GG, Its a good idea to make sure that your equity curve reflects a single position size (every trade = 1000 shares or whatever) otherwise, if you increase your position size, your equity curve will be skewed in your perception and the drawdowns will be more volitile relative to drawdowns earlier in the curve. If your equity curve with a single position size throughout STILL looks like Nasdaq 2000, then you are on your own.. (maybe you will need to determine if YOU are a "Speculative Bubble")
sorry I goofed there, the position size should be an equal dollar amount, not # of shares (unless your trading the same stock all the time)
equal percentage of account equity better than equal dollar amount... equal percentage of account equity w/ percentage adjusted up or down in line w/ total period return better still... this thread will self destruct in five seconds...we should all be killed for treason, don't you know we're giving away too much....
Its like chasing your tail Actually, my thoughts are that if you have a high frequency system that trades a lot, you can implement this approach better than if you are using a longer term strategy. I have looked into how to do this with the longer term systems and one solution has been to 'consider' each day a 'trade'. That is, each day in a long term trade you have 'decided' to be either long/short or flat. Then, use statistics as per this thread: http://www.elitetrader.com/vb/showt...3&perpage=6&highlight=reoptimize&pagenumber=1 to determine if you are statistically confident in your past returns. My solution in testing has been to see if anytime in the last X days have I had a highly confident result. If I have, I keep trading, if not, I stop and wait for the next 'very high' confidence level to resume.