 AB12358
|
|
Total Posts: 70 |
Joined: Apr 2014 |
|
|
I was reading through He, Litterman. The intuition behind Black-Litterman (PDF) the other day, and just saw that there's an active Kelly Sizing thread in which Markowitz mean variance is discussed.
As a non-industry type, I have no idea what's actually going on. I'm curious what the current state of best practice in terms of portfolio optimisation is, particularly with respect to beta neutral portfolios. |
|
|
|
 |
|
The best practice is ... common sense!
Well, from theoretical point of view Kelly criterion is the most consistent approach. I scrutinized it in this paper: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2259133
However, Kelly also is very sensitive to the parameter estimation errors (that are inevitable by limited sample of historical market data and ever-changing markets). This its application is also pretty hard, also possible. But first of all it is to consider as theoretical limit of which performance one can[not] achieve in long-term.
|
www.yetanotherquant.com - Knowledge rather than Hope: A Book for Retail Investors and Mathematical Finance Students |
|
 |
 phopstar
|
|
Total Posts: 10 |
Joined: Jul 2020 |
|
|
Would love to hear more from others on this. From everything I've seen, Markowitz style approaches seem to be quite hard to get right in-production / out-of-sample. Hard to beat a simple inverse variance portfolio-weighting. |
|
|
|
 |
 tomgailey
|
Banned |
Total Posts: 13 |
Joined: Oct 2020 |
|
| |
 |