Aug

14

 For the Kelly investors among us, I found this interview with Edward Thorp very interesting. 

Ralph Vince writes: 

Thank you, Stefan.

Thorp seems to be quite the interesting guy, though I have never met or corresponded with him.

The Kelly Criterion itself was not what Kelly or Shannon thought it was - a fraction (it only appears like a fraction, and equals what is the expected asymptotic growth optimal fraction under certain conditions, ubiquitious in gambling). In trading, this fact is damning.

So even if someone IS attempting to be expected growth optimal, if they are employing the Kelly Criterion, they are NOT employing what they think they are. The solution returned by the Kelly Criterion does NOT equal the expected growth optimal fraction — it is NOT bounded at 1 on the right-hand side (and because it is often < 1, it is often mistaken as a fraction to the great peril of those employing it).

Worse-still is that the formula you can determine the expected growth optimal fraction with gives us the the entire landscape for all values of the fraction (0…1) for all axes. As I pointed out to Duncan on this list a few days ago, this allows one to craft a positioning strategy to satisfy criteria aside from the medieval "expected growth optimal."

In short, in trading, I find there is NEVER any good reason to employ the Kelly Criterion, even when one is seeking to be expected growth optimal.


Comments

WordPress database error: [Table './dailyspeculations_com_@002d_dailywordpress/wp_comments' is marked as crashed and last (automatic?) repair failed]
SELECT * FROM wp_comments WHERE comment_post_ID = '6662' AND comment_approved = '1' ORDER BY comment_date

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search