Firm-Level Input Price Changes and Their Effects: A Deep Learning Approach
August 19, 2025
Summary
The authors of this working paper propose firm-level measures of input and output price changes that they constructed using textual analysis of earnings calls. Their measures cover a broad cross-section of US firms, and they establish several properties of input price changes and their effects.
View PaperWorking Paper 2025-7
Abstract: We develop firm-level measures of input and output price changes using textual analysis of earnings calls. We establish five facts: (1) Input prices increase (decrease) at the median firm once every seven (30) months. (2) Input price changes contain an equal blend of aggregate and firm-specific components. (3) A firm's stock price experiences a –1.15 percent return when our input price change measure is in the top tercile of price increases. (4) Our input price change measure predicts future changes in the cost of goods sold. (5) Firms pass through input price changes to output prices in the same quarter with a magnitude of 0.7.
JEL classification: D24, E12, E44, L11
Key words: deep learning, input price, cost pass-through
Digital Object Identifier (DOI): https://doi.org/10.29338/wp2025-07
A version of this paper was previously titled "Measuring Firm-Level Inflation Exposure: A Deep Learning Approach." The authors thank Viral Acharya, Will Cong, Anthony Diercks, Jan Eeckhout, Norman Guo, Gerard Hoberg, Alex Hsu, Wei Huang, Pete Klenow, Da Li, Jiacui Li, Alejandro Lopez-Lira, Ivan Shaliastovich, Erik Sirri, Wenting Song, Luke Stein, Jerome Taillard, Jincheng Tong, Michael Weber, Michael Weisbach, Nancy Xu, Baozhong Yang, Mao Ye, Tong Yu, Miao Ben Zhang, and participants at ABFER (scheduled), AFA, CICF, CFEA, CETAFE, FMA, MFA, SITE, Swedish House of Finance Conference, UF Conference, UT Dallas Conference, Babson, Florida State, Georgia Tech, Texas A&M, and Miami for comments. The views expressed here are those of the authors and not necessarily those of the Federal Reserve Bank of Atlanta or the Federal Reserve System. Any remaining errors are the authors' responsibility.
Sudheer Chava (chava@gatech.edu) is with the Scheller College of Business, Georgia Institute of Technology. Wendi Du (wendi.du@sc.edu) is with the Darla Moore School of Business, University of South Carolina. Indrajit Mitra (indrajit.mitra@atl.frb.org) is with the Research Department, Federal Reserve Bank of Atlanta. Agam Shah (ashah482@gatech.edu) is with the College of Computing, Georgia Institute of Technology. Linghang Zeng (lzeng@babson.edu) is with Babson College.
To receive e-mail notifications about new papers, subscribe. Under "Publications" select "Working Papers."