feed, so it will hopefully portray the Forex WindFall EA in as realistic a light as is possible. The forward test is using raw spreads and a 0.5 pip commission is being added to each side of a trade, making 1.0 pip total commission per round trip.

Forex WindFall has been set up such that it risks 4.0% of the account free margin on each trade on both of its recommended currency symbols. I would therefore expect to see a similar level of performance to that demonstrated within the backtests which used the same level of risk.

There isn't a slippage parameter within the EA, and I have left the adjustTarget parameter described earlier in this article at its default setting of 1.0 pip.

I am also limiting Forex WindFall's maximum allowed spread in pips as follows:

EURUSD - 2.0
GBPUSD - 3.0

As with my other robots in test, you can monitor Forex WindFall's performance at MellyForex by clicking here.
]]> Fri, 01 Jul 2011 13:33:50 +0100
Morpheus OddBot EA - Expert Advisor Review OddBot EAHaving conjured up images of Greek gods and gun-toting Science Fiction characters within my last Forex article, when I declared that I had put the strangely-named Morpheus OddBot EA into forward test, I guess it's now time to meet the man himself!

He's certainly a funny looking chap, isn't he? :)

Appearances can be deceptive, however, and since putting Mr OddBot into forward test, he's certainly been performing well and has been consistently towards the top of my Live Leaderboard, and is starting to appear like a robot which is both profitable and also very safe to run.

In my latest Forex Expert Advisor review I aim to find out precisely just how good the Morpheus OddBot EA really is!

As it happens, I plan to say some nice things about Morpheus OddBot further on within this review, so it's probably best if I put my witch's outfit on at this stage and get the unpleasant matters out of the way so that we can move forwards.

My biggest complaint relates to the quality of the Morpheus website. I do understand that the OddBot developer is planning to redesign the website in the near future but, at the time of writing this article, it lacks a lot of important information about the EA and it isn't particularly easy to navigate.

Fortunately, there is a support forum on the Morpheus website that OddBot users can use for product information, downloads and support, and this has recently been linked to from the main part of the website. I guess it shows that the Morpheus OddBot developer is moving in the right direction and I'm sure that my relatively minor niggles will all be resolved in the near future.

Hopefully, by the time you read this article, the website issues will have been resolved and my above comments will appear meaningless and irrelevant.

Melly! My second observation is that the OddBot PDF installation manual is probably best described as being succinct and to the point.

It consists of only 4 pages, of which just 2 pages relate to setting the EA up. Personally, I'm not too bothered by this as I know exactly what I'm doing when it comes to setting up a MetaTrader Expert Advisor, but I guess that installation could be a little daunting for some of the Forex-newbie users out there who aren't even entirely sure what MetaTrader MT4 is.

Now that the nasty things have been said, I guess it's probably time to take my black pointy witch's hat off and move on to the positives.

What do you reckon, boys? :)

Well, I guess that the biggest positive of all is that the Morpheus OddBot does actually seem to work, and I'm certainly hoping to fill in some of the empty boxes and cover some aspects of the EA that haven't been explained elsewhere within my review.About the Morpheus OddBot EA

The OddBot EA contains security measures which make it difficult to understand exactly what it is doing. The Morpheus website, unfortunately, doesn't give too much away either, other than to say it "uses internal indicator logic.. harmonic/candlestick patterns.. moving averages to place extremely accurate orders". Hmmm, I'm not entirely sure what all that means but, from studying OddBot's behaviour, I'm guessing that the EA identifies a trend direction, then waits for a price retracement which has formed a particular type of chart pattern before entering the market in the same direction as the trend.

It is a scalping EA which only takes a few pips profit at a time and it works on both the EURUSD and GBPUSD currency pairs on the 15-minute timeframe. I did also notice somewhere on the developer's website that OddBot can run on the USDCAD symbol, although my tests focus solely on the two main currency symbols mentioned, and anybody using OddBot on other symbols should carry out their own tests on those symbols first.

OddBot doesn't use Martingale, and it can trade either fixed lots or it can risk a percentage of the account on each trade. Ordinarily, it will trade through 24/5, although it does include both a time filter as well as a news filter, meaning that trading can be limited to certain times of the day and turned off completely for a while both before and after potentially market moving news is released.

OddBot sets a stop-loss and take profit value when entering the market (which can be hidden if required). From studying the trading pattern, it looks as though OddBot waits until a trade is in profit by a certain amount and then adjusts the stop-loss to lock in a small profit. In fact, users have the ability to adjust the amount that is locked in themselves.

Having locked in a small profit, the obvious hope is that the price will continue its move and the trade will reach the take profit level. In practice, however, the price generally either falls back to trigger the profitable stop-loss or the system closes the trade itself (for a decent profit) ahead of the take profit level being reached.

My tests will show that the approach experiences a high rate of winning trades (over 80%), and that OddBot is profitable both in monetary terms and also in terms of pure pips. Morpheus OddBot works on ECN/STP type brokers and it normally only opens one trade at a time, which means that there are no NFA compliance issues. It does, however, include an option to open multiple trades and I noticed that it will then close the later trades before it closes the earlier ones, meaning that it breaks FIFO rules and therefore isn't NFA-compliant when running under this option.

US-based users will therefore need to ensure that they set the Morpheus OddBot up such that it only has one open trade per currency symbol at any time.

Setting the Morpheus OddBot EA up

Setting OddBot up is reasonably straightforward, although there are still quite a lot of external parameters which the user has the option to adjust himself.

No fewer than 10 external parameters relate to the use of the time filter to determine times that the EA should trade, and 4 parameters relate to the news filter. The default is for the EA to trade through 24/5 and to ignore the impact of all news, and I was quite happy to use these defaults when first setting the EA up. Users should also bear in mind that the OddBot EA doesn't have the ability to automatically detect the broker's GMT server offset, and this will need to be calculated and entered manually for users wishing to use either of these filters.

OddBot also contains parameters to control trade lot sizes and risk level, as well as parameters to adjust stop-loss and take profit levels, and to choose whether the SL and TP are to be hidden from the broker or not. Because the OddBot EA is a bit of a black box, it's difficult to know precisely how it calculates its variable risk, and I'm unable to tell if the calculations are performed on account balance, equity or free margin.

Anybody wishing to let OddBot calculate its own lot sizes (instead of using the fixed lot size option) controls the risk via a parameter called AutoMM . In the absence of a definitive explanation within the manual, I had to start several tests and look at the worst case losses to appreciate how this parameter actually worked. It appears that, in setting its value at 10, I was risking approximately 12% of the account balance (or equity or free margin) on each trade, so I experimented with several levels and discovered that the actual risk per trade percentage-wise was factored at 1.2 times the AutoMM setting, and I eventually settled at an AutoMM value of 3 which I figured would risk 3.6% of the account.

Parameters such as maximum spread, slippage and the SecureProfit parameter are what I would personally consider to be the most important, however, and users should give thought as to their input values. With a scalping EA such as OddBot, allowing too much spread or slippage can mean the difference between a profit or a loss from a trade. The SecureProfit parameter is used to control the amount of profit in pips that is locked in once a trade moves nicely into profit. This is also an important parameter, but more about that later!

The Strategy Tests

Given that spread is always going to be an important factor with any scalping EA, I decided to test OddBot first using a typical broker with a 2 pip EURUSD spread and a 3 pip GBPUSD spread. These spreads can be easily obtained by everybody nowadays, and I happen to think it's a good starting point for testing an EA. As with previous EAs in test, I opted to use the Tadawul broker for my first tests because they not only satisfy the spread criteria, but they're also an older style 4-digit broker and this type of broker isn't generally suited too well to scalping robots. My thinking is such that, if a scalping robot can survive a Tadawul backtest, then it stands a good chance of longer term survival and warrants further investigation.

Anyway, here are my 10-year Tadawul strategy test results.

Tadawul - EURUSD - 2.0 pip spread
Strategy Test

Tadawul - GBPUSD - 3.0 pip spread
Strategy Test

Now that I was satisfied that OddBot was at least capable of survival when using a typical retail broker, I decided to run the same tests again using FX Primus as a broker. I wanted to make use of FX Primus's tighter spreads, so I fixed them at 1 pip for EURUSD and 2 pips for GBPUSD to give some sort of an indication as to how OddBot would be likely to perform with an ECN/STP type broker.

FX Primus - EURUSD - 1.0 pip spread
Strategy Test

FX Primus - GBPUSD - 2.0 pip spread
Strategy Test

Clearly, the better performance was being obtained by using the tighter spread but, after carrying out the latest round of tests I was a little concerned that better performance was being obtained on the EURUSD symbol than on the GBPUSD symbol, and I decided to investigate why this might be.

My first suspicion was that tests on both symbols were being carried out using the same SL and TP values, and it was possibly a case that "one size doesn't fit all". I figured that the TP value probably wasn't an issue simply because the TP was never being hit as trades were being closed ahead of its level. Trades would often run to the stop-loss, however, so I decided to look at some optimisation of the SL values on the GBPUSD symbol. To cut a long story short, my results showed that a little optimisation of the SL value would improve things, but the improvement would only be relatively minor. It certainly wouldn't get the GBPUSD performance up to the same standard as the EURUSD performance.

At this stage, I decided to look at the SecureProfit parameter. As I mentioned above, once a trade moves a certain distance into profit, OddBot modifies the stop-loss to lock in a small profit. The default is to lock in just a single pip of profit, and I decided to investigate if locking in a greater profit would improve performance.

For anybody unfamiliar with MetaTrader optimisation, the following images are screenshots showing the different results obtained for different values of the SecureProfit parameter. It's pretty clear that increasing this parameter by just a pip or two above its default value of 1 is actually detrimental to performance, but increasing it by more than 3 or 4 pips results in increasingly improved performance in terms of both increased profit and reduced drawdown in percentage terms.

I stopped the optimisation with a SecureProfit value of 9, and it's possible that, had I continued to increase its value further, performance might have improved further still. I don't really consider it my job to optimise EAs that I'm testing. Instead, I'd rather just steer people in what I feel to be the right direction, and then leave them to discover the best settings for themselves.

EURUSD - Optimisation of SecureProfit parameter on FX Primus with 1.0 pip spread
Strategy Test

GBPUSD - Optimisation of SecureProfit parameter on FX Primus with 2.0 pip spread
Strategy Test

Having carried out this limited amount of optimisation, I then re-ran the full backtests with a SecureProfit setting of 9 pips to see how it affected the look and shape of the balance curve.

FX Primus - EURUSD - 1.0 pip spread and SecureProfit parameter set at 9 pips
Strategy Test

FX Primus - GBPUSD - 2.0 pip spread and SecureProfit parameter set at 9 pips
Strategy Test

Although, not perfect, the GBPUSD curve is certainly smoother than it was before the optimisation.

The only observation I had from my tests using the higher SecureProfit value is that there were a lot of OrderModify errors within the results. This is down to the broker's stops level being employed which will vary from one broker to the next. FX Primus's stops level in my tests was 4 pips. With a SecureProfit of 9, any attempt to modify the stop-loss while the open profit is anything less than 13 pips (ie. 4 pips plus 9 pips) will result in these OrderModify errors.

It's only a minor problem, however, and certainly shouldn't affect performance of the EA in any way.

At that point I decided to call it a day in terms of carrying out further tests. I mentioned that there are time and news filters but, to my way of thinking, the Morpheus OddBot EA looked capable of trading 24/5 through the impact of news events, so I didn't really see too much point in removing blocks of trades when the EA already appeared to be working in an acceptable manner.

I also decided not to test the option which allows multiple trades to be open concurrently, mainly because of the non-FIFO compliance aspect of this particular feature.


I'm keen to continue with my recent policy of performing a risk simulation and of carrying out Monte Carlo projections based upon the strategy test results. With the OddBot results, I ran a simulation based upon the FX Primus test results for both the default and my optimised settings of the SecureProfit parameter.

Risk simulation based upon original FX Primus results with SecureProfit parameter set at 1 pip
Risk Simulation

Risk simulation based upon original FX Primus results with SecureProfit parameter set at 9 pips
Risk Simulation

The improvement resulting from the optimisation work is clearly visible for the GBPUSD symbol, and OddBot would appear to much safer to run as a result. The risk simulations show that it should be possible to start running Morpheus OddBot with a relatively low account deposit, and users should experience a relatively high level of satisfaction, with around half of all users not seeing their account balance drop lower than their initial deposit.

I also decided to show the modified balance curve achieved from using fixed 0.1 lots using SecureProfit set at 9 pips, simply because you rarely get to see them quite so straight!

Fixed Lot Balance Curve

With everything I've said above having been digested, it's important to remember that Morpheus OddBot is still a scalping EA, and the best performance from any scalper is always going to be achieved when using a broker with a very tight spread such as an ECN or STP type.

As with certain other EAs, Morpheus OddBot also seems works well over a prolonged period of time (10 or 11 years in backtests), but there are still spells of 6 months or so within the equity curves where there has been no overall return or even a loss. Prospective users will therefore need to exercise patience should these situations occur again in the future.

