I think using a) and b) is sufficient. What extra information do you get by including c). Isn't c) derived from a) and b).
Slave: >I develop and trade intraday stock trading systems, and >the only parameter that matters is cents per share traded. >If that can exceed the commission per share, then you have >a winning system. I wish you the very best results in your system trading endeavors. JB PS. But I'm not expecting you to get them.
my first criteria is sharpe ratio. single market figures of 0.7+ are worth a second look. i do not care too much about hit ratio and pay off - it is their combination that counts. for single market patterns i am using a ttest to determine whether the distribution of trades selected by my conditions is significantly enough different from the whole universe of possible trades. a thing that makes be comfortable is a reasonable change of output when i change parameters. for example if a parameter leads to positive results making two hundred trades in ten years than the results should be better per trade when i make that parameter more severe and consequently selecting only maybe hudnred trades now. i think there is no satsfying single number or method. it is not just science to select a trading system. it takes experience. and i recommend to constantly have things on papertrading as well as in realtime. you learn much more by papertrading than by backtesting. i think it is due to a different kind of attention you put on it. peace
another thing is the equity curve. for single market strategies i prefer small but stable wins to huge trades once in a while. and for short term patterns i want to be independent from the market's overall trend. another thing is the correlation between systems. i found it very tempting to always converge towards the same concepts - maybe using copletely different stuff, but essentially trading the same inefficiency. real new research is a tough thing, but it is inevitable to balance out the portfolio of strategies in terms of diversification. this is one of the great advantages of this board: there are many people doing very different things. so there is constant flow of concepts. peace
consistently good sharpe ratio, positive equity curve, and amount risk less than 1% per trade makes for a successful trading system.
Probably not this one: Q Quantifying Edge One of the approaches is to define: Edge=WinProbabilityxWinAmount-LossProbabilityxLossAmount. UQ http://www.elitetrader.com/vb/showthread.php?s=&postid=474371#post474371
Almost everybody does, which is why a tiny minority pursue trading systems with the opposite payoff pattern, i.e. very high % of small losing trades with infrequent but very large payoffs. See the current thread "Taleb's concept of making an (infrequent) killing with options" - the points made there extend to other markets too I believe. A different but related point was made here earlier that searching for methods with eye popping stats is not necessarily a smart thing. Good point.
Turok or JB: thanks for the wish, but in regards to trading, I've found wishing to be a futile exercise. I've been profitably trading my systems since 1998 and do very well, even in the current environment. It's definitely tougher then it used to be, but I've never told anybody that trading is easy. Regards, Slave2Market