Interview on Automated Trading Championship 2012 from 26.12.2012.
Hello Enbo. Please tell us a little about yourself, how you got to the Championship and what ties you with algorithmic trading.
I have a master degree in communication engineering and a major in signal and information processing. I live in Shanghai. Now, I am working in a mobile communication corporation. Trading is one of my hobbies. I also like swimming and reading in my spare time. It was about two years ago when I opened my first account to trade FX. Soon I found it was time and spirit consuming when trading manually, so I started to think about using robots to automate trader's daily job.
It was that time I learned about
MetaTrader 4 and MetaTrader 5. And I find it is amazing that anyone can
get whatever knowledge he wants to design a trading strategy in your
site and forum. Since I am very familiar with C and C++, so there were
no difficulties in learning MQL4 and MQL5.
I also want to note the Step-By-Step Guide to writing an Expert Advisor in MQL5 for Beginners article which was very useful to me at the beginning. In fact, your site has a lot of easy to understand articles that will be useful to any newbie developer of trading strategies.
By the way, what language do you like more?
I prefer MQL5. Testing multicurrency EAs, as well as taking benefits of OOP in designing trading systems - these are the key benefits for me. Some may not like the lack of hedging positions on one symbol. But for me it is not a problem, because I never used bidirectional positions.
What trading strategies have you tried?
I have tested many trading strategies, including trend following, counter-trend, scalping, averaging, strategies based on statistical analysis, etc. Even though it sounds trite, but I have not found a strategy that could win all the time. The market simply does not repeat itself, so it is difficult to predict what will happen next. Trading is a game of probabilities. As of strategies that I prefer the most, I think some classical ones like breakthrough system will always be vital.
Automated trading prevents you from trading emotionally. There is nothing human cannot program, but it is very difficult to make a robot adaptive to all kinds of market. In my point of view, what never changes is the market behavior that affects human psychology of all kinds of people in the market.
What lessons have you learned from the last year? At that time your EA traded not very successfully.
When I learned about the Automated Trading Championship 2011 and began to program my Expert Advisor, it was only a few months left before the start of the competition. Over the past year I have accumulated a lot of experience. However, I believe that trading is not so much about skill as it is about art.
Talking about the art of trading, what masterpieces of Automated Trading Championships you like most of all?
As far as I can remember, some Expert Advisors that use self-learning algorithm like neural network are very amazing. I think this is a very interesting area, and I am planning to develop this kind of EA.
We noticed that this year you have prepared multicurrency EA. Was it difficult to create it compared to a single currency EA?
Yes, this year I decided to send EA that was originally designed to trade more than 5 currencies. Unfortunately, that variant did not pass the test because of the 30 minutes time limit. I had to cut the portfolio up to two currency pairs. As a result, I chose EURUSD and GBPUSD because of their popularity. Multicurrency EAs might be a little more complicated compared to single currency ones if you want to use correlation of some pairs in strategy design.
Last year one of Chinese participants took the third place in the Automated Trading Championship. Are you inspired by the success of your fellow countryman somehow?
Undoubtedly. Actually, it made me believe that everyone from all over the world could have the chance to win the prize. I visit the MQL5.community site frequently. Most of all I like to read articles from which I get my knowledge to solve problems in coding trading strategies. Also, I read every interview and every publication on the Championship site if I have time. It is interesting to read those stories of other developers.
Please, tell us about your trading strategy.
My strategy is simple. It is based on counter trend indicator and uses averaging method with some optimizations. In most cases it will only take a small profit on the basis that the market is rolling most of the time. But if the market has a strong trend then it will add lots to averaging the cost until price recovers.
I use fixed Take Profit and simple money management which does not allow losses to fall below 70% of the balance. 30% of the capital is a very high risk for trading. On my real account I usually risk less than 5% of my balance.
Do you use correlation of the two pairs in your strategy?
No, I did not use correlation in this EA. But I am designing one robot that takes advantage of the correlations between some pairs.
How do you detect a trend? What indicators do you use?
There are many reversal indicators that can be used in this kind of strategy. I use RSI as the only indicator when open the first position.
In your opinion, what chances your EA has to win?
I will be very glad if it can stay in TOP-10 of this year's Championship. Among the competitors I would like to mention trading robots of JPAlonso and FIFO. I think there is no EA that can win all the time, but it is not that difficult to design one that has good performance in a given period of time.
What market your EA is best suited to?
I think it is more suitable for flat, where it has a very high probability of getting small but nevertheless profit. On strong trends it risks either to seize big profit or to suffer large loss.
Was it difficult to prepare for the Championship? What criteria have you used for optimization?
No, I had not got any difficulties except for the limitation of testing time. This one was hard to beat. I hope that next year you will extend the testing time limitation. Not it is too short for some kinds of multicurrency EAs. There were more than 10 parameters that I've used for optimization. But only three or four of them play a major role.
Why have you opted for quite a small M5 timeframe? It is quite risky.
That's because this kind of strategy takes small profits most of the time. If you want to get more profit in a relatively short period, you have to enter the market more frequently. And you have to take risks to get a good ranking, don't you.
What do you think about the common misconceptions in trading?
I only have 2 years of experience in trading, which is very little compared to some other participants of the Championship. There is still much to be learned. There were some painful losses in my real trading, but I have not broken any account yet.
Thank you for the interview, Enbo. What would you like to wish other participants in the remaining days?
Of course, good luck! I want to give my thanks to the Organizers for such an exciting competition, as well as to MQL5 developers for their work. And I want to advise everyone who is interested in algorithmic trading: "Be sure to participate in the Automated Trading Championships, where you can get a truly invaluable experience!"
Translated from Russian by MetaQuotes Software Corp.
Original article: https://www.mql5.com/ru/articles/622