Misspecified Recovery

Working Paper: NBER ID: w20209

Authors: Durvodyrurylnd; Lars P. Hansen; Jos A. Scheinkman

Abstract: Asset prices contain information about the probability distribution of future states and the stochastic discounting of these states. Without additional assumptions, probabilities and stochastic discounting cannot be separately identified. Ross (2013) introduced a set of assumptions that restrict the dynamics of the stochastic discount factor in a way that allows for the recovery of the underlying probabilities. We use decomposition results for stochastic discount factors from Hansen and Scheinkman (2009) to explain when this procedure leads to misspecified recovery. We also argue that the empirical evidence on asset prices indicates that the recovered measure would differ substantially from the actual probability distribution and that interpreting this measure as the true probability distribution may severely bias our inference about risk premia, investors' aversion to risk, and the welfare cost of economic fluctuations.

Keywords: Stochastic Discount Factor; Asset Pricing; Probability Distribution

JEL Codes: D84; G12; G10


Causal Claims Network Graph

Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.


Causal Claims

CauseEffect
Asset prices reflect both stochastic discount factors and transition probabilities (G19)Without additional data or assumptions, these cannot be nonparametrically identified (C14)
The recovered measure of probabilities differs from the actual distribution (C46)Significant biases in interpreting these recovered measures as true probabilities (C83)
Interpreting these measures incorrectly (C20)Biased inferences about risk premia and welfare costs associated with economic fluctuations (D11)
Assumptions about ergodicity and martingale components (C22)Identification process is complicated (F55)
Certain assumptions might lead to identification (C50)Actual probability measures recovered may not correspond to the true underlying distributions (C46)

Back to index