Twenty years ago, the biggest component of executive compensation was cash, in the form of salaries and bonuses. Stock options were just a footnote. Now the reverse is true. With astounding speed, stock option grants have come to dominate the pay—and often the wealth—of top executives throughout the United States. Last year, Jack Welch’s unexercised GE options were valued at more than $260 million. Intel CEO Craig Barrett’s were worth more than $100 million. Michael Eisner exercised 22 million options on Disney stock in 1998 alone, netting more than a half-billion dollars. In total, U.S. executives hold unexercised options worth tens of billions of dollars.
It would be difficult to exaggerate how much the options explosion has changed corporate America. But has the change been for the better or for the worse? Certainly, option grants have improved the fortunes of many individual executives, entrepreneurs, software engineers, and investors. Their long-term impact on business in general remains much less clear, however. Even some of the people who have profited most from the trend express a deep discomfort about their companies’ growing dependence on options. Do we really know what we’re doing? they ask. Are the incentives we’re creating in line with our business goals? What’s going to happen when the bull market ends?
Option grants are even more controversial for many outside observers. The grants seem to shower ever greater riches on top executives, with little connection to corporate performance. They appear to offer great upside rewards with little downside risk. And, according to some very vocal critics, they motivate corporate leaders to pursue short-term moves that provide immediate boosts to stock values rather than build companies that will thrive over the long run. As the use of stock options has begun to expand internationally, such concerns have spread from the United States to the business centers of Europe and Asia.
I have been studying the use of option grants for a number of years now, modeling how their values change under different circumstances, evaluating how they interact with other forms of compensation, and examining how the various programs support or undermine companies’ business goals. What I’ve found is that the critics of options are mistaken. Options do not promote a selfish, near-term perspective on the part of businesspeople. Quite the contrary. Options are the best compensation mechanism we have for getting managers to act in ways that ensure the long-term success of their companies and the well-being of their workers and stockholders.
But I’ve also found that the general nervousness about options is well warranted. Stock options are bafflingly complex financial instruments. (See the sidebar “A Short Course on Options and Their Valuation.”) They tend to be poorly understood by both those who grant them and those who receive them. As a result, companies often end up having option programs that are counterproductive. I have, for example, seen many Silicon Valley companies continue to use their pre-IPO programs—with unfortunate consequences—after the companies have grown and gone public. And I’ve seen many large, sleepy companies use option programs that unwittingly create weak incentives for innovation and value creation. The lesson is clear: it’s not enough just to have an option program; you need to have the right program.
A Short Course on Options and Their Valuation
Before discussing the strengths and weaknesses of different types of programs, I’d like to step back and examine why option grants are, in general, an extraordinarily powerful form of compensation.
The Pay-to-Performance Link
The main goal in granting stock options is, of course, to tie pay to performance—to ensure that executives profit when their companies prosper and suffer when they flounder. Many critics claim that, in practice, option grants have not fulfilled that goal. Executives, they argue, continue to be rewarded as handsomely for failure as for success. As evidence, they either use anecdotes—examples of poorly performing companies that compensate their top managers extravagantly—or they cite studies indicating that the total pay of executives in charge of high-performing companies is not much different from the pay of those heading poor performers. The anecdotes are hard to dispute—some companies do act foolishly in paying their executives—but they don’t prove much. The studies are another matter. Virtually all of them share a fatal flaw: they measure only the compensation earned in a given year. What’s left out is the most important component of the pay-to-performance link—the appreciation or depreciation of an executive’s holdings of stock and options.
As executives at a company receive yearly option grants, they begin to amass large amounts of stock and unexercised options. The value of those holdings appreciates greatly when the company’s stock price rises and depreciates just as greatly when it falls. When the shifts in value of the overall holdings are taken into account, the link between pay and performance becomes much clearer. Indeed, in a study I conducted with Jeffrey Liebman of Harvard’s Kennedy School of Government, we found that changes in stock and stock option valuations account for 98% of the link between pay and performance for the average chief executive, while annual salary and bonus payments account for a mere 2%.
By increasing the number of shares executives control, option grants have dramatically strengthened the link between pay and performance. Take a look at the exhibit “Tying Pay to Performance.” It shows how two measures of the pay-to-performance link have changed since 1980. One measure is the amount by which an average CEO’s wealth changes when his company’s market value changes by $1,000. The other measure shows the amount by which CEO wealth changes with a 10% change in company value. For both measures, the link between pay and performance has increased nearly tenfold since 1980. While there are many reasons American companies have flourished over the last two decades, it’s no coincidence that the boom has come in the wake of the shift in executive pay from cash to equity. In stark contrast to the situation 20 years ago, when most executives tended to be paid like bureaucrats and act like bureaucrats, today’s executives are much more likely to be paid like owners and act like owners.
