Mean-ETL Optimization in HorseRace Competition

Abstract

This article explores the application of a new portfolio optimization approach called Mean-ETL (Mean-Expected Tail Loss). This method combines traditional mean return and the risk measure of expected tail loss to create more balanced portfolios that account for both returns and risks, particularly in the context of financial markets. The study demonstrates the practical implementation of Mean-ETL by applying it to the Horse Race competition, where different portfolios are constructed using three fundamental variables (CTEF, MQ, and REG10) and three distinct stock universes (GL, XUS, and EM). The competition offers a unique opportunity to assess how the Mean-ETL method performs in real-world scenarios, comparing it with other portfolio construction strategies. The article presents the results of the nine portfolios constructed for the competition, showing their performance across various metrics. The analysis provides insights into how Mean-ETL can improve portfolio construction by balancing risk and return, and offers valuable lessons for both financial analysts and investors looking to optimize their portfolio strategies in complex market conditions. The study concludes with suggestions for further improvements and applications of the approach.