Peer comparison and management forecast behavior

ABSTRACT

This study examines whether incentive contracts that compares managers’ compensation to that of stronger peers can influence voluntary disclosure behaviour and improve transparency. We obtain information regarding peer-based compensation arrangements from US public firms and find that managers whose compensation is benchmarked against peers who, on average, have higher managerial abilities, tend to issue management earnings forecasts more frequently. Furthermore, the peer comparison incentive effects are more pronounced when firms depend more on external financing, face greater product market threats, and have more endowed growth opportunities. Overall, our study highlights the effects of pressure arising from comparison with higher-quality peers on firms’ information environment. We contribute to the literature on tournament-based implicit incentives, peer effects, managerial ability, and economic consequences in the context of voluntary disclosures.

KEYWORDS:

• Peer comparison
• voluntary disclosures
• management earnings forecasts
• executive compensation
• tournament incentive

JEL CLASSIFICATION:

• M40
• M41
• D83

Acknowledgment

The authors thank the Editor and an anonymous reviewer for their insights and constructive suggestions for improving the paper. We also have benefited from the useful comments of Jae B Kim and participants from AAA 2018. All errors are our own.

Disclosure statement

The authors declare that they have no known competing financial interests or personal relationships that could have influenced the work reported in this paper. This work has not been published previously (except in the form of an abstract, a published lecture, or an academic thesis) and is not under consideration for publication elsewhere.

Data availability statement

All data are publicly available from sources identified in the text.

Notes

1. They introduce the RPQ index that measures the fraction of peers with higher managerial ability scores (Demerjian et al., 2012) compared to a focal firm’s CEO. We follow their process to construct the variable, as explained in greater detail in the Research Methodology section.

2. As a general term, peers may be used for different meanings in different contexts. In some studies, such as Seo (2021), ‘peers’ indicates firms in an industry (or a product market). Others (e.g. Albuquerque et al., 2013; Francis et al., 2016) use the term to indicate firms that share several economic characteristics, rendering them comparable with a focal firm (and its manager) concerning compensation or performance evaluation. Following this line of literature, ‘peers’ in our study refers to those firms (or their managers) named in compensation benchmarks.

3. The SEC’s proxy statement mandate in 2006 (SEC, 2006) requires all public firms to disclose the peer group(s) they use for compensation and relative performance evaluation purposes (‘compensation peers’) in the Compensation Discussion and Analysis (CD&A) section of the proxy statements (DEF 14F). Our sample begins in 2007 because many firms reported incomplete information about their compensation peers in their 2006 proxy statements.

4. Faulkender and Yang (2010, p. 258) note that compensation peers are likely ‘in the same industry, of similar size, and with a history of observed talent flows between them.’ The selection criteria indicate that peer managers are direct competitors against whom a manager’s performance and ability are compared in a labour market or its narrower segment.

5. In their study, the peer pay effect is measured as the difference between the pay of the firm-selected peers and the economic characteristics-matched (specifically, propensity-score-matched) peers.

6. Peer selection, however, may be biased, particularly under weak corporate governance where managers can influence the board’s compensation decision process (i.e. an entrenched board or compensation committee). Prior studies have corroborated this reasonable doubt, finding evidence that firms include on their list of compensation benchmarks several peers with few similar economic characteristics but whose CEO compensation level is higher than that of their own CEOs (Bizjak et al., 2011; Faulkender & Yang, 2010). Such practice has been criticised for its potential to justify increasing compensation that is not justifiable solely in economic terms (e.Morgenson, 2006).

7. The literature refers to the type of incentive arising from the existence of competition in a labour market as ‘external tournament’ (e.g. Francis et al., 2016) or ‘industry tournament’ (e.g. Coles et al., 2018).

8. They also suggest and find evidence that managers learn from their peers to improve their firm outcomes, such as investment decisions and corporate financial policies and structure.

9. Please note that we do not limit the source of incentives to the prize of being the tournament winner (i.e. winning the job with the highest compensation). For example, we consider retaining a current job (i.e. defeating other labour market candidates) as a positive outcome in labour market competition. Thus, like prior studies, we do not make a strict distinction between the tournament-based incentive and the incentive from labour market concerns (e.g. Chowdhury et al., 2020; J. Huang et al., 2019; H. Zhao, 2018).

10. We acknowledge another line of literature that illustrates competition from potential new entrants and predicts the opposite relation, namely, the entry-deterrent hypothesis. This view concludes that firms would make more disclosures to deter new entries into the (profitable) industry (Darrough & Stoughton, 1990). Consistent with this view, Bamber and Cheon (1998) show that MFs are less specific in firms operating in more concentrated (thus, less competitive) industries. Ali et al. (2014) find that firms in more concentrated (less competitive) industries issue fewer MFs. There also exists empirical evidence that reconciles the hypotheses generating contradictory predictions. For example, Chen et al. (2022) report that the frequency of MFs increases under more intense competition. Still, point estimates in MFs are less likely and argue that their findings suggest that MF disclosures are determined jointly by capital market benefits and proprietary cost concerns.

11. We use product market threats to capture the source of competition (Hoberg et al., 2014). The measure indicates how responsive product rivals are in modifying their product strategies and specifications following a focal firm’s product changes.

12. Demerjian et al. (2012) introduce this managerial ability measure. In particular, they explore the data envelope (‘DEA’) technique to measure the efficiency in firms’ utilisation of corporate resources in producing accounting performance. The firm efficiency is regressed on firm characteristics, and managerial ability is any residual from the estimation (i.e. the efficiency not explained by firm characteristics).

13. We obtain the data from http://hobergphillips.tuck.dartmouth.edu/. Hoberg et al. (2014) develop the new measure using firms’ product descriptions provided in their 10-K filings. Specifically, they compute a cosine similarity between a firm’s word use and the aggregated vocabulary used by its product market rivals to indicate the extent to which a firm reacts to the changes in its peers’ products (more specifically, their descriptions).

14. We follow Murphy (2003), classifying new economy firms based on the following primary 4-digit SIC codes: 3570 (Computer and Office Equipment), 3571 (Electronic Computers), 3572 (Computer Storage Devices), 3576 (Computer Communication Equipment), 3577 (Computer Peripheral Equipment), 3661 (Telephone & Telegraph Apparatus), 3674 (Semiconductor and Related Devices), 4812 (Wireless Telecommunication), 4813 (Telecommunication), 5045 (Computers and Software Wholesalers), 5961 (Electronic Mail-Order Houses), 7370 (Computer Programming, Data Processing), 7371 (Computer Programming Service), 7372 (Prepackaged Software), and 7373 (Computer Integrated Systems Design).

15. A higher value of the Abs_Forecast Error variable suggests a less accurate forecast.

16. Among several potential causes of endogeneity, we are particularly concerned about simultaneity and correlated omitted variables because an unobservable firm factor can affect peer selection and voluntary disclosures simultaneously.

17. In untabulated results, we estimate a probit model of a dichotomous dependent variable for firms with high RPQ. Using the first-stage estimation result, we construct a matched sample of high and low groups of RPQ, as in Rosenbaum and Rubin (1983).

18. We repeat the regressions employing RPQ2 and High RPQ to find that the use of these alternative measures does not alter the implication drawn from RPQ (untabulated).