Given the complexity of options, though, it is reasonable to ask a simple question: if the goal is to align the incentives of owners and managers, why not just hand out shares of stock? The answer is that options provide far greater leverage. For a company with an average dividend yield and a stock price that exhibits average volatility, a single stock option is worth only about one-third of the value of a share. That’s because the option holder receives only the incremental appreciation above the exercise price, while the stockholder receives all the value, plus dividends. The company can therefore give an executive three times as many options as shares for the same cost. The larger grant dramatically increases the impact of stock price variations on the executive’s wealth. (In addition to providing leverage, options offer accounting advantages. See the sidebar “Accounting for Options.”)
Accounting for Options
The idea of using leveraged incentives is not new. Most salespeople, for example, are paid a higher commission rate on the revenues they generate above a certain target. For instance, they might receive 2% of sales up to $1 million and 10% of sales above $1 million. Such plans are more difficult to administer than plans with a single commission rate, but when it comes to compensation, the advantages of leverage often outweigh the disadvantages of complexity.
The Downside Risk
If pay is truly to be linked to performance, it’s not enough to deliver rewards when results are good. You also have to impose penalties for weak performance. The critics claim options have unlimited upside but no downside. The implicit assumption is that options have no value when granted and that the recipient thus has nothing to lose. But that assumption is completely false. Options do have value. Just look at the financial exchanges, where options on stock are bought and sold for large sums of money every second. Yes, the value of option grants is illiquid and, yes, the eventual payoff is contingent on the future performance of the company. But they have value nonetheless. And if something has value that can be lost, it has, by definition, downside risk.
In fact, options have even greater downside risk than stock. Consider two executives in the same company. One is granted a million dollars worth of stock, and the other is granted a million dollars worth of at-the-money options—options whose exercise price matches the stock price at the time of the grant. If the stock price falls sharply, say by 75%, the executive with stock has lost $750,000, but she retains $250,000. The executive with options, however, has essentially been wiped out. His options are now so far under water that they are nearly worthless. Far from eliminating penalties, options actually amplify them.
The downside risk has become increasingly evident to executives as their pay packages have come to be dominated by options. Take a look at the employment contract Joseph Galli negotiated with Amazon.com when he recently agreed to become the e-tailer’s COO. In addition to a large option grant, his contract contains a protection clause that requires Amazon to pay him up to $20 million if his options don’t pay off. One could argue that providing such protection to executives is foolish from a shareholder’s point of view, but the contract itself makes an important point: why would someone need such protection if options had no downside risk?
The risk inherent in options can be undermined, however, through the practice of repricing. When a stock price falls sharply, the issuing company can be tempted to reduce the exercise price of previously granted options in order to increase their value for the executives who hold them. Such repricing is anathema to shareholders, who don’t enjoy the privilege of having their shares repriced. Although fairly common in small companies—especially those in Silicon Valley—option repricing is relatively rare for senior managers of large companies, despite some well-publicized exceptions. In 1998, fewer than 2% of all large companies repriced any options for their top executive teams. Even for companies that had large decreases in their stock prices—declines of 25% or worse in the previous year—the repricing rate was less than 5%. And only 8% of companies with market-value declines of more than 50% repriced. In most cases, companies that resorted to repricing could have avoided the need to do so by using a different kind of option program, as I’ll discuss later.
Promoting the Long View
It’s often assumed that when you tie compensation to stock price, you encourage executives to take a short-term focus. They end up spending so much time trying to make sure that the next quarter’s results meet or beat Wall Street’s expectations that they lose sight of what’s in the best long-term interests of their companies. Again, however, the criticism does not stand up to close examination.
For a method of compensation to motivate managers to focus on the long term, it needs to be tied to a performance measure that looks forward rather than backward. The traditional measure—accounting profits—fails that test. It measures the past, not the future. Stock price, however, is a forward-looking measure. It forecasts how current actions will affect a company’s future profits. Forecasts can never be completely accurate, of course. But because investors have their own money on the line, they face enormous pressure to read the future correctly. That makes the stock market the best predictor of performance we have.
But what about the executive who has a great long-term strategy that is not yet fully appreciated by the market? Or, even worse, what about the executive who can fool the market by pumping up earnings in the short run while hiding fundamental problems? Investors may be the best forecasters we have, but they are not omniscient. Option grants provide an effective means for addressing these risks: slow vesting. In most cases, executives can only exercise their options in stages over an extended period—for example, 25% per year over four years. That delay serves to reward managers who take actions with longer-term payoffs while exacting a harsh penalty on those who fail to address basic business problems.
