Executive Compensation: A Tilted Playing Field

7/22/2019 Executive Compensation: A Tilted Playing Field

1/17

Executive Compensation: A Tilted Playing Field

Alan Reynolds

Shortly after the Enron bankruptcy there were several efforts to attribute thecompanys financial crisis to supposedly fundamental flaws in the ways in whichexecutives in general are paid. Two distinct issues soon became intertwined. One was a

broad complaint about the level of compensation being excessive, using Enron as thoughit was a representative example of a much broader problem. The other was a narrowercomplaint about one particular form of such compensation stock options and aboutthe way that form of compensation was accounted for and taxed. The agitation aboutexpensing options seemed to acquire symbolic value, as though that technical andcontroversial accounting procedure had some connection to accusations of actual fraud inthe way Enron accounted for debts or in the way WorldCom depreciated operating costs.

Media and political attention to the level and form of executive compensation in2001-2003 began as an almost exact replay of events ten years earlier. This chaptertherefore begins with a brief history of the controversy about executive compensation inthe early 1990s, showing how some intended reforms of that era contributed toexcesses and imbalances by the end of that decade. The following section includes acritique of recent efforts to place a large share of blame for the Enron crisis on executivecompensation, particularly stock options (a topic dealt with more intensively in latersections section). These first two sections set the stage for a more detailed look into thecomplexities of executive compensation in the context of rents (paying more thannecessary) and agency costs (trying to bring the personal interests of managers intoalignment with interests of shareholders and creditors). Subsequent sections surveystatistics and studies concerning both the level and particular forms of executivecompensation. This is followed by a descriptive analysis of the many ways in whichstock options are or could be accounted for and taxed a topic that has once againgenerated an unwarranted intensity of political and regulatory interest recently, as it didin the early nineties.

Among numerous conclusions two stand out most clearly. The first conclusion isthat federal efforts to influence executive compensation in 1992-1993 provedcounterproductive. Those efforts included SEC disclosure rules, increased tax rates onsalaries and dividends, and a million dollar limit on deductibility of salaries. The secondconclusion is that federal efforts to influence executive compensation policy in the futureare also likely to prove counterproductive. Remedies that appeal to activist and populistsentiment (such as disclosure and shareholder approval of executive pay) may put toomuch authority in the hands of those who have neither the information nor the incentivesto negotiate on behalf of all owners of the firm (including creditors and nonvotingshareholders). This chapter also suggests that if FASB compels public companies to treatas an actual expense the estimated fair value of stock options at the date they are grantedthe main effects will be to (1) reduce the transparency and pay-for-performance


2/17

effectiveness of executive compensation while (2) making GAAP earnings even lessinformative for investors (due to treating unreliable estimates as actual expenses).

Executive compensation is certainly a legitimate concern of major shareholdersand directors, but introducing third parties into the negotiation (such as the FASB, the

IRS or trial lawyers) is quite likely to introduce rigidities that lead to suboptimalarrangements from the point of view of both executives and shareholders. One-size-fits-all accounting rules can become impediment to the flexibility needed to deal with theinherently individualized and subjective process of negotiating customized contracts toattract and retaining highly specialized managerial talent.

Cautionary Lessons from the 1990s

At a Harvard Business Review roundtable of compensation experts, New Yorkattorney Joe Bachelder reminded everyone that, in 1972 we were criticizing CEO pay.We were still doing it in 1982. In 1992, a raft of regulations and legislation were directed

at reining in CEO compensation. Today we are still discussing it. Much of the currentcriticism comes from the fact that . . . nearly 80 percent of the gain in CEO pay in the1990s is attributable to stock options. And stock markets did pretty well in the 1990s.Werent we saying in the 1980s that we should tie CEOs to the market in order to identifythem with shareholder value? We got what we asked for. 1

Toward the end of prolonged economic expansions, such as 1983-90 or 1992-2000, stock prices reach a cyclical peak and so does compensation linked to stock prices.Information about executive compensation is gathered and reported with a lag, with theresult that peak compensation is reported a year or two after the economy has slipped intorecession. Comparing boom pay with bust conditions then becomes irresistible fodderfor media and political outrage about excessive compensation during the previouscyclical boom. In June 2001, for example, a typical Fortune feature d The Great PayHeist, complaining about the highway robbery of the year before. 2 As Joe Bacheldersuggested, however, the raft of regulations and legislation in 1992-93 should remind usthat this latest cycle of indignation could likewise inspire changes in tax and regulatory

policies that end up aggravating the situation they promised to fix.

