I noticed early this week that my FRWC robots were “frozen”. Rather, I should say the MT4 platform on which I was running them was frozen. All of them were showing “Invalid account” when trying to log in. So I contacted FXCM, which I was using to try to replicate the environment that the FRWC used for testing, and they informed me that their demo accounts are only valid for 30 days. Say what?!? Needless to say, I am a little bummed about that because to keep testing I will have to open new accounts now. I wanted to have some “history” to show for each of these robots, but I guess that can’t be done at FXCM. So I think I will open demo accounts at IamFX. They’ve got a low spread, and they aren’t handcuffed by the NFA nonsense, so all the bots should work on their platform.

Looking at the 30 days I was able to test, my pre-testing prediction held true. HiRider was the only robot of the bunch to show any real promise and would be the only one I would consider putting real money on at this point (though I still recommend more testing). It ended up the month with a bit more than 12% in profit. No, it’s nothing like the 40%+ that FRWC showed in their contest (where they were using unreasonable lot sizes), but that is nothing to sneeze at by any means. If you compound 10-12% per month, you will be rich sooner rather than later.

The other FRWC robots didn’t fare as well, ranging from 6-18% in losses for the month. The good news is that the FRWC is continuing to tweak the settings of the robots so that they are optimized for current trading conditions. They have released new settings for some of the robots and a completely new version of HiRider, which is what I will be using with the new demos.

The other robot I have been testing (Pip Android) has regressed, losing all of the profit it had gained. It is now just under break-even after a month and a half of trading.

Fusion V — Struggling with a series of GBP/JPY trades that it opened last week, this bot is currently carrying about a $200 drawdown and will need for GBP/JPY to sell off a bit to be able to get out of these trades with success. The equity on the account (including the current drawdown) is $950, so the account is down about 5% at the moment. Hopefully it can bounce back this week.

HiRider — As predicted before testing ever began, this EA has been the best of the bunch so far. There is only one trade open at the moment, and the account is sitting at a gain of around 8% so far. As mentioned previously, I have lowered the risk as much as possible, trading a lot size of 0.01 per $1,000. So far, so good.

LMD Multicurrency — This one has been a mixed bag. It wins some and then loses some. Thus far it is down about 4.5% from when we began the demo. Amazing that it could do so well in the competition and then be such a bust afterward. Just shows that two months isn’t enough time to demo a robot.

Super Volcano – Nothing much to write home about with this robot either. It is stuck in one losing trade right now, and the account is down about 11% overall.

I still think HiRider and Fusion (in that order) are the best of the bunch. And as I said early on, even if only one of these robots gets a consistent return month after month of between 10-25%, it will have been well worth the purchase. HiRider is getting close to that 10% mark, and Fusion has quite a lot of catch up to do.

For anyone that might have missed it, I should mention that the FRWC did release updates to Fusion and LMD late last week. I am running the latest version on my demo accounts.

And in case anyone is following the Pip Android demo, it has been profitable, but not exactly overwhelming. It’s winning the vast majority of its trades, but it just doesn’t take that many. The account is currently up about 3%.

The LMD-Multicurrency robot is aptly named, as it made trades on eight different currency pairs during the competition. They included: USD/JPY, USD/CHF, GBP/USD, AUDUSD, EUR/GBP, EUR/CHF, EUR/USD, and USD/CAD.

In all, the robot took 96 trades over the two-month period, for an average of 2.2 trades per day, so it is a fairly active EA. When trade conditions were met for a certain currency pair, it would appear from the trade log that a position was opened at midnight GMT. And if positions did not hit the stop loss or take-profit levels, however those are determined, trades were automatically closed out at midnight GMT of the next day (or on Monday if the trade was opened on Friday). This means each trade was held for a maximum of 24 hours.

The EA does not appear to utilize a predetermined, pip-based SL or TP, as the loss and profit levels varied quite a bit. Money management is built-in, as the lot sizes were gradually scaled up as the balance on the account grew.

Breaking down the trades, the EA had 50% winners with 48 wins and 48 losses. The average win was $156.00, while the average loss was $125.67. The largest win was $544.44, and the largest loss was $311.64. Maximum drawdown at any one time was a little over 34%, which means you’ll need a pretty strong stomach to weather the volatility that accompanies this EA. But, at the end of the day, the profits are incredible, and that’s what really matters.

This and the other top performing robots from The Forex Robot World Cup will be available soon (February 16). We’re looking forward to it! Click here for more info.

Here’s the usual routine. A Forex robot vendor creates a glitzy marketing page that explains how his or her robot gained xxx% in the past month or year or whatever. Amazingly, this vehicle to riches only costs $97 or $149 or something ridiculously cheap for such a sure thing. So I buy a copy to test, and amazingly it backtests very well for the most part. A live demo test also shows promise, and I really begin to get my hopes up. So I then move the EA to a live real account to begin the journey to financial independence. Only natural, right? That’s when for some odd reason, it just stops working as well. Either it doesn’t take as many trades as it should or it has much larger or more frequent losses than ever before. The end result is that it either loses money or simply isn’t worth the hassle for the little that it earns. This scenario has played our for me more often than I care to tell you. So, you can hopefully understand my discouragement. That said, it was good for me to take a little break, hit the reset button, and reevaluate where we are.

I have not given up hope that it is possible for an EA to be very profitable in the Forex market. Deep down I still think that if a human can do it, a robot can also be programmed to do it. I am not going to revisit any of the EAs that I had been reviewing (EuroBlaster, Forex Derivative, Forex Grid Bot, and Robominer), as I don’t believe they are the “answer” for various reasons. But I was encouraged a few weeks ago to learn about the Forex Robot World Cup, which was a competition held from December to January to find the world’s best EA. There were some very promising results, so I will be focusing my efforts for the foreseeable future on the robots that did well in that competition. Stay tuned!