I have run some backtests and come up with trades lists for 10 setups. My question is is it better to review the trade list based on one contract per trade over multiple years or apply money management and risk management and position sizing parameters right away? My feeling is that money management parameters(includes increasing position sizing as account grows) would result in overweighting recent performance. Maybe that's a good thing and maybe that's a bad thing? My gut is saying "Look at the performance with one contract to determine win/loss ratio and profitability. Determine the best variables that way, then apply your risk, money management and position sizing rules after that to get a sense of the what the variances in possible risks would be. Any thoughts out there? Other ways to look at it? I would definitely run both through Monte Carlo and check results there as well.
I would check first naked. Then if it's interesting from there, I would denature it with added complexity. But before denaturing it for real, backtest it. Because it does more than adding volatility. I guess it can turn a winner into a loser. Buy1Sell2 a guy out there said : money managent can turn a loser into a winner. Well ... If he's true, it can turn a winner into a loser.
I started not to answer because the answer is valuable and I paid dearly to discover it. But I was impressed with the fact that you realize the question even exists. Most backtesting techniques use a simulated final account balance as the metric with which to compare different systems/parameter-settings. This sounds like what you're doing. You'd be better off using a metric such as the average % gain per day. Using a simulated account balance as the deciding metric will bias your final results towards the system being more correct in the earlier testing time-frame (and more incorrect in recent time-frames). This is very bad. And, first test the signals, then, once the 'best' system/configuration is found; test for the best money management scheme.
Exactly. The first thing you should know is how does the system behave? What are the strong points and what are the weak points of your system. Without that answer you cannot apply a correct risk and money management. The reason why that is impossible is very simple: you should first know where the risk lies, if not you cannot do anything sensible. And you only know where the risk lies after analyzing the results of the system. The main question is: how much $ give you back in losing trades and how big is your drawdown? The number of losing trades are not important, the total amount in $ of these losses are important, as well as the total amount of $ lost in a consecutive row of trades (drawdown). Even if you give back only 10% of your total profits your drawdown can be 30% or more.