Back in 1990, The Harvard Business Review published a provocative andinfluential study by Michael Jensen and Kevin J. Murphy, who had studied compensationof CEOs at 1400 companies from 1974 to 1988. They found that a $1000 change incorporate value corresponds to a [two-year] change in CEO compensation of just $2.59.They also found that CEOs of large public companies are only slightly more likely tostep down after very poor performance. With pay and job security appearing to be soinsensitive to performance, Jensen and Murphy complained that corporate America paysits most important leaders like bureaucrats. Later studies often disputed thatconclusion, but that is largely because the post-1990 shift from short-term bonusestoward relatively long-term stock options has greatly improved the sensitivity of pay tostock performance. By 1998, for example, Hall and Liebman estimated that a 10 percentincrease in a firms market value added $1.25 million dollars to value of a median CEOs


3/17

accumul ated stocks and options (while it also added billions to the value of shareholderswealth). 3

Comparing CEO compensation in the late eighties with the far more generous payof partners at law firms and investment banks, Jensen and Murphy argued that CEO pay

should be higher for superior stock performance. But they also thought probability of being fired should also be higher for inferior performance. In that case, less-talentedmanagers would be replaced by more able and more highly motivated executives whowould, on average, perform better and earn higher levels of pay. To reward increasedsuccess fostered by greater risk taking, effort and ability, they argued, would eventuallymean paying the average CEO more. 4 Among smaller companies, the Jensen-Murphystudy found the most potent pay-for-perfomance incentives by far were paid to the CEOof Berkshire Hathaway, Warren Buffett, a gentleman who has been known to complainthat much smaller incentives for other CEOs are excessive.

As other economists began looking into executive compensation, it seemed that

Jensen and Murphy probably exaggerated the alleged inefficiency of the market forcorporate leadership. Summarizing numerous studies on executive compensation produced for the National Bureau of Economic Research through 1994, Nancy Roseremarked: We find no evidence for the popular view that boards typically fail to

penalize CEOs for p oor financial performance or reward them disproportionately well forgood performance. 5 Regardless of the possible exaggeration of the Jensen-Murphystudy in 1990 and its probable obsolescence today (due to widespread substitution ofstock options for salary and bonuses), the basic idea that CEO pay and tenure were oftenunrelated to performance made CEO compensation appear arbitrary . And the impressionthat CEO compensation was arbitrary was easily be abused by those who, quite unlikeJensen and Murphy, imagined CEOs to be typically overpaid. As a result, twoseemingly quite different ideas that CEO pay should be more closely tied to stock

prices or that CEO pay should be reduced by law -- came together as a hot political issuein 1991 and 1992. The result was federal intervention that was intended to limit CEO

pay but actually ended up having the opposite effect.

In May 1991, two months before the start of recession, Carl Levin (D-MI), thenChairman of a Senate Oversight subcommittee, held hearings on executive compensationin response to inflammatory articles then appearing in the press about CEO earningsduring the stock markets preceding peak. On June 4, 1991 Senator Levin andRepresentative John Bryant (D - TX) introduced the Corporate Pay Responsibility Actto require more disclosure of CEO pay and to require the value of sto ck options to beestimated at the time they were granted and charged against earnings. 6 Congress did notact on the bill, despite subcommittee hearings On October 17. But the political pressuredid appear to affect regulatory policy at the Financial Accounting Standards Board(FASB) and Securities and Exchange Commission (SEC). FASB agreed to look into thematter of expensing estimated option values (as the FASB did again in 2003 inresponse to very similar political and media pressure). In May 1994, however, the Senateeventually passed a resolution rejecting Senator Levins proposal of mandatoryexpensing of the estimated value of employee stock options by vote of 88 to 9. 7


4/17

On October 16, 1992, the SEC amended Item 402 of Regulation S-K to require

much greater public disclosure of executive compensation and to facilitate shareholdervoting on such compensation. This action, noted Prevost and Wagster, was in reactionto intense political pressure from shareholder activists, presidential candidates, the media

and the U.S. Congress. Jensen and Murphy had anticipated this development in 1990,warning that public disclosure of what the boss makes give ammunition to outsideconstituencies with their own special-interest agendas. . . .How often do shareh olderactivists or union leaders denounce a corporate board for underpaying a CEO? 8

Making CEO compensation a political target for activist groups with quitedifferent goals than maximizing shareholder wealth can be risky for shareholders. Prevostand Wagster surveyed several studies suggesting that managers of labor union orgovernment pension funds have often adopted the stance of being crusaders against big

business to further their own personal interests, including running for political office.They examined several events leading up to the SEC rule change and estimated the

impact on stock prices of firms targeted for institutional monitoring by the CaliforniaPubic Employees Retirement System (CalPERS). They found that firms targeted byCalPERS suffered significant wealth losses from the activist threat to CEO incentivescaused by new SEC rules designed to make CEO pay an easier target. 9

What newly empowered activists often did with the new SEC rules, however, wasto use their special interest leverage to restrain CEO salaries and bonuses, not pay-for-

performance (i.e., stock options). As if such external meddling in employment contractswas not dangerous enough, Congress decided to lend a hand.

On August 6, 1993 President Clinton signed into law the Revenue ReconciliationAct of 1993. The 1993 tax law fulfilled two campaign promises in the 1992 Clinton-Gore campaign book, Putting People First . One was to limit corporate deductions at $1million for CEOs [salary and bonus]. The other was to impose a surtax on millionaires(meaning, once in office, joint incomes above $250,000). 10

The million dollar rule, Section 162(m) of the tax code, denies public companiesany deduction for the cost of executive compensation in excess of $1 million a year.This rule was never really about fairness since it does not apply to other occupations,such as actors, athletes or attorneys. And it was not really about revenue either: TheClinton-Gore book claimed it would raise only $400 million a year, which would have

been possible only if the ceiling had been ineffective. The main point, or at least themain effect of this tax law, was to discourage companies from paying using non-salaryforms of pay to attract and keep top managers. That is because the million dollar ruledoes not apply to performance-related pay, such as grants of stock options and restrictedstock. The million dollar rule does apply, however, to any stock options that are in themoney (above the market price) when granted, thus effectively banning that type ofcompensation.


