Demand-System Asset Pricing: Theoretical Foundations

Département: Finance

Intervenant: Daniel Neuhann (U. Texas)

Salle: TBD

Abstract

Recent approaches to asset pricing use structural methods to estimate investorlevel demand functions for financial assets. We show that cross-asset complementarities and price spillovers can significantly bias these estimates: if close substitutes exist, measured elasticities are near one even if true elasticities are near infinite. This reconciles low demand-system elasticities with higher theoretical benchmarks. Biases are smaller for less substitutable assets, such as broad portfolios or asset classes. Control variables lead to estimates of residual demand elasticities which may offer limited information about asset-level demand. We caution against interpreting estimated demand elasticities as structural parameters which remain stable under counterfactuals.