The Radical Experiment in Pay Regulation under TARP

This is the second in a two-part series on executive compensation and financial service regulation. Part one examines the history of executive compensation and how it is influencing today’s policy debate over regulation reform.

Most people in and out of the financial industry believe that perverse incentives were at least partly responsible for the crisis prompting the TARP program, initiated in the Emergency Economic Stabilization Act (EESA). Stronger alignment of managers and shareholders would likely have prevented much of the damage we are now facing, and it’s sensible that financial services reforms address those incentives. EESA includes a provision that calls on the elimination of compensation that creates “unnecessary and excessive risks,” a provision reiterated in the American Recovery and Reinvestment Act (ARRA).

Unfortunately, EESA and AARA depend on the Treasury Secretary to make distinctions about what incentives do or do not contribute to excessive and unnecessary risk. This is not an easy call. Even the most seasoned corporate boards, with considerable experience in managing or overseeing financial services firms, have a very difficult time distinguishing compensation-induced risk, even without the political considerations that might drive directors to conflate governance risk and business risk.

For instance, most people equate compensation plans with steep pay-for-performance lines as “riskier” than plans with weaker incentives. Critics argue that stronger incentives induce reckless behavior. In fact, shareholders have been placed in far more danger when the incentive leverage is zero than when it’s steep. The area below targets or goals where there is no longer any more pay-for-performance is where the trader finds his or her greatest temptation to double down in the hope of getting back into the green. This phenomenon is well known in banking circles as the “trader’s option.” The fundamental problem is not the steepness of the plan when the participants are on the incentive line, but the discontinuous incentive leverage across the spectrum of performance. This problem is reinforced by discontinuities across time, characteristic of bonus plans with specific end dates, like the November 30 fiscal year-end for many banks, which means to a manager that anything that happens after that date is of no consequence to my bonus this year. These discontinuities create asymmetry between the risk preferences of the employee versus the shareholders.

Keeping key executives on the incentive line

Directors would benefit from better understanding where these discontinuities exist in their firm’s incentive structures. Boards should be wary of any incentive plan that suddenly kicks in millions of dollars, or their share-equivalent, for hitting a specific achievement target, something known as performance-based cliff-vesting, which happens to be very popular among governance critics right now. Most governance risk faced by shareholders appears in that area in front of the cliff, where plan participants have powerful incentives to do silly things.

Boards should be wary of any plan that takes key employees they intend to keep off the incentive line. The point where they no longer have upside potential, they also no longer have any downside accountability. Most “annual plans” look like this in bad years. Corporate critics, however, are appalled that anyone should have any bonus opportunity in a range of below-target performance.

In principle, a performance-based claw-back mechanism could help overcome time discontinuity. If a bonus were based on performance that turned out to be unsustainable, a portion of the prior earned amount, presumably held in a deferred account, would be returned in accordance with the performance shortfall. I have actually designed performance-based ‘claw-backs’ like this at client firms. When implemented correctly, such a mechanism eliminates many incentives for short-term behavior, both in reporting misleading results and in pursuing unsustainable strategies. A couple of major financial institutions are now adopting these mechanisms. But they are not costless, in the sense that making bonuses more deferred and conditional requires tolerating a greater retention risk or offering a higher target level of compensation.

The EESA includes a claw-back provision, but this rule merely expands a prior restriction introduced by Sarbanes-Oxley to require repayment of bonuses based on false results. It doesn’t apply to results that were correctly reported, but were simply not sustainable.

Ironically, one can eliminate virtually all “unnecessary and excessive risk” by eliminating all objectivity from the awarding of bonuses. Adopting subjectivity flies in the face of decades of good governance recommendations, but it solves both the problem of discontinuous incentive leverage and the problem of short-term versus long-term. A subjective plan has no identifiable incentive leverage to game. A supervisor can judge the sustainability of performance meriting a bonus far better than could any bonus formula. It might be no accident that the two major financial firms with the most subjective bonus plans, Goldman Sachs and J.P. Morgan Chase, were among the least affected by the subprime crisis, and the firms with the most rigidly formulaic bonus plans were Bear Stearns and Lehman Brothers.