Finally, it is going to be important to strictly control the maximum spread and (preferably) not to allow any slippage whatsoever on the trades. This consideration is going to be particularly important if the SecureProfit parameter is left at its default value of 1, as a potential profit will be lost before you know it!

Forward Tests

For these tests, I've chosen to use a $5k demo Forex Central Clearing account. This is an ECN account and the demo price feed is identical to the live feed, so it will hopefully portray the Morpheus OddBot EA in as accurate a light as is possible. The account uses raw spreads and adds a 0.5 pip commission to each side of a trade, making 1.0 pip total commission per round trip.

The EA has been set up with its AutoMM parameter set at 3, so that it risks 3.6% of the account balance on each trade on both of its recommended currency symbols.

Because FXCC are charging a 1.0 pip round turn commission on each trade, I am having to add 1 to the setting for the SecureProfit parameter.

I originally set this parameter to 2 but, having carried out the tests above, I have elected to increase it to 9.

The majority of trades will therefore result in what is effectively an 8 pip profit once the 1 pip commission is taken into account.

I've also set the slippage parameter to zero and am also limiting the maximum allowed spread in pips as follows:

EURUSD - 2.0
GBPUSD - 3.0

As with my other robots in test, you can monitor the Morpheus OddBot's performance at MellyForex by clicking here.]]> Fri, 24 Jun 2011 10:55:10 +0100
Morpheus OddBot and Forex WindFall - Expert Advisors in Forward Test Okay, before anyone asks, I'd just like to say that's not me floating on a cloud in the picture. :)

The curly haired chap stretched out on the couch in his full glory, however, is supposed to be Morpheus, the Greek God of Dreams.

Somehow I suspect that the subject of one of my forthcoming Expert Advisor reviews, the Morpheus OddBot, has actually got nothing whatsoever to do with Greek gods and is more likely to get its name from one of the main characters in "The Matrix" series of science fiction films. The trouble is, I'm not really a fan of anything to do with SciFi (even "Doctor Who" is beyond me!) and I thought my picture of a Greek god was so much more romantic.

Joking aside, the OddBot developer seems to have been active on a number of different Forex forums for quite a while, and has recently taken steps to market some of his own work which, as I'm led to believe, will only be available to purchase through affiliates such as myself. I'm sure that some people will bemoan the fact that the developer's website is a little basic and lacking of content, but I hope his approach means that he is able to devote his time to producing some good products rather than becoming yet another internet marketer.

Fortunately, you'll be able to regularly check-in at MellyForex to find out whether OddBot is performing or not.

The EA itself is probably best classified as a non-Martingale scalper and it works on two currency pairs, EURUSD and GBPUSD, on the 15-minute timeframe.

I've already run some preliminary strategy tests on OddBot and I have to say that the initial results are quite encouraging. It will take a few more days to complete the tests and prepare my full review of the EA but, in the meantime, I've set OddBot up in forward test on a $5k FX Central Clearing demo account using raw spreads and charging a 1.0 pip round trip commission on the trades.
OddBot has been set up with its AutoMM parameter at 3.0 which, by my calculations, means that it should be risking 3.6% of the account balance on each trade on its two recommended currency symbols.

Although my version 1.5 of the EA includes both a time and a news filter, it has been set up to discount the filters and trade through 24/5. Because I would class it as a scalper, I am also limiting the maximum allowed spread and not allowing any slippage on the trades

As with my other EAs, I am also publishing Morpheus OddBot's trades to the website, and OddBot's detailed performance can be viewed by clicking here.

The second EA that I have recently put into forward test and which I am intending to review is also new to the market, and is called Forex WindFall.

Like OddBot, Forex WindFall also works on both the EURUSD and GBPUSD currency pairs, although a difference is that WindFall works on the 30-minute timeframe, whereas OddBot works on 15-minute charts. Both EAs trade with similar frequency though and it's true that some of their trades do even open in the same direction at identical times. There are differences between the two EAs though, and I'm not for one moment suggesting that they're one and the same.

It'll certainly be interesting to compare the two EAs though, and one of the things I intend to look at in the future is whether both EAs could conceivably work in harmony on the same account.

For now though, I'm continuing with my tests ahead of publishing my Forex WindFall review.

While I'm doing this I've set Forex WindFall up on a $5k FX Central Clearing demo account using raw spreads and charging a 1.0 pip round trip commission on the trades.

I'm risking 4.0% of the account balance on each trade and strictly limiting the allowed spread and not permitting any slippage.

I hope to see a high percentage rate of winning trades (over 80%) from both of these EAs, and Forex WindFall in particular promises relatively low drawdowns. You can monitor WindFall's performance at MellyForex by clicking here.

Before I finish this article, I'd also like to mention a few other EAs that some of you will have noticed in forward test, and I've received a few emails from readers asking questions about them.

I first mentioned that I had put Forex Shocker into forward test in an earlier article a while ago. At the time I suggested that my Forex Shocker strategy tests were showing very mixed results and that I didn't know if I would bother to prepare a review. Well I guess the proof of the pudding is in the eating, because its performance since then has been dire. It has currently lost over 29% of the initial deposit and has drawn down over 44% along the way, so there's absolutely no way that I intend to prepare a full review.

Secondly, I mentioned in another recent article that I had put FxPowerCube into forward test and that I was carrying out strategy tests with the intention of preparing a review. Disappointingly, although the strategy tests show that the EA would have been marginally profitable over the last decade, I have to say that the equity curve is one of the weirdest I've ever seen. :(

I don't, therefore, intend to spend any more time looking at FxPowerCube, despite the fact that it is currently showing a small profit in forward test.

In case anyone's interested, I've attached my FxPowerCube 11-year backtest and report below so that people can get an idea of what I'm talking about when I say the equity curve looks weird.

Tadawul - EURUSD - 2.0 pip spread
Strategy Test

For what it's worth, I suspect that this EA will behave totally differently from one broker to the next. Amongst other timeframes, it uses the 4-hour timeframe in its calculations and, because of their different GMT offsets, one broker's H4 price bar can open and close at a different hour to another broker's H4 price bar and also register different high and low values in between. This can affect the values that are returned by indicators used on the H4 timeframe and result in a completely different set of trades taken by an EA. Indeed, I've even noticed that my forward test trades seem different to others I've seen published on the web. It might, therefore, be possible to get FxPowerCube working successfully on one or two brokers, but I'm afraid that I don't have the time to experiment looking for those lucky brokers.

Despite the fact that I don't intend to prepare full reviews on either Forex Shocker or FxPowerCube, I plan to leave them both in forward test to see how they perform.

Finally, I recently put the Primeval EA into forward test. This EA is quite well known, having been around for a while now and I'm not, certainly at this stage, planning to carry out a full review. I've really just loaded it up in test as a sort of benchmark to see how it compares to some of the other EAs that I'm testing. For anybody interested in Primeval, they can easily follow its performance here at MellyForex by clicking here. The tests are being conducted on a $5k Alpari UK demo account and version 2.1 of Primeval is being run on all 5 suitable symbols with default settings.]]> Fri, 10 Jun 2011 15:28:08 +0100
Forex Real Profit EA - Expert Advisor Review Forex Real Profit EAThe Forex Real Profit EA is a scalping robot which looks to trade at the time of day when the US are closing and before Asia has opened.

I've already got a couple of other EAs in test which could be classified like this as pre-Asian scalpers, and my biggest concern about robots such as this is that liquidity in the market is at its thinnest during these times and there can be issues in executing trades. The inevitable requotes and slippage often means that this type of EA performs well in demo, but fails to live up to expectations when users run the system live. :(

The attractions of the Real Profit EA, however, are that it not only seems to offer a quite attractive risk/reward ratio by comparison to some of the similar robots available, but that the developer is happy to test his EA using real money live accounts as proof that his product works.

The website contains quite a bit of factual information and isn't the typical single sales letter page which is commonplace with many other EAs. There are both live statements and backtest results on the website, together with a forum which is maintained by the Real Profit EA team to provide support for their users. It is also possible to download a FREE demo version of the EA which enables prospective purchasers to test the EA on a demo account to see how it performs before they buy.The Forex Real Profit PDF instruction manual consists of 26 pages, and is clearly written and easy to understand. As is normally the case with these manuals, around half of the pages cover how to install and configure MetaTrader 4.

About the Forex Real Profit EA

Forex Real Profit EA As I've already stated, this robot trades for just one hour each day and it has default settings for 7 different currency symbols - EURUSD, EURCHF, EURGBP, USDCHF, USDCAD, EURCAD and GBPCHF. In addition, users are welcome to run the EA on any other symbol they choose, but doing so is very much at the user's own risk as it won't have been optimised by the vendor and the user will need to perform his own testing and optimisation.

The EA only opens one trade at a time, so NFA non-hedging and FIFO compliance isn't an issue. In addition, the EA contains a neat little feature which will prevent the EA from opening a trade if there is another trade already opened by a different EA on that symbol. As I say, the feature is a nice idea, but it really needs some other EA developers to follow suit with their products in my opinion, otherwise the feature will only work one-way.

Some days, the EA may take several trades on the same symbol within the same 1-hour session but, at other times, it may go several days without taking any trades. The fact that it works on a number of different currency symbols therefore serves to make the EA more appealing, as it normally finds a trade on at least one of the suitable symbols each session.

The Forex Real Profit EA will also trade only on four days of the week. It won't trade on Fridays so as not to leave any open trades over a weekend, and it won't trade on Sundays in case some market moving news has occurred during the weekend which could impair the EA's performance. In addition, all US public holiday dates are programmed into the EA to prevent it trading then.

The EA trades on the M15 timeframe and, although its precise logic is hidden away securely inside its DLL, I have managed to figure that it gets much of its mojo from the use of Bollinger Bands coupled with an underlying directional filter.

The EA uses a static stop-loss and its initial risk/reward ratio based upon that stop-loss is quite high, but the EA generally manages to close trades well ahead of a SL being hit. This has the effect of lowering the risk/reward ratio (expressed in pips) to a very respectable 1.5. This value is very low for a scalping robot.

Setting Real Profit EA up

I've made no secret of the fact in previous articles that I like to see an EA which is "idiot proof" and easy to set up. I just happen think that if there are only a few external parameters to alter, then there is less opportunity for a user to make mistakes.

The Forex Real Profit EA contains no fewer than 27 external parameters, of which the first 3 are merely cosmetic and serve no purpose whatsoever. I guess that having lots of parameters means that some users feel psychologically that they are in greater control, but I would personally rather see parameters such as the trading times hard-coded into the EA to reduce the risk of error. On the subject of trading times, the EA contains an option to obtain the broker's server GMT offset automatically which is OK for live trading, but this needs to be manually over-ridden in the strategy tester.

There are also parameters for controlling various filters and for controlling the stop-loss and take-profit system. I decided to leave these at their default settings for my tests.

The other adjustable parameters relate to the EA's money management system. Users can decide whether they want to risk a predetermined percentage of the account balance, the equity or the free margin, or they can trade fixed lot sizes if they wish. In my forward testing, I also set the EA not to allow any slippage whatsoever, and I was careful to restrict the maximum allowed spread because of the thin liquidity seen at the times the Real Profit EA trades.

While I'm in my "moaning mode", one thing I found confusing about setting this EA up was that the manual suggested it should trade for two hours a day during the Northern Hemisphere winter and for one hour a day during the summer. Seeing as the trading times aren't automatically detected and need to be input manually, I figured the only way to backtest the second "winter hour" would be to run a series of individual 6 month backtests and merge them into one report. I asked the vendor about this and his reply was that the EA's current recommended trading time is for just one hour a day. I hope that this will be made clearer in the future, and the instruction will be suitably edited, because I have to say that I was a bit confused by this aspect of the setup.

The Strategy Tests

It was my initial intention to test the Real Profit EA on all seven of its suitable symbols, but I had a problem in obtaining any history data for two of those symbols (EURCAD and GBPCHF). It may simply be that the MetaQuotes history centre was having difficulty on the day I tried, but I was able to run good tests on 5 of the symbols and I decided that this would be sufficient for the purposes of my review.

For the strategy tests, I opted to use the Tadawul broker as I have done previously. I don't have any particular preferences for Tadawul, but they don't offer fractional pricing and their fixed spreads are mid-range, so I feel their trade performance in tests is likely to be indicative of what most users should realistically be able to achieve.

Anyway, the test results are below. You can click on any of the equity curve images to view the full report.

Tadawul - EURUSD - 2.0 pip spread, 5.0 pip stops level
Strategy Test

Tadawul - EURGBP - 3.0 pip spread, 5.0 pip stops level
Strategy Test

Tadawul - EURCHF - 3.0 pip spread, 6.0 pip stops level
Strategy Test

Tadawul - USDCHF - 2.0 pip spread, 5.0 pip stops level
Strategy Test

Tadawul - USDCAD - 4.0 pip spread, 8.0 pip stops level
Strategy Test

I did notice in the tester logs that there were a lot of order send and close errors in the logs because the price isn't being normalised when the order is being sent. This is a straightforward matter to resolve and I hope the problem is rectified with the next update.


