Efficient Evaluation of Multidimensional Time-Varying Density Forecasts

Arnold Polanski, Evarist Stoja

Research output: Contribution to journalArticle

Abstract

We propose two simple evaluation methods for time varying density forecasts of continuous higher dimensional random variables. Both methods are based on the probability integral transformation for unidimensional forecasts. The first method tests multinormal densities and relies on the rotation of the coordinate system. The advantage of the second method is not only its applicability to any continuous distribution but also the evaluation of the forecast accuracy in specific regions of its domain as defined by the user’s interest. We show that the latter property is particularly useful for evaluating a multidimensional generalization of the Value at Risk. In simulations and in an empirical study, we examine the performance of both tests.
Original languageEnglish
JournalInternational Journal of Forecasting
Volumex
Publication statusPublished - 2011

Fingerprint Dive into the research topics of 'Efficient Evaluation of Multidimensional Time-Varying Density Forecasts'. Together they form a unique fingerprint.

  • Cite this