Hi

i read on internet that a good system needs to give you money (of course i would like to say !) but has to have something like 70% of win's trade...

for my part, i have almost finished my system and for the moment it multiply by 3 my starting investment in 7 months (including the fees but not the slippage) but i have only 43% of win's trade...

what i have to watch to know if my system is a good system or not ?!?

thx

Most Important: Consider Out-Of-Sample test results only.
Second Imortant: Its the ratio of Profit to Risk which counts. The two components alone are pretty meaningless.
Third Important: Average Profit per trade needs to be well above trading costs and expected Slippage.

Conclusion:
Look at Ratio(APR, expected Drawdown), Sharpe Ratio, Average Profit per trade.

Meaningless:
Results from data intervals you used to develop your strategy
APR alone
MaxDD alone
Percent winners
Average winning trade
and so on

I think, that only trend following strategies can make long-t?rm profit. For example - S&P grows 10%, you system grows 9%. S&P falls 10% - your system falls 5%.

If the system trades patterns, than somewhere in the future this patterns will brake. For example - I had 3 strategies for EurUsd, they worked for 5 last years. But this autumn, when crisis began, markets changed and all my strategies go down and they lost all capital.

But this autumn, trend-following strategies loss less than they already made, since January.

Nice to hear from you Dr. Koch. .... I am so very impressed by your WL creation, work, and inputs.

QUOTE:
Second Imortant: Its the ratio of Profit to Risk which counts. The two components alone are pretty meaningless.

I'm looking at a WLP 5.3 Backtest Performance report, which "profit" heading data would you use. Also, I do not see any heading of "Risk"; do you mean "Exposure" or some other heading, for calculating the "Profit to Risk ratio"?

If the appropriate measurements are not available, perhaps they should be added into the Performance report?

Risk can mean market exposure or percentage of time in the market (the time risk) and on the other hand Drawdown, and in this case there's a number of derivative measures like MAR ratio (CAGR/DD), Calmar ratio etc.

Another interesting technique for estimating risks is projecting drawdowns (see article by Tushar Chande in July, 2001 Active Trader Magazine ): he expresses the worst-case DD as a multiple of the standard deviation of the monthly return.