A1211
Title: Dynamic dependence of equity market factors and empirical analysis of tail risk parity factor portfolio
Authors: Kakeru Ito - Nissay Asset Management (Japan) [presenting]
Naoki Makimoto - University of Tsukuba (Japan)
Abstract: Since the global financial crisis 2008, factor investing has become popular among investment managers. However, not much is known about dependence across factors. Furthermore, several papers have recently pointed out that factor returns tend to deteriorate simultaneously, leading to worse portfolio performance than expected. The dynamic dependence of Japanese equity market factors with dynamic skewed t copula is estimated, and the results with those estimated by traditional multivariate copulas are compared. The impact of capturing the dependence across the factors on the performance of the tail risk parity factor portfolio is also considered. Four important results are found. First, capturing the dynamic linear correlations enhances the performance of the risk parity factor portfolio compared with the static risk parity factor portfolio. Secondly, the tail risk parity factor portfolio tends to outperform the risk parity factor portfolio, especially Sortino ratio is improved. Third, considering the asymmetric tail dependence improves the maximum drawdown of the tail risk parity factor portfolio. Fourth, introducing the CVaR targeting strategy also improves the maximum drawdown of the tail risk parity factor portfolio. These findings indicate that managing the downside risk of the tail risk parity factor portfolio with dynamic skewed t copula is useful.
