Combining the expectations of agency theory and earnings management theoretical models, we investigate whether publicly held companies engage in real earnings management (REM) as a result of peer effects.
We employ a sample of US listed firms over the period 1991–2018, where firm-specific and time-varying peers are identified based on product market similarities. We rely on a two-stage least squares instrumental variable approach (2SLS-IV), where the average idiosyncratic equity shock returns of peers in the previous year serve as an instrument for peers’ REM.
We find a statistically and economically significant association between firms’ use of REM and peers’ REM, consistent with both capital market and product market mechanisms. Our analysis also shows that the peer effect on these activities is mitigated when managers’ expected costs, proxied by the likelihood that specialised auditors and institutional investors would question their actions, are higher.
Our analysis is focused on public firms. Future research may verify if our findings extend to private firms, as they have different reporting incentives.
Our findings may inform policies aimed at strengthening external monitoring to mitigate peer effects in REM.
We contribute to the literature on corporate peer effects by documenting that peer influence extends to firm activities that are potentially harmful in the long term, such as real earnings management. Further, we highlight that firms do not completely ignore the costs associated with emulating suboptimal actions, though these costs reduce but do not eliminate imitation.
