Financial econometrics - a new discipline with new methods

摘要

Financial econometrics is simply the application of econometric tools to financial data. For many years, least-squares techniques provided satisfactory tools. Stock market forecasts, efficient market tests, and even tests of portfolio models such asthe CAPM and APT were essentially implemented with least squares on cleverly manipulated data sets. More recently, however, the field has developed its individual character as new statistical tools have been invented to analyze new questions. In this short overview, I would like to suggest a framework that includes much of the recent literature and important tools of financial econometrics. Let P_t be a vector of asset prices observed at time t, and let F_t be the information set known to the econometrician at time t which automatically must include these prices. Corresponding to the price change and dividend payment from t to t + k is a return vector R_(t+k,t). A central concern of financial econometrics is to discover the joint conditional density f(P_(t+k)/F_t). Estimates of the conditional mean, #mu#_t + k/t = E_t(P_(t+k)) of this density were used to test the efficient markets hypothesis, which, in its simplest form, supposed that expected excess returns should be zero.