Saturday, July 17, 2010

Forex Robot World Cup - results

It has been a long time since I blogged about the results of my experiment with Forex Robots, and I sort of left you hanging. The reason I stopped recording my results in real time was that I was about to say some pretty negative things about the sponsors of the Forex Robot World Cup, and they had a thousand dollars of my money that I wanted refunded.  They had guaranteed it, but I had no certainty that they would honor that guarantee in the face of negative publicity I might generate.  But they did.

Okay, here's what I found. The "Fusion" robot that they offered does not come anywhere close to living up to the gains that they implied that it would.  In the roughly three weeks that I was actively trading it, I lost about 40% of my original, real live, cash investment ($400 went to $240) - something they loudly proclaimed in their enticing ad copy wouldn't happen. But, when I fulfilled the terms of their guarantee, they did, indeed, refund my $999 purchase price without a hassle.

I started this series with the question "Forex Robots: Scam or Opportunity?".  So, was it a scam?  I think that all depends on your definition of "scam".  Did they deliver what they promised (or strongly implied)? No.  But did they take my money and run, or cheat me out of something to their gain? No.

If they had delivered a robot that was just trash and then disappeared or otherwise refused to honor their guarantee, that would be a scam.  Or if they had delivered the robot and used it to install a virus that could get access to my personal information (something I feared when I saw the virus warnings), that would be a scam.  Or if they used it to promote a Forex broker with whom I invested and then he ran off with my money, that would be a scam.  They didn't do any of these things.

They DID way over-promise and under-deliver, and after seeing the details and analyzing their strategy, I believe they did it deliberately.  I don't think they lied about the results they achieved. But I don't think they have told the whole story.  I think they manipulated the results to achieve one purpose: being able to quote a return number so astronomical (356% return in 19 days) that it would be compelling. And they did this by trading the robots with a risk factor that no sane trader using his own real money would ever do.

The Fusion robot was actually made up of 5 robots that performed well in their 2 month competition. (Note - successful traders say that 2 months is far too short a time to adequately judge any track record - robot or manual). The one that traded most frequently and with the best success was called High-Rider.  It's strategy appeared to be fairly simple - take small gains, referred to as scalping, with very few losses. 

One of the things people look at in evaluating a robot is its won/lost percentage - how many of its trades end up as profitable, vs. how many lose money.

It turns out that it is actually fairly simple to have an astoundingly high won-lost percentage.  All you have to do is set the stop-loss point so high that it is almost certain not to be hit. (The stop-loss point is the price point that you select when entering a trade such that, if the trade goes against you and hits that price point, you will "stop your losses" and get out.)

Trading the forex market is different than trading the stock market.  In the stock market, if you're buying individual stocks, you're buying ownership in a company whose value can go up or down by large amounts, depending on their success in the business - and can keep going.  In forex, you're trading one form of a commodity - money - for another form.  And there will be small differences and changes in the relative value of the currencies you're trading, but in the end, it's all money.  And so you're making money on the very, very small changes - changes in the 4th decimal place of the relative valuation, equivalent to 1/100th of a cent.

The result of this is that prices of one currency in terms of another tend to move up and down in relatively small ranges.  Even when there is a relatively large change in valuation, it occurs somewhat gradually, over longer periods of time.

So, if I enter a trade believing that the relative price of a currency is going to go up, and I'm willing to take a small gain,  it's very likely that the price will move to give me that small gain before it moves against me in a big way - even if it moves somewhat against me at first.  So if I'm willing to exit the trade on a gain of 12 pips (which High-Rider was using), and I set a stop-loss point of 200 pips, it's much more likely that the price will move up 12 pips before it moves down 200 pips, even if it might initially move down by 50 or 100 pips before it does. [A pip is the smallest price change that  will be quoted - usually the 4th decimal point of the price.)

And this is what High-Rider was doing.  It had a VERY high success rate during the competition. But here's the downside - if it DOES have a loss (and, eventually, every trading system will have a loss), the loss will be huge.

The promoters advised that you set your trading size  for High-Rider to one-tenth of a mini-lot (one-one hundredth of  a full lot) for each thousand dollars of account valuation.  At that point,  if the trade went against you, you would have lost 200 pips times $.10/pip, or $20, out of a $1000 account.  That's two percent, which is what I have read is the best standard for limiting your risk on each trade.  So that sounds reasonable.

EXCEPT that, as part of its strategy, when High-Rider found a trade, it would trade it up to 10 times, usually within a few minutes of each other.  And if the price moves against it, ALL of those trades are going to go down.  Which means that now your 2% risk is actually a 20% risk. In fact, that was what happened in my live account - a trade did go against it, and it wiped out $80 of my original $400 account in one day.

Now, all of that explanation was to get this:  When the promoters built their 356% return in 19 days, they were trading High-Rider with not one tenth of a mini-lot, but a full mini-lot.  That means each trade was risking 20% of the account balance, and the ten trades that High-Rider entered were risking 200% of the account balance! If the very first set of trades that was entered had gone against it (something that MIGHT happen), the account would have been  wiped  out.  That is why I say that no sane trader using his own real money would have traded the robot that way. In fact, in their case, it worked out, and, if I remember right,  the account gained in excess of 10% the very first day.

But they weren't using "real" money. Yes, it was a live account (supposedly) with real money, but if that trade had gone against them, they would have simply started over with a new investment of $1000, until they got an account record that they could advertise, and never tell you about the attempts that they had made that had lost money. In fact, it would surprise me if, behind the scenes, there weren't multiple attempts to get a good track record, and they only used the best one in the advertising copy.

So, was it a scam? No, I still wouldn't say it was a scam.  Was the advertising misleading? Definitely.  Was it false to the point of being fraudulent? Probably not.  A lot of the conclusions they wanted  you to come to were implied rather than stated.

Just because this robot did not represent a real opportunity doesn't mean that none of them do.  Tomorrow (I hope!) I'll tell you about one that I have been using that, while  not living up to its advertising copy, nevertheless has given me some very solid gains in a real-money account.  Stay tuned!

No comments: