PDF File: 

cfo3_4_118_wp_corplibrary_hodgson.htm

Corp. Library

When setting executive compensation, boards should consider a variety of incentive options.

First, a cautionary statement. The following maxims are not laws. If there is a good argument to break with the advice set out in this paper, boards should follow their own arguments. But in the interests of shareholder communication and disclosure, the board should explain exactly why it has taken this course of action.

Maxim #1: There is no one-size-fits-all incentive plan.

The question that most often follows any criticism of an incentive plan is: What should it look like? There is no single answer to this question that would apply to all companies. Companies have different strategic aims, different views on how to achieve those aims and different views on how best to incentivize management to achieve those aims. It is up to the directors on the board, who are, or should be, fully aware of strategy, to decide on the most effective incentives and the metrics to use to measure performance. While there is no one answer, there are universal pieces of advice that most plans would do best to follow. These comprise the remainder of the performance pay maxims.

Maxim #2: Not all CEOs should be eligible for annual incentives.

Rather than accepting the universal adoption of annual incentives as a standard part of the executive compensation package, boards should question it.

The typical executive compensation package builds on the foundation of base salary, with annual bonus as the first floor and long-term incentives the second, third and, in some cases, fourth story. While this reflects common practice, it is worth re-examining this structure. The objective of a board is to build long-term shareholder value. Therefore the foundation of an incentive compensation policy should be a plan that rewards the long-term decisions that create this value. The annual incentive plan, if there is one, can then become a vehicle for rewarding managers who achieve steps on the way to long-term shareholder value growth. Rather than supporting it, however, many existing annual bonus plans discourage focus on the company’s long-term strategy.

For example, if the annual bonus plan is based on achieving earnings targets, but earnings must suffer in one year in order for the company to increase its research and development (R&D) budget, managers may be unlikely to accept such a decision because they will lose out on their annual bonus, despite any notional long-term gains. If, on the other hand, a long-term incentive plan designed to reward long-term value growth is the dominant incentive program, then the annual bonus can be adapted each year to reward progress toward this end. In the previous example, rather than the annual bonus rewarding high earnings, managers would be rewarded for:

• Minimizing the effect on operating expenses of investing high amounts of capital in R&D;
• Managing the recruitment of R&D staff that adds value to the company; and
• Driving R&D processes that will bring new products to market.

At many companies, long-term decisions and long-term strategies are already the dominant thought processes. Yet many annual incentive plans undermine this long-term focus. Compensation policy should support long-term strategy by constructing programs where long-term compensation dominates both long-term and annual incentive payments. Only in this way will incentive plans cooperate with each other and not encourage conflicting behaviors.

Maxim #3: It is not possible that a single form of long-term incentive is effective at 99 percent of all companies.

It does not make sense that a single form of long-term incentive – the stock option – should be appropriate for 99 percent of all companies. This is not to say that stock options are never effective. Stock options can be extremely effective, but there is so much more to the stock option than the simple, time-vested, market-priced stock option that is the current favorite. It is time for boards to wake up to these other opportunities: performance-vesting stock options; premium-priced stock options; and index-linked stock options.

During the bull market of the 1990s, stock options appeared to be very effective at rewarding executives for rising stock prices. Management and shareholders were making money, so no complaints were heard. But when the market turned downward, executives continued to make money while shareholders lost money, so the system seemed to be broken. In fact, the system was broken all along. The link between pay and performance gave the appearance of being a close one. But a rising tide lifts all boats, and all executives could benefit from the award of stock options, whether or not their actions had directly or indirectly led to the increase in the company’s stock price. In best practice, only those whose boat’s rise exceeded others should have benefited.

In most circumstances, a mix of different types of award will be the most effective. In some situations, where, for example, a stock option award may lose value, a performance share grant will reward exceptional performance under difficult circumstances, and a grant of career shares – restricted stock that will only vest on retirement – could tie an executive to the company over the long term, ensuring that the company does not lose executive talent when it is most needed.

Just as important is the effective disclosure of such a set of awards. Consider the following examples:

• Stock options are underwater: The company should reaffirm that they are not going to be repriced.
• Performance shares are being awarded and may pay out: A detailed explanation of the choice of the performance metrics that might lead to these payouts should be made.
• Restricted stock is being awarded: It should be made very plain that there are no circumstances under which the award will vest early, or at all, if the officer leaves before retirement.

Maxim #4: Metrics should not be either absolute or relative; they should be both.

Metrics should be both absolute and relative ... in the majority of cases. If metrics are solely absolute, then a situation has been created where a company could pay out for a level of achievement that met or exceeded the goal but was, nevertheless, far below peers. On the other hand, if metrics are solely relative, then a situation has been created where a company could pay out for negative returns just because its peers had even worse returns. Neither of these situations is desirable. In some cases, obviously, the choice of either an absolute or a relative measure can be effective, but the philosophy behind this choice should be carefully considered in the boardroom and carefully explained in the proxy statement.

Maxim #5: Performance assessment should measure achievement of strategy.