5/17

Hall and Liebman found that the million dollar rule led firms to adjust thecomposition of their pay and toward performance related pay but did not decrease thetotal level of compensation. 11 They focused on the short-term impact, and thus found arelatively minor tax-related shift from salary to options. Yet the million dollar cap

becomes more binding with each passing year, as William McDonough pointed out,

because that million dollar figure has never even been adjusted to keep pace withinflation, much less with the real increase in other professional incomes.

The 1993 tax law added a new marginal tax rate to 36 percent, up from 31 percent, on taxable income between $140,000 and $250,000. And an extra 10 percentsurtax (the millionaires surtax) raised the top rate to 39.6 percent on incomes above$250,000. That 28 percent hike left the top individual rate far above the new 35 percentcorporate rate (also increased by 1 percentage point). The steep new tax on marginalsalary income contributed to a shift toward deferred compensation, including stockoptions. And the steep tax on marginal dividend income contributed to excessivecorporate debt and inefficient corporate acquisitions.

Increasing the individual income tax to 39.6 percent from 31 percent meant thedouble tax on dividends after 1993 was increased from about 55 to 60 percent at thefederal level alone. That increased tax on dividends made stockholders even lessinterested in dividends. Ayers, Cloyd and Robinson found conditioned on dividendyield, a negative stock price reaction during the two weeks that Congress passed andPresident Clinton signed the increase in individual income tax rates in August 1993. 12 Individual investors subsequently became more inclined to favor companies that usedearnings to increase assets per share, through acquisitions or share repurchases, thusgenerating capital gains. The individual tax on capital gains was far less punitive than thetax on dividends, particularly since 1997 when the tax on long-term gains was reduced to20 percent from 28 percent.

The higher tax rate on corporate profits, particularly when paid out as dividends,also increased the incentive for corporate managers to use increased debt as a tax shield.Enron and WorldCom created the appearance of rapid revenue growth largely by usingretained earnings and debt to acquire companies. That may have been clever tax

planning, but it proved to be dangerous business planning.

Another relatively minor effect, explained in Taxes and Business Strategy byMyron Scholes and others, was that because the corporate rate was now l ower than theordinary income tax rate, the corporate form offered a deferral advantage. 13 Executivesfacing much higher tax rates on salaries and bonuses after 1993 could negotiate to lettheir companies use the new deferral advantage by switching a larger share of CEOcompensation toward the future through larger stock options, deferred income plans andretirement perks.

It is one thing to recommend to corporate boards that they should make executivecompensation more dependent on stock prices, as scholars such as Jensen and Murphywere doing in the early nineties (and many other scholars still do). But it is quite another


6/17

thing for the government to use tax policy or accounting mandates to tilt the playing fieldeither in favor of stock options or against them.

The SEC in 1992 and the Clinton Administration in 1993 acted on a belief thatregulations and tax policy should entice companies to tilt executive compensation toward

risky stock options and away from the security of salaries and perks. Recent complaintsthat most of the increases in executive pay from 1993 to 2000 came from options ratherthan salaries are ironic: That was precisely what the SEC and Clinton Administrationwere trying to accomplish. But they surely never anticipated and neither did mostcorporations -- what a roaring bull market would do to the value of stock options in thelate nineties when even the benchmark S&P 500 index was rising by nearly 30 percent

per year.

That bull market was over by April 2000, and two years later Busi ness Week wasreporting that average CEO pay declined 16% [in 2001], to $11 billion. 14 A year afterthat, in April 2003, Fortune reported that ave rage CEO compensation among the 100

largest companies fell by 23 percent in 2001.15

A different survey in The New YorkTimes found total compensation of chief executives declined 20 percent. 16 Suchsurveys differ in ways that can be quite misleading. Some surveys bravely attempt toestimate the value of new options granted during the previous year (such as the highlyunlikely $76 million estimate for AOL options granted to Steve Case in 2001). Otherscount the value of old options that were finally exercised, and then write an irate storycomparing such CEO compensation to a drop in the companys profits. But sales ofstock during a year are not compensation earned during that year, so this is an absurd wayto compare pay with performance. In many cases the stock sold by executives was

purchased rather than earned, yet some news reports have even referred to proceeds fromstock sales as if that was extra compensation. Selling financial assets is no different thanselling a home -- it does not make executives more wealthy and it is not income.

In 2002, as in 1992, popular articles complaining about stock-related pay began toappear long after the values of stocks and stock options had collapsed. In October 2002,for example, Paul Krugman wrote in The New York Times Magazine that CEO salariesfrom Fortunes top 100 had risen from $1.3 million in 1970 to $37.5 million in 1999.The latter figure, he cl aimed, was more than 1000 times the pay of ordinary workers(supposedly $35,864). 17 Krugmans 1000-to-1 figure even outbid the AFL-CIO, whichused similar sta tistical magic to claim CEO pay in 2000 was 411 times average hourlyworker wages. 18 Fortune s 1999 figures on CEO pay greatly exaggerated the estimatedvalue of options granted in that boom year, but the more obvious point is those 1999figures were inexcusably antique by October 2002. It should have been obvious to Mr.Krugmans readers, if not to the author, that CEO pay had been hugely reduced by themarkets long and horrific decline.

