Evaluating Asset Pricing Models in Absence of Arbitrage:Econometric Approach and Empirical Applications
We develop asset pricing tests that explicitly require that a true asset pricing model has to be arbitrage free, i.e., its stochastic discount factor has to be strictly positive. The no-arbitrage restriction is especially important in applications that involve dynamic trading strategies and derivatives. One prominent example is performance evaluation of actively managed portfolios,such as mutual funds and hedge funds. Empirical results show that the no-arbitrage restriction makes significant di.erences in asset pricing tests involving either the traditional Fama-French size and book-to-market portfolios or hedge fund style portfolios. Ignoring the no-arbitrage restriction, we reach the misleading conclusion that certain models can satisfactorily price both sets of assets.However, these models are overwhelmingly rejected by our tests because their stochastic discount factors take negative values with high probabilities. The no-arbitrage restriction also leads to better identified parameters, better behaved models, and more insights on potential sources of model misspecifications.
Stochastic Discount Factor Models Asset Pricing Tests Hansen-Jagannathan Distances Arbitrage.
Haitao Li Yuewu Xu Xiaoyan Zhang
Li is from the Stephen M. Ross School of Business, University of Michigan, Ann Arbor, MI 48109. Xu is fromthe School of Business, Fordham University, New York, NY 10017. Zhang is from the Johnson Graduate School of Management, Cornell University, Ithaca, NY 14850.
国际会议
西安
英文
2006-07-17（万方平台首次上网日期，不代表论文的发表时间）