Generalized arbitrage conditions for factor pricing models
Title: Generalized arbitrage conditions for factor pricing models
DNr: SNIC 2022/22-1244
Project Type: SNIC Small Compute
Principal Investigator: Iñaki Rodríguez Longarela <inaki.rodriguez@sbs.su.se>
Affiliation: Stockholms universitet
Duration: 2022-12-21 – 2024-01-01
Classification: 50202
Keywords:

Abstract

Most asset pricing theories can be formulated as, or approximated by, a system of moment conditions that are linear in latent model parameters and subject to linear parameter restrictions. The dual alternative to these moment conditions and parameter restrictions is a system of linear restrictions on the systematic risk profile that ensure that a portfolio of base assets is Unambiguously Undervalued. Portfolio opti- mization subject to these portfolio restrictions can be used to measure the economic significance of pricing errors. A Distributionally Robust Optimization approach can achieve robustness to estimation error in addition to robustness to the pricing model parameters. An empirical application will analyze the joint cross-section of monthly SPX options and monthly VIX options since the introduction of the latter type of options in 2006.