I am often asked about day trading systems. Usually what I’m referring to are stock index systems (E-minis & S&Ps etc). I do not have much positive to say about these types of approaches. I do not think I am being unfairly biased; I’ve spent over 15 years investigating all types of trading systems.
Day trading seems to satisfy the want for action and excitement in many traders. Sometimes I think these traders are not looking to make money but keep constant adrenaline pumping through their body! From my perspective,
I cannot think of even one day trading system from five years ago that is still performing today, that is right NOT ONE! Maybe a few have had occasional “comeback” periods, but I’m talking about 5 years of solid performance. A well respected trading system developer who developed some popular short term index trading systems has reportedly told some of his customers that even he thinks they only are valid for 2 or 3 years at best (he’s already removed one from his offering and slashed the price on another). Often, the day trading systems that still look terrific hypothetically are not realistically factoring in slippage and commission costs that eat up performance. I’ve seen some vendor’s factor in zero slippage! When adding realistic slippage the systems go from looking splendid to looking bad.
It is logical that these day trading systems could break down. The same thing that can cause them to look so compelling is the same thing that breaks them down. When working with one market (S&P’s) or sector, it becomes easy to “optimize” performance. Traders can “force” the computer to show them exceptional performance just from pure curve-fitting of that past data on that one market or sector, but when dependant on the market characteristics of just that one market or one sector what’s going to happen when that market sector changes? It reminds me of the research that showed that the drop in the S&P in 1987 should have only been a once in a several hundred year occurrence based on the current data, yet it happened in the first few years of the index trading! Markets change constantly and traders need robust systems.
On the other side of the spectrum, let’s look at trend following approaches.
They are not nearly as “sexy” as day trading. Traders may go through extended drawdowns or flat periods before making money, but think about this; Richard Donchain developed some basic trend following rules popularized back in the 1960′s. Those methods still work today, more than 30 years later!
I’m not saying I would trade those Donchain methods now. I think there are far better reward-to-risk systems and approaches available (such as ours), But it strikes me as significant that longer-term trend following methods popularized in the 60′s still work today. Yet, I can not think of one day trading system from even 5 years ago that is still working today. Does that say something? I invest my own money in the commodity markets with methods that would be considered mid to long term trend following, but, I do not invest even one dime in day trading methods.
Now, all that being said I do have something positive to say. My research has shown that short term (not day trading) systems can have low correlation to longer-term systems. So the right, short term system could help smooth out the performance of a suitable longer term system. Even if, that short-term system is marginal on its own, it may possess a synergistic effect when properly combined, but if that low correlation is the result of a curve fit system that is certain to break down, then there is no gain.
I do continue to devote time and research to short-term systems. Maybe someday I will have something that I believe is worth releasing. There are FAR more people interested in a short term index trading system than almost any other commodity trading system.
Owning an excellent short term system would be in my best interests, but so far I’m not convinced that I should commit any of my own money to those methods because of the limitations I outlined above. I do not want to release something and then have to embarrassingly do away with it a few years later! I’m afraid I’ve seen others go through this already. Personally I’m sticking to trading systems that have worked for a long time, and what I believe will continue to work.
I cannot think of even one day trading system from five years ago that is still performing today, that is right NOT ONE! Maybe a few have had occasional “comeback” periods, but I’m talking about 5 years of solid performance. A well respected trading system developer who developed some popular short term index trading systems has reportedly told some of his customers that even he thinks they only are valid for 2 or 3 years at best (he’s already removed one from his offering and slashed the price on another). Often, the day trading systems that still look terrific hypothetically are not realistically factoring in slippage and commission costs that eat up performance. I’ve seen some vendor’s factor in zero slippage! When adding realistic slippage the systems go from looking splendid to looking bad.
It is logical that these day trading systems could break down. The same thing that can cause them to look so compelling is the same thing that breaks them down. When working with one market (S&P’s) or sector, it becomes easy to “optimize” performance. Traders can “force” the computer to show them exceptional performance just from pure curve-fitting of that past data on that one market or sector, but when dependant on the market characteristics of just that one market or one sector what’s going to happen when that market sector changes? It reminds me of the research that showed that the drop in the S&P in 1987 should have only been a once in a several hundred year occurrence based on the current data, yet it happened in the first few years of the index trading! Markets change constantly and traders need robust systems.
They are not nearly as “sexy” as day trading. Traders may go through extended drawdowns or flat periods before making money, but think about this; Richard Donchain developed some basic trend following rules popularized back in the 1960′s. Those methods still work today, more than 30 years later!
I’m not saying I would trade those Donchain methods now. I think there are far better reward-to-risk systems and approaches available (such as ours), But it strikes me as significant that longer-term trend following methods popularized in the 60′s still work today. Yet, I can not think of one day trading system from even 5 years ago that is still working today. Does that say something? I invest my own money in the commodity markets with methods that would be considered mid to long term trend following, but, I do not invest even one dime in day trading methods.
Now, all that being said I do have something positive to say. My research has shown that short term (not day trading) systems can have low correlation to longer-term systems. So the right, short term system could help smooth out the performance of a suitable longer term system. Even if, that short-term system is marginal on its own, it may possess a synergistic effect when properly combined, but if that low correlation is the result of a curve fit system that is certain to break down, then there is no gain.
I do continue to devote time and research to short-term systems. Maybe someday I will have something that I believe is worth releasing. There are FAR more people interested in a short term index trading system than almost any other commodity trading system.
No comments:
Post a Comment