Specific TARP limits on compensation

Congress has always hated golden parachutes. “Golden parachutes” is actually slang invented by the press. No executive agreement actually uses the term. Nevertheless, Congress banned “golden parachutes” for top executives of TARP firms in the EESA. In the ARRA, they amended the law to actually include a legal definition of that term.

This ban is the clearest illustration both of the degree to which Congress legislates based on the critics’ depictions of compensation, and the general belief that such features -- whatever they are called --are gratuitous benefits born of managerial power, simply tossed into a pile by complacent boards rather than incentives that are actually worth more than the actual cost to the company. At the very least, this ban will force boards to renegotiate with their CEOs with one less benefit to offer that will eventually need to be replaced by another.

The boldest proposal in EESA is the $500,000 tax limit on the top corporate officers. Unlike the current $1 million limit on tax-deductible pay Congress passed in 1993, this $500,000 limit has no performance-based exceptions. On the one hand, it eliminates the artificial preference for variable compensation, which may restore some balance between fixed and variable pay. But we can soon expect to see the first multi-million dollar salaries in the financial services sector, albeit with much smaller bonuses or equity grants for these executives. All of the resulting tax penalty, aimed at the executives, will initially hit the shareholders, and eventually workers and customers, too. Congress will have to learn another lesson in tax incidence.

ARRA includes some additional provisions in direct reaction to the headlines since the passage of EESA, specifically the $18.4 billion in bonuses paid out to Wall Street employees for 2008, which still added up to below-target compensation for most of the recipients. ARRA’s key provisions include:

• Bonus limited to 50 percent of salary for top earners, which bonus may be taken only in restricted stock, i.e., vested when they repay TARP funds,
• Prohibition on any incentive plan that would encourage manipulation of the reported earnings,
• Board committee to review expenditures on “luxury” items,
• General requirement for Treasury to judge if compensation is “unreasonable or excessive” or “in the public interest.”

The law nominally imposes its bonus limits on up to the top 20 “most highly-compensated employees” (for the largest recipients) plus the five “senior officers.” But this restriction poses a logical quandary. Somebody has to be the 26th highest paid person. One possible consequence is that every employee at a major bank becomes limited to a bonus that brings their total compensation no higher than 150 percent of the salary of the 25th highest salaried employee. If that is how the rule gets implemented, the first thing we should expect is a rapid increase in salaries at TARP recipient firms, which I believe we are starting to see. There are other interpretations to this rule, but they don’t make any more sense than this scenario.

A reasonable interpretation of the second item would be that earnings can no longer be used as a bonus metric, since any incentive to achieve is indistinguishable from an incentive to cheat. This would also rule out most forms of profit sharing, and probably any financial metric, since all of them affect earnings and are, in principle, subject to manipulation. That basically leaves non-financial metrics, which, while notoriously manipulable, are allowed only as long as they can’t influence earnings. The safest bet would be to award bonuses based purely on subjective assessments that, as mentioned earlier, are something governance critics generally oppose.

The clearest evidence of popular envy being the root of these provisions is the clause requiring the board to review “luxury” expenditures—a kind of Optics Committee—for items like office decorations and corporate jets. Clearly, these expenditures are far more material to the press and public than they are to the shareholders. Does it really matter to the shareholders if the company’s most valued employees benefit from nice offices? Sure, there is a crude sense of entitlement that drives an executive to spend a lot on lavish appointments, but such appointments tend to be personalized, which arguably creates a retention benefit, especially in situations where bonuses are otherwise being suppressed. Corporate jets might be a frivolous expense, or they may be a cost effective way to provide efficient and secure transportation to busy executives. The ARRA law leaves it up to the board to make such determinations on behalf of the shareholders, but this is a canard; boards already make these determinations. The real purpose of this clause is use public sentiment to eliminate perks on no firmer grounds than the grade school “chewing gum rule”—if I can’t have some, nobody else can.

