In this statistical report we analyze the TOP 10 Expert Advisors as of the 10th of December, 2007. The analyzing method is the same as in the Statistical Report #3 of the Automated Trading Championship 2006. Like a year ago, Expert Advisors are based on different trading strategies. Even working under similar strategies, EAs can differ in their parameters. We will try to estimate the changes in such properties of leading EA's as risk level and stability.

Risk level is the parameter that shows what part of its deposit an Expert Advisor is ready to lose if the situation develops unfavorably. The Expert Advisors' risk levels will be estimated by two factors: maximum amount of lots in positions opened in the same direction and StopLoss levels. Stability of an Expert Advisor shows how steadily its deposit increases. To estimate this parameter, it will be necessary to evaluate the maximal drawdown experienced by the EA.

Risk Level

Risk level is the maximal amount of lots ever opened by an EA. If an EA opens several positions in the same direction, the maximal amount of lots is equal to the total volume of trades. If positions are opened in different directions, the difference between their volumes is calculated.

It should be noted that the values of this parameter changed significantly. Leaders of the ATC 2006 opened no more than 5 lots in one direction. The only exception was ldamiani, who in some moments had 10 lots opened in the same direction. This year there are almost no such "modest" Expert Advisors. Seven EAs out of the TOP 10 opened from 10 up to 15 lots in the same direction as of the 10th of December 2007.

Partly this fact can be explained by a higher average Equity of this year participants as compared to the previous year. All the 10 top EAs traded only one currency pair; the TOP 10 does not contain any multicurrency EAs.

Five Expert Advisors, i.e. half of the TOP-10, opened up to three positions simultaneously; each position could be of the maximal volume in lots. They are: Better, winwin2007, Yuraz, zyx40 and wackena. While Better's EA contains three independent trading systems that open trades though in the same direction but in different times, the EA of wakena opens 3 trades simultaneously at each trading signal. The EA of Yuraz often opened positions in the same direction, actually following the tactics of averaging. The necessity of placing the protective StopLoss cannot be doubted. The further The StopLoss level is, the larger losses we are ready to bear in case the situation develops unfavorably. The absence of a protective StopLoss actually means that the Stop level is equal to the level of a possible Stop Out (a forced closing of loss trades, if the Margin Level falls lower than the Margin Call level). The average StopLoss in points is shown in the blue color, the TakeProfit level (also in security points) is in the red color.

Two Expert Advisors from Indonesia - dwijayanti and admadinata - do not use StopLoss. The Participant dwijayanti opens position without specifying StopLoss and TakeProfit levels. The Participant admadinata executed 7 profitable trades, and no one of them had StopLoss, but had a placed TakeProfit level, upon which the trades were closed.

The EA of the Participant TeamSky from Japan used the Trailing Stop, that is why his average StopLoss level in points is not shown, because the average value is positive, i.e. his risk level cannot be defined. The Expert Advisor of the Participant wackena also used the Trailing Stop, so the average StopLoss value is positive (a little less than 10 points) and is not shown in the picture.

The evaluation of the stop level in points is not always vivid, so additional calculations were made: the maximal risk for a trade calculated in percents of the current Equity. Here we also can see some changes as compared to the values of the Statistical Report #3 of Automated Trading Championship 2006. This year the TOP 10 does not contain any EAs that risks more than 25% of their deposit. Last year the risk level often exceeded 50 and sometimes even 100% of the deposit. The only exception now is the Participant dwijayanti that trades without Stops. This Expert Advisor opened only Sell positions and the risk is limited to the Margin Call value, it is equal to 50%.

The three presented diagrams prove the conclusion made in the two previous statistical reports - Statistical Report #1 and Statistical Report #2 - the Participants of the current Championship ATC 2007 have presented more aggressive and more cautious Expert Advisors at the same time. EAs trade larger numbers of lots simultaneously in the same direction on one security, at that the risk level of the most top EAs is reasonably limited per one trade. The Participants of the ATC 2006 placed protective Stop Levels larger than the deposit size.

One more conclusion can be made from the presented data: only by controlling risks one can gain large profits at an aggressive trading style. Most EAs from the TOP-10 list control the maximal risk and at that open large positions.

