#024: HOW SWITCHING SYSTEMS CAN IMPROVE YOUR TRADING
The beginning of the last year I spent traveling in really inspirational places, to get some distance from my trading and to come up with new, “out-of-the-box”, ideas. What affects me really much is Singapore - everytime when I stay there, I come up with lots of new ideas. In this article, I will share with you one of the ideas that I got when I stayed in this inspirational city last time.
In Singapore, I have done an unbelievable amount of work, among others I have created my first breakout automated trading system that generally works in 27 different markets. But this is not the topic of this article.
It has all begun when I tried to create a system for NIKKEI index, to diversify my trading activities also outside the markets in the USA and in Europe. While traveling, I spent a lot of time trying to develop a system for Nikkei, but I wasn't successful in the beginnings. First of all, I had to move from intraday systems to swing ones, as intraday systems had too low average profits and after I included commissions and slippage, there wasn't much left.
And, as it showed up, the results of swing systems were even worse! Whatever I did, the equity of my systems looked always like this:
I was constantly thinking about any new idea that could improve these results. Similar situations always encourage me and release maximum power of creative part of my brain. Even though coming up with new ideas might be sometimes painful, the results are (mostly) worth it.
I have created 10 testing systems for Nikkei and they all had one common problem - the equity curves looked always, even after reoptimization, as the one above.
Finally, the idea came: What if instead of regular annual optimization of parameters (which is part of my successful workflow of building breakout trading strategies), I try to switch one system for another? In other words: I take all 10 systems that I have created for Nikkei (despite the fact that all equities look as the one above) and I create a rule to switch, in regular intervals, one system for another. It is similar to regular annual reoptimization of a system, but in this case it means to connect 10 systems into one and switch on an annual basis one entry rule for another.
It took some time to develop a working system from this idea, but the result was breathtaking. This idea, when tested for the first time, has brought this great result:
It is an amazing improvement and further testing has just proved that this idea is really robust and can be used for real live trading. There are several practical challenges, like to create robust rules for system selection that we will want to use for the next trading interval (for understandable reasons, I will not share this rule with you). This is how it usually works - with every new idea there also comes a lot of practical aspects that need to be solved and the ability to manage this differentiates successful traders from unsuccessful ones.
This interesting “discovery” in Singapore has brought yet another point - it has confirmed me that in these days of dynamical markets it is absolutely essential to come up with new ideas that constantly adapt to markets changes. In other words, even if I hesitated last year whether the regular reoptimization is really the best way how to improve my way of trading style and workflow or not, today I am fully convinced that it really is. After all, when coming back home from this trip I have run several verification tests and some of the results were really shocking (like finding out that using 3D optimization charts for optimization is the best way to trading hell).
According to my experience, using regular optimization is really necessary even for most of the intraday discrete strategies when the trader has to be able to react to constant volatility changes and change the stop-loss and profit target accordingly. In this case, it is not an area that is specified exactly, but I think that every discretionary trader already knows how volatility changes can be bothering - so he must have an according plan how to adapt to these changes. This is, from my point of view, another step that can be called reoptimization. Switching one system for another is, of course, quite advanced technique and I still work on different variants how to use it. Moreover, this is no holy grail, it is just one of many ideas and another improvement of my portfolio of breakout automated trading systems. Even so, for me, it is one of the very promising ways that are really worth finishing. Above-shown tests speak for themselves.