Despite being unable to test the EA on two of its symbols, I strongly suspect that the equity curves for those symbols would look very similar to the curves of the other five symbols. My only concern is that the Tadawul fixed spread for both of those symbols is a massive 8 pips which is about the same as the average profit per trade of the other 5 symbols. I have to question, therefore, the logic in using those two symbols on anything other than an ECN/STP broker and I suggest the maximum allowed spread should be controlled very tightly, even if this were to result in some trades not being taken.

The equity curves in the strategy test reports above very much speak for themselves, but my biggest concern from the outset has been whether a similar level of performance can be replicated going forwards and market liquidity may be an influencing factor in this.

Another concern is going to be the fact that some trades look to have been opened and closed within a minute (which is, by nature, the definition of scalping). Many brokers impose little known rules such as "the 90-second rule" on their clients, meaning that trades need to be open for a minimum length of time before they can be closed. This is one method employed to prevent clients from scalping, and it's quite possible that the Real Profit EA could fall victim to it on some trades with brokers using a minimum trade duration policy.

My final concern relates to the importance of using a broker with a low stops level. The Real Profit EA likes to close trades ahead of the take profit level and, if the stops level is too high, it will be unable to do so and performance of the EA will be impaired. If your broker employs too high a stop level, the solution may be to set the "UseTakeProfit" parameter to false, but I haven't tested this setting to compare performance.

Clearly, the choice of broker is going to be very important for the Real Profit EA, as it will be necessary to choose one with a tight spread and low stops level, who really does allow scalping.

I also ran my backtests above through a risk simulator, to see what else I could learn about this EA, and those results are tabulated below.

Risk Simulation

It's pretty clear that, provided the backtests can be replicated live, users of this EA are likely to experience a very high level of satisfaction. The Monte Carlo Simulation suggests that only around one in five users is likely to see their account drop below the initial deposit, and the deposit needed to start running the EA with zero risk of ruin looks to be very low by comparison to the vast majority of EAs. The fact that Real Profit EA can be used on a number of symbols qualifies it as a frequent trader, and the suggestion is that recovery from a drawdown should be quite quick.

Before proceeding, and before I get an inbox full of emails from readers of this article, I should perhaps point out that the maximum drawdown of a couple of the symbols in the table is less than the minimum deposit which the simulation suggests is necessary to start running the EA. This is because the probability of hitting the maximum drawdown straight after loading the EA is very remote, but it is a possibility nonetheless and users should at least be prepared for this possibility and deposit sufficient funds to cover the consequences of it happening.

The proof of the pudding with this EA is clearly going to be whether it can replicate the great backtests results going forwards. If it can do so, then it will certainly be worth every penny spent, and I recommend that users check back here regularly to monitor its ongoing performance in my forward test.

Forward Tests

For these tests, I've chosen to use a $5k demo Forex Central Clearing account. This is an ECN account and the demo price feed is identical to the live feed, so it will hopefully portray the Real Profit EA in as accurate a light as is possible. The account uses raw spreads and adds a 1.0 pip commission per round trip trade.

The EA has been set up to risk 3.5% of the account balance on each trade on all 7 of its recommended currency symbols. In my opinion, a higher risk level would inflict too much damage if multiple SLs were hit together.

I've set the slippage parameter to zero and am also limiting the maximum allowed spread in pips as follows:


You can monitor Forex Real Profit EA's performance at MellyForex by clicking here.]]> Wed, 01 Jun 2011 13:25:00 +0100
FxPowerCube and Real Profit EA - Expert Advisors in Forward Test
Forex Real Profit EA has been around for a little while now and has already started to forge a good reputation for itself. It trades on seven different symbols for just one hour each day at around the time that New York is closing, and looks to scalp a few pips at a time.

Normally, I'm not a fan of this type of robot, but this EA certainly seems to work and I wanted to see if it's possible to replicate this performance.

Because of its trading nature, I've set the Forex Real Profit EA up on a $5k demo FX Central Clearing account. This is an ECN/STP account and the demo feed is identical to the FXCC live feed with the added benefit that demo accounts have the option of either treating the commission separately or of adding the commission to the spread which, in turn, has the effect of widening the spread and possibly affecting performance. I've therefore chosen to handle the commission separately in order to keep the spread at its raw level to achieve the most meaningful and accurate results when testing this EA.

Because it's trading on seven symbols, Forex Real Profit has been set up to risk no more than 3.5% of the account balance on any single trade and, because of the general lack of market liquidity around the time of day that it trades, I'm strictly controlling the maximum allowed spread and not allowing any slippage. FXCCFor anybody who may be unaware of FX Central Clearing, they're based in Cyprus and are regulated by CySEC. They're also authorised by the FSA, compliant with MiFID and are a member of the Investor Compensation Fund which was established to protect retail clients in the event a member firm is unable to settle its claims.

In addition to the MetaTrader platform, FXCC also offer the Currenex platform, together with a multi-account management application for the use of fund managers and a mobile platform for traders who like to trade with their iPhone while they're playing tennis. :)

FxPowerCube The minimum FXCC deposit for a live account is $500 and they allow hedging which, unfortunately, prevents them from accepting US clients. You can find out more about FXCC and set up a demo account of your own to try before going live by clicking here. The demo accounts are non-expiring provided they're used regularly, but it's important to remember that you will need to register your new account via the hub on the FXCC website rather than trying to open the account through the MetaTrader terminal, as that particular MetaTrader function has been disabled.

What I'm trying to say here is that FXCC are a professional outfit who are a cut above the average retail Forex bucket-shop.

Oh, one other thing, although you'll have probably realised that I'm acting as an introducer for FXCC, I promise that I don't charge anything on top of FXCC's raw spread, so please be sure to use one of my links if you like MellyForex and find my reviews helpful.

Seriously, it's what helps to keep the site free for your enjoyment.

I'm currently carrying out tests and my review of the Real Profit EA will follow soon but, in the meantime, you can monitor its progress on both the Live Leaderboard and at the EA's detailed analysis page by clicking here.

My second EA under test is called FxPowerCube and it claims to operate using three independent strategies.

Being brand new, I haven't managed to run any proper strategy tests on FxPowerCube yet, but I can say that it works on the EURUSD symbol on the 1-hour timeframe.

While I'm carrying out my strategy tests and evaluating the results, I've put the EA onto a $5k FXCC ECN demo account to see how the robot performs.

Judging from the vendor's website, the EA seems to backtest well over a period of the last couple of years, and the recent live performance shows much promise, so I plan to find out just how good it really is.

FxPowerCube is risking 5.0% of the account free margin on each trade. You can monitor its progress and keep track of its trades at MellyForex by clicking here.]]> Thu, 26 May 2011 09:35:00 +0100
Forex Growth Bot - Expert Advisor Review

When I decided to write my review about Forex Growth Bot, I had to think long and hard about what to say in the review.

The fact is that the EA has been around for a while now, and there's a lot about it that's already been written.

I didn't just want to simply repeat somebody else's words, therefore, and wanted to come up with something unique for this review.

Fortunately, while I was in the process of carrying out the various tests for my review, I learned something about Forex Growth Bot totally by accident and, hopefully, my analysis of the EA fulfills that "unique criteria", but you'll need to read on to find out more....The Forex Growth Bot website consists of the typical sales letter page that everyone is doubtless familiar with by now. Unlike some EA sites, however, there are backtests on the website dating back to 2002 plus a rather interesting forward test statement (which is verified at MyFxBook) dating back to the beginning of November 2010 which shows some pretty impressive gains since it was started.

About the EA

Forex Growth Bot comes in its own self-installer and includes a licence management program which enables users to activate the EA for different account numbers.

It works primarily on the EURUSD currency symbol on the M15 timeframe, and is probably best described as a "pyramiding" trend-follower. For anybody who is unaware, the process of pyramid trading involves using the extra margin which is made available from an open winning position to open additional positions in the same direction.

This technique can produce explosive profits, but it can also be a little risky if several positions are open simultaneously which don't amount to anything, as the system can then experience higher than normal losses. It's therefore important to use a strict approach towards trade exits with this type of system, in order to stop the losses from mounting up.

I'm not entirely sure what criteria Growth Bot applies for deciding when to add to a position, and I'm pretty sure that it doesn't just use the free margin generated from open winning trades in the true sense of a pyramid trader, but it certainly seems to work! Although its win rate is relatively low (around 40% in backtests), the average winning trade size is a little under twice the size of the average loss size, demonstrating the effectiveness of Growth Bot's pyramiding strategy.

Setting Forex Growth Bot up

The Growth Bot installation is quite straight forward and the PDF instruction manual is easy to understand. There are 10 external parameters in the standard version of the EA but, for most users, it's probably only the lot size parameter and FIFO setting which will need to be altered. The other parameters relate to such things as whether a SL and TP are to be declared and some hocus pocus to do with optimising the EA for use on symbols other than EURUSD. I chose to leave the other parameter values at their defaults when I began testing.

It's probably also worth noting that there are "Power" and "Advanced" versions of the EA which are available as an upgrade and contain even more external parameters that can be adjusted, but I'm only evaluating the standard version here.

Growth Bot It's difficult to know precisely how Growth Bot works, as much of the trading logic is hidden away inside the EA's DLL, but it's no secret that the EA uses something to do with market volatility when entering its trades. I would guess that it enters the market once the volatility starts to increase marking the start of a trend, and exits its trade(s) as volatility starts to decrease. It seems to only trade at the start of a new 15-minute price bar and can have up to 5 open trades running at once as part of its pyramid strategy that I described earlier.

By default, Growth Bot won't send either a SL or a TP to the broker and, instead, it monitors its trade exits internally. It also seems to "phone home" regularly, as it seems to monitor and compare its trades against similar trades taken by the EA on the vendor's server, and is able to report if any signals to close out of a trade have been missed. I'm sure there will be many users who are prone to server outages who won't be happy that a SL isn't declared, so the EA also includes an option to set one anyway. The ability to compare trades against those on an external server should also act as some form of comfort to those worried that a SL isn't being declared.

Money management within the EA is, in my opinion, a little primitive as the EA only trades fixed lot sizes. Personally, I'd rather see an EA risk a predetermined percentage of the account balance on each trade and calculate its lot sizes based upon the distance away that the stop-loss is located. I guess that must be a little difficult seeing as the EA doesn't ordinarily set a SL as mentioned above but, even if the SL is managed internally, it shouldn't be too difficult to incorporate this sort of risk management system. The manual does set out guidelines for adjusting lot sizes according to account balance which, quite frankly, is a bit basic.

The Strategy Tests

I've previously used the Tadawul broker for my strategy tests and saw no good reason to change on this occasion. They're a 4-digit broker with a 2.0 pip EURUSD spread, and I figured this would produce some fair and representative results.

Tadawul don't apply NFA FIFO rules, so my first backtests were with Growth Bot's FIFO parameter set to false. For anybody reading this who is unaware, FIFO stands for First-In-First-Out and it means that, when several trades in the same instrument are open together, the first trade to have been opened must always be closed first before any other trades can be closed.

Anyway, the test results are below. You can click on any of the equity curve images to view the full report.

Tadawul - 2.0 pip EURUSD with FIFO false
Strategy Test

The first thing I noticed in the results is that some of the trades were taken with lot sizes which were 2 or 3 times the size I had entered. I assume that, seeing as it's a pyramid trader, this is an intentional part of the EA's strategy rather than being a bug in the code. I haven't asked the developer about it though, and you can see the trades in question by looking at the detailed statement above where trade no 57 is the first trade of several which portrays the strange increased lot size. The giveaway sign is the fact that some of the green lot size bars in the equity curve GIF image are higher than others. If every trade were the same lot size, then there wouldn't even be a green lot size bar across the bottom of the image.

Having run a backtest with the FIFO parameter set to false, I decided to re-run the same test with the parameter set to true. It's my guess that around 40% of MT4 users around the world are based in the US and subject to NFA regulations, so I figured it was only fair that I looked at a system that those guys could use.

The test results are shown below, but I have to say straight away that there's something a little odd about them. The net profit with FIFO true and false seems to be identical, but other aspects of the report such as number of trades, percentage of winners, and gross profit and loss values is different. I really can't see why this should be (and don't have the time to dwell too long on the matter) so, seeing as the net profit for FIFO and non-FIFO seemed the same, I figured the best thing to do would be to use and analyse the results with FIFO set to true.

Tadawul - 2.0 pip EURUSD with FIFO true
Strategy Test

At the start of this article, I mentioned that I had discovered something about Forex Growth Bot by accident. When I was setting my backtests up, I entered the wrong symbol and ran a pass on the GBPUSD symbol by mistake. I quickly realised the error of my ways, and stopped the test so that I could continue testing on the EURUSD, but I couldn't help but notice that Growth Bot appeared to work OK on the GBPUSD. The Growth Bot instruction manual says that the EA can be used on other symbols so, with this in mind, I decided to run a full test on the GBPUSD. I set FIFO to true, and the results are below.

Tadawul - 3.0 pip GBPUSD with FIFO true
Strategy Test

I should point out that the GBPUSD settings are "straight out of the box", and I guess that the above performance could probably be improved upon with a bit of optimisation.