For 2002, Fortune s estimate of average top 1000 CEO salaries and benefits was$15.7 million nearly 60 percent smaller than the $37.5 million figure Krugman used.His figures on salaries of ordinary workers certainly did not refer to average salarieswithin the same top 100 firms; it was a national average diluted by hourly wages of many


7/17

fast food and farm workers. Unlike figures on CEO compensation, the pay of ordinaryworkers excluded increasingly important health and pension benefits and, in manycases, stock options. Krugmans measure of average salaries (and wages) would have

been $44,613 in 2000 if basic benefits were properly included, but $69,559 in finance and$75, 679 in communications. 19 Industries with high CEO pay also offer higher salaries to

ordinary workers and top 100 firms offer even higher salaries.The Krugman statistical exaggerations of October 2002 were very similar to

equally misleading articles written in 1992 about allegedly excessive CEO pay in 1989.Such politicized misinformation was not helpful then either.

What began in 1990 as a plausible academic complaint (from Jensen and Murphy)that CEO pay was insufficiently linked to the firms performance soon turned into achange in regulations and taxes that was designed to designed to encourage companies to

pay less in salaries and more in stock options or stock. Nobody imagined that optionsand restricted stock granted in 1993-95 would end up being hugely valuable after only

three or four years of vesting. Thanks to the bull market, the populist tax and regulatoryreforms of 1992-93 had the ironic effect of making CEOs far more wealthy than theywould have been if salaries and bonuses had been permitted to remain a larger share oftotal pay packages.

An Imaginary Link to Enron

In the wake of the Enron bankruptcy, a flood of press stories about supposedlyexcessive executive pay echoed the media agitation of a decade earlier. Enrons failurewas often attributed to corporate greed, said to be a result of what Paul Krugman labeledpermissive capitalism. Senators Carl Levin and John McCain (R-AZ) quickly seizedon this media opportunity to revive their moribund 1997 proposal to arbitrarily limitemployer deductions for the cost of exercised employee stock options.

On February 13, 2002, Senator McCain attempted to connect his recycled tax planto Enron by arguing that according to a recent analysis [from Citizens for Tax Justice],Enron issued nearly $600 million in stock options, collecting tax deductions whichallowed the corporation to severely reduce their payment in taxes. The Senatorapparently failed to realize that the estimated value of options when issued issued tells usnothing about the amount of tax deductions for options that may or may not be exercisedseveral years later. He also failed to understand that comparing corporate taxes paidwith income during the same year, as his cited analysis had done, is meaningless

because the corporate tax is largely based on accruals rather than immediate cash flow.Tax liabilities are sometimes deferred and operating losses are carried forward or back intime. 20 In any event, Senators McCain and Levin never explained how paying largertaxes would have made it any easier for Enron to pay its bills and thus avoid bankruptcy.The concocted link between Enron and the McCain-Levin assault on executive stockoptions was political opportunism. Yet many reporters hungrily leaped for the bait.


8/17

The proposed Levin-McCain tax penalty on options was explained in thefollowing Press Release from Senator Levins office, also dated February 13, 2002:

[The bill] would restrict the compensation deduction that company could claimfor the exercise of a stock option by limiting the stock option deduction to the

amount that company has claimed as an expense in its financial statement. Thissection would also make it clear that the deduction cannot be taken prior to theyear in which the employee declares the stock option income. . . .

Companies would have to expense the estimated value of options in the year theywere granted, under the McCain-Levin plan, but could deduct the cost of employee stockoptions only after such options had been vested, exercised and reported as taxable income

by employees. Yet the cost deducted by the employer would not equal the incomereceived by employees, as is the case today with this and every other form ofcompensation, but it would be limited to the amount that had previously been estimatedto have been the fair value of those options when granted years before. If the actual cost

exceeded the early estimates of fair value (which, as explained later, are not evenintended to forecast actual cost), the difference would be taxable to employees but not deductible to employees. This would have been an indefensible bias in tax policy. Yetmany press reports about this scheme in early 2002 completely ignored the tax aspects ofthe McCain-Levin plan and instead focused on alleged improprieties of executive stockoptions at Enron and stretched that political complaint to stock options in general. Whatnew reforms would do, claimed a Wall Street Journal reporter, is shine a spotligh t oncompanies that overuse options, in particular high-tech concerns in Silicon Valley. 21 Proposing to arbitrarily limit tax deductions for one form of compensation was neverabout shining a spotlight.

Trying to connect Enron to political complaints about stock options requiredconsiderable journalistic creativity. The first approach was to suggest that failure toexpense employee stock options in one particular way (i.e., when granted rather thanwhen vested, in-the-money or exercised) had caused Enrons earnings to be seriouslyoverstated, with the result that investors were thought to have been deluded into payingtoo much for Enron stock.

