Black-Litterman Portfolio Allocation Stability and Financial Performance with MGARCH-M Derived Views

Sammanfattning: 2 Abstract This paper deploys methodology typically utilized in financial econometrics, namely univariate and multivariate GARCH-M forecasting techniques, as inputs into the Black-Litterman asset allocation process. While previous works have examined the usefulness in deploying select GARCH specifications as a source for the required Black-Litterman views vector, to the best of our knowledge, this is the first such work comparing the effects of select GARCH specification on asset allocation volatility. This paper draws parallels with Beach and Orlov (2007) and Duqi, Franci, and Torluccio (2014) in finding improved portfolio financial performance after the incorporation of GARCH-derived views relative to market equilibrium weighting. Financial performance is further improved with the incorporation of the multivariate DCC models. While this increase in performance is accompanied by an increase in asset allocation instability, the multivariate portfolios provide a better return-to-risk relationship for the associated degree of allocation volatility. In both the univariate and multivariate specifications the more simple GARCH(1,1) provides superior performance relative to the asymmetric GJR model.