by Burch T. Kealey, University of Nebraska at Omaha, and Vance Lesseig, Texas State University, San Marcos
In this descriptive study we examine the compensation of the highest-paid named executive officers (NEOs) of the Dec. 31 fiscal year-end utilities that filed proxy statements with the Securities and Exchange Commission (SEC) by the April 30 reporting deadline for accelerated filers.
Most academic research on the components and determinants of public company compensation generally has excluded utilities and firms from other regulated industries–banks–from their samples because of the extent of regulatory oversight to which companies in this industry have been subjected. Researchers have been concerned that including utilities subject to rate regulation in their samples would bias their results downward. Researchers also have interpreted the existence of rate regulation as providing assurance of a fixed or minimum level of assured profits to the utility, which suggests incentives for managerial performance are not as meaningful when compared with the pay-performance relation found in analysis of companies in less regulated industries.
Given the significant changes that have shaped the electric utility industry during the past 10 years (see the Edison Electric Institute 2006 report “Electric Utilities and Risk Compensation”), it is reasonable to try to understand whether the standard methodologies used to analyze the pay-performance relation should be applied to companies in this industry.
General Determinants of Compensation
Past academic research generally has established that in unregulated industries, size (as measured by total assets) and measures of accounting returns, as well as stock market returns, are all important determinants of NEO pay. The argument for the relation between firm size and compensation is driven by early analytical work that has hypothesized that the market for managers rationally places managers with higher abilities at larger firms. Accounting returns have been shown to be associated with compensation to provide NEOs a share in the growth of their firms that is independent of the risks imposed when compensation is tied too tightly to market measures such as stock returns, which are affected by both macro events (such as the financial crises of the past several years) and investor sentiment about firm performance. The relation between compensation and stock market returns has been hypothesized to exist because managers act as agents for shareholders whose gains are reflected in stock price changes. A long line of academic literature links these measures with observed compensation practices. This study takes the initial step of testing the association between proxies for these measures and the compensation of the highest-paid electric utility executives.
Because we need measures of size, accounting income and market values, we must start our sample selection with Research Insight, a data product that provides financial and market data coverage of a significant subset of publicly listed U.S. companies. We required our sample to have common stock traded on the New York Stock Exchange, be a Dec. 31 fiscal year-end firm and have data available on the financial measures needed for this study. We used directEDGAR to extract and normalize the summary compensation table for the 29 firms that had data available from Research Insight. From the extracted data, we identified the NEO with the highest compensation for each firm and discarded the rest. We focused on the most highly compensated officer to be consistent with prior research.
Most of the following analysis is reported in percentage terms. To get started, we first report the total compensation for 2008-10 in Table 1.
The average total compensation increased approximately 13 percent from 2008 to 2009. The average increased only 4 percent from 2009 to 2010. The SEC requires that registrants report the compensation components in the summary compensation table. Instead of summarizing the amounts, we report the percentage of total compensation reflected in each component in Table 2. We focus on the percentage instead of amounts to allow better comparisons across firms.
Table 1 shows a fairly significant increase in total compensation across the three years. Table 2 suggests the increase is driven by shifting more weights to riskier components of compensation rather than increasing base salary. This is supported by analysis of average total salary levels, which remained virtually unchanged across the three-year window. Table 2 shows that annual stated cash bonuses are an insignificant component of compensation in this industry. Only one executive in the sample received a cash bonus in each year analyzed. Further, this executive was the only NEO who received any bonus in 2009 and 2008. Two NEOs received a bonus in 2010; one appeared to be the result of a post-retirement board retention agreement.
Nonequity incentive payments is compensation awarded for achievement of specific annual performance goals. We examined the nature of the goals disclosed. Most companies set targets for financial performance measures (EBITDA, EPS and cash measures), safety outcomes such as lost-time incidents and employee, customer satisfaction and operational reliability measures. The most weighting was on the financial performance measures with between 10 to 20 percent of the weighting placed on safety and other nonfinancial measures. The average amount paid increased by almost one-third across the three-year window (from about $900,000 to $1.2 million). This increase explains almost half of the total increase in compensation.
Most companies in the sample reported two components of equity-based compensation: option awards and stock awards. The proportion of total compensation represented by options decreased almost one-third. This decline is not simply the result of an increase in total compensation. We analyzed the average option award across the three years and found that it declined about $250,000. The average increase in stock awards (which include outright stock and nonoption-based stock derivatives such as restricted stock) almost matched the decline in options in dollars and percentage. Many payouts reported in the stock awards column represent payouts for performance over a multiyear horizon.
Nonqualified deferred compensation generally represents retirement benefits earned based on longevity rather than performance achievements. According to the proxy statements, the amounts generally are set by an executive’s employment agreement. This component accounted for an increasing share across the three years; however, the increase almost is explained completely by a few spikes rather than a systematic overall increase across the sample.
The amounts reported for others include amounts attributed for personal use of aircraft, security services, membership fees and car allowances. The average amount declined from $195,000 in 2008 to $190,000 in 2010. For most NEOs in the sample, this category represented a negligible amount of total compensation.
Our analysis suggests a gradual but steady shift across this period from less-risky pay packages to those that include more at-risk compensation. This shift is most evident in the decline in the proportion of salary-based compensation and the increase in the proportion of total compensation represented by nonequity incentive payments.
Association Between Compensation and Measures of the Firm
Given this is an exploratory study, we test for association using correlation analysis. An alternative would be to use regression models and control for alternate explanations of compensation determinants. This might be worthwhile later, but it would require a much more refined data set than is available now.
We first test for an association between changes in total compensation and changes in firm size (as measured by total assets). The null hypothesis of no association cannot be rejected because while the correlation between changes in total compensation and total assets is approximately 0.25, the t-value is only 1.36, which is not significant at regular confidence levels for small samples. This result seems consistent with the shifting in weights in the compensation components we observed in the sample.
Despite increasing weight on at-risk compensation, the association between changes in nonequity incentive payments and measures of accounting performance were not significant at conventional levels. We tested the association between changes in EBITDA, ROE, EPS, operating income and stock returns over the three-year window with changes in the amounts of nonequity compensation. We attribute this lack of statistical significance to three features of this data. First, the sample size is fairly small. Second, the measures used by the compensation committees to determine payouts varied. That is, some proxies specifically mentioned the use of EPS, others mentioned EBIT or operating income, as well as others. Third, almost all of the sample reported the use of nonfinancial measures in determining appropriate payouts. We can only analyze changes in financial measures as reported in financial statements.
Generally, past research has justified excluding utilities and other regulated firms from studies because of the impact of rate regulation generally limiting profits and weakening the expected relation in nonregulated industries. We find a significant move toward more financial performance-based compensation when we examine the data, but this finding is not statistically significant at conventional levels. This is more a result of a lack of power of our tests (small sample and noisy measures not necessarily correlated to performance targets). These findings justify the decision of prior researchers to exclude these firms from large sample tests. If the trends observed during the past three years continue, however, we suspect the same analysis performed in the future might lead to statistically significant associations.
Research Insight website: http://www.compustat.com/Research_Insight/
Burch T. Kealey, Ph.D., is a William C. Hockett professor of accounting and director of the Master of Accounting Program at the University of Nebraska at Omaha. Reach him at firstname.lastname@example.org.
Vance Lesseig, Ph.D., is an assistant professor of finance at Texas State University, San Marcos. Reach him at email@example.com.
Past EL&P Issues