I performed a walk forward optimization today and here are the results. Although this is not exactly identical to my live strategy, some interesting observations may still be made.
I chose one-year periods for in sample as well as out of sample data. I felt this would demonstrate better the responsiveness of the strategy to changing market conditions over a relatively short period instead of averaging out performance over several years.
I like how my strategy is "alive," adapting to differing market conditions, much as a living organism adapts to its changing environment. I particularly like the consistency of the average bars held. I can know within a few days if the trade is profitable, and can turn around and deploy the cash elsewhere rather than wishing I had exited when I had the chance.
The net profit and net percentage profit may be a bit unrealistic, because Amibroker is not looking at intraday data. This may result in "cherry picking" during broad market declines over a period of days. The discrepancy to live trading is yet to be seen. So far this year, I am very close to paper trading percentage wise at an 85% APR return. I have included a penny slippage on the limit entries and 5 cents slippage on the market exits, as well as $2 round trip comm. for the optimization.
The maximum system draw downs, typically around 10-15%, should be quite manageable going forward, with the worst being the '08-'09 period of around 35%. It is interesting that the years with the largest draw downs also had some of the greatest returns.
My current live profit factor and percentage winners are both just about on target to the historical averages.
My # of trades per year will likely be higher than paper trading, but if I can keep it to 1000-2000 per year, I feel I will not be overtrading. Around 40 trades per week would be around 2000 trades per year.
Some years the OOS results actually outperformed the IS results, most notably the '06-'07 period. My "realistic" expectation is to maintain a 72% APR. This would be a doubling of account size every 12 months when rolling profits back into the system.
). In addition, unpredictable fonts are a metaphor for the day-to-day unpredictability of the markets. 