presentvaluenotes091911 哈佛大学博士资产定价课件（Harvard University PhD Asset Pricing Lecture Notes）.pdfVIP

Ec2723, Asset Pricing I Class Notes, Fall 2011 Present Value Relations and Stock Return Predictability John Y. Campbell1 First draft: October 20, 2003 This version: September 19, 2011 1 Department of Economics, Littauer Center, Harvard University, Cambridge MA 02138, USA. Email john_campbell@. Market e¢ ciency The general intuition of market e¢ ciency is that Önancial markets are competitive, so there should not be an easy way to proÖt by trading Önancial assets. The di¢ culty is to translate this compelling intuition into a testable hypothesis. Fama (1970) famously deÖnes a market as e¢ cient if ìprices fully reáect all avail- able informationî. Malkiel (1992, New Palgrave Dictionary of Money and Finance) usefully expands this deÖnition. ìA capital market is said to be e¢ cient if it fully and correctly reveals all available information in determining security prices. Formally, the market is said to be e¢ cient with respect to some information set, , if security prices would be una§ected by revealing that information to all participants. Moreover, e¢ ciency with respect to an information set, , implies that it is impossible to make economic proÖts by trading on the basis of .î In practice this means Ri;t+1 = it + Ui;t+1 , where it is the equilibrium return on asset i generated by some economic model, and Ui;t+1 is a fair game with respect to the information set at t. Given the economic model, market e¢ ciency is equivalent to rational expectations. The joint hypothesis problem is that market e¢ ciency is not testable except in combination with a model of expected returns. The reverse is not true, however, because models of e