Title: Smart stochastic discount factors
Authors: Alberto Quaini - University of Geneva (Switzerland) [presenting]
Fabio Trojani - USI (Switzerland)
Sofonias Alemu Korsaye - University of Geneva (Ethiopia)
Abstract: A unifying theoretical framework is developed for selecting model-free Stochastic Discount Factors (SDFs) in arbitrage-free markets under general convex constraints on pricing errors, and show that such SDFs arise in a wide range of economies featuring, e.g., various forms of frictions, ambiguous asset payoffs, asymptotic no-arbitrage conditions under Ross' Arbitrage Pricing Theory (APT), or a need for regularization in large asset markets. We introduce a new family of minimum variance SDFs incorporating APT pricing error bounds, which are designed to optimize the tradeoff between pricing accuracy and the SDF ability to comove with systematic asset return risks. Empirically, we find that a model-free adaptation of an SDF under the CAPM, which exactly prices market risk but otherwise constrains the amount of mispricing across assets with a model-free APT pricing error bound, generates an optimal tradeoff.