Interview with Nikolas Ehrenberg (eniksoft)
Hello, Nikolas. We congratulate you on excellent results in the Championship. Are you surprised with these results or did you expect everything would be so successful? Many thanks for congratulations. But it is too early to exult. The Championship is a marathon, and we have run only about 100 meters. Well, I hoped to have positive results, but I didn’t expect the market to make such a surprise. The system showed much lower results during backtesting. The number of trades increased nearly two times. However, the market is slowly calming down and we’ll see how my system will behave further. You are not so long in the automated trading. Have you tried to use your developments on live accounts? The extent of my interest in MQL4 and program development proceeds by a sine curve. Everything depends on results. Ultimately, a good idea to be implemented is necessary. I can’t invent a more or less adequate program for the market yet. The EA that competes in the Championship basically is not able to work for a long time without human intervention. That is why I do not use such programs on live accounts. I tried to use one of my Expert Advisors, however a higher volatility that has existed lately in the market made me disable it for a while. We’ll see how things turn out. How do you search for the trading ideas and strategies? Perhaps, in forums, books, or observing the market? What is the most beneficial in your opinion? I use different sources. Well. The best way is to analyze the market. In this case there is no extrinsic influence and chances to find something unique are high. Furthermore, a system should be in tune with its user primarily. It is necessary to understand what the program does to be able to correct errors and inaccuracies in time. Generally, I observe the market while trading manually. Unfortunately, I do not always have the opportunity, because I do not have much free time to sit in front of the monitor. I haven’t seen any interesting or books useful (considering the development of Expert Advisors). Well, everything is in forums, both useful things and harmful ones. Please, tell us about your EA functioning. How does it open positions? My Expert Advisor calculates levels for entering. When price reaches this level an order is opened and Stop Loss and Take Profit are placed. And during operation the Stop Loss level is modified. That’s all. When placed for the first time, stops are quite far from the current price. Do you want to reduce the risk of getting a loss? Too far stops? In my opinion, they should be even further in the current situation. And risks should be well calculated. Otherwise it turns into playing roulette, where profit depends only on your luck. You can correctly set the size of the lot knowing the maximum loss level, so money will be used to their optimal advantage. You have written that your EA moves the Stop Loss to the lossless level if price moves in the necessary direction. Could you please describe it in more details? The Stop Loss and Take Profit levels are calculated when a new order is placed. If price moves in the “right” direction and crosses a special limit (in this case, it is the half of the profit) then the stop will be moved to a small loss area. Unfortunately, I had to shift lossless level to the small loss because the EA did not pass the automatic test in that case. It was accounted as a pipser (25 % of trades within spread). In any case, a small defect is better than disqualification. Besides, the number of trades is not large and the profit obtained from a trade is much bigger than these small losses. Have you used any standard indicators? And have you ever tried to develop your own custom indicators? Surely, I have. I wrote maybe about 30 of them for manual trading. I still search for the one that will be able to forecast price movements in the best way. Unfortunately, I do not have much time for Forex because of my work and study. That is why the number of programs I’ve written is not large. Please, tell us more about the use of the dynamic channel formed by indicators. Does it have any peculiarities? This channel was not invented by me. Long ago I noticed that the price “spins” around the MA. It means, what we need to do is to find the boundaries of this movement. And we obtain a dynamical channel. In my case, I took Envelopes as a delimiter of this movement. Its values act as the level for the Stop Loss. Advantages of such channel, in my opinion, are its mobility and objectivity. There is no need to search for the maximums and minimums to draw the channel frontiers. Indicators will make everything themselves and they will change together with price changing. However, there is a problem in choosing indicator parameters. I work on this problem at present. Can you share some interesting thoughts in this respect? The idea (I can even say not an idea, but a desire) is to make a fully dynamical system. I don’t know how to achieve it yet. I will try to add a static analysis. The market is chaotic in its movements but there is an order in any chaos. What are the maximal possible losses of your EA? What did backtesting show? Surely, it is out-of-limit. The amount of losses in case of initial Stop Loss triggering is 20% of the balance. The risk is reduced with the growth of the balance (with the maximal lot size). It is about 10% at the moment. Such a high risk is explained by the Championship rules, first of all. The winner is the one who earns a lot, and not the one who earns constantly. I looked through the results of the preliminary tests of some participants (that were published in the forum) and understood that the profit level has grown significantly this year. In no case it should be done on a real account. You can lose everything. How many currency pairs does your EA use? I use 3 currency pairs: EURUSD, AUDUSD and USDCHF. I’ve tried to use, but there is not much benefit from it when the number of orders is limited. Movements in the market usually appear synchronously on different currency pairs. I.e. the orders are often opened with a short interval in my system. As a result, there was no activity by some pairs if many of them were used, and the risks could increase. Trades are performed for the USD currency pairs. And have you tested your strategy with the cross-courses? I’ve checked it on all 12 pairs. Perhaps, their movement just has another structure (form). I have not analyzed this problem yet. And there is no need to use many currency pairs in the Championship for the present. It could bring a benefit if only the large amount of trades were performed within a day. Have you performed optimization of the EA on every tradable currency pair? Which pair is the most preferable for your strategy? Yes, I’ve performed the optimization on every currency pair. A special set of parameters was chosen for every pair. Now trading is performed using these parameters. The best result was achieved on EURUSD. However, the movement of the other currency pairs did not remain far behind from the leader. Could you please explain what you mean by the “special set of parameters”? Since I use standard indicators the parameters were chosen accordingly: the period, percentage deviation (for Envelopers) and the type of price for the calculation of MA. Essentially, the number of parameters is not so large. But I think even one parameter can be over-optimized. What parameter was the most important one for you during the optimization of the EA: maximum profit, a good profit factor, minimal drawdown or something else? Generally, it is a good ratio between the profit and the drawdown. The drawdown didn’t exceed 15% at the risk of 5%. That was the thing that allowed increasing the risk by 4 times not exposing the main deposit to the full devastation. Although I think that I optimized it too much anyway. It is only left to hope that the market will not change very much (this explains why I am looking for a fully dynamical system). After a brief analysis of trades on different currency pairs it can be assumed that the trading is lead by a trend and exits are performed from the channel frontiers during corrections. Is it really so? Quite so. They say, trend is your friend. So, as long as there is a movement up or down with small corrections the system must work. If there is a tight flat then the trades either won’t be performed at all or will be kept opened for a long time (because of the large Stop Loss and Take Profit levels). And if a trend begins to change frequently the EA will bear a lot of losses. Does it mean that a flat is dangerous for your EA? All depends on the width of the flat. If it lies between the Stop Loss and Take Profit (about 300 points) then it won’t be something terrible. However, we can face the situation when an order is opened and then the price goes directly to the opposite direction not reaching the place of setting the lossless level. Such a situation is really dangerous – it will produce a lot of losses. I hope it won’t occur during the Championship. Or if it still occurs, I hope it won’t last long or be frequent. Thank you for the interview, Nikolas. We wish you the best of luck! Created: 2008.10.15 Author: MetaQuotes
|
|





