Firm Complexity and Information Asymmetry: Evidence from ML-based Complexity to Measure Information Processing Costs

50 Pages Posted: 25 Mar 2024

See all articles by Brian J. Clark

Brian J. Clark

Rensselaer Polytechnic Institute (RPI)

Sai Palepu

Rensselaer Polytechnic Institute

Akhtar R. Siddique

Government of the United States of America - Office of the Comptroller of the Currency (OCC)

Date Written: March 18, 2024

Abstract

We investigate the role of machine learning (ML) model complexity in capturing the information processing costs that lead to information asymmetry in financial markets. The basic idea is that informed traders are better suited to process complex, non-linear relations between observable characteristics and future returns. As such, we propose and compute an ML-derived complexity metric to capture the magnitude of the relative advantage informed traders have over noise traders. We hypothesize that increased model complexity leads to increased information asymmetry. To this end, we show that our model complexity metric is positively associated with several well-known proxies of information asymmetry. Specifically, we find positive relations between firm complexity and future return volatility, wider bid-ask spreads and elevated probabilities of informed trading (PIN).

Keywords: Information Asymmetry, Stock Returns, ML, Model Complexity

Suggested Citation

Clark, Brian J. and Palepu, Sai and Siddique, Akhtar R., Firm Complexity and Information Asymmetry: Evidence from ML-based Complexity to Measure Information Processing Costs (March 18, 2024). Available at SSRN: https://ssrn.com/abstract=4763575 or http://dx.doi.org/10.2139/ssrn.4763575

Brian J. Clark

Rensselaer Polytechnic Institute (RPI) ( email )

Troy, NY 12180
United States

Sai Palepu (Contact Author)

Rensselaer Polytechnic Institute ( email )

110 8th St
Troy, NY 12180
United States

Akhtar R. Siddique

Government of the United States of America - Office of the Comptroller of the Currency (OCC) ( email )

400 7th Street SW
Washington, DC 20219
United States

Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Downloads
64
Abstract Views
245
Rank
648,192
PlumX Metrics