I did also try Growth Bot with a few other currency symbols, but the results weren't good and I didn't have much time to spend experimenting unfortunately. If the EA is to work on other symbols, it's my guess that quite a bit of time would need to spent carrying out optimisation to find the best settings. It would probably also be better to purchase one of the enhanced versions containing more external parameters if you intend to experiment with other symbols.


Although I didn't mention it above, I also ran Growth Bot through the tester once again using some 99% modelling quality tick data from April 2007 using a real variable spread. Normally, I would expect the tick data results to show worse performance than the history centre data results. In the case of Growth Bot, however, the tick data results were actually much better than the history centre results. Not only that, but the history centre tests weren't particularly impressive for the first couple of years of the full 10-year test, and it's my suspicion, therefore, that Growth Bot's optimisation has been weighted more towards the last 3 or 4 years. Whilst this is good in so far as the EA should deliver well in the shorter time, it needs to be remembered that the current settings could go out of tune at any stage, and the EA's performance would then revert closer towards its longer term average.

I decided to run the history centre results through a risk simulator to get an idea of what users might expect, using the history centre test results in preference to the tick data results, as I felt they would be more indicative of the EA's longer term performance. I also ran the simulator on just EURUSD as well as on the combined results of both EURUSD and GBPUSD for comparison. The risk simulation results are set out below.

Risk Simulation

What's interesting is that running Growth Bot with both of the symbols results in well over twice as many trades, but the minimum deposit needed to run both symbols safely with zero risk of ruin is only an extra 35%. In addition, the time in days before users are likely to see their return exceed their drawdown is likely to be considerably less when running both symbols (around 15 days, compared to around 30 days for EURUSD alone) because of those extra trades. Remember as well that the GBPUSD test results used in my risk simulation are out of the box. With a bit of optimisation, it's possible that even better performance could be achieved.

Users should also be prepared to accept the rather unusual product licencing system employed by Growth Bot, and the possibility that some strange or buggy behaviour may occur resulting from Growth Bot's comparison of local trades against trades taken on its own server.

Prospective users should also note that the balance curves in my strategy tests above are prone to prolonged periods where they can flatline. Those periods may well last a year or two so, whilst Growth Bot has certainly been performing recently, longer term users may need to exercise considerable patience.

In conclusion, therefore, I would have to say that Forex Growth Bot certainly looks to have great potential, although the minimum deposit needed to run it safely is somewhat on the high side compared to other EAs. In addition, I would like to see an improved money management system implemented which uses variable lot sizes.

Otherwise, I would strongly suggest that users try running it with the GBPUSD symbol as well as its standard EURUSD symbol, and that they also consider looking at one of the enhanced versions to enable better optimisation of some other symbols.

Forward Tests

As part of my testing, I've set the Forex Growth Bot up on a $5k Alpari UK demo account trading only the EURUSD symbol on the 15 minute timeframe. Alpari UK don't have to comply with NFA rulings, so I've set the FIFO parameter as "false". That said, I wouldn't expect much overall difference in performance if I had set it at "true". Otherwise, I'm using the default settings and am trading 0.1 lots per trade.

You can monitor Forex Growth Bot's performance at MellyForex by clicking here.

]]> Tue, 17 May 2011 08:33:47 +0100
Risk Simulation and The Monte Carlo Method - Forex Article There do not seem to have been any new EAs released recently of sufficient interest to capture my attention, so I thought I'd write an article about how to use either backtest or forward test results to gain further information about how good or bad a trading system actually is.

I guess this article could be viewed as a follow up to my previous article asking "Why do all Expert Advisors Fail?". In that article I suggested that some or many EAs didn't actually fail at all, but instead were discarded or written off by their users because those users had a completely different level of expectation to what the system would be likely to deliver in reality.

Let's face it, wouldn't it be fantastic if you were to buy an Expert Advisor, deposit $10k into an account, and the EA never had a losing trade? Inside of a year, your account balance might be up to $20k, or $100k or even more.

Sadly though, life just isn't like that, and stories like that very rarely actually happen in reality.Readers with a keen eye might like to note that in my last paragraph I wrote "very rarely". The implication here is that there are instances where this does actually happen. The point I'm trying to make here is that it would be good to know at the very outset what the probability is of that scenario actually happening. Or, turning that scenario on its head, what are the chances of your $10k deposit being eroded to zero within that first year?

Believe it or not, there's a wealth of information such as this which can be obtained very quickly from some basic backtest results by using something called a Risk Simulator.

For instance, it's a straightforward task to find out how much minimum account deposit is needed to run a system safely with zero risk of ruin. It's also possible to find out how many trades a system will probably need to make (or how many days it will need to run) before the profit which is returned exceeds the drawdown that has occurred in the meantime.

It's also very easy to carry out something called a Monte Carlo simulation. Despite having a name which might conjure up images of gambling and casinos, the Monte Carlo Method is simply a method of using computational algorithms to generate models with uncertain inputs. There's more about Monte Carlo on Wikipedia for anybody who's interested, but I'll leave it that a Monte Carlo simulation is used to advise such things as likely returns, the risk of ruin and the probability that the account balance will at some stage drop lower than the initial deposit.

These are things which are impossible to ascertain from a conventional MetaTrader statement, but which are actually more important to know than anything else that the statement might say.

So how does Risk Simulation work?

When you look at any account statement and see that the win rate is 70% this means that, in a sample sequence of 10 trades, there will on average be 7 winning trades and 3 losing trades. The actual trade sequence might be W-W-W-W-W-W-W-L-L-L or it may be L-L-L-W-W-W-W-W-W-W or, more likely, it will be something totally random along the lines of W-L-W-L-W-W-W-L-W-W. The end result will nearly always be the same, but the journey along the way will differ according the exact sequence of those winning and losing trades.

Risk Simulation involves replaying different random win/loss sequences to produce different journey paths and find the best and worst things that could potentially happen along any of those journey paths. The only information that is needed to perform a risk simulation is the size of the average winning trade, the size of the average losing trade and the percentage win rate. It's then possible to generate a series of random trade sequences to find out just how safe a trading system really is.

The information above can be obtained either from a backtest or a forward test statement. Whichever method is used, it's important that there are enough trades in the report for it to be truly representative. In other words, it's no good if the report just covers a period of a few weeks. Secondly, if a backtest is used, it's important that the data used for the test is accurate. That means that data with a 90% modelling quality is the absolute minimum that should be used. If the system is a scalping system with low SL and TP values, significant attention needs to be paid to the spread used in the test and, if possible, 99% modelling quality tick data should be used.

The chances are that there will be either no, or only limited, forward test available available for most new EAs, in which case the only option will be to use the backtest data in a risk simulation exercise.

The other thing that's important to note about risk simulation is that trade lot sizes are generally fixed within the simulation test, and the compounding effects achieved by using any form of money management are ignored. This means that the system NEEDS TO BE PROFITABLE IN TERMS OF PIPS WON, not just in terms of money. Many Martingale type systems don't fulfill this criteria, and a risk simulation will simply show straight away that there is a 99.9% risk of ruin. Risk simulation certainly shouldn't be ignored!

Risk Simulation Results and Example

As an example, I've entered some imaginary results into a risk simulator to demonstrate how it works.

The average win size of this imaginary system is 4.4 pips, its average loss size is 30.3 pips, and the win rate is 89%. I've set it up to simulate 2,000 random trades using 0.1 lot sizes with an initial $2,000 deposit. Using fixed lot sizes, a conventional MetaTrader report would show the system to be profitable with a Profit Factor of 1.17.

Monte Carlo Simulator

The simulator very quickly makes 5,000 random passes and tells me that the likely return is 58.1%, the likely drawdown is 12.9%, there's a 1.0% risk of incurring a 30% drawdown (I chose this drawdown level in the test), and that only 14.6% of users will not see their account balance dip below their $2,000 deposit (meaning that 85.4% of users will see their deposit eroded at some stage).

I could obviously have set the maximum drawdown level at any level I wanted, but I chose 30% because experience dictates that this is the level at which most people can take no more pain and switch the system off. The mere fact that there was a 1.0% chance of hitting a 30% drawdown is not good enough and renders the system as being unsafe, so the next stage of the risk simulation process involves finding out what minimum deposit level is needed to run this imaginary system so that there is zero chance of incurring a 30% drawdown.

Fortunately, the simulator provides that information as well. It just continues increasing the deposit amounts and running fresh sequences of random passes until it reaches an amount where the risk of ruin is 0.0%.

Minimum Deposit

In the case of our imaginary system, the simulator suggests that a minimum deposit of $2,4333 would be necessary to start trading 0.1 lots with a 0.0% risk of hitting a 30% drawdown. Similarly, a minimum deposit of $243 would be needed to start trading the system with micro lots.

Having established the minimum deposit which is needed to run the system safely, the final task is to get an idea of how long it will take to see a return from the system which is higher than the drawdown which has been incurred along the way.

Minimum Trades

The simulator tells us that it would probably take around 261 trades before the return exceeded the drawdown. Assuming the system is a relatively high frequency trader which takes, say, 5 trades a day, it's worth bearing in mind that it might take 10 or 11 weeks to see this return. You would, therefore, need to be quite patient if you were to use this system live, as most people expect to see a return in a far shorter time. This is precisely the sort of scenario I was referring to in my previous article entitled "Why do all Expert Advisors Fail?"

Hopefully, the above will give some insight of what can be gained from a test report and it underlines the importance of backtests using accurate data. Even so, you'll notice that I've used the words "probably" and "likely" in the article above. That's because there are no cast-iron guarantees when it comes to trading. It is fraught with risks which can never be removed, and successful traders will simply use Probability Theory to control those risks.]]> Thu, 05 May 2011 13:37:00 +0100
Forex Growth Bot - Expert Advisor in Forward Test Forex Growth BotI was first attracted to the Forex Growth Bot because it offers what appears to be a very attractive risk/reward ratio. Many of the EAs that I look at offer a very high win rate (often 80% or more), but they only achieve this high win rate through taking small profits and operating large stop-losses. This gives them a very high risk/reward ratio, and it means that users can sometimes find a single stop-loss with a high risk/reward robot will wipe out weeks or even months of gains. :(

Forex Growth Bot, however, appears to be different to these types of robot. The backtests on the vendor's website (and also the forward test which has so far been running over 5 months) suggest that the system is likely to win less than 50% of its trades, but its risk/reward ratio seems to be somewhere between 0.4 and 0.5, meaning the average profit is likely to be over twice the size of the average loss. The EA is made even more attractive by the fact that its average win size is around 60 or 70 pips which should theoretically mean that its performance is unlikely to change much regardless of which broker it's run on.

The system works on just one currency symbol, the EURUSD, and runs on the 15-minute chart. I wanted to test it for myself to see if it can live up to its promises.
Unfortunately, the Forex Growth Bot system is very much a black box and it's difficult to fully understand what it's doing. It seems to be indicator driven and, if its indicators align, it will open several positions at 15 minute intervals, so it's not uncommon to have 4 or 5 trades open at the same time.

I'm hoping to understand a bit more about how it works from my tests and, while I'm carrying out those tests and preparing my review, I've put Forex Growth Bot onto a $5k Alpari UK demo account to see how it runs.

The EA is running on the 15-minute EURUSD chart using the default settings as recommended within the manual, and is trading 0.1 lots as is suggested in the manual for account balances over $2,000.

You can monitor Forex Growth Bot's performance at MellyForex by clicking here.]]> Fri, 15 Apr 2011 12:14:00 +0100
Forex Morning Trade - Expert Advisor in Forward Test Forex Morning Trade

Forex Morning Trade is an EA which is based upon a manual trading system of the same name.

The system works on just one currency symbol, the GBPUSD, and there's a very strong clue as to how it works in the name. :)

For all budding detectives out there, you've probably guessed that it takes one trade at the same time each morning. Unfortunately, the time it trades is a time when most Europeans are still in their PJ's and most Americans are either already in bed or enjoying their evening.

There's a very strong background case, therefore, for converting a strategy like this into a robot.Being derived from a manual trading system, the Forex Morning Trade system is very transparent. It is indicator driven and the manual explains precisely how it works, so the EA is not a black box in the same way that some EAs are.

While I'm carrying out some tests and preparing my review, I've put Forex Morning Trade onto a $5k Alpari UK demo account.

The EA is running on the 15-minute GBPUSD chart and is risking 2% of the account on each trade using the default settings.

You can monitor Morning Trade's performance at MellyForex by clicking here.]]> Tue, 12 Apr 2011 10:02:00 +0100
Wall Street Forex Robot - Expert Advisor Review Wall Street Forex RobotThe Wall Street Forex Robot is a scalping robot which looks to bank just a few pips of profit at a time. There is no shortage of robots like this available on the market, but they generally tend to scalp during thin Asian trade when market liquidity can be a major issue which impairs the robot's performance on live accounts.

So what makes the Wall Street Forex Robot any different from all the others?

Well, the Wall Street EA is certainly different in that it has no such trading time restrictions and it will trade at all times of the day, hopefully when there is greater liquidity within the market and there is no problem in getting trades away.

It's also different in that it fetches its settings from the vendor's server each time that it starts. This means that the vendor can issue newly optimised settings which the EA will automatically detect and use, enabling it to stay at the top of its game without the user needing to manually update his own copy of the robot.

