Corporate earnings are an important source of information for many market participants, yet managers have a certain degree of discretion over the earnings that they report. Given this discretion, there is a large literature that tries to understand whether firms manage their earnings and why. Much of this literature focuses on individual characteristics of firms, such as operating and financial characteristics, and how they affect the decision to manage earnings. In contrast, in a current working paper, we focus on the idea that a firm’s peers might play a role in that decision as well. Specifically, we investigate whether there is peer pressure in earnings management.

When we look at earnings management within peer groups, which we define as groups of firms that operate in the same industry, it quickly becomes apparent that earnings management is positively correlated within these peer groups. In other words, there are peer groups in which many firms manage their earnings, and peer groups in which few firms do so. At first glance, this seems to confirm the idea that there is peer pressure in earnings management. However, observing that firms in a peer group all manage their earnings in a similar way does not necessarily imply that they do so out of peer pressure. After all, these firms are exposed to similar market forces and tend to have similar characteristics, both of which influence the decision to manage earnings. Therefore, each individual firm might just be responding to the market forces that affect all firms in the peer group, without taking its peers’ decisions into account.

In order to truly identify peer effects, we need a shock that changes the earnings management of some firms in the peer group but leaves the earnings management of the other firms unchanged. Then, we can observe how the other, unaffected, firms in the peer group respond to the changes of their affected peers. Ideally, this shock should be unrelated to the characteristics of firms, so that we can circumvent the problems outlined in the previous paragraph. We identify two such shocks: (1) passive mutual fund flow-induced selling pressure and (2) pilot stock status in the Securities and Exchange Commission’s (SEC) Regulation SHO experiment.

The first shock refers to situations in which many mutual funds are simultaneously forced to sell a firm’s stock, which causes the stock price to tank. In response to the sudden drop in the stock price, managers at the affected firm adjust their earnings management. The second shock refers to an experiment carried out by the SEC, in which certain firms were suddenly exposed to a higher threat of short selling. In response to this increased threat of short selling, affected firms significantly reduced earnings management, due to a disciplinary effect resulting from higher monitoring by short sellers.

Using the changes in earnings management induced by these shocks, we show that other firms in the peer group – firms that were unaffected by the shocks – change their earnings management in the same direction as the firms that were affected. This result provides clean evidence that firms do take their peers’ decisions into account when they make their own earnings management decisions.

It remains unclear, however, why firms mimic their peers. It is possible that firms are simply unsure about the “right” level of earnings management and are trying to learn from their peers. Such behavior can lead to herding within peer groups. In line with this idea, we find that firms are especially sensitive to changes in earnings management of large, profitable, and close peers (geographically and in terms of product similarity). These findings support a herding story, as firms tend to follow industry leaders and peers in close proximity.

It is also possible that managers are benchmarked against the performance of their peers. In this case, a manager may decide to ramp up earnings management precisely when peer firms increase their earnings management. Indeed, we find that firms without relative performance benchmarks in their compensation contracts do not mimic their peers’ earnings management. This finding suggests that incentives – in the form of compensation concerns – are an important driver behind peer effects in earnings management.

Finally, to get a feel for the long-run implications of these peer effects, we analyze an event that is plausibly related to earnings management: financial fraud. We show that, like earnings management, fraud also clusters within peer groups. We also show that our measures of earnings management are positively related to fraud, both at the industry- and firm-level. We interpret this as suggestive evidence that peer effects in earnings management might manifest themselves in higher rates of financial fraud within peer groups.

This post comes to us from Constantin Charles at the University of Southern California’s Marshall School of Business, Felix von Meyerinck at the University of St. Gallen’s School of Finance, and Markus Schmid at the University of St. Gallen’s Swiss Institute of Banking and Finance and School of Finance. It is based on their recent article, “Peer Pressure in Corporate Earnings Management,” available here.