[]

Last edited by Omar on August 8th, 2003, 10:00 pm, edited 1 time in total.

FishSauce,http://satyrican.tripod.com is a bunch of crap. It gives a bad name to TA.kr, Nonius - Keep trying to solve problems on a Stiefel manifold....N

Omar,You seem like a great person but you need to polish your math skills a bit before commenting.Regards,Newton

QuoteOriginally posted by: NOmar,You seem like a great person but you need to polish your math skills a bit before commenting.Regards,Newtonyeah Omar, didn't you know that a covariance matrix isn't symmetric?

QuoteOriginally posted by: Nonius<bryeah Omar, didn't you know that a covariance matrix isn't symmetric?Of course not. It's a Stiefel manifold.

Of course not. It's a Stiefel manifold.Sh!t, elan. Stop giving hints to Omar.

Last edited by N on August 7th, 2003, 10:00 pm, edited 1 time in total.

QuoteOriginally posted by: NOf course not. It's a Stiefel manifold.Sh!t, elan. Stop giving hints to Omar.Dude, just trying to keep people honest.

elan,So what type of map would you use to get this stiefel sucker into something Nonius can work with?N

QuoteOriginally posted by: Nelan,So what type of map would you use to get this stiefel sucker into something Nonius can work with?NGoing out for a session with my personal trainer. Maybe he has a good idea.

Out of curiosity: How many of you actually read the article? Be honest ...The article was not meant (I think) to be mathematically-challenging. Is it after all a financial newsletter and not a Quant newletter like wilmotts'.I just think it's interesting that he or she use simulation to test the profitabilty of technical analysis under a pure random walk state.I have been messing around with the excel file and the results are similar to the articles.For those who DIDNt read it: the thesis was that random walk proneness to trends might explain why TA works.(Which is logically considering the notion of spurious correlation due to trends). As the writer said "rudimentary" TA was used for simplicty sake.Just my two cents.

Last edited by FishSauce on August 7th, 2003, 10:00 pm, edited 1 time in total.

Here's a little financial maths quiz:The price X(t) of a tradeable asset follows a random walk such that at time t the expected value at time T E[X(T)|X(t)] = X(t) for all T > t.There exists a tradeable riskless asset whose price is B for all times t > 0.Show that the expected profit from any trading strategy involving only the asset and the riskless asset is zero.

Last edited by Johnny on August 7th, 2003, 10:00 pm, edited 1 time in total.

fishsauce - I read it twice. It's still bs.Johnny - Can't be shown, unless you show it to math midgets.

- matthewcroberts
**Posts:**295**Joined:**

Quotefishsauce - I read it twice. It's still bsI'm going w/ N on this one. Holy cow, that's entertainment!Fish, if you don't see the problems here, let me give you a few places to start:1) transactions costs2) statistical significance

"fishsauce - I read it twice. It's still bs."Newton I'll go with you on this one ..."Johnny - Can't be shown, unless you show it to math midgets."... but not this one. Of course it can be shown. And that's why the paper is bs.

"Johnny - Can't be shown, unless you show it to math midgets."... but not this one. Of course it can be shown. And that's why the paper is bs.Johnny, Johnny - you were going so good there for a while.I'm afraid it can't be shown. It's only true for Wiener processes. Wiener processes can't have fat tails, ever, period, end of story.

Last edited by N on August 7th, 2003, 10:00 pm, edited 1 time in total.

GZIP: On