That's the theory anyway. I wanted to test the EA thoroughly to see if it could live up to this high level of expectation.The Wall Street Forex Robot website is a typical single-page sales letter which makes plenty of claims about how good the EA is and takes forever to scroll down from top to bottom. I'm no fan of websites such as this but, as they go, the Wall Street Forex Robot website is one of the better ones, as it is well written and contains impressive graphics.

Like many robots nowadays, Wall Street Forex Robot comes in its own self-installer for quick installation and there is even an MT5 version of the EA included free within the purchase price for anybody wishing to run this platform

About the EA

The Wall Street Forex Robot works on the 15 minute timeframe and seems to take its trading signals from a combination of several indicators which align together. The EA's main trading logic is protected inside its DLL, so it's difficult to know precisely what it's doing. Anyway, I'm more concerned about whether Wall Street Forex Robot works or not.

Each copy of the Wall Street Forex Robot can be used with unlimited demo accounts and up to three live accounts. The EA has the option of either trading fixed lot sizes, or of its own money management system which seems to risk a fixed percentage of the free margin on the account depending upon the distance away of the stop-loss.

The robot sends a fixed pip SL and TP to the broker immediately after opening an order. Once a trade is in profit by a certain amount, it then moves the SL to the order entry level so that the risk on the trade becomes zero. It only seems to adjust the stop-loss once to protect the position and it isn't a trailing stop-loss as such. Most trades are closed well ahead of the SL being hit, and the tests below will show that its average win rate is around 80%.

Setting Wall Street Forex Robot up

This is where I initially got confused because the manual starts by saying that the robot works on three currency symbols (EURUSD, GBPUSD and USDJPY), so I chose to run tests on all three of those symbols as per the manual. Having run those tests, and whilst in the process of preparing this review, I then noticed towards the end of the manual that the EA also seemed to work on the USDCHF symbol as well. I cross-checked on the website where I also found reference to the USDCHF symbol, so I decided to also run some tests on this symbol and include their results within my review.

The manual consists of 18 pages, of which half relate to installation guidance. There are a lot of user-adjustable external parameters within the EA (which I personally feel would make it a little bit confusing for novices) but, for the vast majority of users, only a few of those parameters actually need to be adjusted. This is because the EA seems to have two modes - an automatic mode where it sets everything using its default internal settings and the user just needs to enter the risk in percent per trade - and a manual mode where the user can over-ride the default settings and specify his own SL, TP and lot size, etc. For my purposes, I was happy to use the automatic mode and let the EA use the vendor's settings.

In addition, on the EA parameters page of the manual, it states that the EA can actually be used with any currency symbol and the user can choose which of the default settings from another symbol he wishes to use. For instance, if the user wanted to run Wall Street Forex Robot on the EURGBP symbol, he could apply any one of the four default settings from a choice of GBPUSD, EURUSD, USDJPY and USDCHF.

As I mentioned above, the Wall Street Forex Robot also fetches its settings from the vendor's own server each time it is loaded onto its chart. This gives the vendor the opportunity to keep the EA tuned to current market conditions behind the scenes. The downside of this approach is that the user becomes reliant upon the vendor updating those settings at the appropriate times. Also, the EA only fetches the settings when it is first loaded, so I'm not entirely sure what would happen if the EA was just left to run on a VPS 24/7 and the vendor updated his settings. How would the user know that the settings had been updated? I'm presuming that the vendor would email his userbase to say that the settings had been updated and to suggest they reload the EA on the charts. It's a small point I know, but it needs to be considered nonetheless.

Given that the EA fetches the settings from a server, I wanted to ensure that they weren't just being curve fitted to current conditions, so the 10 year test was going to be more important than ever.

The Strategy Tests

Given that it's a scalper, I was careful to choose a broker with a realistic spread , so I chose Tadawul for the tests. Scalpers tend to prefer fractional pip (3/5-digit) brokers because their take-profit levels are small and easier reached using a fractional pip broker. Tadawul are a 4 digit broker, so I figured this broker should show the EA in as bad a light as possible for the purposes of getting some fair and representative test results. Tadawul's spreads are also fairly indicative of what the majority of users could expect to get throughout the times that the EA trades.

Given that the EA will work with several symbols, I chose to test with a risk of 2% as I didn't want the combined drawdown from running multiple symbols to be too excessive. In hindsight, I could have probably used a slightly higher risk level, but I was happy to err on the side of caution.

My 10 year backtest results are shown below. You view the full report in a new window by clicking any of the images.

Tadawul - EURUSD - 2.0 pip spread
Strategy Test

Tadawul - GBPUSD - 3.0 pip spread
Strategy Test

Tadawul - USDJPY - 2.0 pip spread
Strategy Test

Tadawul - USDCHF - 2.0 pip spread
Strategy Test

Despite working on other symbols, I elected to only test the four publicised symbols.

I did also test Wall Street Forex Robot with a couple of other brokers who use higher spreads than Tadawul. I haven't posted the results into this review, but I did notice that there was reasonable consistency in the same number of trades being taken between the different brokers. As would be expected, the net profit obtained from higher spread brokers was lower but, interestingly, the drawdown level and win rate didn't seem to particularly change as a result of using a higher spread brokers.


Despite being a scalper, the Wall Street Forex Robot is not a prolific trader. On each symbol, it only seems to trade around 150 trades a year. This isn't very frequent for a scalper by any means, and I would have expected to see it trade several times a day, implying that there is some pretty extensive filtering going on behind the scenes to only select trades with the highest upside potential.

If it were only being used on one symbol, I would therefore have to say that it is not particularly good value for money. The worst performing symbol clearly seems to be the USDJPY, where the net profit was a rather modest 84% over the full 10 year backtest. Seeing as users would find themselves risking a 7% drawdown in exchange for this return, I suspect any user running this symbol alone would quickly find themselves frustrated, especially as it looks that each symbol can endure lean spells of being in drawdown for a few months at a time.

Despite this negative aspect, Wall Street Forex Robot looks to really come into its own though through its ability to use multiple symbols concurrently.

Using multiple symbols can be either a very good thing or a very bad thing. If every symbol tends to drawdown at the same time, then there's obviously a heightened risk of blowing an account which would obviously be very bad. :(

If, however, each symbol tends to drawdown at a different time, using an EA with multiple symbols can be a very good thing as the positive effect from symbols which are outperforming more than offsets the negative effect of underperforming symbols, and the balance curve tends to be a lot smoother.

To find out how Wall Street Forex Robot would perform using multiple symbols, I used a 3rd-party script to merge all four of the above reports into one single report. The merged drawdown of around 11% is no worse than any of the individual drawdowns, which makes this a major positive. One of the side effects of merging reports in this manner when an EA is using its own money management, is that any compounding effects are discounted because the EA would have taken higher lot sizes than in each individual test. This means that the performance would most likely actually be better than the merged report suggests.

To demonstrate the point I'm trying to make, the USDJPY monthly return is 0.51%, the EURUSD return is 0.92%, GBPUSD is 1.35% and the USDCHF return is 0.81%. Combined, the total average monthly return is therefore around 3.59% for all four symbols. The annual compound return would be about 52% and, applied to the $50k start balance that I've been using for my tests, the 10 year net profit would be $3.39m. This is is considerably more than the $426k which is suggested in the merged report where compounding is discounted, and it makes the EA look very good value for money!

Tadawul - 4 symbols merged
Merged Test Results

I did note that the EA does features a recovery mode, which is presumably based on some form of Martingale. Individual tests showed that worst case consecutive losses varied between 4 and 7 (depending upon the symbol), and the average win rate is 80% which probably isn't high enough for Martingale. Besides which, the EA looks to be profitable in terms of pips, so I wouldn't personally be inclined to use the recovery mode.

The EA does include a parameter to permit slippage. Given the nature of scalpers having a low TP value, I would personally set the slippage to 0 (zero), as the slightest amount of slippage could seriously impact profits going forwards.

Wall Street Forex Robot has been well and thoughtfully coded. It only runs one open trade at a time, so NFA hedging and FIFO regulations are no problem which will suit US-based users of the EA. In addition, it also works automatically with both ECN and 3/5 digit brokers.

The only limitation I can see is that the SL and TP values are fixed. Markets change constantly and I would personally prefer to see a more dynamic stop-loss and take profit system which is adaptive to the market. I guess that's why the vendor is employing the server-side optimisation approach, as this will enable him to regularly adjust the SL and TP values to the current market.

It's difficult to know how effective this approach will prove to be in the long term but, even without any optimisation whatsoever, it's worth noting that the EA certainly looks profitable over a sustained period using its current settings.

In the UK, we have a saying, "If it ain't broke, don't fix it." Based upon what I've seen of it so far, Wall Street Forex may not need any fixing. :)

Forward Tests

As part of my testing, I've set the Wall Street Forex up on a $5k Alpari UK demo account trading all four suitable symbols on the 15 minute timeframe.

I'm using the default settings using the EA's internal money management system and I am only risking 2% of the account on each trade.

You can monitor Wall Street Forex's performance at MellyForex by clicking here.]]> Sat, 02 Apr 2011 16:17:10 +0100
Forex Shocker and Wall Street Robot - Expert Advisors in Forward Test Forex Shocker is an EA which many will be familiar with already as it's been around for a couple of years now since 2009. Reviews have tended to be mixed, so I wanted to see for myself how good it actually is.

It's a scalper which just takes a few pips at a time for a couple of hours each day in what I call "The Twilight Zone". That's the time after the US has closed for the day and before Asia has really got going.

At these times, trade can be very thin and liquidity can be a big issue, so one of my concerns is how Forex Shocker would perform with the wider spreads that you tend to see around these times.

So far, my testing has shown mixed results and I'm still undecided as to whether to continue and prepare a full review or not.

Nonetheless, I've put Forex Shocker into a forward test on a $5k Alpari UK demo account trading all 6 of its suitable currency symbols.

You can also monitor Forex Shocker's performance at MellyForex by clicking here.

The other EA which I've chosen to test is brand new and it's called Wall Street Forex Robot. It is also a scalper, but supposedly with a big difference.Wall Street Forex Robot Most scalping robots tend to trade during Asian sessions. There's an age-old theory that markets very rarely change direction during Asian sessions, and that direction changes nearly always occur during European or US sessions. Instead, markets tend to trade within an easily defined price range during Asian sessions, which is easy for scalping robots to track.

The difference with Wall Street Forex Robot is that it's happy to scalp at all times of the day.

Liquidity, therefore, shouldn't be an issue and spreads should be tight which is really a prerequisite for any scalping robot.

While I'm preparing the strategy tests for my review, I've set Wall Street Forex Robot up on a $5k Alpari UK demo account trading its three standard currency symbols - GBPUSD, EURUSD and USDJPY with a low 2.0% level of risk.

The EA trades on the 15 minute timeframe and trades all hours of the day.

You can also monitor the Wall Street Forex Robot's performance at MellyForex by clicking here.]]> Tue, 29 Mar 2011 06:23:21 +0100
PipRider - Expert Advisor Review PipRiderWhen I first started to look at the PipRider EA I hadn't realised that I was going to have to jump through hoops in order to test the EA and write this review. :o

The EA is heavily protected against piracy and, as a result, it runs very slowly in the Strategy Tester. Whereas a normal 10 year test may take anything between 1 and 3 hours, a 10 year PipRider strategy test takes anything upwards of a week.

Fortunately, the PipRider support is pretty responsive and, when I explained to them what I was trying to achieve, they were able to assist in speeding the process up considerably.

Of all the Expert Advisor websites I've visited, the PipRider website is one of the more comprehensive and informative sites. It consists of several pages and explains in reasonable detail how PipRider works. There is both forward and strategy test information, together with a page dedicated to a free indicator and another page detailing a partnership broker offer which enables prospective PipRider purchasers to get the robot for free.About the EA

Probably the main marketing approach behind PipRider is how it turns $100 into $1,000,000 in a 10 year strategy test and still works live today. I'm very sceptical of outlandish claims like this, and it was certainly something that I planned to look at in closer detail when I started my review.

PipRider trades the GBPUSD pair on the 1 hour timeframe and, to be honest, I'm not really sure how to best describe its trading style. It is certainly trend following, but it operates in a unique way by waiting patiently for a retracement in the price before entering the market in the same direction as the trend. Its profit target and stop-loss seem to vary from trade to trade and, although the TP targets are relatively small (in the region of 11 to 13 pips), they're large enough for PipRider to fall outside of the scalper category in my opinion. The other important point to note is that PipRider avoids trading the European and US morning sessions and seems to only trade onwards of 3pm UK time through to 7am the next day.

I asked the vendors about this, and their response was that PipRider avoids awkward price spikes and trend changes caused through news releases. Once the newsflow for the day is finished, markets return to a more orderly fashion and trend more gently.

Setting PipRider up

Of the EAs I've tested so far, PipRider is probably about the easiest to set up. Aside from coming with its own self-installer, the number of external parameters that PipRider offers its users to adjust is just three. And one of those three parameters is the unique licence key!

Although this certainly makes life simple and there is no possibility for any user-related errors to occur, there's also a part of me that worries about trusting money to what is effectively a black box.

