Li Dai and Peng Zhou
Most of the theoretical and empirical literatures are dedicated to selecting a more empirically sound model in economics and finance. When choosing from competing alternatives, the most popular criterion is econometric goodness of fit of models against structured data (cross-sectional, time-series and panel). It is the data that determines whether a theory is valid or not, but in many scenarios, structured data are not available or not precise measures of the reality. Therefore, this paper attempts to propose a complementary criterion in model selection in terms of investment practice—we are not trying to find an econometrically “true” model, but a practically useful model. A series of models are examined by the latest data in the FOREX market. Monetary CAPM turns out to be the “best” model for practical purpose.
PDFShare this article