On March 26, 2002, a seminal front-page Wall Street Journal feature by GreggHitt and Jacob M. Schlesinger was titled Stock Options Come Under Fire In the Wakeof Enrons Collapse. Citing Senator Carl Levin, a Michigan Democrat, the articleclaimed executive options pump up the earnings figures and thus deceived investorsinto paying too much for Enron stock. In 2000, Enron issued stock options worth $155million, according to a common method of valuing options [the Black-Scholes model].Had accounting rules forced the company to deduct the cost of those options from its2000 profit . . . Enrons operating profits for the year would have been 8% lower. Thistheme was subsequently echoed uncritically throughout the business press and broadenedto lump most companies together with Enron, eventually producing a thunderousdrumbeat on behalf of deducting estimates of the value of options in the year when theyare granted (like that $155 million). What somehow escaped the Wall Street Journals


9/17

reporters and their editors, however, was the fact that stock options Enron issued in 2000were not worth $155 million but were, in fact, completely worthless. Had accountingrule forced the company to deduct that $155 million estimate from its 2000 profit, thenthe resulting earnings figure would have been erroneous by $155 million.

Several news reports echoed the claim of Senators McCain and Levin that taxdeductions for the cost of stock options were the main reason Enrons U.S. taxesappeared small relative to (grossly overstated) domestic and foreign earnings. Thatclaim was even inconsistent with the flawed Hitt-Schlesinger calculation that theestimated value of newly granted options amounted to no more than 8 percent ofearnings. Yet journalists nonetheless seized on the alleged unfairness of employersdeducting the cost of options as though this was another indictment of options in general.

On March 28, 2002 two days after the misleading Wall Street Journal piece --Washington Post staff writer Charles R. Babcock wrote, During the boom year of 2000,some highly successful companies issued so many stock options to their executives and

employees that they paid little or no corporate tax because the options were deductible.That statement was completely wrong. There is no corporate tax deduction during theyear in which options are granted, so options granted during the boom year of 2000resulted in no tax deductions. Indeed, most options granted during 2000 were so farunder water by March 2002 that they were unlikely to ever produce taxable income foremployees, without which there can be no deductible expense for employers.

On March 27, 2002, one day before the other erroneous reports, Washington Post staff writer Alec Klein wrote that, Despite a tough financial year, AOL Time WarnerIncs chairman and chief executive was rewarded in 2001 with stock options worth anestimated $76 million. Klein at least said estimated, which is more than many similarreports on CEO pay have done. Half of that estimated $76 million depends on AOLstock rising higher than $49 a share about four times the level of early 2003 --otherwise the options will be worthless. The other $38 million of Mr. Cases supposedwindfall depends on AOL stock somehow exceeding $61-73. That is the trouble withestimates they often look ridiculous within a year. Yet proponents of expensing have

been seriously insisting that estimates of exactly this sort should (in some Calvinistsense) be treated as actual expenses and subtracted from earnings. We will have more tosay about this later. Meanwhile, keep in mind that subtracting Mr. Cases estimated $76million booty from AOLs 2001 earnings would indeed have made that companys $4.9

billion loss look even worse. But that certainly does not mean that treating estimates asreality would have provided investors with more accurate information.

Nonqualified options become deductible only after they are vested after three orfour years, in-the money and exercised. At that point, any gain on options becomestaxable to the employee and deductible to the employer. This raises income taxed by theindividual tax just as much as it lowers income taxed by the corporate tax. And since thetop two individual tax rates were higher than the corporate rate after 1993, the net effectwas a windfall for the Treasury. Yet the Washington Post mistake about tax deductions

being taken in the year 2000 for options granted in that same year, and the other


10/17

Washington Post story about Steve Cases options being worth $76 million, and the WallStreet Journal blunder about the estimated value of Enrons worthless options grantedduring 2000, became part of a groundswell of accumulating folklore that, in turn,supported a confused and greatly exaggerated fascination with reforming (in one

particular and rigid way) the manner and timing of stock option bookkeeping. And those

three news reports are merely a two-day sampling from the flood of mediamisinformation about executive stocks options.

In February 2003 the Joint Committee on Taxation (JCT) issued a massive reportin response to political charges that Enrons problems may have been somehowconnected to its executive compensation programs. They had virtually nothing to sayabout stock options, except that many were exercised when the stock was high and manyothers lost value when the stock fell. When it came to effects of taxation on CEOcompensation, however, the JCTs strongest and most relevant recommendation was thatthe limitation on the deduction for compensation in excess of $1 million should berepealed. 22

It could nonetheless be argued that there really were unique peculiarities aboutEnrons compensation plans that might have contributed to executives incentives tomanage earnings in increasingly devious ways. But the fact that these features wereunusual and not confined to stock options is the reason they cannot be plausiblyconverted into a generalized complaint about options. Such successful companies asMicrosoft and Cisco make much greater use of stock options than Enron ever did.

If there is a specific lesson to be learned from Enron about compensation policy,it concerns the pace and duration of vesting rather than the particular form of incentive

pay. Some critics of typical stock option plans claim that restricted stock is a superiorincentive, for example, yet Enron used equal measures of both. Other critics say

performance pay should be indexed that is, made more valuable if company stock gainsexceeded those of some stock index-- but Enron did that too.

