Pricing Kernel Monotonicity and Conditional Information
指出现有非参数定价核估计方法不一致，提出新估计量，利用投资者条件信息，通过模拟和S&P 500、FTSE 100期权数据表明定价核谜题源于估计方法缺陷。
A large literature finds evidence that pricing kernels nonparametrically estimated from option prices and historical returns are not monotonically decreasing in market index returns. We argue that existing estimation methods are inconsistent and propose a new nonparametric estimator of the pricing kernel that reflects the information available to investors who set asset prices. In simulations, the estimator outperforms existing techniques. Our empirical estimates using S&P 500 index option data from 1996 to 2014 and FTSE 100 index option data from 2002 to 2014 suggest that the “pricing kernel puzzle” is due to flaws in existing estimators rather than a behavioral or economic phenomenon. Received August 2, 2015; editorial decision April 15, 2017 by Editor Andrew Karolyi.