The most vexing limits associated with ARRA will be the ones executives didn’t know existed because the Treasury Secretary was later embarrassed by some aspect of the plan and decided that paying under a contract, even one approved by the government, was “contrary to the public interest,” a term necessarily left undefined in the law. This provision was not even broad enough to catch most of the infamous AIG bonuses, which were otherwise protected by contract protections of the law. But these impediments to Treasury’s recovering bonuses not “in the public interest” are up for grabs in proposed legislation.

An uncertain trend

In all of these pronouncements, Congressional leaders are asserting that they don’t want taxpayer money to be “wasted” on certain forms of compensation. But if the only difference between TARP and other troubled firms is the composition of the ownership, i.e. whether or not taxpayers are among the shareholders, then this claim is tantamount to saying that the proscribed forms of compensation are always wasted. What, then, will prevent Congress from using that rationale to more broadly apply these kinds of standards?

As long as new laws continue to be written in response to the newest headlines, it’s difficult to anticipate what the government will do next. In the immediate aftermath of the AIG-FP bonus flap, the House passed an incredible 90 percent tax on the bonuses of all financial services employees making over $250,000 in family income. The Senate and White House were lukewarm to this, on Constitutional grounds as well as common sense, and the measure has since been watered down to merely expand the Treasury Secretary’s power to determine what is “unreasonable or excessive” under older agreements as well as new ones.

The ARRA requires of all TARP companies annual approval of executive compensation in a non-binding shareholder vote. The intent is for shareholders to keep boards honest in awarding compensation. Looking ahead, we will almost certainly see this “say on pay” requirement expanded to all public corporations, very likely in this session of Congress. Since this requirement is entirely underpinned by the managerial power thesis, the least likely outcome of its adoption will be a reduction in CEO pay. Its most likely effect will be greater politicization of executive compensation, which is how the union pension funds that are its main proponents appear to define corporate governance.

It is fashionable to joke about how much bankers can really be worth, given the mess in which banks now find themselves. But for a sector as important as financial services, especially to the extent that any specific group is responsible for our current crisis, doesn’t that make it all the more necessary to get the right talent into key positions? Isn’t it critical to insure that this talent is motivated to do the right things? By belittling or ignoring the need to attract and retain talent in financial services, the prospect for future legislation does no bode well for a balanced treatment of executive compensation in this sector.

Conclusion

The government certainly has the prerogative to create crazy restrictions on the companies it owns, including with regards to pay. But it can’t argue that such restrictions are good for the taxpayers-as-owners when they’re not. In thinking about a world after TARP, it is important for regulators to broaden the discussion of executive compensation beyond an incomplete assessment of alignment and an obsessive focus on cost. To the extent that government must address cost for political reasons, policy makers are more likely to gain traction on controlling cost by abandoning the “managerial power” theory in favor of one that better explains why costs have soared. They must realize that permanently increased politicization of compensation and increased complexity are not the shareholder’s friend, and consider how their direct involvement has and might affect these vitally important issues.

These suggestions assume that Congress is truly motivated by a desire to improve corporate governance for the benefit of the shareholders, when, in fact, its behavior often demonstrates otherwise. The “g” word we continuously hear in their debates is not “governance” but “greed.” Greed is distinguished solely by the raw size of the paychecks being discussed, which obliterates any discussion of the governance trade-offs associated with pay.

Author Bio

Marc Hodak is Managing Director of Hodak Value Advisors, a firm specializing in the finance and compensation issues of corporate governance. Marc teaches corporate governance at New York University’s Leonard N. Stern School of Business. He can be reached at mhodak@hodakvalue.com.