Equity Prices And Cartel Activity
Tufts University – Department of Economics
Warren Buffett: If You Own A Good Business, Keep It
Tufts University – Department of Economics
October 1, 2015
We use a new data set to examine the equity price impact of announced cartel investigations. Unlike prior research, we estimate normal returns using the Fama-French (1993) three-factor model. We find that cartel investigation announcements have a long-lasting negative share-price effect of two percent, but one near zero for firms receiving leniency. The two percent loss is notably less than the estimated present value of profits lost due to cartel termination, implying that cartel participation is profitable. However, the results also suggest that the no-cheating stability condition for cartels may often not be satisfied, especially given the incentive to seek leniency.
Equity Prices And Cartel Activity – Introduction
The revelation by antitrust authorities that a firm is involved in collusive price-fixing can be expected to impact the firm’s net worth for at least two reasons. First, if the investigation subsequently leads to legal prosecution and conviction, the firm will likely be subject to sizable fines. Second, assuming that the antitrust action ends the collusion, the announcement of such action reveals to all market participants that the firm will no longer generate the extra profit that motivated the cartel in the first place. In addition, the revelation of cartel activity may cause customer relations problems and other issues that can also negatively impact a firm’s anticipated profit and share price.
In this paper, we use an event study methodology to isolate the impact of cartel investigation announcements on a firm’s equity value. In this respect, our work is similar to previous work such as Ellert (1976), Garbade, et al (1982), most particularly Bosch and Eckard (1991), and more recently, Günster and Van Dijk (2012) who focuses on European Union cases. Yet while these studies all investigate stock market reactions either to antitrust violations in general, or illegal price-fixing, in particular, they each rely on the equity returns predicted by a simple one-factor model, essentially the well known, Capital Asset Pricing Model (CAPM), to identify the irregular returns due to antitrust announcements. In contrast, we use the more modern three-factor model proposed by Fama and French (1992, 1993) in the long span of 1994 to 2010. In addition, we also explore how the market response to announcements that a firm is under investigation for cartel behavior differs between those cartel members who cooperate with authorities by seeking leniency and those that do not.
In brief, we find that the announcement that a firm (or its subsidiary) is under investigation for price-fixing leads to an unanticipated decline in the firm’s equity price of roughly two percent in the first few days following the revelation. While there is some subsequent recovery, over 80 percent of this decline or over 1.7 percentage points, appears to be long lasting. The dollar magnitude of these losses is greater than the subsequent fines and damage payments. At the same time however, the equity loss net of fines and damages is generally less than estimates of the discounted value of cartel profits no longer earned unless demand is relatively elastic, the precartel markup is relatively high, and the remaining cartel life is anticipated to be short. Even when these conditions are met, the profits earned during the time of successful cartel operation (eight years on average in our data) remain un-offset. In turn, this raises some question regarding the deterrent effect of anti-trust enforcement.
We provide a brief discussion of our framework in the next section. We then describe the data and our empirical analysis. A summary and concluding remarks then follow in Section 4.
2. Event-Study Framework
As noted, we use an event study framework to identify the impact of antitrust price-fixing announcements on a subject firm’s daily stock returns. This is a relatively common approach to evaluating the impact of “news” on financial markets, e.g., Brown and Warner (1985), Campbell et al (1997). The essential elements of an event study are straightforward. The first requirement is to identify the “event”. For this we use two possible definitions—the date the firm is first announced as the subject of an antitrust price-fixing investigation and, as an alternative, the date that the firm reaches a final settlement with the antitrust authorities regarding the case.
See full PDF below.