Stability

The value of the gained profit cannot fully characterize a trading system. Very often, important is the information about the drawdown experienced by a trading account within the analyzed period of time. This year we added the possibility of seeing the statistical characteristics of Participants on the Championship website page. The description of the characteristics was published in the article "Mathematics in Trading: How to Estimate Trade Results". In this Report we gathered information about the following drawdown types:

• Relative drawdown is a ratio between the maximal drawdown and the value of the corresponding local maximum of the equity. This coefficient shows losses, in percents of equity, experienced by an Expert Advisor;
• Maximal drawdown is the maximal difference between one of the local maximums and the subsequent minimum of the equity;
• Absolute drawdown is the difference between the initial deposit and the smallest value of the equity. In other words, this parameter shows how much below the initial deposit level the Participant's equity has ever fallen.

Fig.4 shows that most Participants experienced the relative drawdown from 30 up to 40%, which is not so significant as compared to the acquired profit from 150% to 1000% (Better). But some of the Participants exceeded the 50%: dwijayanti (79%), Saloght (65%) and maxfade (65%). The relative drawdown of the Championship leader Better is 29%.

The higher the value of the acceptable drawdown is, the higher is the possibility of losing money in future. But after two months from the competition start it is impossible to enter the TOP avoiding large drawdown. One has to risk anyway. That is why all the leaders have quite a large relative drawdown and this is a conscious decision of the Participants.

Also interesting are the changes in the maximal drawdown values as compared to the previous year. Remember, in the Statistical report #3 of the ATC 2006 the majority of EAs described in the report had the maximal drawdown no more than $6000. This value is considerably higher in the current Championship.

As seen in Fig. 5, only winwin2007 has the maximal drawdown less than $5 000. Other Participants' drawdown is in the range from $7 204 (zyx40) to 19 020 (dwijayanti). What is interesting, Better whose profit is the largest is only in the fifth place according to the maximal drawdown value. The drawdown is equal toое большое значение собственных средств на счету (Equity) $13 552. The Prize-winners of the ATC 2006 did not have the largest drawdown. It was equal to $ 9 770 in Rich's EA (the 1st place), $11 532 in ldamiani's EA (the 2nd place) and $5 890 in the EA of GODZILLA (the 3rd place).

The rule that the maximal profit requires maximal risks is fully proved here. None of very cautious Participants entered the TOP-10. At the same time the TOP-10 does not contain Participants with a too large maximal drawdown, because such EAs are more likely to be placed at the bottom of the list.

Let us view the next index - the absolute drawdown. It shows by how many dollars the equity has decreased as compared to the initial deposit ($10000). Here we observe a good conformity: the majority of participants have the absolute drawdown from $1 000 to $2 000. Only dwijayanti stands out - his absolute drawdown is equal to $7 433. It means that in some point of time his equity was equal to $2 567. We recall that earlier we have found that this Participant does not use protective stops and can tolerate unlimited drawdown values. And the appearance of such an Expert Advisor in the TOP-10 is quite a rare exception.

Using the data obtained, we can calculate the "stability coefficient" or "reliability coefficient" of Expert Advisors. Another accepted definition is the Recovery Factor.

For this, it is necessary to find the ratio between the Expert's profit and its maximal drawdown. The higher the coefficient is, the more stable the Expert's work is. For example, if an Expert has stability coefficient of 10, it means that this Expert can have profits 10 times exceeding its maximal drawdown.

It can be seen in Fig.7 that most Participants have the recovery factor value about 2. It means that the gained profit is only twice higher than the experienced maximal drawdown. The stability coefficient of the Participant dwijayanti is equal to 0.99; it means the confidence in his future fortune is minimal. And only three EAs have the stability coefficient more than three. They are winwin2007, zyx40 and Better. And if we view the 25 EAs analyzed in the Statistical Report #3/2006, 17 of them (more than 50%) had the stability coefficient more than 3. We cannot say for sure, whether zyx40 can become a prize-winner, but it is obvious that taking into account all the described characteristics, Better has the best indexes. And it is quite natural that he is leading in the Championship for already several weeks.

Conclusions:

1. The Participants have made necessary conclusions and are trying to limit risks at opening positions. The majority of Expert Advisors have built-in Money Management systems and limit the amount of losses either by StopLoss or closing by market;
2. The Participants of this Report have quite a low absolute drawdown value. However maximal drawdown reaches 40% and more. It means, the Participants use a more aggressive reinvestment system;
3. Most Expert Advisors have low stability coefficients (recovery factor). It is necessary to improve this index, so that an EA could gain stable profits not only during testing, but in real trading as well.