This book analyzes the verification of empirical asset pricing models when returns of securities are projected onto a set of presumed (or observed) factors. Particular emphasis is placed on the verification of essential factors and features for asset returns through model search approaches, in which non-diversifiability and statistical inferences are considered. The discussion reemphasizes the necessity of maintaining a dichotomy between the nondiversifiable pricing kernels and the individual components of stock returns when empirical asset pricing models are of interest. In particular, the model search approach (with this dichotomy emphasized) for empirical model selection of asset pricing is applied to discover the pricing kernels of asset returns.
Part I Asset Pricing Models: Discussions and Statistical Inferences
1. Asset Pricing Models: Specification, Data and Theoretical Foundation
2. Statistical Inferences with Specification Tests
3. Statistical Inferences with Model Selection Criteria
Part II The Alternative Methodology
4. Finding Essential Variables in Empirical Asset Pricing Models
5. Hypothesis Testing with Model Search