ORCID

Jeremy Bertomeu, https://orcid.org/0000-0001-6746-5767

Language

English (en)

Publication Date

2019

Abstract

We develop a model of voluntary disclosure and production decisions and use it to establish that firms will tacitly collude by disclosing when current market demand is low and when the decision horizon is long. Low demand helps sustain tacit collusion because deviation from tacit collusion yields only a limited increase in profit when demand is low. Similarly, longer decision horizons give firms incentive to receive the benefits of collusion over a longer period. Using monthly production forecasts issued by the Big Three U.S. automobile manufacturers, we show that the frequency, horizon and accuracy of the production forecasts increases when demand decreases and when the firms focus more on long-term profit. Collectively, the evidence suggests that firms use voluntary disclosures to tacitly collude. ∗J.BertomeuisfromtheRadySchoolofManagement, UniversityofCaliforniaSanDiego, 9500GilmanDr, LaJolla, CA 92093. J. Evans and M. Feng are from the Joseph M. Katz Graduate School of Business, University of Pittsburgh Mervis Hall, Pittsburgh, PA 15260. A. Tseng is from Indiana University, 1309 E. Tenth Street, Bloomington, IN 47405. We thank workshop participants at the ARW in Zurich, the CAPANA conference, Columbia University, University of Iowa, the University of Pittsburgh, the LBS accounting symposium, the University of Texas, Austin, the 26th Annual Conference on Financial Economics and Accounting (CFEA) at Rutgers, and the University of Washington for constructive comments. This paper won the best paper award at the CAPANA conference. We also thankWard’s for insight concerning theWard’sdata sharing process, Jing Wu, Marshall Han, and Federico Damasceno for their assistance on the project, as well as Cathy Schrand, Doug Skinner and Georg Schneider (discussants) for helpful advice. 1. Introduction There is a long history of regulatory concerns with respect to strategic disclosure to facilitate tacit collusion, including anti-trust actions against information dissemination by trade associations (e.g., American Column and Lumber Co. v. United States (1921), United States v. Container Corp. of Am. (1969), FTC v. Airline Tariff Publishing Company (1992), FTC v. National Association of Music Merchants (2009)). The informational role of trade associations remains, to this date, an area of contention, especially given that tacit collusion cannot be directly observed. We follow Ivaldi, Jullien, Rey, Seabright and Tirole (2003) by distinguishing tacit from explicit collusion on the basis that explicit collusion requires direct communication between the firms and is generally illegal. In contrast, tacit collusion involves no direct communication between firms and is generally legal. In this paper, we ask whether patterns of public disclosure by trade associations can suggest tacit collusion. We develop a multi-period model of tacit collusion that predicts firms are more likely to disclose during periods of low demand and when executives face a longer employment horizon. We then empirically test these two predictions using public disclosure in the form of monthly production forecasts by the Big Three U.S. automobile manufacturers (General Motors, Ford, and Chrysler). Our empirical results are generally consistent with our predictions. To our knowledge, our paper is among the first to provide theoretical and empirical support for the notion that firms not only can, but actually do, use their public disclosures to collude. Our model examines a duopoly in which the firms engage in repeated Cournot competition. In each period, each firm observes a public and a private signal about the level of current industry demand. The public signal might represent a macroeconomic forecast that is likely to reflect demand for industry products. The private signal captures each firm’s unique information about the strength of consumer demand. Firms commit whether to disclose their private signals with the other firm prior to making their production decisions for the period. Public disclosures have two effects on product market competition. First, it gives both firms the ability to adapt their production quantities more precisely to current demand, thereby coordinating the industry toward more profitable production plans. Second, disclosure creates a potential pro- prietary cost by allowing each firm to better forecast the quantity chosen by its competitor, which increases the potential payoff from expanding production to capture a greater market share. In a one- period interaction, the proprietary cost effect dominates the coordination effect and firms are better- offnotdisclosingtheirinformation(Verrecchia1983,Gal-Or1985,Wagenhofer1990,Darrough1993). A multi-period tacit agreement preserves the coordination effect while providing a means of re- ducing proprietary costs. Coordination captures the potential benefit of using more information when making production plans for any period and will continue to hold within the tacit agreement. In contrast, increased competition to gain market share in the current period can now be disciplined by the threat of future retaliation, which takes the form of reverting to an equilibrium without tacit agreement, featuring no disclosure and lower profits. This threat discourages firms from deviat- 1

Document Type

Working Paper

Author's Department

Accounting

Author's School

Olin Business School

Download

Included in

Business Commons

Share

COinS