The PipRider PDF instruction manual is 18 pages in length, although around half of that is general preamble, disclaimers and gubbins explaining how to attach PipRider to a chart. Once past that bit, I was able to find out that PipRider risks a fixed percentage of the account balance on each trade and that the lot size will vary according to the balance and the distance of the stop-loss from the order opening price. Encouragingly, the default PipRider risk is set at 3.5% and the maximum allowed risk is 10%. Clearly this is promoting safe trading which is something I'm in full agreement with.

The Strategy Tests

I mentioned earlier that, when I first tried to run a Strategy Test on PipRider, I found myself banging my head against a brick wall. :x

I asked questions and discovered that, when it is first run, PipRider gathers some information about your broker's server time offset in relation to GMT and writes this to a file. This information is then used to automatically detect when PipRider starts and finishes trading each day, but it also means that, unlike many EAs, it can be used in the Strategy Tester at weekends without the user needing to input an over-riding manual server offset. The problem is that, to collect the time offset information and write it to file in the first instance, the market needs to be open. Not realising this, I got up with the larks to set PipRider up one Saturday morning, only to have to wait nearly 40 hours for the markets to re-open before PipRider could complete its registration and be used in the tester. :cry:

The PipRider website explains in reasonable detail the effect that a broker's spread can have on an EA's performance, and suggests that PipRider users select a broker whose spread is no greater than 3 pips. I've been testing EAs with FXCM, and as I was able to fix their GBPUSD spread at 2.9 pips, I decided to continue with that theme. I used the default risk of 3.5% and set the maximum allowed trading lots to 100. Although in practice I doubt for one second that anybody would attempt trading 100 lots, I figured from my "back of an envelope" calculations that I'd need to set it this high to be able to get a truly indicative set of results in a 10 year strategy test.

FXCM - 2.9 pip spread
PipRider Strategy Test

So far, I was happy that PipRider was ticking all the right boxes and seemed capable of surviving 10 years. Its long term win rate was around the expected 80% and its worst case drawdown was also a respectable 17.5%.

The only disappointing thing was that there seemed to be several 6 month periods during the backtest when the equity wouldn't have made any progress. Personally, I'm not too bothered by spells of sideways movement, but I did want to look a bit closer at this, so I ran a second test over the last 18 months from mid-2009 through to the end of 2010 to get an idea of how PipRider had been performing more recently. For this test, I decided to switch broker to Tadawul to get an idea of different broker compatibility.

The Tadawul GBPUSD spread is 3.0 pips, but they offer different lot sizes and they're a 4 digit broker, so I figured this test would also portray PipRider in as realistic a light as possible.

Tadawul - last 18 months - 3.0 pip spread
PipRider Strategy Test

If you look closely at the equity curve, it's pretty easy to spot the 6 month or so lean spell towards the end of 2010 when PipRider's equity curve didn't progress, and this is when the 13% drawdown in the report occurred. In the 10 year test, I could see three or four similar periods.

If you ran PipRider live, you'd need to be prepared to accept that spells like this may happen at some stage. They're in exchange for a good, robust overall return with a low drawdown though, so it's important to sit with them, even if they might seem frustrating at the time.


PipRider looks to be well coded, it didn't report any errors in the logs and it seems to trade a solid strategy which is capable of withstanding the ravages of time. It shows on screen what its trading intentions are and at what price it plans to trade, and internally it is coded not to allow any slippage so I would expect to see the good backtest performance being replicated going forwards.

It declares a SL and TP with the broker, but generally uses some internal strategy to exit trades ahead of an SL being reached. Although the initial risk/reward ratio is typically a quite high level of about 9, the internal exit system reduces this to a more acceptable level of about 3. Even so, I would expect a full stop-loss to be hit on average about once every three weeks.

PipRider only enters one trade at a time, so there are no issues with NFA hedging or FIFO regulations. I noticed also that it is ECN compatible and can handle 5 digit brokers automatically.

It seems to trade around 350 times a year which is an average of about 7 trades a week. It doesn't seem to be too broker dependent, but I think it would be best to follow the vendor's recommendation and select a broker whose GBPUSD spread is 3 pips or less. There are plenty around, so it's not a major restriction by any means.

PipRider's biggest drawback, in my opinion, is that it could be prone at some stage to a 6 month spell on the sidelines where it merely breaks even. I'm quite patient and tolerant of these things, but I guess there are some out there who would become frustrated and remove the EA from its chart just before it recovers.

What I'm trying to say is that PipRider isn't a sexy EA, and it won't have you hanging on to the edge of your seat or biting your nails down to the bone. Instead, it's clearly been designed to last and it's built for comfort, not speed. :)

Forward Tests

As part of my testing, I've set PipRider up on a $5k Alpari UK demo account trading the GBPUSD symbol on the 1 hour timeframe. I'm using the default settings which only risk 3.5% of the account balance on any single trade.

You can easily monitor PipRider's performance at MellyForex by clicking here and there is a PipRider discussion thread on the MellyForex forum here.]]> Thu, 17 Mar 2011 12:58:00 +0000
MetaPRO Forex - Expert Advisor Review Having decided to review the MetaPRO Forex EA I realised instantly that running my normal 10 year backtest might be counter-productive in this case.

The MetaPRO website makes no secret that their EA is optimised for 2010, and I wasn't of a mind to believe anything else.

Incidentally, the MetaPRO website is quite informative and factual and consists of several linked pages rather than the typical one-page salesletter which seems to be the norm in the EA business. The website contains both backtests and forward tests, together with an explanation of what the EA actually does and what it's looking to achieve.

MetaPRO Forex is delivered in its own self-installer which makes for a nice and easy set up and it comes with its own 9 page PDF instruction manual which, in my opinion, isn't particularly clear. That said, I know enough about these things to be able to get it up and running to find out what it's really made of.About the EA

MetaPRO Forex works on just the EURUSD currency pair on the 1 hour timeframe and it claims to be a "swing trader". For anybody who's unaware, swing trading involves detecting turns in the market at support and resistance levels and either buying or selling the market as close as possible to the turning points. Swing traders tend to stay in the trade until the next turning point, at which time they tend to close out and reverse the position in the other direction.

You would therefore expect the trade pattern of a swing trader to be buy -> sell -> buy -> sell, and so on. It is generally rare to see several successive buy trades or several sell trades in succession.

Setting MetaPRO Forex up

Although there are a number of external parameters within MetaPRO Forex that the user can adjust, the EA comes with its own set files.

The main user inputs involve the risk/money management employed by the EA. It seems to use 3 methods for money management. Method 1 involves allocating a percentage capital to the EA and then using some mumbojumbo to calculate the lot sizes, method 2 involves trading with fixed lots and method 3 involves risking a user-determined percentage of the account on each trade which is based upon the distance of the stop-loss. If I were to use this EA live, method 3 would be my approach.

The user has a few other options to do with stop-loss and take-profit levels. There is also an overriding-parameter which allows the EA to set and adjust the SL and TP, and seeing as this is supposed to be a swing-trader, I opted to let the system calculate the SLs and TPs.

The Strategy Tests

Having mentioned previously that I suspected MetaPRO forex was heavily optimised for 2010, I began by running a couple of tests for just that year to see how it performed. The first test was with FXCM using a 3.0 pip spread, and I then repeated the test with Tadawul using a 2.0 pip spread.

Both of these tests were performed using the money management method 1 which I described above. This wouldn't be my preferred money management approach, but it was the method employed within the set file and I chose to run with it.

FXCM - 2010 - 3.0 pip spread
Strategy Test

Tadawul - 2010 - 2.0 pip spread
Strategy Test

Although the test results showed the returns to be pretty healthy, the drawdowns were a bit too high for my liking and I would expect to see them reduced if the risk was lowered.

Looking at the results, it wasn't difficult to see that there is a fine dividing line between success and failure with this EA. The rate of winning trades is not too far above 50% and the risk/reward ratio is very close to 1. You would only need to see a slight worsening of either of these two values for MetaPRO Forex to turn from being profitable into being loss-making.

Having already suggested that the EA was optimised for 2010, I decided to see what would have happened in previous years and ran a test from 2008 through 2010 using Tadawul.

Tadawul - 2008 through 2010 - 2.0 pip spread
Strategy Test

Although the 3 year results are obviously not good and confirm my initial thoughts that MetaPRO has been heavily optimised for 2010, they also highlight my previous assertion that there is a fine line between success and failure as far as this EA is concerned. The report shows the win rate dropping from 58% in just 2010 to a 3 year average of 48%, and this is the main reason the EA is unprofitable.


Despite being unprofitable over any sustained period, MetaPRO Forex has been profitable recently. Seeing as we are currently trading in the present and not 3 years ago or 10 years ago, I guess that fans of this EA will argue that current profitability is the most important thing, and offer a case for buying and using the EA in the short term while it still remains profitable. The fact remains though that the EA is unable to demonstrate any long term profitability and it is likely only a matter of time before it surrenders its recent profits and goes into decline.

I also can't help but think that with some better optimisation, MetaPRO could be made into a more robust product which would survive the test of time. It's internal coding is reasonably clean, it's not overly reliant upon broker spreads, it's ECN compliant and it is also 3/5 digit broker compatible. All of these things are positives, but they're offset by one major negative which is a lack of long term profitability. In addition to the profitability aspect, MetaPRO also doesn't appear to be NFA compliant. In test it appeared to open both long and short trades simultaneously, rendering it unsuitable for an army of US-based users who are restricted by non-hedging rules.

Unfortunately, the user is only really able to optimise TP and SL values himself. The so-called swing entries and exits are found using standard indicators and the EA does not include any option to optimise the indicator settings which I suspect would be necessary to get MetaPRO working profitably over any prolonged period. I'm not even convinced that the EA is a swing-trader in the true sense of the term - instead, I believe it simply follows a set of lagging indicators. Anybody purchasing the EA is therefore reliant upon the vendor continuing to provide timely updates with freshly optimised settings.

Forward Tests

As part of my testing, I've set MetaPRO Forex up on a $5k Alpari UK demo account trading the EURUSD symbol on the 1 hour timeframe. I'm using the H1 default settings with the exception that I am only risking 4% of the account on each trade.

You can monitor MetaPRO Forex's performance at MellyForex by clicking here.]]> Sat, 12 Mar 2011 13:10:00 +0000
Let the trend be your friend till the bend at the end In this article I'm going to attempt to explain a little bit about my personal trading style.

The thing about trading is that it isn't an exact science, and there is no such thing as a "right" way or a "wrong" way. Oh, if it were only that easy, we'd all be doing it. :wink:

Everybody seems to be looking for something different from their trades, and there are obviously many different approaches that you can employ.

Some people will look to scalp 3 or 4 pips at a time, while others will be happy to stay in their trades for days or more in the hope they can rack up gains of hundreds of pips. Some people will follow chart patterns, some people will follow indicators, some people will trade ranges and others will probably just flip a coin.

The best way of describing my personal approach is that I trade with the trend. My most useful tools are, believe it or not, a ruler and a pencil. That's right, the sort of thing which you can find in any corner shop for about 50p.There's a saying in trading that you should "trade what you see, not what you think". Simply "thinking" that a particular stock or instrument such as a forex pair is about to rise is no justification whatsoever to purchase that instrument. Just because something looks cheap at its current price, doesn't mean that it can't get cheaper.

Statistically, the chances are that, if the price is still falling, it will probably continue to fall further.

So how do you know whether something is rising or falling in price? Simply by looking at the chart, of course. If you take your 50p ruler and pencil, you should be able to draw a line through the recent highs and the recent lows on the chart. If the last low was higher than the low before and your line is going up, then this means that the trend is up.

Similarly, if the last high was lower than the high before and your line is going down, then the trend is down.

These lines are called trendlines and my personal trading style is to ONLY TRADE IN THE SAME DIRECTION AS THE TREND.

Here's an example of what I mean. I've drawn parallel trendlines on the chart to form what are called trend channels. You can click on the image below to view a larger version in its own window.

Trendlines Drawing trendlines and channels as I've done in my example makes it easy to identify a direction in which to trade.

Having identified the direction, it's easy to also manage any trades because stop-losses can be stationed just beyond the trend line. Also, if you know how far away your stop-loss is, it's easy to calculate the size of the trade and manage your risk. For instance, if you were entering a long trade close to a rising support line, you would take a larger position than if you were entering the trade in the middle part of the channel. This is because your stop-loss would be nearer to the trade entry price.

The other benefit of trading with the trend is that the trend provides strength if you get your entry wrong. If you enter a trade and it goes against you initially, there is a greater chance of the trade coming back in your favour, than if you are trading against the trend.

Finally, by producing parallel channels as I've done in my example, you've got an idea of where the price is going to halt and either consolidate for a while or retrace a bit. I generally stay out of the market completely while the price consolidates or retraces. It's a dangerous time to enter the market, so I use those times to go shopping or get my nails done. :)

As I said at the beginning of this article, trading isn't an exact science and it won't always be possible to clearly identify the trend. There will also be times when the price quickly breaks a trend and this method won't be right 100% of the time.

