Comments on: Stats… and Bill Eckhardt

By: Jez

Jez — Wed, 17 Feb 2010 10:12:13 +0000

Thanks “Correction”… ;-)

I updated the post – although I tend to take wikipedia info with a pinch of salt (I recently found out by way of experiment that you dont even need a login to it to change the pages!)
I guess the fact that he does not mention PhD on his fund website is a good enough double-check confirmation…

In any case, thanks for taking the time to point to that factual mistake – and I would tend to agree with you on the fact that NOT getting a PhD is probably a favourable factor to trading success ;-)

By: Correction

Correction — Wed, 17 Feb 2010 07:38:19 +0000

William Eckhardt does not have a PhD in Mathematics.
http://en.wikipedia.org/wiki/William_Eckhardt_(trader)

Maybe that is exacty why he is such a successful trader!

By: Jez

Jez — Thu, 11 Feb 2010 09:24:57 +0000

Thanks for the recommendation RiskCog! I’ll check them out

By: RiskCog

RiskCog — Wed, 10 Feb 2010 06:05:16 +0000

A good book for some practical statistical theory such as robust estimators and interpolation is Numerical Recipes. Old books are available online: http://www.nr.com/oldverswitcher.html

See chapters 3, 5, 14, and 15

Also, a body of theory that is rich with applications for speculators is Robust Non-Linear Control Theory.

By: Milktrader

Milktrader — Mon, 08 Feb 2010 13:02:59 +0000

He has hit it on the head. Markets do not conform to Gaussian distributions because they lack stationarity. They are also not iid (independent and identically distributed). Treating them as such only leads to trouble. I’m also impressed with the reference to Mandlebrot, who is the father of chaos theory.

Another area that may be worth pursuing is machine learning, which is the other side of the statistics coin. Inferential statistics and neural network predictions both attempt the same thing, but use different terms and techniques to arrive at their answers.