实证资产定价Empirical Asset Pricing复习资料.pdf

Doctoral Seminar in Empirical Finance Topic 4 The Conditional (Consumption) CAPM Lecturer: Lars A. L¯chst¯er London Business School January 30, 2008 1 Main Concepts Lecture 4 Can a Conditional (C)CAPM explain the failure of the unconditional (C)CAPM? ñ Math of conditional models (Cochrane Ch. 8). The second coming of standard models: Two papers that claim the success of the CAPM and the Consumption CAPM ñ Jagannathan and Wang (1996) ñ Lettau and Ludvigson (2001b) A forceful critique of the test methodology used by these papers ñ Lewellen and Nagel (2006) Let me know of any errors, please. These notes draw on the excellent PhD course in empirical Önance I took at Berkeley, taught by Gregory R. Du§ee. Contact info: Lars L¯chst¯er, P222, London Business School, Sussex Place, Regentís Park, NW1 4SA, London, United Kingdom. E-mail: LLochstoer@london.edu 0 2 Implications of time-varying expected excess re- turns Mertonís derivation of CAPM does not hold ñ Assumption of constant investment opportunities is false What kinds of cross-sectional models can accomodate this? 1. CRRA representative agent using consumption data in SDF: consumption- CAPM 2. Log utility retains most (not all) features of CAPM 3. Mertonís Intertemporal CAPM 4. Others to be considered later. . . How do we compare these alternative models? ñ Linearize SDF and test using standard regression-based tests. ñ Then compare how well di§erent models can price interesting models Pros and cons of linearization Why linearize? ñ Easy to test l