Interview with Harry Brinkhuis (Hendrick)

Harry, you are developing Phoenix 5! Tell us, please, why did you decide to improve something in your rather successful Expert? What is new in Version 5?
I think Phoenix, the contest version, has two problems. Phoenix is doing quite well in a trendless market. The problem is, however, that we have a trendless market for only one third of the time. For the rest of the time, the trend is upwards or downwards and Phoenix is in trouble because it is a countertrend EA and it places most of the time trades against the trend.

The other problem is that Phoenix is overoptimized for the last 6 months. I did it for the purpose of trying to achieve the most profit. You have to change the parameters a lot in order to keep Phoenix on the winning side. Phoenix 5 is still doing the same. I had 2 forums for Phoenix 5 but that got pretty out of hand. It took me a lot of time to maintain these forums: answer questions (the same questions over and over again) and study proposals to improve Phoenix. One time, I decided to quit with these forums because there was no more time left for developing. So I gave Phoenix to the community. There is now a group of people trying to further develop Phoenix in a very organized way. I think they are doing a great job, but I do not participate in this group any longer.

Why are your patterns true for only one or two months?
Because I used only backward testing for the last six months to find the optimal settings for Phoenix. Now, I think this is the wrong way to do it. A good Expert Advisor should have the same results for every period you are testing. 1 month, 6 months, 2 years, 4 years, whatever. Your results should be always the same.

You said on your forum that you would test parameters for Phoenix yourself. If it is not a secret, what does this testing technique consist in?
Yes, they asked me over and over again how to find the right settings for Phoenix. I promised them to make a manual about how to find the right settings. But that was very time consuming so I decided to drop that. To be honest, I really can't say how the testing has to be done. I think it's more a "feeling".

It looks like indicators are not the main thing in your Expert. Can practically any indicators be used in it or only those you specified?
There are so many indicators in MetaTrader 4. I've tried them all, but for Phoenix the OSMA and EMA seem to work quite well. I think you're right about the optimizing. The contest version of Phoenix has been optimized and maybe overoptimized. In other words, I have been using curve-fitting. I think this works, but only for a short time. Curve-fitting is ok, but you have to test a lot afterwards to find the right settings again. That's why I don't use this in my new Expert Advisors any longer.

Very interesting, Harry. So, if it is not a secret, what is the main principle of your new Experts?
I'm working at 2 new Expert Advisors: Unicorn and Kraken. Unicorn is a combination of a countertrend and a trend-following Experts. It's not using a Stop Loss or Take Profits. As soon as a trade is made, some code constantly watches how it is developing. It's using a special combination of EMAs and EMA levels to determine whether the trade has to be closed or it can continue. In a trendless market, it is working as a countertrend Expert. When the market becomes trending, it will follow the trend all the way. Backward testing for several periods (back to 1999) gave me always the same results with the same settings. Kraken? Well, Kraken is my secret and I think I will keep it this way for the moment.

How do you plan to avoid the short-term efficiency of the Unicorn and make it long-term efficient?
Unicorn is mainly based on the development of price and a combination of Moving Averages. Because of this and the fact that it's based on the development of price there are much less settings than I've used for Phoenix. I think that's the main reason why Unicorn gives the same results for different periods.

So, Harry, are you planning to use only Moving Averages (MA) in your Unicorn?
Only MAs and MA levels for the trend-follower part. I think MAs for a trending market are unbeatable. For the countertrend part, I'm still not sure what to use. Maybe an adapted version of Phoenix.

Harry, it is quite interesting to know about your new Money-Management system. Please tell us about it.
For money-management or RISK-management, I've used the code you can find in the example Expert "MACD sample" when you install MetaTrader 4. I think it's a very useful piece of code. I'm still trying to make my own piece of code for this, but I have not been satisfied with the results yet.

At what development stage are your Expert Advisors now, Harry?
Unicorn needs other 2 or 3 months before it can be used on live accounts. By the way, Unicorn will not be available for free. Kraken is highly experimental (it's based on pattern recognition). I really can't say when this EA will be ready. But it's a lot of fun to work on this one!

By the way, can you explain your desire to make your Phoenix public? Didn't you worry of possible competitors?
Well, maybe it wasn't a good idea after all to make Phoenix public! I even got an e-mail from some guy telling me that he wanted to sell Phoenix on e-bay! Maybe I am too naive for this business. On the other hand, I've learned a lot about all this in the meantime. I will not make this mistake a second time.

Did you use your Phoenix in real?
No. Before I go live with an EA I have to be 100% sure that it is performing in such a way that it will make a constant profit. I'm a bit surprised that other people are using Phoenix for their live accounts! Phoenix is ok, but you can't trust it in a trending market. I received several e-mails from people telling me that they are going to use Phoenix for their live accounts and asking me which pairs and settings they should be using. Unbelievable!!

Did the Championship influence you?
No, not really. All I can say that this Championship was (and still is) a very exciting adventure! I have my own ideas about developing Expert Advisors and the contest had no influence on that. On the other hand, I started the contest with the idea that my Phoenix would end somewhere at the bottom. So it was very surprising to see that it kept its place in the Top Ten all the time! So, yes, it changed my opinion about myself as a programmer of Experts. It gave me confidence!

There is not very much time left till the end of the Championship, but you still have not reached 30 thousand, Harry. Can you tell us why? Do you think you will take one of the first three places?
I think the first place will be for Rich. He's now protecting his profit by decreasing his lot size. Very good! On the other hand, his Expert has a maximal drawdown of more than 90%! So this is not an Advisor I would be using for my live account! I think there will be a heavy battle for the second and third place. Will Phoenix be part of this - I really don't know!

Will you participate in the next Championship? In Automated Trading Championship 2007?
Oh yes, I will! If I'm not ending in the top three this year I will have my revenge in 2007.

Thank you for the interesting Interview, Harry. We wish you luck in the Championship and in your developments!

Championship Report: Eleventh Week (10-17 December) The eleventh, the last but one week of the Automated Trading Championship 2006 is over. There are only 5 trading days left to the end of the Championship. The first place is being tightly held by Rich. Tough struggle for the second and the third place is being carried on.	Championship Winners’ Prize The first annual Automated Trading Championship 2006 is coming to its end. There are only two days left, which means there is plently of time still for the first 3 places to keep changing. As it was declared before, the Winners will get 80 000 dollars and much more to come.
Previous	Next