Interview on Automated Trading Championship 2010 from 01.12.2010.
Tomasz Tauzowski (ttauzo) is a long-standing member of the top ten on the Automated Trading Championship 2010. For the seventh week his Expert Advisor is between the fifth and the seventh places. And no wonder: according to the report of the current Championship leader Boris Odinstov, ttauzo is one of the most stable EAs participating in the competition.
How did you start trading?
As far as Forex is concerned - about 5 years ago, but earlier I was trading from time to time on the stock exchange. Generally, MetaTrader "encouraged" me to try my luck in Forex. I knew I would not have enough time to trade manually so I was very happy to see scripting option functionality in MetaTrader 4.
I wrote my first EA in 2005. I gained some technical analysis background during my studies - Finance and Banking majoring Capital Markets. As for the technical side - programming had always been my hobby. So I tried to combine those two skills and here I am. After making my first EA and testing it on demo accounts, ATC 2006 was announced and I thought I'll give a try. I entered every edition of ATC with various luck. As far as trading on real accounts is concerned I tried my previous EA on a mini account. At first it was 400% profitable but after 2 months, as you may expect from Forex, I went nearly to 0. This year I decided to write the second one, first for MetaTrader 4, then I translated it for MetaTrader 5. Anyway since my first EA was not successful on live account I decided to focus on participating in ATC. But still I trade successfully on stock exchange which is totally different and much less risky I guess.
What do you do besides trading?
I graduated from Warsaw School of Economics, where I studied economy -
finance and banking - my master thesis was on capital markets which
have always been of my interest. I work for a consulting company in
Business Intelligence group which specializes on data warehouses,
reporting systems. So apart economic background I have data processing,
IT background - a good mix, I guess, for developing Expert Advisors for
What was your experience in previous ATCs?
I participated in every edition. In 2006 I had no luck – I finished on place around 200. In 2007 I finished on place 66 – this one was the best of all my starts. But in 2008 it was a total disaster. In 2006 the battle looked promising - at the end of November I was in the first 8 with 100% won trades but because of my aggressive, scalping EA I had significant drawdown and lost the position soon after.
The participation gave me some results. I learned that when you base
only on historical price changes, Forex is unpredictable in longer
period of time. At least for me. My and, I guess, all other EAs need to
be optimized for certain period of time in order to work well and be
profitable. I guess we have such examples - some participants from top 5
lost their money in the following editions. I can't remember names but
was checking that - at least found one or two. So my lesson is not to
have too aggressive money management.
Before registration for
ATC 2010 edition I had one day to convert my EA from MetaTrader 4 to
MetaTrader 5 and did not have enough time to develop a proper money
management module for my EA. Who knows, maybe in the end it may appear
to be its biggest advantage but on a certain level of "fight" I have
little chance to win with such Goliaths as bobsley or even forez.
forez for example opens positions almost the same as my EA, but of twice big
volume - then all I can do is pray for a loss position - then I will
lose less and have bigger balance than the competitor. So next year I
need to focus on better money management.
What is the principle of your current Expert Advisor? What is it based on?
It is based on Bulls Indicator. It analyses its trend and, what’s most important, magnitude. When significant change of trend occurs my EA opens proper positions. I decided to implement fixed Stop Loss and Take Profit, Take Profit twice as large as Stop Loss. In this way I need only 33% of profitable trades in order to have profits. I like a sentence "trend is your friend", which is very common on capital markets, and modified it to "New trend may be your new friend" - I assume, that first direction of trend change, even if it is followed by a small correction, in majority cases appears to be a good direction to follow. I have over 47% of profitable trades - not much one might say, but take into consideration that for my EA profitable trade is the one that reaches 80 pips - that makes a real difference.
Another feature of my EA is that it allows only one opened position at a time but, thanks to quite frequent trade signals, is in the market for almost all the time.
Last feature of my EA is also a total mess in the code. I learned how to convert from MetaTrader 4 to MetaTrader 5 in one evening - thanks to changing the deadline for registration I managed to do that – otherwise I would be late and out of the ATC.
What technical analysis tools do you use in your developments in
addition to Bulls Indicator? And why have you chosen exactly Bulls
I use mainly oscillators, moving averages, RSA, Bulls Power, momentum. I didn't try to implement formation recognition as it is quite a difficult programming task.
Why exactly Bulls Power? Maybe I take advantage of the psychological effect. If price breaks highs and lows something happens in peoples’ minds and my EA shows that it's not a coincidence. I mean my EA hunts for high values of Bulls Power indicator because they might signal new trend up or down. As every EA developer, I guess, I have such a hypothesis. Then I optimized my EA to perform well on a certain set of parameters. No one can really guarantee that his predictions will come true.
