Working Paper: NBER ID: w2054
Authors: Charles Engel; Anthony Rodrigues
Abstract: We propose and implement a Wald test of the international capital asset pricing model. Ex post asset returns are regressed on asset supplies. CAPM requires that the matrix of coefficients from a regression of n rates of return on n asset supply shares be proportional to the covariance matrix of the residuals from those regressions. We test this restriction in the context of a model that aggregates all outside financial assets for each of ten countries. We do not find strong support for the restrictions of CAPM.
Keywords: International Capital Asset Pricing Model; CAPM; Wald Test; Asset Returns; Asset Supplies
JEL Codes: G12; F36
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| international CAPM does not support strong evidence (P00) | weak relationship between asset supplies and returns (G19) |
| large number of restrictions imposed by the model + limited dataset (C24) | low statistical power (C46) |
| testing a subset of the restrictions (C52) | CAPM model is strongly rejected (C59) |
| expected returns do not align as predicted by the CAPM framework (G17) | weak relationship between asset supplies and returns (G19) |
| test has low power + many restrictions simultaneously complicates analysis (C90) | weak relationship between asset supplies and returns (G19) |