Berkshire Hathaway Inc.’s Warren Buffett, whose $100,000 base salary long ago became uncompetitive with his peers, ranked as the most undercompensated executive in the industry, according to a new study of 2004 executive pay released by SNL Financial’s insurance research department.
And to read Buffett’s annual letter to shareholders last March, he deserved to be there in 2004.
According to the legendary investor and executive, even as the company’s book value gain fell short of the rise in the S&P 500, CEOs of Berkshire’s operating units “pulled more than their share of the load” a year ago in terms of running their businesses and sending excess cash to Buffett for deployment.
“I didn’t do that job very well last year,” Buffett wrote, noting that he “struck out” in attempts to make “several multibillion-dollar acquisitions” and that he found few attractive securities to purchase.
Buffett’s total compensation of $311,000, including his base salary and $211,000 in income from directors’ fees and/or deferred phantom equity interests from certain nonsubsidiary companies in which Berkshire has significant investments, ranked in magnitude alongside CEOs of companies that are but a small sliver of the Omaha firm’s size.
Joining Buffett at the top of the SNL undercompensated rankings, or the bottom, as the case may be, was American Financial Group Inc.’s Carl H. Lindner Jr., who retired as CEO of the company in January 2005. Lindner earned a base salary of $990,000 and asked to forgo a bonus based on “his desire to have shareholders benefit from the retention by AFG of any bonus amount that could otherwise have been paid to him,” according to a company filing.
Loews Corp. CEO James Tisch ranked No. 3 on the undercompensated list, according to SNL’s criteria, while Stephen Lilienthal of majority-owned subsidiary CNA Financial Corp. was No. 6 on the “overcompensated” list; it would appear that Loews’ considerably higher return on average equity and better share-price performance resulting from its diversified set of businesses were the primary reasons for the discrepancy. Tisch’s total compensation in 2004 was $2.4 million, including a base salary of $1.3 million and bonus of $1.1 million. Lilienthal earned $3.3 million.
According to Forbes’ 2004 list of the richest Americans, Buffett was No. 2, the family of Carl Lindner Jr. was No. 124, and Preston Robert Tisch, uncle of James Tisch, was No. 60.
Former American International Group Inc. CEO M.R. “Hank” Greenberg” ranked one spot lower in Forbes’ 2004 list, but he finds himself on the overcompensated side of the ledger in SNL’s rankings.
Penalized, according to SNL’s criteria, by AIG’s below-average returns between 2002 and 2004 and total compensation that was supplemented by $10.1 million in long-term incentives from Starr International Co., Greenberg finished behind only Allmerica Financial Corp. CEO Frederick Eppinger and XL Capital Ltd. CEO Brian O’Hara on the overcompensated list.
Of course, CEO pay that might be considered too high or too low relative to SNL’s set of performance criteria could be viewed in an entirely different light by company shareholders and compensation committees of various boards of directors, depending upon which specific metrics they emphasize. XL’s board, for example, based O’Hara’s $1.8 million bonus and $1.5 million restricted stock award on the company’s achievement of record net operating earnings, double-digit book value growth and ROE (relative to net income) that compared favorably to its peers. Certain intangible factors also came into play.
“The committee and the board continue to believe Mr. O’Hara’s leadership is strong, effective and appropriate,” XL’s compensation committee said in its report. “The company’s growth objectives are focused and the profitability of current-year business reflects the strong underwriting culture driven by Mr. O’Hara. In areas where objectives were not being met, appropriate corrective actions were initiated, including management changes where necessary. The committee believes that the company is well positioned, financially strong and focused on enhancing shareholder value.”
SNL’s rankings also do not consider an executive’s compensation relative to the previous year.
In Lilienthal’s case, his 2004 base salary was unchanged and his bonus was $95,000 smaller when compared to 2003, even though CNA swung from a net loss to a profit over the course of those two years and its consolidated GAAP combined ratio fell to its lowest point in at least a decade (albeit still in excess of 100%) in 2004. But since his actions to shore up reserves and exit non-core and unprofitable lines of business continued to result in profitability that fell short of property & casualty peers, the CNA CEO was penalized by SNL’s criteria.
Indeed, three of eight most overcompensated CEOs in SNL’s rankings found their companies (Allmerica, CNA and Phoenix Cos.) in the middle of turnarounds and restructurings in 2004, meaning long-term stock performance and profitability measures were penalized for actions that might well have occurred under someone else’s watch.
Such imperfections, the varying circumstances in which CEOs find themselves and the backward-looking nature of compensation disclosures complicate the process of developing an effective model for gauging whether or not executive pay packages are reasonable. Nonetheless, they can serve as a useful catalyst for identifying where change needs to occur.
As shareholder activism continues to grow across industries and boards of directors gain greater independence, one would hope that compensation committees will move more swiftly than has been the case historically to address gaps, both positive and negative, between pay and performance.
SNL defines total compensation as the sum of base salary, bonus, other annual compensation and long-term compensation consisting of annual cash, stock or performance unit awards paid or credited to executives during the fiscal year in accordance with multiyear performance goals. It includes restricted stock awards, long-term incentive plan awards and other long-term compensation, but excludes stock options. Option-adjusted compensation, or the sum of total compensation and the fair value of options issued to the executive during the given fiscal year, also factored into SNL’s rankings.
SNL’s insurance research department constructed a basic model for the projected compensation that assumed all inputs are independent of one another and assigned each input to a function that included both the input and a tuning parameter in its argument. To tune the model, they varied the tuning parameters for the inputs until they maximized the overall correlation between their prediction and actual values.
On the input side of the coin, the SNL insurance research department said, the model is relatively basic and considered four factors when evaluating the compensation level of an executive: size, total shareholder return and profitability, in particular, return on average assets and return on average equity.
The connection between company size and compensation reflects the relative difficulty of managing a larger institution, the SNL insurance research department said. To weigh profitability metrics and total shareholder return, SNL’s insurance research department measured these terms as the relative variation from the peer mean. For example, the measured item for total return is the difference between the average annual return for the company and the average annual return of the relevant peer index all as a fraction of the average annual return of the relevant peer index, whether it be multiline, life and health or property & casualty.
The bottom line is that the model compares the relative ranking of compensation among the companies to the ranking of its performance (adjusted on a peer basis) within the entire group. Those whose ranking are the most different are put into the overpaid and underpaid categories.
John McCune contributed to this article. All data is according to SNL Financial, Charlottesville, Va.