It shouldn't take too long to determine if you strategy is workable. Then you need to test it in the real world. There is nothing like live trading to prove whether your system really works.
I have found a method of paper trading that works best for me as follows:
1) Develop rigid entry and exit rules for your budding new trading strategy. If possible, find a way to automate these rules to make it easy to find and record numerous "trades".
2) Review the results of your paper trading over several weeks. Be sure to include commission costs when calculating returns. Are you satisfied with the average trade profit, drawdowns, etc. Typically, if the results don't 'knock your socks off', then it won't work very well in the real world.
3) Once you've found a paper trading strategy which is successful over several weeks time and has given you confidence, it is time for some real trading. The key here is to trade a moderate size at first AND continue to paper trade the same trades as your real trades.
4) After about 10 - 20 trades, it's time for evaluation. What were your real results? More importantly, what was the difference between your average real profit (loss) per trade and your average paper profit (loss) per trade? This is where your extensive prior paper trading becomes very valuable.
Let's assume that your prior paper trading over several weeks and 100+ trades gave you an average profit of $150 per trade. Furthermore, let's assume that after 20 real trades, your average real profit is $100 per trade, while the simultaneous paper trades show an average profit on the same trades of $300 per trade ($200/trade slippage real versus paper). With these assumptions, you should be very cautious trading this system. Your prior extensive paper trading data indicates that the system is only capable of averaging $150/trade (on paper). During your real trading, however, you were fortunately to experience an unusually good time for your system, averaging a PAPER profit of $300/trade. This allows your real trades executed to show a net profit of $100/trade. However, recognizing that the system is only be expected to product a PAPER profit of $150/trade over the long run, AND the fact that you have seen an average slippage of $200/trade versus paper trading => Means that this system should be expected to LOSE $50/trade over the long run.
I realize that this is oversimplifying the situation, and statistical significance sometimes requires more data than you have available. However, this is a good way to integrate the knowledge gained in your paper trading with the results from your real trading to make intelligent conclusions early and cheaply.
I've used this strategy as an effective tool to lose as little as possible while finding strategies that work. Once identified and validated, there will be plenty of time to profit from your new system. Just don't lose the farm trying to find it...
Good luck,
-Eric |
