Saturday, September 27, 2014

The Kelly Capital Growth Investment Criterion: Theory and Practice

This is from Taleb's review of The Kelly Capital Growth Investment Criterion: Theory and Practice:

"There are two methods to consider in a risky strategy.

1) The first is to know all parameters about the future and engage in optimized portfolio construction, a lunacy unless one has a god-like knowledge of the future. Let us call it Markowitz-style. In order to implement a full Markowitz- style optimization, one needs to know the entire joint probability distribution of all assets for the entire future, plus the exact utility function for wealth at all future times. And without errors! (I have shown that estimation errors make the system explode.)

2) Kelly's method (or, rather, Kelly-Thorpe), developed around the same period, which requires no joint distribution or utility function. It is very robust. In practice one needs to estimate the ratio of expected profit to worst- case return-- dynamically adjusted to avoid ruin. In the case of barbell transformations, the worst case is guaranteed (leave 80% or so of your money in reserves). And model error is much, much milder under Kelly criterion. So, assuming one has the edge (as a sole central piece of information), engage in a dynamic strategy of variable betting, getting more conservative after losses ('cut your losses') and more aggressive 'with the house's money'. The entire focus is the avoidance of gambler's ruin.

The first strategy was only embraced by academic financial economists --empty suits without skin in the game -- because you can make an academic career writing BS papers with method 1 much better than with method 2. On the other hand EVERY SURVIVING speculator uses explicitly or implicitly method 2 (evidence: Ray Dalio, Paul Tudor Jones, Renaissance, even Goldman Sachs!) For the first method, think of LTCM and the banking failure.

Let me repeat. Method 2 is much, much, much more scientific in the true sense of the word, that is rigorous and applicable. Method 1 is good for 'job market papers'"
I have a copy which I haven't read yet. It's an academic volume; the layman and investor's book on Kelly is the excellent Fortune's Formula.

No comments: