Skip to main content
Top
Published in: Review of Quantitative Finance and Accounting 3/2021

09-08-2020 | Original Research

Earnings management surrounding forced CEO turnover: evidence from the U.S. property-casualty insurance industry

Authors: Jiang Cheng, J. David Cummins, Tzuting Lin

Published in: Review of Quantitative Finance and Accounting | Issue 3/2021

Log in

Activate our intelligent search to find suitable subject content or patents.

search-config
loading …

Abstract

In this paper, we investigate earnings management surrounding forced CEO turnover for U.S. property-casualty insurance companies with differing organizational forms. We analyze the three principal organizational form types in the industry—publicly-traded stocks, closely-held stocks, and mutuals. We utilize a unique measure of earnings management, the loss reserve error. Multivariate results show that all ownership types over-state earnings during our sample period whether or not forced turnover occurs. Over-statement is highest for publicly-traded stocks, followed by closely-held stocks and mutuals. Organizational form matters in constraining managerial opportunism in the presence of forced turnovers. Incumbent CEOs of publicly-traded stocks manage earnings upward prior to forced turnovers, consistent with the cover-up hypothesis, but this hypothesis is not consistently supported for mutuals or closely-held stocks. The univariate results support the big-bath hypothesis for closely-held stocks, but the multivariate results do not support the big-bath hypothesis for any organizational form. Finally, corporate governance matters—high board independence and large board sizes are associated with less income over-statement.

Dont have a licence yet? Then find out more about our products and how to get one now:

Springer Professional "Wirtschaft+Technik"

Online-Abonnement

Mit Springer Professional "Wirtschaft+Technik" erhalten Sie Zugriff auf:

  • über 102.000 Bücher
  • über 537 Zeitschriften

aus folgenden Fachgebieten:

  • Automobil + Motoren
  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Elektrotechnik + Elektronik
  • Energie + Nachhaltigkeit
  • Finance + Banking
  • Management + Führung
  • Marketing + Vertrieb
  • Maschinenbau + Werkstoffe
  • Versicherung + Risiko

Jetzt Wissensvorsprung sichern!

Springer Professional "Wirtschaft"

Online-Abonnement

Mit Springer Professional "Wirtschaft" erhalten Sie Zugriff auf:

  • über 67.000 Bücher
  • über 340 Zeitschriften

aus folgenden Fachgebieten:

  • Bauwesen + Immobilien
  • Business IT + Informatik
  • Finance + Banking
  • Management + Führung
  • Marketing + Vertrieb
  • Versicherung + Risiko




Jetzt Wissensvorsprung sichern!

Footnotes
1
CEO turnover usually is classified as two types, routine turnover and nonroutine or forced turnover, based on whether the incumbent CEO voluntarily leaves the firm or is forced out. Incumbent CEOs in routine turnovers might over-state firm performance to maximize performance-based compensation, e.g., profit-based bonus and retirement plans, during their final year(s) at the firm (Murphy and Zimmerman 1993; Kalyta 2009). This so-called horizon hypothesis is beyond the scope of this research.
 
2
Another example of factors that could motivate managers to report low earnings, similar to “big-bath” behavior, is managerial compensation contracts. Eckles et al. (2011) find that insurance managers receiving larger bonuses and stock rewards are more likely to manage earnings downward at the contemporary year.
 
3
For example, loss reserves accounted for 54.4% of total P–C liabilities in 2015 (A.M. Best Company 2016).
 
4
Loss reserves have been extensively applied to study P–C insurer earnings management in the accounting literature (e.g., Petroni 1992; Petroni and Beasley 1996; Beaver and McNichols 1998; Nelson 2000; Gaver and Paterson 2004; Gaver et al. 2012). However, ours is the first study of earnings management in insurance surrounding CEO turnover events.
 
5
Corporate governance and CEO turnover information in insurance are not readily available in any public database, although detailed financial data are available in the National Association of Insurance Commissioners (NAIC) annual statement database.
 
6
Other organizational forms exist in the insurance industry, such as reciprocals, U.S. Lloyds, and risk retention groups (RRGs). Modern reciprocals are virtually indistinguishable from mutuals, and our tests reveal that mutuals and reciprocals behave similarly. Thus, we combine these two organizational forms and refer to the joint category as “mutuals.” We do not include Lloyds due to their relatively small and idiosyncratic role in the P–C insurance industry (U.S. Lloyds account for less than 0.1% of total P–C premiums). We do not include RRGs because they are small (less than 1% of premiums) and are an atypical organizational form.
 
