I know what keeps me going - it's the possibility that the next system will be IT. And IT is the system that will just keep on winning. It will be simple, easy, and obvious. It will make complete sense and when I discover it I will say, "Gee, why didn't I think of THAT before?" - a lot like discovering the wheel, or the flush toilet (arguably the most important discovery of all-time).
And, over the years, I have said "GEE" many times. And these systems HAVE worked - some fairly well, and some for quite a long time. At some point they run into a string of losses - and record drawdown, and after that they don't seem quite as exciting.
What I have learned to do over time is to always trade my best system, while continuing to either improve it, or come up with something better. But every day, I call in my trades with my best system, even when I'm losing money, scared, depressed, or just not comfortable with anything I am doing.
I have found that my systems are smarter than I am. After all, they contain all the knowledge I have acquired to date about the markets, and are the result of many, many long hours of thinking, testing and making or losing money. I can't recall just when I figured out that my systems were smarter, but one day it hit me - probably after some failed attempt to pull off some brilliant intraday strategy.
The trouble with actually making money by overriding your system is that you'll actually begin to think you can make money this way all the time. It just sucks you into a losing strategy. Don't do it.
Another saying I have developed over time goes something like this, "If you can't test it, don't trade it." I came to this conclusion after backtesting many great ideas - ideas that seemed to work all the time in recent months. What I usually found out was that lurking in the past somewhere was a combination of market moves that would have murdered my system.
I want to know what the worst drawdown is likely to be. And, of course, the REAL drawdown I will experience trading any system is always going to be much worse than any drawdown I can calculate by optimizing.
The reason for this is obvious: an optimized system is the very best that could have been. Reality says that I won't get that result. I have found that in real-time trading I can get perhaps 30-40% of optimized results. Or, to put it another way, my real-time drawdown at some future point will always be two to three times higher than my worst optimized drawdown. This rule seems to be a universal constant.
Another trap - everything looks easy in retrospect. And, here's a biggie - my computer doesn't feel any pain during drawdown. It's one thing for my printout to tell me that my worst drawdown was $5,000 - it's quite another thing to experience that drawdown trading several contracts with real money.
If you can't survive the drawdown, you won't last long in the market. Whatever you do, first plan for the worst drawdown. And I don't mean add the drawdown to the required margin. I hear this all the time but it's just too scary to do with real money. In real-time trading you'll often get near-record drawdown. This means that you'll experience huge equity drops on a regular basis. Nobody can take this kind of punishment for long.
So, over time, I have developed a methodology for deciding how much money to allocate to trade one contract. I have decided that I get upset if I have a drawdown of more than 10% and REALLY upset if it gets beyond 20%. At 30% I am a zombie. You may be different.
So, to make sure I won't have a drawdown of more than 10% I take my worst calculated drawdown (let's say this is $5,000) and I divide that by .10. This gives me $50,000, and tells me to allocate $50,000 to trade one contract. If I get my $5,000 drawdown next week, my $50,000 will drop to $45,000 and my account will be down 10%. Simple, isn't it?
If you only have $5,000 to trade, this means you'll have to trade something small. Forget the S&P. My broker tells me some of his "clients" are trading 100% of their equity on margin. This means that to trade one S&P they are allocating $16,000 per contract. This means - two bad days, and you're out of here. That's too scary for me.
Of course, you can always just hang on to Bill Gates' shoe laces and buy Microsoft stock as a "long-term" investment. Seems so easy, doesn't it?
Or, perhaps you can test buying Microsoft leaps, or perhaps....GEE, so many systems, so little time. Back to the computer.
---Tom Loffman
Copyright, 1997
Tom Loffman, Equity Systems