Enrons compensation plan involved the usual mix of salary, annual bonuses andlong-term incentive grants which in recent years consisted of one-half restricted stock andone-half stock options. What was unusual about Enrons restricted stock was that it hadan accelerated vesting feature. That meant an executive would be allowed to sellrestricted stock sooner than the usual four year cliff vesting requirement if Enronsstock outperformed the S&P 500 average. What was unusual about Enrons stock optionsis that they vested in only three years but expired in five. Other companies stockoptions plans vest only gradually after three or four years and most employee options donot expire for ten years. Leaving only two years between the time options could beexercised and the time they expired, as Enron did, meant there was a narrow window ofopportunity to either gain much or lose it all. In some cases, stock options awards hadaccelerated vesting if Enron achieved 15% compounded growth in earnings-per-share.

Although it appears arguable that accelerated vesting was an imprudent policy, itwould be equally imprudent to excuse executive activities that resulted in criminal


11/17

charges, and the absence of effective ethical supervision by directors, as merely asensible response to financial incentives. Leaving jewelry on the dresser when yourhouse is being cleaned is an incentive to theft too, but housecleaners who value theirintegrity and reputation never steal.

Gillan and Martin of the Center for Corporate Governance note that $100invested in Enron stock in 1995 grew to $474.61 in 2000, compared to $227.89 for theS&P 500. Enron executives, like Enron stockholders, had comparable increases inwealth in those years, but that is the point of performance-related pay. Enron executivesdid not realize all those paper gains, however, so they faced substantial downside riskright up to the end. Lay and more than 140 senior executives at Enron apparently heldsome $430 million of stock before the firm declared bankruptcy, of which Lay accountedfor $50 million, according to Gillan and Martin; they stood to loose a substantialamount in the event of the companys demise. They also stood to ruin lucrative careers,which is probably the best explanation why accounting tricks accelerated after the houseof cards began to tumble. In the end, Gillian and Martin conclude that many would

argue that Enrons personnel and compensati on programs were not only innovative, butin concert with recommended best practices. 23

Whether or not Enrons compensation plans were what experts regard as best practices, they surely had little to do with the worst practices of many other professionals who are supposed to act as watchdogs and gatekeepers directors, auditors,credit-rating agencies, and stock analysts.

The populist crusade against executive pay during the early 1990s ended upsaddling the U.S. with IRS and SEC regulations that had the unintended consequence ofover-promoting stock options, thus generating huge windfalls for CEOs when the stockmarket boomed.

There is a serious risk that the latest populist crusade against the stock options promoted by the last populist crusade could likewise end up generating unanticipated anddecidedly unpleasant consequences. Executive compensation is a complex and criticaltopic, and one in which the best arrangements are inherently subjective, differing forevery firm and every executive. There is a large body of evidence suggesting thatvoluntary compensation agreements are remarkably efficient in terms of the interests ofcorporate managers and owners. Third parties, such as government regulators and self-styled activists, are generally unwelcome at this bargaining table.

In a recent survey of the evidence on executive compensation for the FederalReserve Bank of New York, Core, Guay and Larker of the Wharton School note that incontrast to the allegations of many media pundits . . . who assert that incentive levels arerandom, arbitrary or out of equilibrium, empirical evidence suggests that, on averag e,firms base their equity incentives on systematic and theoretically sensible factors. 24

Efforts to impose any one-size-fits-all regulations, including FASB rules aboutaccounting for stock options, risk adding many costs and few benefits. When it comes to


12/17

meddling with contracts between consenting adults, the best rule to follow is normallynot IRS, SEC or FASB but MYOB (mind your own business).

Theory and Evidence of Executive Compensation

Theoretical problems of executive compensation involve agency costs and rents.Agency costs arise because corporate managers are not the firms owners but merelyagents ostensibly acting on behalf of principals who supply the firms equity and debtcapital (stockholders and creditors). Unfortunately, agents have their own interests whichcan be the opposite of the interests of principals.

In this context, rent essentially means executives being paid more than wasreally required to attract, retain and motivate them. Legal theorists Bebchuck, Fried andWalker

Facts and Conjectures about Executive Stock Options

Stock options give some employees the right to buy a certain number of companyshares in the future, almost always at todays price. According to a Bureau of LaborStatistics survey, about 78 percent of ESOs are nonqualified, which means any gainswill be taxed at ordinary income tax rates. Qualified or incentive stock options (ISOs)were provided to m or e than 31 percent of all employees, notes the BLS, but withminimal overlap. 25 Incentive stock options must be held long after they are exercised,so they are ostensibly taxed at the lower 20 percent rate for long-term capital gains (butcan actually be taxed by the 28 percent alternative minimum tax whether there is any gainor not).

The current controversy is entirely focused on nonqualified options. Such optionsare granted at the market price at the time of the grant, called the exercise or strike

price. They usually require 3-5 years of vesting before they can be exercised (whichmeans buying the stock at the previous strike price). Vesting can be all at once (cliffvesting), but often requires more gradual exercises, such as one-fourth per quarter. Theoptions usually expire within ten years of the grant and are most often exercised soonafter they are in the money that is, as soon as the stock price is significantly higherthan the strike price. If the market price stays below the strike price, options are said to

be under water. Employee stock options become worthless and are forfeited if theyare not in the money between the time of vesting and expiration, or if the employees quitsor is retired (fired).


