Estimating Demand Systems with Bidding Data
Working Paper 34774
DOI 10.3386/w34774
Issue Date
Revision Date
We introduce a framework for estimating demand systems across multiple assets using bidding data, such as price-quantity limit orders. Our approach does not require price instruments, which are difficult to obtain, and allows for fully flexible substitution patterns without imposing logit-style restrictions. We establish identification for both price-taking and strategic investors, describe the data requirements for implementation, and illustrate the framework using data from Canadian Treasury bill auctions. We find that dealer demand is elastic and that bills of different maturities act as weak substitutes. We demonstrate how recovering demand allows policymakers to isolate cross-asset spillovers from strategic bid shading.
-
-
Copy CitationJason Allen, Jakub Kastl, and Milena Wittwer, "Estimating Demand Systems with Bidding Data," NBER Working Paper 34774 (2026), https://doi.org/10.3386/w34774.Download Citation
-