How long do you forward test a discretionary system?

Hi,

After spending ages backtesting various mechanical systems I'm now in the new territory forward testing a daily discretionary system.

What I would like to know from the more experienced members is how long would you forward test a discretionary system before trusting it with real money?

Or for discretionary intra-day trading systems, how many number of trades would you want as a confirmation it's a possible valid system to risk real money on?

And do you have any cautionary war stories of how long you have forward tested a system and then as soon as you traded it live it completely bombed!!!

Many thanks
Alan

You need 400 trades to represent what the system will do in the future to 95% accuracy.

That is a good question.

I think 3 months is the least, eh?

What can adversely affect outcomes on a live system is your discipline. When you are risking real money, a whole load of emotions kick in that were never there in your demo account... greed, fear, disappointment, regret...

How did you establish these numbers?

Well, it's not just the number of trades but also the duration. Market moods change over time, so if you have a bajillion trades within a small time frame that does nothing to prove the value of a system over the larger time frame.

IMO 50 trades and 3 months of consistent profitability is a reason to go live with 50% of one's trading capital to start, then scale in the remainder after 90 days of profitable live trading. No real science to back any of that up... just a combination of advice I've heard from lots of people here.

I think you have to test out the system until you have ran into a great period (where you made lots of profits) but also in a bad period (where you lost an amount of money) if your system is positive after these two periods than you have something good in your hands.

The downside of this is that it probablly will take a few months before you have completed this cyclus.

I would not necessarily put a certain number of trades as a mark but rather a percentage win loss ratio over a certain period of time. I would feel comfortable if I reached at least a 60%win to loss ratio before considering live money. Now what I mean by this is that If I am letting my wins run and cut my losses short over a period of two or three months. Then I would use a min account trading with small number of lots and trading with pennies nickels and dimes Make sure I am always using TP and SL. Again this percentage is based on my bare minimums before I risk very little money on it initially. The other factor involved int eh psychological effect real money has on the human mind even if it is small amounts. That is a big part of the bombing effect beginners and experienced traders face when trading live with a new strategy or system. Hope this helps