It's certainly a good start though, and trading with the trend will definitely stack the odds more in your favour than if you are trading against it. Trends can be strong, so why bother to fight them?

Good luck!]]> Fri, 11 Mar 2011 19:01:00 +0000
Forex Cash Protector - Expert Advisor Review Disappointingly, the Forex Cash Protector website doesn't tell an outlandish story in the way that some EA websites do and is generally more factual than most sites I've come across.

The perverse side of my nature does enjoy reading the success stories of former bin men who have discovered Forex and now own a Hollywood mansion and drive three Ferraris and a Bentley. :)

The website also has separate pages to show both forward and strategy test results which does make a refreshing change, even if the backtest does only cover a one year period so possibly isn't too indicative of longer term capabilities.

Forex Cash Protector is delivered in a compressed ZIP folder which contains the EA, a DLL and some preset files, so it is necessary for users to navigate to their MetaTrader installation folder and copy the files across manually.

There are also three separate PDF manuals which are sent with Forex Cash Protector - a 5 page manual explaining how to install the EA, a 3 page manual explaining the different EA user inputs and a 2 page manual explaining some irrelevant nonsense about a trading strategy using a moving average.About the EA

Forex Cash Protector is a trend following EA which is intended to work on the EURUSD currency pair. The manual states that it will work on both the 1 hour and 5 minute timeframes.

It uses a combination of several indicators for its trade signals and runs a stop-loss which is declared with the broker. Declaration of a take profit level is optional and the user can choose how to handle exits of the EAs open trades.

Setting Forex Cash Protector up

Most EAs nowadays come with their own self-installer, so it's a bit of pain having to navigate Windows in order to copy the files across into the correct folders.

The user inputs are quite straightforward and include 3 different options for risk management. Users can choose between fixed lot sizes, or they can risk a pre-determined percentage of the free margin on their account or they can choose a third method which is something to do with moon cycles and witchcraft. :)

Users can decide how they want to handle stop-loss and take profits and there is an interesting option which enables users to close a fraction of an open trade at a certain level and to then trail a stop-loss on the remainder of the position.

There are also 3 variables which control the settings for some of the indicators which are used. These are simply described as "smp", "fmp" and "rp" within the manual, with no explanation of what they might actually be doing.

There is also a default preset file included which can be applied to make Forex Cash Protector easy to set up and use. I'm not really sure why the default variables couldn't have been set at the levels of the preset, but what do I know?

The Strategy Tests

I began by testing Forex Cash Protector on the H1 timeframe using the default preset values. The 10 year strategy test showed it to be profitable which I considered to be a positive start. The balance curve is decidedly rocky though, and probably breaches most people's level of pain tolerance.

A return of 116% over 10 years equates to a compounded 8% per year for which you'd need to tolerate a 30% drawdown along the way.

Strategy Test

Encouraged by this reasonably positive start on the 1 hour timeframe, my next step was to test Forex Cash Protector on a 5 minute chart. Unfortunately, this was where it started to get a bit messy. I initially tested over just 2010 using the default set file and was encouraged even further still by a 37% return in just one year in exchange for a 37% drawdown along the way. When I then ran the same test over 10 years, Forex Cash Protector seemed to go straight into drawdown before crashing and burning after just a few months. No matter what I tried, it was impossible to produce any satisfactory long term results on the M5 timeframe.

I saw no real point in continuing with the 5 minute timeframe, so reverted back to the 1 hour timeframe which had showed initial success. Although profitable over 10 years, the equity curve was quite rocky and the drawdown wasn't too great as mentioned above. My FXCM test had been carried out using a 3.0 pip spread, so I decided to switch broker to Tadawul where the EURUSD spread is only 2.0 pips and run the test again.

With Tadawul, the end results were surprisingly quite similar even though the two equity curves have a different pattern. The 10 year net profit and relative drawdown were about the same, despite the win rate increasing from 70% to 84% with the tighter Tadawul spread. Forex Cash Protector was taking around 25% more trades with Tadawul than with FXCM, and these extra trades seem to have the effect of halving the size of the average winning trade somehow to impair performance.

Strategy Test

FXCP does contain various options to optimise performance. I did try some optimisation, but I wasn't able to improve on the performance of the default settings to any worthwhile degree.


It's pretty clear that any decent performance obtained on the M5 timeframe is going to be short lived. I suspect that the 5 minute settings have been heavily curve fitted to the current market conditions, so anybody choosing to run Forex Cash Protector on this timeframe would need to do a fair bit of regular optimisation work to keep the EA in tune. If you don't optimise it regularly, it's my guess that it will go into drawdown after a while and not recover.

Despite this drawback, FXCP certainly looks to be robust over time on the H1 timeframe where it is profitable. It's not a frequent trader on this timeframe (30 or 40 trades a year), and you'll need to accept that its drawdown could go up to around 30% in return for a compounded return of around 8% pa.

My personal feeling is that using profits and losses in terms of a fixed number of pips is not the way forward because markets constantly change and you need to adapt with those changes.

I'm also not a great fan of EAs that don't set a stop-loss which is what happens with FXCP using its defaults. There is an option to set a stop-loss, so I would strongly recommend that users experiment to find a stop-loss value that works.

I noticed that FXCP uses the same magic number for the 5 minute chart as on the 1 hour chart. I didn't investigate too thoroughly, but I would therefore suspect that users will encounter problems if they try to run FXCP on both timeframes simultaneously on the same account. I suspect that the M5 EA would end up interfering with trades taken by the H1 EA, and vice-versa.

The version I tested was sending its stop-loss with its order to open a new trade, and therefore wasn't ECN broker compliant. I also got a substantial number of OrderModify and OrderClose errors in the tester. These are minor bugs to fix, but some work is needed by the vendor to fix them nonetheless.

Forward Tests

As part of my testing, I've set Forex Cash Protector up on a $5k Alpari UK demo account trading the EURUSD symbol on the 1 hour timeframe. I'm using the H1 default settings as used in the strategy tester and discussed above. I've chosen not to run the EA on any other chart timeframe.

You can monitor Forex Cash Protector's performance at MellyForex by clicking here.]]> Tue, 08 Mar 2011 11:26:00 +0000
Martingale - Son of Satan or a Winner in the Trading Race? Many people will have heard of Martingale in the context of Forex trading, but I've often wondered how many people are actually aware of how it works and the dangers contained within.

There is an excellent article which explains Martingale on Wikipedia, but for anyone not wishing to study the mathematics in too great a detail, the concept is probably best explained simply by considering a flip of a coin.

There are two possible outcomes, heads or tails. Let's suppose you bet $1 that the coin will land heads up. If it does land heads up you win $1. If it lands tails up, then on the next flip you apply Martingale and bet $2. If it lands the right way, you win $2 but remember you just lost $1, so you really only win $1.

But what if it lands the wrong way? Then you simply bet $4 on the next flip. If it lands the right way you'll win $4 less the $3 you've already lost, leaving the net gain at $1. If it goes wrong, you'll be down $7 overall, so you'll bet $8 on the next flip to claw the loss back and still end up with a $1 profit.

At the end of a losing run, you'll always recoup your losses and win $1. How cool is that?

So what's the catch?Well actually there are several catches.

The first mistake that people make is believing that what has happened in the past will statistically influence the future. Just because you've called the coin flip wrong 10 times in a row, the odds of getting the next flip right are still exactly 50:50. They're no higher nor are they any lower because of past performance.

The second mistake that people make is that they don't realise quite how quickly the money gets eaten up when you double up after every losing trade. When you are betting $1 on the flip, it doesn't sound like a lot of money and it isn't. But after 5 losing flips, the stakes will have increased to $32. And they will increase to $64 after the sixth losing flip. And so on ....

Just ask yourself for one moment .... Would you ordinarily bet $64 on a single flip of a coin where the odds of success are evens? It's my guess that most people would answer no to that question because they're considering it in the cold light of day and can think about a rational answer. But, in the heat of the moment after a few losses, people tend to act irrationally and end up regretting it later.

Of course, it goes even further than that, because if you do bet $64 and call it wrong, the stakes will increase further still. Eventually YOU WILL RUN OUT OF MONEY AND WON'T BE ABLE TO AFFORD TO TAKE THE NEXT BET.

When you apply Martingale to Forex the same situation applies, with the major difference that people don't appreciate quite how the leverage on their account makes them risk far more than they think. Whereas it's easy to withstand several successive losing coin flips, it's a lot more difficult to withstand the same number of successive losing Forex trades where Martingale is at work. Ironically, the players who tend to rely most on Martingale tend to be those players with the smallest accounts and the maximum leverage who can least afford the extra risks.

Martingale is clearly a game which is best suited to players with very deep pockets. I'm not saying that people shouldn't use Martingale, but I do believe it's very important to run proper tests beforehand to discover exactly what the worst case scenario really is likely to be. Hopefully some of the testing work that I publish to this site will be of use to all the "Martingalers" out there. :)]]> Sun, 06 Mar 2011 15:05:00 +0000
Forex Hippo - Expert Advisor Review Forex Hippo website consists of yet another long sales letter page. 8O

Forex HippoIt tells the story of Matt Robinson who was fed up with his job, fed up with everyone at work, fed up with losing money and fed up with fraudsters. In fact the poor chap seemed so fed up with his lot in life that I'm quite surprised he didn't top himself. :)

Instead though, Matt decided to quit the rat race and start selling his Forex Hippo robot which he describes as "The Real Deal". Since becoming Hippo-tised our hero has become a full-time forex trader and now "earns as much money as a doctor". Wow!

Forex Hippo allegedly gulps up profits, so I sincerely hope that it doesn't gulp up any of my money by mistake! Let's try to find out.....

Forex Hippo comes with its own self-installer. The PDF instruction manual is easy to understand. The manual is 23 pages in length, but only its last few pages are of any real relevance, as most of the manual explains how to install MT4 and the Hippo EA.About the EA

Forex Hippo works on eight different currency symbols, EUR/USD, EUR/CHF, EUR/GBP, EUR/JPY, USD/CHF, USD/JPY, GBP/USD and GBP/CHF. Forex Hippo is attached to 1-hour charts and seems to be a trend following system which acts on a combination of several different indicators to enter its trades.

The user is required to input a number of external parameters to control things like stop-loss and take-profit levels, as well as being able to decide whether to close positions out before the market closes on a Friday, or whether to even bother trading on a Friday at all. I guess now that Mr Robinson is earning as much money as a doctor, he can probably afford to take Fridays off. :)

Setting Forex Hippo up

Forex Hippo users have the choice of trading either fixed lots or variable lots which risk a specified percentage of the free margin on the account. The default setting is to risk a whopping 15% on each trade - something that I'm not happy with, especially if all 8 symbols are being traded simultaneously. Risking this much could theoretically produce astronomical returns, but this sort of risk could also blow the account right out of the water if the stop-loss on several symbols were hit at the same time, so I don't intend to even go there. I was taught to never risk more than 4% on a single trade, so that's the setting I intend to use in my tests.

There are several other settings that the user can adjust including a couple of filters for trade entries and exits and an option to use "Stealth Mode". The Stealth Mode sets a dummy protective stop-loss and take-profit level on the broker's server and the EA will close out of trades automatically once the true SL or TP is reached.

Despite offering all of the different parameters to its users, the Hippo manual then advises that all of the settings be left at their default value with the exception of 3 or 4 values which should be altered for the different symbols that Hippo can be used with. If something ain't broke, I don't believe in trying to fix it so, aside from lowering the risk level, I dutifully followed the instructions and left everything alone unless otherwise stated.

The Strategy Tests

Following the recommended settings within the manual, I set up 10 year backtests on some of the suitable currency symbols. I didn't encounter any problems in the tester, so I've published my results below as I suspect they'll speak for themselves. I also noted the spread which was used for each test.

The win rates and SL and TP levels vary widely from one symbol to the next, so you'll need to look at each test report in closer detail for a better understanding.

GBPUSD - 2.9 pips spread
Strategy Test

EURUSD - 2.8 pips spread
Strategy Test

USDJPY - 2.6 pips spread
Strategy Test

EURGBP - 2.6 pips spread
Strategy Test

Although Forex Hippo can be used with 8 currency symbols, I decided to give up testing after just 4 symbols as the results weren't exactly blowing me away.


Clearly Hippo's strategy test results show that some symbols work better than others and that, over the course of 10 years, some years will be good and some years will be bad. I suspect the biggest limitation is that Hippo is inflexible in its use of a rigid SL and TP. The fact is that market conditions are constantly changing, so it's no real surprise that the best SL and TP levels one year won't necessarily apply the following year.

With this in mind, I suspect that Hippo could be better optimised for the last 6 or 12 months and it will probably continue to perform if it is regularly re-optimised.

As an exercise, I decided to use a script to merge the four test reports into a single report. The merged report isn't entirely accurate because it discounts the compound money management effects of running multiple symbols concurrently. Strangely enough, the report does show an overall profit for the 10 years.

With a bit more time and effort than I'm prepared to put in, I suspect that Hippo could be improved upon and I would describe it as "work in progress". Anybody who chooses to purchase Forex Hippo could conceivably get some good returns through spending a bit of time and effort in selecting the best settings.

