Statistical Report #1

The contest has been lasting for one month already, so we can see some figures on Participants and their Expert Advisors, like previous year. The most Expert Advisors started opening positions on the first day of the Championship. There are 371 of them (i.e. a bit more than 61% of the total amount). The number of such EAs has increased to 574 (over 95% of Participants) by the end of the month. Only 19 (3.2%) Expert Advisors have not opened any positions yet. Last year, the amount of “sleeping” Experts had made about 5% by the end of the first month. As we can see, the results are rather close to each other.

There are much more losing Expert Advisors this time – 423 (70%), which significantly exceeds the last-year result (62%). Perhaps, Expert Advisors submitted for this-year Championship Perhaps, the Expert Advisors submitted for the Championship this year are based on riskier strategies. This can also be indirectly confirmed by the fact that AlexG was the first to finish the contest on the 31st of October 2006 and resulted in $117.96, whereas the first EA that had finished this-year contest, iyong, needed less than 10 trade days to reduce its deposit to $100.64. The Expert Advisor closed its last trade at 15:00 on the 12th of October 2007. This Participant does not have to worry anymore about what fate awaits his contest account. However, higher volatility of the market this October as compared to the last year is entirely possible.

The amount of profitable Expert Advisors has not essentially changed within the first month. Although the number of profitable EAs made 172 on the first day of the Championship, they have made 161 (approximately 26.7% of all Participants of the Automated Trading Championship 2007) by the end of the first month. This is much less than it was a year ago. The amount of profitable Expert Advisors made 33% at that time. Perhaps, this effect can be explained through a more representative sample, a higher amount of Participants of the actual Championship. But then the amount of losing Expert Advisors increased from 199 to 423. We can resume that the most EAs that had started trading flew into the number of losing ones.

Fig. 1. Expert Advisors Grouped by Equity.

Changes in Average Equity

A kind of barometer for Expert Advisors on the Championship is, traditionally, the mean value of equity. The average equity of all Expert Advisors participating in the Championship is decreasing. The blue curve shown in Fig. 2 indicates how the average equity changed in all Expert Advisors. Here we can see again that the current Championship differs from the preceding one. The average equity of Expert Advisors in the Automated Trading Championship 2006 had made just a bit less than $9 000 by the end of the fourth week, whereas the mean equity has made $8 026 by now.

The green curve shows the changes in equities of profitable EAs, whereas the red one is for losing EAs. Equity of profitable Expert Advisors grows quicker this year as compared to the same period of the last-year Championship. The average equity of this group of Expert Advisors made, as of the 31st of October 2007, $13 867 to $12 300 of the last year. At the same time, the average equity of losing Expert Advisors falls quicker than it did last year. As of the 31st of October 2007, it is $5 691, which is considerably less than it was after four weeks of the Championship in 2006 ($ 6 800).

Thus, we can conclude that Participants of the Automated Trading Championship 2007 show in total riskier and more aggressive strategies in this contest.

Fig. 2. Average Equity of Expert Advisors.

Distribution of Expert Advisors by Equity

We tried to study distribution of Expert Advisors by equity last year. The following procedure was carried on for this:

All experts are divided into groups depending on their equities. For example, the first group includes the Expert Advisors, equities of which range from 0.00 to 1000.00. Those having equities from 1000.01 to 2000.00 will be in the second group, etc. Some of these groups will contain many EAs, some will be empty. However, having calculated the amount of Expert Advisors in each group, one can imagine how the things are going.

Fig. 3 represents distribution of Expert Advisors by equity as of the end of the first month of the Championship. The distribution pattern differs of that it was a year ago. The dissymmetry of distribution is determined by the fact that Expert Advisors couldn’t lose more than $10 000, whereas the potential profit is not limited in any way. Considering the higher aggressiveness of this Championship’s trading strategies, it’s no wonder that we see even stronger skewing of the distribution to the left, into the range of losing Expert Advisors.

The region located to the right of the level of $10 000 shows us the distribution of profitable Expert Advisors. So we can see that there are 87 accounts in the interval from $10000 to $11000, whereas there are only 44 of them in the interval from $11000 to $12000. The maximal equity did not exceed $23 000 in the last-year Championship, whereas there are 8 Expert Advisors this year that have already overstepped this level. Time will tell how many of them will keep this result by the end of the Championship. It should be reminded that many Expert Advisors, which finished Championship 2006 being in the first page of it, had already been in Top 10 after four weeks. So the probability of that the future Winner of the Automated Trading Championship 2007 is already among today’s leaders is very high.

Fig. 3. Distribution of Expert Advisors by Equity as of the End of the 4th Week.

How Position Size Impacts Equity Fluctuations

Each Participant of the Automated Trading Championship 2007 has a page with statistic data based on executed trades. The article "Mathematics in Trading: How to Estimate Trade Results" contains the description of Money Compounding that characterizes the risk factor of strategies. The higher the value of an EA’s Money Compounding is, the more fluctuations experiences the account managed by such an EA. The Participants of the Championship were divided into 5 groups based on the value of Money Compounding:

1. Money Compounding ≤10
2. 10 < ;Money Compounding ≤20
3. 20 < Money Compounding ≤30
4. 30 < Money Compounding ≤40
5. 40 < Money Compounding ≤50

The average Equity value was calculated for each group with the purpose of checking the correlation between the Money Compounding parameter and the changeability of Equity.

Fig.4. Equity Graphs for Different Values of Money Compounding

We see, the higher the Money Compounding value is, the more changeable a group Equity is. One may have an impression that Expert Advisors with higher risks (Money Compounding ≥ 30) "feel" better than others, because the purple and orange Equity lines are above the value of the initial balance of a trading account. But in reality it is different, and groups with higher risks include Expert Advisors that have not born great losses up to the moment.

As the Championship goes on, the number of Expert Advisors in risky groups will become lower, while loss trades will shift aggressive strategies into groups with lower risks. They will lack free margin for opening trades with the large number of lots.

Conclusion

1. Expert Advisors participating in the actual Championship seem to be “psyched up” for a riskier trading. As a result, the amount of losing Expert Advisors exceeds the last-year data on percentage. Due to automated pre-testing system, the amount of trade errors in EAs decreased. Trading tactics became, judging by the published interviews of Participants, more sophisticated.
2. The percentage of profitable Expert Advisors has not changed much as compared to those of the last-year Championship. However, we can feel as if this group progresses in a more cohesive way and it is thinning off slower than during the previous Championship. The overwhelming majority of Expert Advisors has more sophisticated mechanisms of money management and trades management, which raises such EAs to the top of the “standings” if the situation in the market is favorable. However, the obverse case of such complications and elevated risks is that positions are lost as quickly as they have been taken if the situation is unfavorable.
3. At a higher risk level, the Equity value changes greatly, and Expert Advisors become less stable.

Jury Session Record of the 6th of November 2007 At the session, the Jury discussed disqualification of three more Participants.	Interview with Tai Elliott (MiltonWaddams) At present, I'm interested more in something which has proven results over a long period without any intervention, rather than an EA which requires frequent optimization.
Previous	Next

It seems my trade was lost somwhere:

2007.11.08 16:35:45	'500031': request in process
2007.11.08 16:35:45	'500031': request was accepted by server
2007.11.08 16:35:45	'500031': instant order buy 5.00 EURUSD at 1.4662 sl: 1.4562 tp: 1.4692

?? where is it?

66

ttauzo

2007.11.09 19:29