Working Papers

  • Selection Bias and the Cross-Section of Expected Returns

    - with Tom Zimmermann

    To be presented at AFA 2018

    Paper (ssrn)  [March 2017], Cite (Bibtex)

    Data mining bias is surprisingly small, accounting for a modest 10-15% of the typical in-sample return.  This small bias implies that the decline in anomaly returns post-sample is due to traders eliminating mispricing.  Our bias adjustment is robust and works in a wide variety of simulations of publication bias. 

  • A Likelihood-Based Comparison of Macro Asset Pricing Models

    - with Rebecca Wasyk and Fabian Winkler

  • Paper (ssrn)  [April 2017], Cite (Bibtex)

    We Bayesian-estimate a model where price/dividends depend on long run risks, habit, and a persistent residual.  The residual is dominant, showing that something other than long run risks and habit is the primary driver of market volatility. 

  • Semi-Parametric Restrictions on Production-Based Asset Pricing Models

    Paper (ssrn)  [April 2016]  Slides

    Matching the data on asset prices requires either extremely volatile IST shocks or huge capital adjustment costs.  Those parameters imply a very low EIS.  These restrictions apply regardless of many other details of the model. 

  • A General Equilibrium Model of the Value Premium with Time-Varying Risk Premia

    Paper (ssrn)  [July 2017]  Slides Code

    This paper presents a quantitative GE model of the value premium and equity premium.  Value firms
    (1) have higher cash flow growth and thus (2) are more exposed to consumption volatility fluctuations.  Empirical evidence supports (1) and (2).