Back-testing system performance vs actual performance

[bongoclub]

I am currently executing a trading system manually. The backtesting data has R:R of 1:3 and win ratio of 50%, but based on same methods my actual performance is 1:2 and 40%.

Can any experienced traders comment on how this gap between these two can be narrowed?

 

[saakk]

Hi,
1. On how much data u have back tested your system (no. of trades).
2. instrument used for back testing (example: nifty spot or future).

If sample is too small say less then 500 trades then it is of no use truly, any thing less then 500 trades is too small, future is more volatile then cash (read: nifty future) hence discrepancies.

 

[bongoclub]

Hi,
Backtesting is done on trades that I have made based on the system, around 300. So in a sense it is forward testing. Instruments are all futures stocks.

My concern is that looking back there are always mental errors such as entering too early or exiting too quickly, inconsistent position size and so on reducing the performance of live trade significantly. how do you tackle these problems?

 

[saakk]

Hi,
Backtesting is done on trades that I have made based on the system, around 300. So in a sense it is forward testing. Instruments are all futures stocks.

My concern is that looking back there are always mental errors such as entering too early or exiting too quickly, inconsistent position size and so on reducing the performance of live trade significantly. how do you tackle these problems?

If you have a system in place then it should provide you wit Entry, exit and position size. If there are premature entry and exit along wit no proper position size then u don't have a system.

If u have a system which provides you wit all the above then u simply not following it.