No it is not an academic pursuit. I have tried to put it into the simplest terms possible.
I can't see any other way to do it .. that is why I am going into so much detail in the discussions .... to draw out all of the nuances in evaluation. I am learning new things from the different opinions. Cheers. brian_z --- In [email protected], "Mubashar Virk" <mvir...@...> wrote: > > Hi Brian, > > Please understand that I intend no disrespect. I am a fan of yours (I have > almost have all of your AFLs) and other friends who are engaged in this > higher order discussion. > It is just that I feel that an ordinary trader: > > 1. needs to study the market over time > 2. have a couple of simple systems (trend lines, averages, chart or price > patterns). > 3. have a money management system (stop-loss, equity commitment & allocation > rules, etc.) > > Optimization, forward testing, oosings and aasings, and expectancies are more > of academic pursuits, and results have limited application because of the > "market is stationary" assumption. > > > > From: brian_z111 > Sent: Sunday, May 10, 2009 6:10 AM > To: [email protected] > Subject: [amibroker] Re: Expectancy - and related--specifically K-rato > > > > > > Hi Mubashar, > > I am pretty certain that you have had a better math education than moi. > > I am surprized that you didn't answer the question for the benefit of the > forum yourself. > > On a case by case basis your frequency distribution is probably not normal. > > However cumulative distribution is asymptotic and so if you reference the CD, > for any trading system, it will give you the prob of receiving any particular > value on your next trade. > > (This assumes that the market is stationary). > > It is very hard to find any firm footing, anywhere, in system evaluation but > the CD is one place were we can start to approach something akin to certainty > (as N --> oo) > > oo infinity > > In fact my .xl file on BiSim, posted at the Zboard) relies on this principle > to do a quite good job of predicting the distribution of equity outcomes for > a trading system with a non-N dist of the trades. > > --- In [email protected], "Mubashar Virk" <mvirk67@> wrote: > > > > Sorry to barge in, if I make 100 points on a system after oosing and > > aasing, what is the probability that I will make 100 points in actual real > > life trade? > > > > > > From: Howard B > > Sent: Saturday, May 09, 2009 5:45 PM > > To: [email protected] > > Subject: Re: [amibroker] Re: Expectancy - and related--specifically K-rato > > > > > > > > > > > > Hi Keith, and all -- > > > > Exactly my point. > > > > Thanks, > > Howard > > > > > > > > > > On Fri, May 8, 2009 at 10:26 PM, Keith McCombs <kmccombs@> wrote: > > > > > > > > > > I've been told that no question is a "dumb question". So here goes: > > If I have a system with good OOS performance, why should I care what the IS > > performance is? And similarly, why should I care what the OOS/IS ratio is? > > > > Couldn't it be more important that I have a high OOS/BH (buy and hold) > > ratio, so that I don't "confuse brains with a bull market"? Or at least > > something that gives me confidence that I haven't just accidentally > > stumbled on a once in a lifetime event, that, of coarse, will disappear the > > minute I start trading real money? Does OOS/IS ratio somehow help? > > -- Keith > > > > > > > > brian_z111 wrote: > > > I also create a t-test of the ave returns. > > > > How do you do that? > > > > --- In [email protected], Rajiv Arya <rajivarya87@> wrote: > > > > > > > > > I also create a t-test of the ave returns. > > > > > > The in-sample is almost always significant > > > > > > And try to have the out of sample t-test greater than 1.64, which happens > > > for about 50% for the out-of sample results. > > > > > > > > > > > > > > > > > > > > > > > > To: [email protected] > > > From: dloyer123@ > > > Date: Sat, 9 May 2009 03:03:16 +0000 > > > Subject: [amibroker] Re: Expectancy - and related--specifically K-rato > > > > > > > > > > > > > > > > > > > > > > > > --- In [email protected], Rajiv Arya <rajivarya87@> wrote: > > > > > > > > > > > > I like to compute a ratio of the out-sample metric and divide it by the > > > > in-sample metric. > > > > > > > > And I like to look for multiple runs of out-sample/in-sample ratio to > > > > be above 0.5 and with little fluctuation. > > > > > > > > > > That is similar to Pardo's WFE (Walk forward efficiency), or a measure of > > > how much curve fitting inflated test results. Pardo suggests taking the > > > concatenated out of sample returns and divide by the result treating the > > > entire combined data set as in sample. Anything below 0.65 will probably > > > not trade well live. The higher, the better. > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > __________________________________________________________ > > > Hotmail® has a new way to see what's up with your friends. > > > http://windowslive.com/Tutorial/Hotmail/WhatsNew?ocid=TXT_TAGLM_WL_HM_Tutorial_WhatsNew1_052009 > > > > > >
