Rashid, you suggested to introduce interactive reporting for the forthcoming Championship. Why you suppose this analysis to be necessary?
It’s a good question. The last-year Championship was “the first bird”, nobody ever saw a beat of it. So I understand how challenging it was to organize the first Automated Trading Championship in the world. Of course, all effort was hurled into creation of the new resource, supporting and purchasing of equipment. I do understand this now, when I’m working in the company.

Everything was new, everything was interesting. However, it was natural that some things were missing. For me personally, at least, they were. For example, I was not mostly interested in profits obtained on an account, let me say so. I wondered much more how that profit was obtained, what its nature was. This is why the first parameter I was looking at last year was the amount of trades made. I mean that profit obtained as a result of three or four trades does not inform us about anything. I primarily considered accounts, at which at least ten trades had been made. This new reporting system was interesting for myself, so I suggested to use it in the new Championship.

For example, if we take those current developments, the Sharpe Ratio characterizes the risk level of a strategy. Z-score helps to detect whether there is a relationship between trades. Parameters MAE and MFE show possible drawdowns per trade, the profits traders fall short of.

I mean that all these parameters are very important, in my opinion, and one is part and parcel of another one. Besides, it was suggested to consider correlation between these parameters. Correlation, as a measure of relation between two sequences of random values, is the fundamental notion to be known by every trader, I think. Using the example of calculating a trend line by the least square method, a view of deviation minimization (estimate of balance smoothness) was given, as well. So I hope that this will be interesting and useful for not only me, but also all others.

The most of traders will benefit from this. This will allow them to re-interpret both their own and others’ accounts when developing a trading system. Traders will be able to emphasize some new things, maybe those omitted before. I hope that the general level of developed Expert Advisors will become higher. This is the main purpose to introduce this interactive reporting.

What were the results of the analysis of Expert Advisors that had participated in the previous Championship?
Well, another “unchildish” question. The main conclusion I drew is that money management plays a significant role in obtaining the final result (the final profit). I must say that I analyzed all accounts of the last-year Championship. For this purpose, I wrote a script in MQL4, then I connected to each account using investor’s password and launched my script.

It processed about 80 parameters (it means that I was not sure which ones may be helpful and included many), calculated them for about a second, and then saved in csv files. After I had taken all the accounts (over 200 ones), I got such a large file containing all the data. I hoped that processing of that global report would help me to obtain some additional knowledge. As a result, about ten parameters remained that become the basis for new reporting system of the forthcoming Championship.

Thus, the main conclusion is that it is insufficient to have a profitable strategy: it is necessary to strengthen its advantages and smooth its disadvantages. This is the point in money management: You should gain the largest profit from your approach, your Expert Advisor, your system with minimal risk and in maximal amounts, without being short of it. This is the most important thing, to my mind.

Considering the results of the reporting, would you now call the Top Ten of 2006 really good Expert Advisors?
Well, I did not precisely analyze each account, but my overall impression is that they are all good and it is logical that they got into the Top Ten. Yes, there is an element of chance there. What is it? Well, each of them could be in the other’s place. They could change places what we actually observed in the last days of the Championship 2006. However, they got into the Top Ten, mainly in absolutely natural way.

They are good because Expert Advisors submitted to participate in an automated trading competition within three months are, in fact, F1 bolides. There is a small difference, though: there are no pit-stops for them – if an EA starts, it must just go, that’s all. The fact that they finished (and finished with profits) means that they inherently cannot be bad.

You described a year ago what a profitable Expert Advisor should look like. Tell us, please, did you change your mind after the first Championship and after your researches?
No, in general, I do not think otherwise. I’m still sure that a good Expert Advisor must use general regularities of the market, not the nature of a specific currency pair. The matter is that these regularities show up a bit different for different currency pairs. In some of them, the regularities can be expressed strongly, whereas in other ones they are not so noticeable. Anyway, regularities can always be found. In other words, a good Expert Advisor must detect the moments when the market is ineffective and explore these moments. The inefficiency itself may appear on various time horizons. However, I do understand that it is much easier to say this than to do. I myself have not written such an Expert Advisor yet. Nevertheless, I hope that some of Participants of the forthcoming Championship will manage to do this.