How have you selected the Stop Loss and Take Profit levels?
SL and TP levels were also optimized for the highest balance and reasonable drawdown - I guess majority of short moves/breakdowns on EURUSD were close to 80 pips and were sometimes followed by corrections (sometimes by 40 pips) - this is my hypothesis, I didn't check that actually. TP was set 2x SL in order to discipline my EA not to mistake too much - I don't like big drawdowns as they show weaknesses of your strategy.
This year I had many positions that lacked 2-5 pips (this confirms, in a
way, my 80 pips hypothesis) to be close in plus and then reverse
occurred - If I had some position closing system I would secure those
profits then. But I don't have such unfortunately.
Why do you trade EURUSD? And why don't you use multicurrency strategies?
My strategy works quite well on EURUSD. I guess because of high liquidity of this pair. I can't really explain why but the fact is that I didn't manage to optimize it for other pairs. If I could I would trade other pairs as well. Another thing is that my EA was developed for MetaTrader 4 which didn't allow testing of multicurrency EAs so I was not able to check whether my EA performs well.
I would definitely use other pairs to make sure signals on EURUSD are significant. I mean - I would definitely use currency correlation in order to confirm my signals. The next thing I plan to do with my MT5 EA is to take advantage of the fact that some currencies might signal a little bit earlier direction of move of other pairs correlated to them. Next year I'll definitely try to do that and see If such things really happen. However it is really likely to turn out that Forex is so close to efficient market theory and such things are not possible. Anyway I'll give a try.
The Z-score of your EA is very high, you have a certain
dependence: win is always followed by another win, loss is followed by
loss. Why don't you use this dependence in your Expert Advisor?
Well, if I knew how to take advantage of that I would do that. I guess it might me "polluted" by problems with requotes. Some positions are not opened and it might turn out that if they were opened that z-score would be different. If you mean that I should for example skip 1 signal after loss occurs, it's not as easy as it seems. Otherwise my EA would be a holy grail then. But I'll try to work on it now and prepare it for the next ATC 2011.
What do you think about the recent disqualification of participants from Belarus?
Well, to tell the truth I am not totally for that rule that various versions of the same EA can't be registered in ATC. However, I also understand that the aim of ATC is the competition of EAs that are regarded by their authors as the best ones and most profitable - not the place for testing and "live" optimization.
It would be a sin not to mention that this disqualification also provided a place in the top 3 for such EAs as forez or like mine – everybody wants to win and this was a surprise for me. Without that it would be really very difficult to catch up with the top 3. Anyway, if they broke the rules they should take consequences of that but I guess that this decision must have been tough and probably discussions between jury members lasted for long.
You have used the MQL5 standard libraries in EA. Do you prefer to use a ready-made one or write your own?
As I said I had only one evening to convert my EA from MT4 to MT5 so I put the whole signaling engine inside some template that used those libraries. It's great that finally you have predefined libraries that handle generic operations.
How can you evaluate your chances of winning the Championship?
Winning? 0.5%. bobsley sometimes looses but doesn't go below $100,000. My EA with no money management can earn, I guess, maximum - 80.000. So there's a chance for 3rd place and theoretically 2nd, but very favorable coincidences must happen to allow my EA to be 2nd. Every EA around my EA would have to start losing. Getting on the podium will be a tough work to do for my EA but I don't give up.
Does your EA's behavior now differ much from its performance during tests?
To tell the truth yes. The profit it made during 2 months of ATC was, as far as I remember, the same as its profit for qualification period which was much longer – 8 months. So It was a nice surprise for me to get to 4th or 3rd place today with such a balance.
Do you like any of participating Expert Advisors?
avoitenko and bighope are EAs that have really nice equity curves, especially bighope made a great job - earning so much money in a couple of trades. Also forez, which is my greatest competitor performs quite well - its signals are quite similar to mine - whenever I open a position it opens the same position but with twice as much lots as mine - then i have to pray for this position to be a losing one – it’s the only way for my conservative EA to be ahead others. kgo (greetings for friend from Poland) has also been doing a great job until 2 weeks ago - it had almost 100% of profitable trades but then a disaster happened - I hope it will recover but not too much so that It won’t overcome me. I don't have to mention leaders which require admiration just because they managed to be top 3.Thank you for the interview and good luck in the Championship!
Translated from Russian by MetaQuotes Software Corp.
Original article: https://www.mql5.com/ru/articles/537