Fully Flexible Extreme Views

Abstract

We extend the Fully Flexible Views generalization of the Black-Litterman approach to effectively handle extreme views on the tails of a distribution.

First, we provide a recursive algorithm to process views on the conditional value at risk, which cannot be handled directly by the original implementation of Fully Flexible Views.

Second, we represent both the prior and the posterior distribution on a grid, instead of by means of Monte Carlo scenarios: this way it becomes possible to cover parsimoniously even the far tails of the underlying distribution. Documented code is available for download.

Keywords: Entropy Pooling, Kullback-Leibler, Black-Litterman, VaR, CVaR, grid-probability pair, Monte Carlo, Gauss-Hermite polynomials, Newton-Raphson, kernel estimator

JEL Classification: C1, G11