Merged Strategy Test

Anybody using Hippo needs to bear in mind that each currency symbol will only take around 70 or 80 trades a year and it isn't what I'd describe as a prolific trader. As Hippo only has one open order at a time, NFA hedging and FIFO compliance isn't a problem.

The one thing I did notice is that the spread is deducted from the TP and SL levels when using the stealth mode. In other words, if you set a TP of 30 pips and the spread is 2 pips, then the trade will close when the trade is 28 pips in profit. I ran a quick comparison test with the stealth mode switched off and discovered that the spread wasn't being deducted, so I would suggest that there's a bug in the code. It's therefore also possible that my test results would have improved if I had ran the tests with stealth mode switched off.

Forward Tests

As part of my testing, I've set Forex Hippo up on a $5k Alpari UK demo account trading all eight suitable currency symbols variable lots and risking 4% of the account on each trade. The exception is the EUR/JPY cross where I'm only risking 2% as per the guidelines in the Hippo instruction manual.

The other EA settings are all as per the guidelines within the manual.

You can monitor Forex Hippo's performance at MellyForex by clicking here.]]> Sat, 05 Mar 2011 13:12:00 +0000
Forex Crescendo - Melly's First EA Review Forex Crescendo.

Forex CrescendoThe Crescendo website seems to consist of a typical long sales letter suggesting its audience should listen carefully because the vendor is about to give away his best kept trading secret that made a fantastic wedge of dough in under 8 months. Personally, I don't get anything whatsoever from that sort of spiel, but I guess other people must do otherwise vendors wouldn't bother to use that approach.

The Crescendo EA itself comes with its own self-installer which installs the necessary files into the MetaTrader installation folder(s) that you choose. The PDF manual is comprehensive and consists of no fewer than 37 pages, although about 1/3 of those pages are just pre-amble explaining how to use the self-installer which I did think was a bit overkill.About the EA

Forex Crescendo works on two currency symbols, the GBPUSD pair and the GBPJPY cross, using what is commonly called a "Grid" strategy. Grid strategies will enter their initial trade with a profit target in mind. If they hit their profit target then great, they will bank the profit and start over again. If they don't hit the profit target, and the trade moves out of the money by a certain amount, a new trade will be opened. The idea is that the price now needs to move to the average value of the two trades to break even, so a new profit target can be set which is marginally better than break even.

If the price moves even further out of the money, a third trade will be opened which alters the break-even level further still. And so on... You could theoretically have umpteen trades open for days, weeks, or even months while you wait for the price to retrace to the average profit target so that you can exit for a net overall profit.

The issue I have with this type of strategy is that the process of averaging down is frowned upon by professional traders. The assumption is that the price will retrace but, in practice, there will be times when the price doesn't retrace and the trader won't be able to keep margining new positions while he hopes for the best.

I was taught to get out of my losing trades before the losses got worse - not to risk increasing the losses further still. Anybody who's ever heard of Nick Leeson wil be aware that he broke Barings Bank by averaging down his losing positions (OK, he also margined his new positions by exceeding his limits without his employer's authority) in the hope that they would come good.

Putting those prejudices aside, I want to see whether this type of system, IF USED CAREFULLY, does actually have a role to play.

Setting Crescendo up

My first observation here is that setting Crescendo up can require far more user interaction than is probably necessary. The list of external parameters that the user is able to enter is quite long and, I couldn't help but think, is a bit pointless. I can only think that users must like being able to alter settings because it gives them a sense of being in control.

Personally, I believe in keeping things as simple as possible because too much user interference normally ends in tears. Also, I'm presuming that the Crescendo vendors must know what their optimum settings are, so why not just hard-code the best settings into the EA and remove some of those external parameters completely?

The fact that there are so many external parameters that the user is able to modify rings alarm bells, quite frankly. Rightly or wrongly, it sends a message that the EA has problems and the user is being invited to test an unfinished product in the hope that he might find something that works.

The Strategy Tests

It's got to be said that my first attempt at testing Crescendo was a disaster, although that was of my own making. :oops:

Wherever possible, I like to use default or recommended settings to test an EA. In the case of Crescendo the default risk setting uses fixed lots to trade. Because of this, I would expect to see a straight(ish) line equity curve rather than an exponentially rising equity curve which is typically produced by EAs which increase the trade size as the account balance grows.

Crescendo also includes an option to use different forms of Martingale (more about this later), but it doesn't use this by default so I left it switched off.

I first set Crescendo up on a $5,000 FXCM UK account where the minimum lot size was 0.1 on the GBP/USD currency pair. The profit target was default $4 and $1,000 of the account was allocated to Crescendo with a stop-out set at 30%. In other words, as soon as the open drawdown hit $300, all trades would close to crystalise the loss before it got any worse. Trading 0.1 lots with a $4 profit target meant that trades would close once they were just 4 pips into profit, so I expected the win rate to be high.

An important thing to note about systems like Crescendo is that, because of their tendency to hold multiple open losing positions, they don't use stop-losses at all. Instead their stop-loss system relies upon them closing out once the aggregated open loss exceeds the maximum drawdown level that I referred to above. The lack of a declared stop-loss means that Crescendo needs to be kept running 24 hours a day while it has open trades so that it can manage them. It also needs to be running on a system which isn't susceptible to power or IP outages, otherwise there could be a problem with Crescendo being unable to exit its trades.

Strategy Test

In the the 10 year test from 2001 through 2010, I got as far as 6 months and 136 trades before deciding to abort the test. Although the win rate was 90%, there were 4 separate occasions when the 30% stop out was hit and the stop outs took away all of the gains and then some.

It was pretty clear that I was doing something wrong, so I decided to restart the test having multiplied everything by a factor of 10 in compensation for the minimum lot size being 0.1 lot. In other words, I increased the account size to $50k, set the profit target to $40 (ie. 40 pips) and allocated $10,000 to the Crescendo EA.

Ordinarily, I probably wouldn't even have mentioned my mistake, but it occurred to me that, if I could make this mistake (and I'm familiar with EAs), then other people could easily make the same mistake too. Seeing as the consequences could be very costly, I therefore decided to mention it in my review. So, if you're one of the many people out there whose broker doesn't permit micro (0.01) lots, please take note of the above and adjust your target profit accordingly!

Having run the test again, the first thing to note was that the larger profit target results in Crescendo taking far fewer trades because it stays in its trades for much longer. In addition, the higher profit target in pips meant that the win rate dropped significantly to around 70%. The reduced number of trades aside, Crescendo still was not profitable over a 10 year backtest. Sure, it had its moments when the equity curve would progress upwards in a relatively straight line, but there would always be an accident waiting to happen just around the corner. You can see what I mean in the curve below.

Strategy Test

I mentioned previously that Crescendo also includes options to use Martingale. The effect of Martingale is to bring the target profit closer towards the later trades, thereby increasing the chances of hitting the profit sooner (or even of hitting the target at all). Personally, I'm not a fan of Martingale type systems but I ran a test nonetheless using Crescendo's "Linear Martingale" setting. This has the effect of increasing lot sizes in the sequence 0.01, 0.02, 0.03, 0.04, and so on.

You can see the effect of using Martingale below. The equity curve pretty much speaks for itself and, in addition, there seems to be some sort of bug in the code because it reached a point where Crescendo was opening a trade and then immediately close it for a loss, only to then immediately re-open another trade and do the same over again.

Strategy Test

When this started to happen I chose to stop the test because I'd seen enough to draw a conclusion.


Crescendo is always going to be susceptible to taking large drawdowns when it gets caught on the wrong side of a trend and the tests show that it will be loss making over any sustained period of time. There are nonetheless periods which can last months on end where it will make good progress before drawing down, and it could theoretically be used with a high level of risk to grow a relatively small account by a significant amount in a short space of time. This is pure gambling, however, and it's not of interest to me.

I did also note that, whilst Crescendo is NFA hedging compliant (it won't try to open trades in the opposite direction to any existing open trades), it isn't FIFO compliant. In other words, it closes the most recent trades first when it has a basket of trades open. It should be a straightforward task for the vendor to modify the code to make it compliant. The user is also invited to enter his own slippage value, so I'm sceptical that Crescendo would deliver the same performance live as in demo. If anyone plans to use it live I think that I would suggest setting the slippage to 0 (zero).

My only other observation is that, because the profit target is a monetary value, Crescendo will probably work best with 3/5-digit brokers because the profit target can be reached quicker.

I didn't bother to look at testing Crescendo on the GBP/JPY cross, mainly because I didn't expect it to show anything new and I didn't want to waste my time.

Forward Tests

As part of my testing, I've set Forex Crescendo up on a $5k Alpari UK demo account trading fixed 0.01 lots with $1,000 allocated on both the GBPUSD and GBPJPY symbols. I've set the maximum drawdown limit at 25% and the order step at 50 on the GBPUSD symbol, although all of the other settings are their default value.

On the GBPJPY symbol I've left everything at default.

You can monitor Forex Crescendo's performance at MellyForex by clicking here.]]> Tue, 01 Mar 2011 16:14:00 +0000
Why do all Expert Advisors fail? Good question but, for what it's worth, I don't believe that ALL expert advisors do actually fail.

Sure, some commercial robots are just out and out scams. For instance, an unscrupulous vendor might realise that his EA would have blown up an account on a certain day, so he will hardcode any bad dates into the EA to prevent them from taking trades in the strategy tester on those dates.

Unsuspecting buyers will see an equity curve produced by the tester showing a steady rise because it won't have taken any of the fatal trades. Anybody caught by a scam like this risks not only losing the purchase price of the EA, but also the contents of their account when the EA inevitably blows it up at some stage in the future!

In my view, this is just downright deception and such products rightfully need to be exposed for what they are!Fortunately such instances tend to be few and far between, although there becomes a tendency to tar every EA vendor with the same brush. This is unfair because there are some good guys out there.

The worst enemy facing EA purchasers in my opinion though, happens to be the purchaser himself. I believe that most EA purchasers tend to have too high a level of expectation and wrongly think that they'll be able to make their million from an EA within a few weeks.

Life isn't like that though. Most novice traders don't understand that losses can happen regularly. They don't understand that losing spells may last for several weeks or months on end. And, above all, they don't fully appreciate the emotional turmoil that people face when losses inevitably happen.

Realistically, a professional trader should be looking to double his money each year. To do this, he will invest wisely and he will not risk too much. His primary aim will be to ensure that he still has sufficient funds to return to the market. Let's face it, most banks are paying savers peanuts for interest at the moment, so anybody who doubles up in a year will have easily beaten The Street.

The same should be said of expert advisors and this is probably because they tend to be marketed towards inexperienced traders. Most are deemed to be failures because the user has an unreasonable level of expectation. Users will frequently trade them with too high a level of risk, and won't do the necessary tests before putting them on a live account.

Tests are very important because they show whether the robot is likely to endure lean spells. They will show the true consequences of a few losing trades in succession. They will also show the right level of risk to apply when setting the robot up. They will even show whether the robot is suited to a particular broker.

Forewarned is forearmed, as they say, and if all robots are tested properly from the outset, it should be seen that many don't really fail at all. All that is often needed to make them work profitably over a sustained period is a little patience.]]> Mon, 28 Feb 2011 10:41:00 +0000
Welcome to Melly's Forex Reviews and Articles Page As this is my first article, I thought it probably best if I spent a few moments explaining what I'm trying to achieve here and what the review section is all about.

First and foremost, I'm trying to find out whether the robots that I choose to review are any good. By that, I obviously want to know if they're profitable, but I also want to know if they're going to be reliable long term and I want to know if they're safe. It's all very well finding a robot which doubles your account balance inside of a week but, if it blows up the account in the second week, it's not a lot of use.

So, as part of each of my reviews, I intend to cut through the vendor marketing gobbledegook and discover the truth.To do this, the starting point is to run a 10 year backtest. If an EA fails the backtest, then it won't have a hope of surviving for any real time live.

Wherever possible, I'll use either the default settings or settings recommended within the manual. I don't have the time to optimise EAs trying to find the best settings - this is something the vendor should have done before putting his product on the market!

I want to find out how the robot handles things like slippage, 5th digit pricing and ECN/STP brokers. I need to know how it handles money management and controls and limits the risk it takes. I was taught to understand that disciplined money management was the key to trading success, and I don't intend to throw those beliefs away. I'd also like to get an idea if an EA might be broker dependent.

Finally, I'll put the robot onto a $5,000 non-expiring demo account and just allow it to run 24/7 without any interference. I'd never contemplate putting a robot onto a live account without first seeing how it performs on a demo account for a while. I'll also use the wonders of modern technology to publish the forward tests results to MellyForex for all to see.

I can assure my audience that my reviews will be honest and forthright and I'll be sure to mention any negative points that I notice as I try to navigate through the hyped-up marketing speak. I might make money through an affiliate link, but I'd never knowingly mislead a reader by publishing a dishonest review just so that I could cream the commission.

In addition, I'll try to keep the reviews as subjective as possible in the hope that some of you find them useful and, ultimately, benefit financially.

Enjoy!]]> Fri, 25 Feb 2011 18:08:00 +0000