Relative Entropy and Minimum-Variance Pricing Kernel in Asset Pricing Model Evaluation.

摘要: Recent literature shows that many testing procedures used to evaluate asset pricing models result in spurious rejection probabilities. Model misspecification, the strong factor structure of test assets, or skewed statistics largely explain this. In this paper we use relative entropy kernels provide an alternative framework for models. Building on fact law one price guarantees existence a valid kernel, study relationship between mean-variance efficiency model’s factor-mimicking portfolio, as measured by cross-sectional generalized least squares (GLS) R 2 statistic, and determined Kullback–Leibler divergence. regard, suggest entropy-based decomposition accurately captures divergence portfolio minimum-variance kernel resulting from Hansen-Jagannathan bound. Our results show that, although GLS are strongly correlated, approach allows us explicitly decompose explanatory power model into two components, namely, corresponding its correlation with returns. This makes versatile tool designing robust tests pricing.