Thanks southall, The problem I have notice with my friends is when they start trading a system they may some experience losses. Then they say "this system does not work cause I am losing money now, how can I make money every month, if I just lost $1000 or so" They think of it as "this trading needs to produce money for me every month" I believe it will be too stressful asking friends or family for money to trade, cause they will ask "so how much money you made so far". Then I may have to say " the trading system is losing money right now cause its it's drawdown" They may not understand there may be months of negative return. They want money every month. So it's learning curve teaching when dealing with other people's money. I don't want that headache.
Steer well clear of people who expect profits every month. You need to show realistic back test results, how the profit has varied every month in the past. And a few months of real forward trading as well. Having said all this, you need to be really confident you have a good system yourself. Dont risk other peoples money if you are not sure you can actually win in the long run.
Thanks southall, Good comments your written. I am nowhere near risking my money or anyone else market at my stage in my trading career.
Backtest, backtest, backtest and then forward test....at least 200 trades in the backtest and another 100 minimum going forward
Thank you Bellwether1 for your response. Very good comment. Is there a specific number of trades to back test and forward test for an intraday trading system? In other words , does 200 trades of backtest qualify for an intraday system that have about 2 to 3 setups per day? Most people suggest back test for 10 years. Thank you
I think 10 years is worth of data is irrelevant, think about how many times the market changes, think about how the Russell changed from $10/tick to $5/tick, too many variables to account for...I personally look no further than 2 yrs back and I use a 200 trade minimum, maybe in your case you might want to include more. The point should be that you're looking to learn the "personality" of the system: Avg $/trade, profit factor, % winners vs % losers vs %BE, Expectancy, Statistical Expectancy, Avg Win and Avg Loss. Also I do not auto-backtest, doing it manually lets me feel what its like to actually trade the system and sort of experience the ups and downs. I hope this was helpful for you.
Bellwether1 thanks for response, Good response. I also believe 10 years of data is irrelevant "dependent" on type of trading system. If its an intra-day (2-4 trades per day) trading system, no way someone can manual back test for this amount of years. For this case, (and I could be wrong) one year of testing is decent, IMO. I say this because with an intra-day system (rather mechanical, auto, or discretionary, etc.) in one year that system should experience near every market condition that could happen. Logically, I would think so. But if can auto back test, sure back test it 10 years, BUT from what I am study, its best to make sure that back testing simulation software is accurate as practical. Your was very helpful. I agree with you, manually back testing for about 100-200 trades puts you in rhythm with the system and gets all the gotchas or "I never thought about that". But I tell you one thing about manual back testing, better make sure all variables and system design is clearly defined. Cause I half-ass (or just poorly kept records) back test for nearly 6 month and realize "oh crap, I didn't think about this condition". LOL