Discretion in Financial Reporting : The Use of Self-Constructed Peer Groups in Proxy Statement Performance Graphs

We examine how firms exercise reporting discretion when preparing the performance graph required under the SEC's new compensation disclosure rules. Results from a sample of 449 Samp;P 500 firms, 31.4% of whom self-constructed a peer group, indicate that three sets of factors -- the firm's visibility in the compensation debate, the level of monitoring, and the suitability of the public index for use in peer group comparisons -- explain both the decision to self-construct a peer group and the performance of the self-constructed peer group relative to the corresponding public industry index. We further find that self-constructed peer groups are quot;opportunisticquot; in that self-constructed indexes enhance the reporting companies' relative performance compared to performance measured relative to the company's Samp;P industry index. Paradoxically, self-constructed peer groups are also quot;informativequot; in that they explain cross-sectional variation in firm performance, firm size, and firm risk incremental to that explained by public indexes. Thus, a characterization of self-constructed peer groups as indicative of self-serving behavior on the part of managers is overly simplistic