Stock options are, in short, the ultimate forward-looking incentive plan—they measure future cash flows, and, through the use of vesting, they measure them in the future as well as in the present. They don’t create managerial myopia; they help to cure it. If a company wants to encourage a more farsighted perspective, it should not abandon option grants—it should simply extend their vesting periods.1
Three Types of Plans
Most of the companies I’ve studied don’t pay a whole lot of attention to the way they grant options. Their directors and executives assume that the important thing is just to have a plan in place; the details are trivial. As a result, they let their HR departments or compensation consultants decide on the form of the plan, and they rarely examine the available alternatives. Often, they aren’t even aware that alternatives exist.
But such a laissez-faire approach, as I’ve seen over and over again, can lead to disaster. The way options are granted has an enormous impact on a company’s efforts to achieve its business goals. While option plans can take many forms, I find it useful to divide them into three types. The first two—what I call fixed value plans and fixed number plans—extend over several years. The third—megagrants—consists of onetime lump sum distributions. The three types of plans provide very different incentives and entail very different risks.
Fixed Value Plans.
With fixed value plans, executives receive options of a predetermined value every year over the life of the plan. A company’s board may, for example, stipulate that the CEO will receive a $1 million grant annually for the next three years. Or it may tie the value to some percentage of the executive’s cash compensation, enabling the grant to grow as the executive’s salary or salary plus bonus increases. The value of the options is typically determined using Black-Scholes or similar valuation formulas, which take into account such factors as the number of years until the option expires, prevailing interest rates, the volatility of the stock price, and the stock’s dividend rate.
Fixed value plans are popular today. That’s not because they’re intrinsically better than other plans—they’re not—but because they enable companies to carefully control the compensation of executives and the percentage of that compensation derived from option grants. Fixed value plans are therefore ideal for the many companies that set executive pay according to studies performed by compensation consultants that document how much comparable executives are paid and in what form.2 By adjusting an executive’s pay package every year to keep it in line with other executives’ pay, companies hope to minimize what the consultants call “retention risk”—the possibility that executives will jump ship for new posts that offer more attractive rewards.
But fixed value plans have a big drawback. Because they set the value of future grants in advance, they weaken the link between pay and performance. Executives end up receiving fewer options in years of strong performance (and high stock values) and more options in years of weak performance (and low stock values). To see how that works, let’s look at the pay of a hypothetical CEO whom I’ll call John. As part of his pay plan, John receives $1 million in at-the-money options each year. In the first year, the company’s stock price is $100, and John receives about 28,000 options. Over the next year, John succeeds in boosting the company’s stock price to $150. As a result, his next $1 million grant includes only 18,752 options. The next year, the stock price goes up another $50. John’s grant falls again, to 14,000 options. The stock price has doubled; the number of options John receives has been cut in half. (The exhibit “The Impact of Different Option Plans on Compensation” summarizes the effect of stock price changes on the three kinds of plans.)
Now let’s look at what happens to John’s grants when his company performs miserably. In the first year, the stock price falls from $100 to $65. John’s $1 million grant provides him with 43,000 options, up considerably from the original 28,000. The stock price continues to plummet the next year, falling to just $30. John’s grant jumps to nearly 94,000 options. He ends up, in other words, being given a much larger piece of the company that he appears to be leading toward ruin.
It’s true that the value of John’s existing holdings of options and shares will vary considerably with changes in stock price. But the annual grants themselves are insulated from the company’s performance—in much the same way that salaries are. For that reason, fixed value plans provide the weakest incentives of the three types of programs. I call them low-octane plans.
Fixed Number Plans.
Whereas fixed value plans stipulate an annual value for the options granted, fixed number plans stipulate the number of options the executive will receive over the plan period. Under a fixed number plan, John would receive 28,000 at-the-money options in each of the three years, regardless of what happened to the stock price. Here, obviously, there is a much stronger link between pay and performance. Since the value of at-the-money options changes with the stock price, an increase in the stock price today increases the value of future option grants. Likewise, a decrease in stock price reduces the value of future option grants. For John, boosting the stock price 100% over two years would increase the value of his annual grant from $1 million in the first year to $2 million in the third. A 70% drop in the stock price, by contrast, would reduce the value of his grant to just $300,000.
Since fixed number plans do not insulate future pay from stock price changes, they create more powerful incentives than fixed value plans. I call them medium-octane plans, and, in most circumstances, I recommend them over their fixed value counterparts.
Now for the high-octane model: the lump-sum megagrant. While not as common as the multiyear plans, megagrants are widely used among private companies and post-IPO high-tech companies, particularly in Silicon Valley. Megagrants are the most highly leveraged type of grant because they not only fix the number of options in advance, they also fix the exercise price. To continue with our example, John would receive, at the start of the first year, a single megagrant of nearly 80,000 options, which has a Black-Scholes value of $2.8 million (equivalent to the net present value of $1 million per year for three years). Shifts in stock price have a dramatic effect on this large holding. If the stock price doubles, the value of John’s options jumps to $8.1 million. If the price drops 70%, his options are worth a mere $211,000, less than 8% of the original stake.
