[SIUYA]

Originally Posted by riwaseeka »

Many trades per day does your trading system generate? It’s not important how long you backtest a trading system; it’s important that you receive enough trades to make statistically valid assumptions*: If your trading system generates three trades per day, i.e. 600 trades per year, then a year of testing gives you enough data to make reliable assumptions*. But if your trading system generates only three trades per month, i.e. 36 trades per year, then you should backtest a couple of years to receive reliable data.

I completely disagree. You need more than just enough trades, you need enough trades over a valid period of time that incorporates all types of market conditions. Otherwise you are merely cherry picking.

(Given you have a number of asterisk in your post it seems more of a disclaimer )

[BlowFish]

Having said that with smaller granularity systems (smaller 'time frame' if you like) you will experience all market conditions over the course of a much shorter time. If you are sampling data every 1 minute (a 1 minute chart if you like) the system will experience most types of market conditions within a week, certainly a month. Even in a 'bull' market stuff moves up down and sideways if you 'zoom in' a bit. There is an argument that you should trade different systems (or have a system that trades differently) dependant on market conditions.

[OptionsMike]

Blowfish, SIUYA, and riwaseeka,
Thanks for your posting on this thread. What software, if any, do you use for backtesting?

OptionsMike
Safe Options Trading Income

[SIUYA]

a combination for me.....but I view backtesting as a way to test ideas more than to actually run a system on
I have used ninja, multicharts, trading blox., and my own excel system

[OptionsMike]

SIUYA,
Many thanks. At the moment I'm somewhat simialr to you in that I like to test ideas reather than actually do pure mechanical trading.

Great trading to you.

OptionsMike
Safe Options Trading Income

[afalde]

I like tradestation for programming back tests. Multi-charts also works good. If you don't know how to program code for signals, someone can do that for you, build a combination of signals, and optimize the results... example: Less Than Random ... Less Than Random

[onesmith]

Andrew, please elaborate on what the phrase "less than random" means. TIA

[BlueHorseshoe]

Originally Posted by SIUYA »

a combination for me.....but I view backtesting as a way to test ideas more than to actually run a system on
I have used ninja, multicharts, trading blox., and my own excel system

Is Trading Blox the Curtis Faith software?