Vishal Gupta

 

The Black-Litterman (BL) model is a widely used asset allocation model in the financial industry. By replacing its statistical foundations with ideas from inverse optimization, we significantly expand the scope of the model. Unlike the original model, our approach is able to incorporate investor information on volatility and market dynamics. Moreover, using our approach, we can also construct ``BL"-type estimators for general convex risk measures, thereby significantly extending beyond the mean-variance paradigm. Overall, these results provide a richer formulation.

Computationally, we study the performance of two new "BL"-type estimators constructed from our new formulation. Based upon numerical simulation and historical back-testing, we show both methods often demonstrate a better risk-reward trade-off than their traditional BL counterpart and are more robust to incorrect investor views.

This work has been submitted to Operations Research for publication.