13/17

Estimating Options Means Estimating Earnings

In 2003, the seven-member Financial Accounting Standards Board (FASB) voted

unamiously to reconsider requiring US corportions to treat employee stock options as anestimated expense when the options are granted. Londons International AccountingStandards Board was contemplating similar regulations.

Many business reporters have described this topic as merely a moral battle withgreedy executives and lobbyists on one side and heroic defenders of honesty and virtueon the other. The issues are more complicated than that, but they can be boiled down totwo simple points: First of all, a possible future expense is not the same as an actualcurrent expense. Second, the estimated fair value of options to employees when they aregranted is not the same as the actual expense to employers if and when those options arelater exercised.

On the surface the issue of how and when to account for employee stock optionsmay look like a trivial bookkeeping detail. Yet the idea has generated a heated debate,even among academics. Writing on this page last year, economists Burton Malkiel ofPrinceton and William Baumol of New York University warned that this politicizedaccounting crusade risks destruction of equity compensation instruments that have beenengines or innovation and entrepreneurship.

An entirely different perspective was offered in a March 2003 Harvard Business Review article For the Last Time: Stock Optio ns Are an Expense by Professors ZviBodie, Robert S. Kaplan and Robert C. Merton. 26 The title and much of the text seemedto challenge an argument no serious critic of expensing ever made namely, to suggestthat options are either worthless to employees or a free lunch for employers. The realdebate has to do with just what that expense is, when it occurs, and how and when itshould be reported on company books.

Companies with a highly mobile labor force and a highly uncertain future like tooffer stock options precisely because there is an excellent chance those options will costthe firm nothing. Executives put up with that downside risk for the same reason people

buy lottery tickets options offer a slim chance of becoming very rich.

Before the Bodie, Kaplan and Merton article appeared, the December 2002 Harvard Business Review published Expensing Options Solves Nothing by William A.Sahlman of the Harvard Business School. Sahlman emphasized that the issue ofexpensing options has been a huge distraction from serious issues of deceptiveaccounting. He noted that options have the earmarks of an investment in the futurerather than an operating expense. On balance, Sahlman concluded, options should not

be expensed, particularly if it entails disclosing less information in their footnotes. 27 Under current FASB rules, companies that expense options are not required to providethe information that others do about how many options were granted, outstanding,


14/17


15/17

by share repurchases, as a reduction in cash on hand). The fact that all estimates prior toexercise had to be revised to conform with the actual cost at the time of exercise clearlydemonstrates that all previous estimates were incorrect and that basing reported earningson such estimates therefore made the earnings figures incorrect.

Proponents of expensing estimates at the time options are granted say grantingoptions to an executive is no different from granting stock. But options, unlike companyshares, have big strings attached: They cannot be sold until after a few years of vestingand they become worthless if the stock falls or the excutive quits or is fired. Indeed, thisis why options are so considered so important by employers, particularly in firms with ahighly uncertain future.

Granting an employee the right to buy shares at the current price after a few yearsof vesting creates a contingent liability to provide those shares to employees if (1) thestock price goes up and (2) the employee remains with the firm long enough to be vested.If and only if those two contingencies are met, the company must either issue more

shares, or it must dip into cash to repurchase shares in the stock market. Dilutionobviously reduces earnings per share -- the only measure of earnings that matters when itcomes to stock prices per share. And share repurchases bviously reduce company cashand (because that cash would otherwise have been invested) future earnings. Thecommon claim that the cost of employee options to shareholders is not reported inearnings per share is flatly untrue.

If we could agree that the cost of options consists of dilution (or sharerepurchases) at the time of exercise, then there would be no reason to expect that cost to

be even approximated by a Black-Scholes estimate of value at the time options aregranted. Estimating the value of options when they are granted is not at all the same asforecasting the cost if and when they are exercised. The Black-Scholes model is based oninformation available at a particular moment in time. It was never intended to forecastthe future value of stocks or options. The value of options today is not the same as thefuture cost of deferred and contingent compensation, and estimates of the former are notforecasts of the latter.

Potential dilution from options that are worth anything is prominently displayedin every earnings report as diluted earnings per share. Bodie, Kaplan and Merton,however, prefer a more academic definition of cost, one which is not at all comparable tothe way companies measure other costs. They say the opportunity cost of grantingstock options consists of the cash the company could otherwise have received by sellingthose options to investors. This is, they claim, is exactly like giving stock to employeesrather than selling it. But is it? No option market is remotely comparable to employeeoption grants. Giving options to employees is not at all comparable to giving them stock

because the options (unlike stock) will have value only if the employee remainsemployed at least long enough to be vested and (if the stock has not risen before then)

possibly for ten years.. No investor purchasing stock or options has to worry about beingfired before the option can be exercised, but employees holding stock options certainly


16/17

do. Besides, selling options to buy a companys stock several years from now at todays price would be like announcing that management expects the stock to fall.

The main problem with estimating earnings by estimating the cost of employeestock option is not that simply that Black-Scholes estimates are inherently inaccurate

(which they are), but that such estimates are essentially irrelevant to the actual cost ofexercised options.

