Does anything beat parametric portfolio policies?

Abstract

This thesis’ objective is to test the parametric portfolio policies (PPP) approach to asset allocation developed by Brandt, Santa-Clara and Valkanov (2009) on an investment universe of large stocks. I enlarge the number of conditional variables to include volatility and tail risk alongside value, size and momentum. I introduce a novel approach by using industry specific standardization when normalizing the characteristics. I also model the stocks for both the unconstrained and the long-only portfolio of stocks. Using a power utility function as representative of the investor’s preferences I test this approach using the Standard & Poor’s 500 as a market proxy. I include a sensibility analysis to different risk aversion coefficients. I conclude that despite the overall good performance of this strategy it should not be seen as a way to hedge the market exposure, but as a way to ’ride’ the market with high risk adjusted returns. I find that an investor always prefers small stocks and past winners. The preference between value and growth stocks depends on the models specifications.

Date of Award	27 Oct 2016
Original language	English
Awarding Institution	Universidade Católica Portuguesa
Supervisor	Bruno Gerard (Supervisor)