7
Mayers and Smith (2013) argue that the co-existence of various ownership structures is due to their relative efficiency in controlling agency costs. For example, the agency costs of controlling managers are significantly higher for mutuals than stock insurers, such that mutuals concentrate in less complex lines of business where managerial discretion is less important. A firm’s ownership structure is predetermined with regard to operational decisions from an econometric perspective due to the infrequency and difficulty of changing the ownership structure (Mayers and Smith 1990). No firms in our sample changed their ownership structure during the sample period.
 
8
We acknowledge that the availability of proxy fights, the take-over market, and widespread analyst monitoring are likely to mitigate the agency problems to some extent.
 
9
Fama and Jensen (1983) argue that when a stock insurance company is closely held, the monitoring from the owners on managers is direct and simple. Ke et al. (1999) posit that closely-held insurers should have more direct monitoring of management by owners. Nagar et al. (2011) indicate that a key feature of closely-held firms is that shareholders are typically few in number, and they are also are involved in management.
 
10
It is possible that the argument of Hermalin and Weisbach (2012) does not apply in insurance. All types of insurers are required to disclose the same detailed information to state regulators and thus transparency is similar for all organizational types.
 
11
Hope et al. (2013) note that “these (higher) reporting qualities of public firms are mitigated or eliminated in settings where public firms are more likely to manage earnings or faced reduced demand for their financial information.”.
 
12
At the time the sample data were collected, we were limited to using data years of 2006 and earlier, because of our usage of the 5-year loss reserve error, which is required in calculating the reserve errors.
 
13
The CEO turnover of foreign firms might be due to the foreign firm characteristics rather than performance. For example, the corporate policy of Nissan Fire and Marine Insurance Company is to replace the CEO of its U.S. subsidiary every 3 years. Further, the performance of the subsidiary of a multinational company partly depends on its parent company’s global strategy. The incentives of incumbent or incoming CEOs involved in M&A events might be different from those in non M&A firms.
 
14
Blackwell et al. (1994) argue that it is the holding company management rather the boards of subsidiary banks who play a dominant role in deciding the retention or removal of subsidiary executives. They find evidence that it is quite common that one person simultaneously holds several executive positions in different subsidiary banks.
 
15
Our sample firms represent 65% of total industry premiums in 2000 and comparable percentages for other years.
 
16
As a robustness test, we also scale the reserve error by the revised estimate of total incurred losses in calendar year t+5, following Gaver and Paterson (2004). Our results are not sensitive to the choice of scaling variables.
 
17
Usually, Best’s Insurance Reports and proxy statements have the names of company officers with the following titles: Chairman of the Board, President, CEO, Senior Vice President, Secretary, CFO, Vice President, and Treasurer. All companies in our sample have at least one of these three titles: CEO, President, or Chairman of the Board. If an insurer has no individual listed as CEO, then the executive who has the title of president is selected. If no individual is listed as either CEO or president, then we define the chair of the board of directors as the top executive. In a limited number of association-related closely-held stock firms and mutual insurance firms, the officer with CEO title is only responsible for the daily administration and the president of the company, usually the president of the association, is the decision maker. In this case, however, we still code the CEO as top executive in order to be consistent. Results do not change if we code the president as the top executive in these firms.
 
18
Best’s Insurance Reports, published each year in July, does not consistently report CEO turnover events in the publication year (i.e., from January to June). Furthermore, Best’s Insurance Reports does not always report the exact month of a CEO turnover event, which further adds potential inaccuracy to the results.
 
19
As in Chen et al. (2013), we retain cases where a CEO is replaced by an interim CEO, but exclude cases where an interim CEO is replaced by a new CEO. Kalyta (2009) keeps only retiring CEOs who have held the position for more than two full years in his sample to alleviate “the potential impact of the horizon problem associated with the departure of the previous CEOs.” We repeat our analysis using 2 years to remove interim CEOs, and the results remain qualitatively the same.
 
20
After collecting the management and board information from Best’s Insurance Reports, we find that an insurance company’s management information and board information may not be updated at the same time. An update of board information may lag 1 year behind management information. Thus, we define the turnover event, in which previous CEOs stay on the board for at least 2 years, as routine turnover to avoid possible bias. We also search for the reason for a CEO’s departure from the company’s website and news articles on LexisNexis Database and the Internet. This helps us to further identify whether the CEO departs the company due to normal retirement, death, health issues, or a comparable appointment elsewhere. Such events are classified as routine turnover. As in prior research, our approach does not perfectly classify routine and forced turnover. Our definition of forced turnover is different from that of Pourciau (1993) on cases where managers take a comparable or better position elsewhere. We define these cases as routine turnover. Routine turnover observations are dropped from our sample.
 
