Minimum volatility indices. Free lunch or risk-based anomaly?

This thesis examines the relation between MSCI Minimum Volatility indices performance and underlying risk factors during the 2001-2011 period. The majority of minimum volatility indices outperforms its capitalization weighted benchmark in terms of Carhart four-factor risk-adjusted return. However, a large share of this outperformance is attributed to a bundle of risk sources.This implies the return premium on these indices is just a compensation for risk. The findings indicate minimum volatility portfolios are biased to low-beta stocks. Also, it is shown thatconcentration risk and absence of implied protection during extreme market events partly explain the Carhart four-factor outperformance. This study adds to existing literature by examining these risk sources into detail. After adjusting for all risk factors, the outperformance of the global minimum volatility portfolio compared to its capitalization weighted benchmark turns out to be statistically insignificant.