The reason this matters is that

1 Whats Wrong with Executive Compensation?: A Roundtable Moderated by Charles Elson, The Harvard Business Review (January 2003) p. 69.2 Geoffrey Colvin, The Great CEO Pay Heist, Fortune (June 25, 2001) p. 64.3 Brian J. Hall and Jeffery B. Liebman, The Taxation of Executive Compensation National Bureau ofEconomic Research Working Paper No. 7596 (March 2000) p. 3.4

Michael C. Jensen and Kevin J. Murphy, CEO Incentives Its Not How Much You Pay, But How,The Harvard Business Review (May-June 1990) pp. 138-153.5 Nancy L. Rose, Executive Compensation, National Bureau of Economic Research, NBER Reporter,(Winter 1994-95) p. 11.6 Joseph E. Bachelder, Proposed Tax Changes: Stock Options, New York Law Journal , March 29, 1993.7 FASB Rule Rejected, The San Francisco Chronicle , May 5, 1994, p. D2.8 Jensen and Murphy, op. cit. 9 Andrew K. Prevost and John D. Wagster, Impact of the 1992 Changes in the SEC Proxy Rules andExecutive Compensation Reporting Requirements, mimeo, Wayne State University School of BusinessAdministration (September 1999).10 Governor Bill Clinton and Senator Al Gore, Putting People First , New York, Times Books (1992) p. 31.11 Brian J. Hall and Jeffrey B. Liebman, The Taxation of Executive Compensation, National Bureau ofEconomic Research, NBER Working Paper No. W7596 (March 2000).12 Benjamin C. Ayers, C. Bryan Cloyd and John R. Robinson, Capitalization of Shareholder Taxes inStock Prices: Evidence fro the Revenue Reconciliation Act of 1993, mimeo, Tull School of Accounting,University of Georgia (October 20, 2000).13 Myron S. Scholes, Mark A. Wolfson, Merle Erickson, Edward L. Maydew and Terry Shevlin, Taxes and

Business Strategy, Prentice Hall (2 nd edition, 2002) p. 77.14 Louis Lavelle, Executive Pay, Business Week (April 16, 2002), p. 80.15 Jerry Useem, Have They No Shame? Fortune , April 28, 2003. The article describes an increase inmedian compensation more telling than the mean average, but that is dubious spin. The only way themedian could have risen while the mean fell was for the largest pay packages to have been deeply cut.16 Patrick McGeehan, Executive Pay: A Special Report The New York Times (April 6, 2003).17 Paul Krugman, For Richer: How the permissive capitalism of the boom destroyed American equality,The New York Times Magazine (October 20, 2002) p. 64. On the widespread misuse of income distributionstatistics, by Krugman and others, see Alan Reynolds, Economic Foundations of the American Dream inLamar Alexander and Chester E. Finn, Jr., eds., The New Promise of American Life , Indianapolis, Hudson

Institute (1995) pp. 194-220.18 AFL-CIO, Trends in Executive Pay http://www.alfcio.org/corporateamerica/paywatch/pay/index.cfm ? Accessed May 15, 2003.19 U.S. Census Bureau, Statistical Abstract of the United States (2002). Table No. 607 p. 400.20 For . . . Citizen for Tax Justice purposes, the effective tax is defined as taxes paid currently divided bynet income before tax. The numerator includes not only implicit taxes but also tax deferrals. Scholes, op.cit. , p. 495n.21 Janet Whitman, Stock Option Awards Undergo Scrutiny In Wake of Outcry for More Accountability,The Wall Street Journal , April 3, 2002.
http://www.alfcio.org/corporateamerica/paywatch/pay/index.cfmhttp://www.alfcio.org/corporateamerica/paywatch/pay/index.cfmhttp://www.alfcio.org/corporateamerica/paywatch/pay/index.cfmhttp://www.alfcio.org/corporateamerica/paywatch/pay/index.cfm


17/17

22 Joint Committee on Taxation, Report of Investigation of Enron Corporation and Related EntitiesRegarding Federal Tax and Compensation Issues, and Policy Recommendations (February 2003) Volume1, p. 20.23 Stuart L. Gillan and John D. Martin, Financial Engineering, Corporate Governance and the Collapse of

Enron, University of Delaware Center for Corporate Governance, Working Paper WP 2002-001(November 6, 2002) pp. 29-32.24 John E. Core, Wayne R. Guay and David F. Larcker, Executive Equity Compensation and Incentives:A Survey, Federal Reserve Bank of New York, FRBNY Policy Review (April 2003), p. 32.25 Beth Levin Crimmel & Jeffrey L. Schildkraut, Stock Option Plans Surveyed by NCS [NationalCompensation Survey], U.S. Bureau of Labor Statistics , Compensation and Working Conditions (Spring2001) p. 5.26 Zvi Bodie, Robert S. Kaplan and Robert C. Merton, For the Last Time: Stock Options Are an Expense,The Harvard Business Review (March 2003) pp. 63-71.27 William A. Sahlman, Expensing Options Solves Nothing, The Harvard Business Review (December2002), pp. 91-96.

Documents

Executive Compensation: A Tilted Playing Field