Performance metrics should focus executives’ attention on the company’s strategic plan, not simply on the most common, quantitative measures like earnings or total shareholder return (TSR). Just because everyone else chooses earnings and TSR does not make it right. It may be the case that, at every board meeting, directors and managers sit around discussing where the stock price is going and nothing else, but it is unlikely.

In addition to these strategic measures, the range of metrics used should generally include some form of value-growth measure. Furthermore, in addition to purely quantitative metrics, boards should consider the use of qualitative measurements. It is often claimed that qualitative measures are more difficult to test and more open to abuse than quantitative measures. It may be true that it is harder to detect abuse for qualitative measures, but the many examples of revenue and earnings manipulation suggest that they may not be more open to abuse than quantitative ones. Nevertheless boards and compensation committees should be very careful to explain the methodology of measurement and the results. In many cases, it will be simpler to employ an outside agency to measure these achievements.

Maxim #6: Single performance metrics are, by their very nature, more open to abuse than two or more metrics.

Two examples should demonstrate why this maxim is true.

The board of Widget Inc. sets a simple revenue target for its annual incentive plan. The CEO and the COO have been pushing sales all year and the company has been very successful securing new customers, but much of the revenue from those widget sales is set to come in during the first quarter of the next fiscal year. The CEO and the COO take the decision to prematurely recognize the income from those sales in order to meet the revenue target set by the compensation committee. They “meet the target,” and bonuses are paid out. A year later, the SEC launches an investigation to uncover premature booking of revenue at the company. They discover evidence that it has taken place and the accounts have to be restated. The CEO and the COO both resign and are required to pay back the bonuses that were based on the misstated accounts. Damage is done to the company’s reputation, and the stock price drops because of the investigation and the sudden loss of two senior managers. The whole situation could have been avoided if the annual bonus had been based on the achievement of more than just a revenue target.

Disenchanted with revenue, the board of Widget Inc. decides to set earnings as a target for the next annual bonus. Halfway through the year, the new CEO and the senior management team realize that, not only do they have no chance of meeting the earnings target under current business conditions and strategy, but that it was unrealistic in the first place. They find a way to meet it, nevertheless. Earnings can be increased in the short term by drastically reducing the workforce, cutting the research budget to the bone and disposing of profitable businesses. The management team takes these actions in the name of “efficiency.” But what happens in the long term? Talent is lost, so business development is harder; there are not enough new products coming to market because the R&D department has been virtually closed down; and the one-time gains from the disposals disappear – along with the annual earnings – in the next fiscal year. Earnings targets have been met, but value has been destroyed. Again, this situation could have been avoided had the board made the annual bonus dependent on other measures as well as earnings.

Maxim #7: Concentration of too much compensation in one form of incentive is, by its very nature, more open to abuse than spreading compensation over a variety of incentives.

This maxim needs little explanation. If 90 percent of an executive’s compensation is derived from stock options, then there is an enormous temptation to manipulate stock price – either by backdating options or timing the grant or exercise of options to coincide with abnormal lows or highs, respectively, or simply by lying to the markets. If compensation is spread around three or four different forms of award, each based on a different set of metrics, then the temptation to manipulate any one of these is significantly lessened.

Maxim #8: It is not good policy to use the same performance metric for more than one incentive plan.

It would seem common sense, but it is startling how many companies base the payout of an annual bonus and a three-year performance share plan on achieving, say, earnings targets. The board is effectively saying, “We’ve paid you three times for achieving these earnings targets over the last three years, and now we’re going to pay you again.” In the same way, if stock options are awarded, any other type of long-term incentive should not then use TSR, because TSR is just another stock-price-based metric.

Maxim #9: Performance periods should accord with the company’s strategic plan.

A quick analysis of the operation of 95 percent of all long-term incentive plans seems to indicate that the U.S. economy consists of companies whose strategic plans cover the next three years. This, of course, is not only unlikely, it is impossible. There is significant resistance, however, to measuring performance over longer periods. For example, it is said that, with CEO tenure at its current levels, attempting to incentivize someone over longer periods is pointless. This is a weak argument at best. It might even be because CEOs are so rarely incentivized over longer periods that they so rarely stay to see out long-term strategy. It is also claimed that stock options that expire after 10 years are effectively incentivizing executives over this period. This is also unconvincing. Firstly, there is nothing long term about an incentive that begins to vest and become exercisable after a single year and is wholly exercisable after four to six years, at the very longest. Secondly, if executives typically hold on to their options for 10 years, then 99 percent of Black-Scholes valuations are significantly underestimating the grant-date value of stock options.

There is no silver bullet for revising the three-year standard. If it takes six years to get a pharmaceutical product to market, if it takes 15 years from finding an oil field to filling the first barrel – these are the performance periods that should be built into plans. If strategy is planned out for 30 years, it is unreasonable to expect executives to wait this long for compensation, so boards must set goals along the way. Not every three years, however; these should be set every five years at the least, but preferably every 10.