Disney’s Michael Eisner is perhaps the best known CEO who has received megagrants. Every few years since 1984, Eisner has received a megagrant of several million shares. It is the leverage of these packages, coupled with the large gains in Disney’s stock during the last 15 years, that has made Eisner so fabulously wealthy.
The Big Trade-Off
Since the idea behind options is to gain leverage and since megagrants offer the most leverage, you might conclude that all companies should abandon multi-year plans and just give high-octane megagrants. Unfortunately, it’s not so simple. The choice among plans involves a complicated trade-off between providing strong incentives today and ensuring that strong incentives will still exist tomorrow, particularly if the company’s stock price falls substantially.
When viewed in those terms, megagrants have a big problem. Look at what happened to John in our third scenario. After two years, his megagrant was so far under water that he had little hope of making much money on it, and it thus provided little incentive for boosting the stock value. And he was not receiving any new at-the-money options to make up for the worthless ones—as he would have if he were in a multiyear plan. If the drop in stock value was a result of poor management, John’s pain would be richly deserved. If, however, the drop was related to overall market volatility—or if the stock had been overvalued when John took charge—then John’s suffering would be dangerous for the company. It would provide him with a strong motivation to quit, join a new company, and get some new at-the-money options.
Ironically, the companies that most often use megagrants—high-tech start-ups—are precisely those most likely to endure such a worst-case scenario. Their stock prices are highly volatile, so extreme shifts in the value of their options are commonplace. And since their people are in high demand, they are very likely to head for greener pastures when their megagrants go bust. Indeed, Silicon Valley is full of megagrant companies that have experienced human resources crises in response to stock price declines. Such companies must choose between two bad alternatives: they can reprice their options, which undermines the integrity of all future option plans and upsets shareholders, or they can refrain from repricing and watch their demoralized employees head out the door.
Adobe Systems, Apple Computer, E*Trade, Netscape, PeopleSoft, and Sybase have all repriced their options in recent years, despite the bad will it creates among shareholders. As one Silicon Valley executive told me, “You have to reprice. If you don’t, employees will walk across the street and reprice themselves.”
Silicon Valley companies could avoid many such situations by using multiyear plans. So why don’t they? The answer lies in their heritage. Before going public, start-ups find the use of megagrants highly attractive. Accounting and tax rules allow them to issue options at significantly discounted exercise prices. These “penny options” have little chance of falling under water (especially in the absence of the stock price volatility created by public markets). The risk profile of these pre-IPO grants is actually closer to that of shares of stock than to the risk profile of what we commonly think of as options.
When they go public, the companies continue to use megagrants out of habit and without much consideration of the alternatives. But now they issue at-the-money options. As we’ve seen, the risk profile of at-the-money options on highly volatile stocks is extremely high. What had been an effective way to reward key people suddenly has the potential to demotivate them or even spur them to quit.
Some high-tech executives claim that they have no choice—they need to offer megagrants to attract good people. Yet in most cases, a fixed number grant (of comparable value) would provide an equal enticement with far less risk. With a fixed number grant, after all, you still guarantee the recipient a large number of options; you simply set the exercise prices for portions of the grant at different intervals. By staggering the exercise prices in this way, the value of the package becomes more resilient to drops in the stock price.
Many of the Silicon Valley executives (and potential executives) that I have talked to worry a lot about joining post-IPO companies at the wrong time, when the companies’ stock prices are temporarily overvalued. Switching to multiyear plans or staggering the exercise prices of megagrants are good ways to reduce the potential for a value implosion.
Sleepy Companies, Sleepy Plans
Small, highly volatile Silicon Valley companies are not the only ones that are led astray by old habits. Large, stable, well-established companies also routinely choose the wrong type of plan. But they tend to default to multiyear plans, particularly fixed value plans, even though they would often be better served by megagrants.
Think about your average big, bureaucratic company. The greatest threat to its well-being is not the loss of a few top executives (indeed, that might be the best thing that could happen to it). The greatest threat is complacency. To thrive, it needs to constantly shake up its organization and get its managers to think creatively about new opportunities to generate value. The high-octane incentives of megagrants are ideally suited to such situations, yet those companies hardly ever consider them. Why not? Because the companies are dependent on consultants’ compensation surveys, which invariably lead them to adopt the low-octane but highly predictable fixed value plans. (See the exhibit “Which Plan?”)
The bad choices made by both incumbents and upstarts reveal how dangerous it is for executives and board members to ignore the details of the type of option plan they use. While options in general have done a great deal to get executives to think and act like owners, not all option plans are created equal. Only by building a clear understanding of how options work—how they provide different incentives under different circumstances, how their form affects their function, how various factors influence their value—will a company be able to ensure that its option program is actually accomplishing its goals. If distributed in the wrong way, options are no better than traditional forms of executive pay. In some situations, they may be considerably worse.