Say you want to trade two systems that anti-correlate with each other, such as a trend system, "system A", and a counter-trend system, "system B". You have backtested and found that when one is losing money, usually the other is making money. What if when system A is in a drawdown, you reallocate a percentage of system B's resources to system A? You would be counting on a reversion to the mean. Alternatively, you could increase resources to the currently profitable system at the expense of the system currently in a drawdown. This would be a trend strategy in itself. Does anyone do this? Would this be a money management technique or is this another strategy altogether?
I know this sounds simple, but it seems very much like hedging to me. I am assuming that you are talking about two instruments that regularly trade in "opposite" directions though. This seems like a correlation in itself. If you are following up with a what do you think is a better method question, I would go with allocating out of the drawdown into the winning trade, rather than adding to the losing position.
No, same instruments, different strategies. E.g. trend system and anti-trend strategies on index futures. Since same instrument and opposite strategies can come close to anti-correlation.
This will sound harsh but it seems like neither system works. Becasue you never know when the drawdown will end, re-allocating the money is a trade in and of itself. What if you take money out of the system that is in drawdown..right before that ends and move it into the system that is just about to start its own drwadown?
You could do this fairly simple on a simulator with one stock, going long and short, using money management adding to one position with the trend.
I considered that many many years ago what you will realize is that EVEN if it makes you money it won't make you enough there are better ways
Assume this is a mechanical backtested system. In plan #1, a regression strategy, as System A approaches x% of max historical drawdown, you divert resources from system B to A.
I am not sure but it sounds like you must first see how each system does individually. Are they profitable individually? If so you can just trade both and one hedges the other. Then you can "balance" your portfolio but thats another subject. However, if you are telling me the systems are perfectly negatively correlated then I dont think it will work b/c its like two parties in a zero sum game ending with zero profits. In the end commissions, ect.. will kill you. You can reduce the position sizes for one during a drawdown and increase it in the other but that reduces it to speculative trading b/c how do you know when the drawdowns end and the other begins? It becomes effectively another strategy and not a money management system b/c you are still speculating/betting. Its like options where you are betting on future volatility except you are betting on future trendiness/ranginess. The big problem is how do you know when to switch strategies? What do you guys think?