21
For some publicly-traded firm year observations in which proxy statements are unavailable, we take their board information from the proxy statement in the nearest available year. Weisbach (1988) indicates that this approximation is reasonably accurate since board composition remains stable over time.
 
22
See Cummins and Weiss (2014) for an analysis of systemic risk in the insurance industry.
 
23
Insurance is a highly regulated industry. Regulation by the state insurance commissioner constrains managers from aggressively pursuing opportunistic behavior and other non-value-maximizing actions. Regulatory monitoring attenuates managerial earnings management, although regulation is not expected to eliminate all opportunistic behavior. Furthermore, loss reserves have tax implications, suggesting oversight of earnings management by the U.S. Internal Revenue Service. Warfield et al. (1995) find that the positive relationship between managerial ownership and earnings’ explanatory power for returns is weaker in regulated industries compared to non-regulated firms, suggesting regulation serves as an alternative monitor of managers’ accounting choices.
 
24
We also account for regulation by including a capital adequacy measure, the risk based capital ratio, as in Cornett et al. (2009). The results are similar, but the number of observations is greatly reduced due to the unavailability of the RBC ratio. All firms in our sample have RBC ratios well above 200%—the threshold that triggers regulatory intervention. This allows us to evaluate the use of reserve error to manage earnings by CEOs rather than their use as a last resort to avoid regulatory intervention (Dechow et al. 1995; Kothari et al. 2005).
 
25
An alternative non-exclusive explanation is that claims settlement in long-tail lines covers a longer period and hence a higher proportion of losses incurred has to be based on estimates rather than paid claims.
 
26
The fraction of net premiums written (NPW) from commercial long-tail business lines is the proportion of NPW in long tail lines (all of the commercial lines for which the full runoff triangle is reported in Schedule P in the NAIC annual statement) to total NPW. The fraction of NPW from personal lines is the proportion of NPW in personal lines (farm-owners multiple peril, homeowners multiple peril, automobile physical damage, and personal automobile liability) to total NPW.
 
27
Because P–C insurers are required to submit their annual statements to the NAIC by the March 1 of year t + 1, it is plausible to argue that the incumbent CEO has the main control of the financial statement (thus loss reserve reporting) in year t − 1. It is possible, although the chances are low, that the incoming CEO has some discretion in the preparation of the financial statement of year t − 1 if she assumes the CEO post before the finalization and submission of the statement deadline (first 2 months of year t).
 
28
The financial statement for year t is not completed until the end of the second month of year t+1. According to our definition of turnover, a manager resigning as of year t would not likely have had significant input into the determination of loss reserves and other discretionary accounting decisions affecting the financial statements of year t that are issued during the first 2 months of year t+1. Incumbent CEOs’ impact is only limited to real practice in the earlier periods of the year (Best’s Insurance Reports is published in July each year). In contrast, the incoming CEO has full control of a firm’s practice after assuming the position.
 
29
In unreported results, the means of loss reserve error have similar patterns. To mitigate the impact of outliers, all continuous variables in this study are winsorized at the 1% and 99% levels (Kothari et al. 2005; Cornett et al. 2009). In robustness tests, we winsorize or trim outliers at the 5% and 95% levels. Our results are robust to these variations. Non-parametric Kolmogorov–Smirnov tests provide similar results.
 
30
Incoming CEOs also over-reserve slightly more than firms without a CEO turnover in years t+1 and t+2, although the difference in magnitude is even smaller than that for transition year t. This is also consistent with the argument that incoming CEOs “reverse accrual” due to outgoing CEOs’ under-reserving.
 
31
The difference in mean errors with firms not experiencing a CEO turnover is significant at the 5% (10%) level in year t − 2 (t − 1).
 
32
Accrual reversal generally means that income-increasing (decreasing) discretionary accruals initiated in a prior period reverse to become income-decreasing (increasing) accruals in the next period (e.g., Baber et al. 2011).
 
33
The Breusch-Pagan Lagrange multiplier tests suggest that fixed/random effects models are preferred for our sample. Hausman tests indicate that fixed effects are preferred to random effect models. The modified Wald test rejects the null hypothesis that there is no group-wise heteroskedasticity (p < 0.0001). Wooldridge (2002) tests also indicate serious autocorrelation in our panel data. Beaver and McNichols (1998) also report positive serial correlation in reserve errors indicating multi-period reserve management. A few firms are dropped from the estimation in various models, because they are present for only 1 year of the sample period. We also estimate OLS and the firm fixed effect model using robust standard errors clustered at the firm level as well as additional median regressions. These results are not materially different from results obtained using FGLS and are available from the authors.
 
