Hi Steve, I know it's been a while since you wrote your reply/comment below, but you really touched upon something essential and I wanted to thank you for pointing out what you are pointing out: there must be an overall edge to the strategy, a positive expectancy. Aah, the elusive edge...;-)
I can statistically prove that "keeping your losses small and your winners big" is in spite of all the repetitions of it as a mantra, NOT superior to the opposite "Keep your losers and sell for quick gains". BOTH are breakeven strategies assuming ultimate randomness. The proof I can go into separately, but it simply involves tossing a coin, say, 4 times and having a reward of +1 on heads and -1 on tails. Then a mechanical strategy of: "stay in while heads come up, get out on first tail" or respectively: "get out first time heads come up, stay while (losing) heads come up" are the equivalents of the two stock strategies above. Both, of course, result in break- even. The difference is, as you point out, that with stop losses, say 4x the sizes of winners one needs to be right over 80%. The dificulty of applying statistics is that with larger stop lossess one WILL have more winners. The size of the stop and the percentage of winners are, unfortunately, very correlated. So, it boils down to (as you point out) a) psychological predisposition (do you prefer to be RIGHT more often, albeit with small gains, or do you prefer to be rarely right, albeit with huge gains), and b) can you come up with an empirical expectation for whatever strategy you are using. My problem #1 is: I do not have a good handle how to come up with the expectancy. My question to you: how did you come up with the, indeed, very attractive expectancy numbers you have for your strategy? Is it based on actual experience? Or, can you also theoretically show it to apply? My problem #2: psychology. I have had diffulties with sticking to either strategies, always "adapting" (that's a too kind of a word) them in the middle and in the end losing money. In either case, I would love it if you could share your E2 strategy with me and further comments. I am in my 6th year (and several $100k in the hole), looking for the elusive "succesful trader/investor/mentor"... Cheers, André > --- In [email protected], "scourt2000" <[EMAIL PROTECTED]> wrote: > > > > I wanted ("buy at MACD crossing, sell at 3% profit or 10% loss"), > > This is a troubling statement to me. Why would anyone in this world > want to take over 3 times the loss vs. a win? Your win rate has to > (realistically, before transaction fees) maintain an 80% rate in > order for you to have a winning system. > > It's NOT about how often you win. It's about this: > > Expectancy = (prob. of a winner) * (avg winner) - (prob. of a loser) > * (avg loser) > > If the expectancy is not a positive value, you will lose at this > game. Look for systems that give you a 2:1 plus advantage of your > winners over your losers. The higher that ratio, the more losses > you can take. The trick is to take those small losses relative to > the big gains in order to finish profitable after a long series of > trades. > > For example, I have a an ER2 swing system that wins, on average, > around 40% of the time, but the avg winning trades maintain a 3:1 > advantage over the avg losing trade. Believe me, it's no fun > emotionally watching a system lose multiple times in a row, but when > you have this kind of expectancy: > > Expectancy = (0.40 x 3) - (0.6 * 1) = 0.60 > > [here, "3" represents $3 for an avg. winner and "1" represents $1 > for a avg. loser] > > It means, in relative terms, that you can expect to see 60 cents > returned for every dollar invested. > > I would never, ever, EVER be willing to lose more than I intend to > gain. It is a rookie mistake in this business to focus on a high > winning percentage as a primary factor in being successful. You > should focus on taking good risk/reward opportunities and be willing > to accept losing (a lot of the time) in order to win overall. > > I realize this is an Amibroker thread, but please, remember that it > is only a tool and if no one ever tried to step in and get you to > think beyond the mechanics of using this tool, it would be a greater > disservice than saying nothing at all. > > Steve > Please note that this group is for discussion between users only. To get support from AmiBroker please send an e-mail directly to SUPPORT {at} amibroker.com For other support material please check also: http://www.amibroker.com/support.html Yahoo! Groups Links <*> To visit your group on the web, go to: http://groups.yahoo.com/group/amibroker/ <*> To unsubscribe from this group, send an email to: [EMAIL PROTECTED] <*> Your use of Yahoo! Groups is subject to: http://docs.yahoo.com/info/terms/
