Evaluating Value-at-Risk Forecasts: A New Set of Multivariate Backtests
38 Pages Posted: 13 Apr 2015 Last revised: 6 Dec 2015
Date Written: December 4, 2015
Abstract
We propose two new tests for detecting clustering in multivariate Value-at-Risk (VaR) forecasts. First, we consider CUSUM-tests to detect first-order instationarities in the matrix of VaR-violations. Second, we propose x2-tests for detecting cross-sectional and serial dependence in the VaR-forecasts. Moreover, we combine our new backtests with a test of unconditional coverage to yield two new backtests of multivariate conditional coverage. In all cases, a bootstrap approximation is possible, but not mandatory in terms of empirical size and power.
Keywords: Model Risk, Multivariate Backtesting, Value-at-Risk
JEL Classification: C52, C53, C58
Suggested Citation: Suggested Citation