34
Following the best of the prior literature (e.g., Grace and Leverty 2012), we do not include firm fixed effects. Previous literature and tests conducted on our sample indicate that loss reserving errors are serially correlated (Beaver and McNichols 1998; Grace and Leverty 2010, 2012) and heteroscedastic (Grace and Leverty 2010, 2012). Feasible generalized least squares (FGLS) rather than full fixed effects is commonly used when these problems exist (Grace and Leverty 2012).
 
35
Exogenous economic factors including unexpected inflation, regulatory changes, and changes in court attitudes are out of management control (Petroni 1992). Gaver et al. (2012) document that the P–C industry over-reserved from 1993 to 1997, under-reserved from 1998 to 2002, and returned to over-reserving in 2003–2004.
 
36
Interestingly, the coefficient of board size is positive for stock subsample regressions, but mostly negative for mutual subsample regressions. This suggests that board members in stocks have a higher degree of coordination than that of mutuals in promoting conservative accounting. We do not find a significant difference in the impact of board size on reserve error between closely-held stocks and publicly-traded stocks.
 
37
Results are similar when we use the logarithm of assets as an alternative proxy for size.
 
38
We also use the tax indicator variable equal to one if the insurer has a high tax rate, and zero otherwise, and find similar results. The variable is defined in Petroni (1992) and Nelson (2000).
 
39
Several studies find that CEO turnover is significantly higher for firms filing material financial restatements (Arthaud-Day et al. 2006; Desai et al. 2006). Farber (2005) argues that a board is more likely to replace a CEO accused of fraud to restore a firm’s financial reporting credibility, which is crucial in building consumer confidence in the financial industry. Hennes et al. (2008) also find higher CEO and CFO turnover rates for restatements due to intentional irregularities than those due to unintentional errors. In other words, managers will be punished for misleading accounting reports involving fraud, through reduced compensation, decreased credibility, and in the most severe cases, loss of employment and even criminal charges.
 
Literature
go back to reference A.M. Best Company (2016) Best’s aggregates and averages—property-casualty: 2016 edition, 2016th edn.M. Best Company, Oldwick A.M. Best Company (2016) Best’s aggregates and averages—property-casualty: 2016 edition, 2016th edn.M. Best Company, Oldwick
go back to reference A.M. Best Company (annual a) Best’s insurance reports—P/C edition. A.M. Best Company, Oldwick A.M. Best Company (annual a) Best’s insurance reports—P/C edition. A.M. Best Company, Oldwick
go back to reference A.M. Best Company (annual b) Best’s key rating guide: property/casualty. A.M. Best Company, Oldwick A.M. Best Company (annual b) Best’s key rating guide: property/casualty. A.M. Best Company, Oldwick
go back to reference Dechow PM, Sloan RG, Sweeney AP (1995) Detecting earnings management. Account. Rev. 70:193–225 Dechow PM, Sloan RG, Sweeney AP (1995) Detecting earnings management. Account. Rev. 70:193–225
go back to reference Mayers D, Smith CW (1990) On the corporate demand for insurance: evidence from the reinsurance market. J Bus 63:19–40CrossRef Mayers D, Smith CW (1990) On the corporate demand for insurance: evidence from the reinsurance market. J Bus 63:19–40CrossRef
go back to reference Mayers D, Smith CW (2013) On the choice of organizational form: theory and evidence from the insurance industry. in Handbook of Insurance, 2nd edn. Springer, New York Mayers D, Smith CW (2013) On the choice of organizational form: theory and evidence from the insurance industry. in Handbook of Insurance, 2nd edn. Springer, New York
go back to reference Wooldridge JM (2002) Econometric Analysis of Cross Section and Panel Data. MIT Press, Cambridge Wooldridge JM (2002) Econometric Analysis of Cross Section and Panel Data. MIT Press, Cambridge
Metadata
Title
Earnings management surrounding forced CEO turnover: evidence from the U.S. property-casualty insurance industry
Authors
Jiang Cheng
J. David Cummins
Tzuting Lin
Publication date
09-08-2020
Publisher
Springer US
Published in
Review of Quantitative Finance and Accounting / Issue 3/2021
Print ISSN: 0924-865X
Electronic ISSN: 1573-7179
DOI
https://doi.org/10.1007/s11156-020-00910-z

Other articles of this Issue 3/2021

Review of Quantitative Finance and Accounting 3/2021 Go to the issue