Out-of-sample comparison of copula specifications in multivariate density forecasts

摘要: We introduce a statistical test for comparing the predictive accuracy of competing copula specifications in multivariate density forecasts, based on Kullback-Leibler Information Criterion (KLIC). The is valid under general conditions: particular it allows parameter estimation uncertainty and copulas to be nested or non-nested. Monte Carlo simulations demonstrate that proposed has satisfactory size power properties finite samples. Applying daily exchange rate returns several major currencies against US dollar we find Student's t favored over Gaussian, Gumbel Clayton copulas. This suggests these are characterized by symmetric tail dependence.