The Holy Grail: What LTI Measures Drive Corporate Performance?

Choosing the right metrics, and applying them consistently, are keys to improving executive and corporate performance.


Like Richard Wagner’s epic opera, “Parsifal”, which depicts King Arthur’s knights in the quest to find the Holy Grail, companies strive heroically to find the right performance measures to improve corporate performance.

Based on our recent analysis, we discovered two very interesting points. First, the companies that used the same performance measures in each of the past five years outperformed others that changed measures. Second, the best performance measure is Earnings per Share (“EPS”), followed by Capital Efficiency, and Total Shareholder Return (“TSR”) is the worst measure.

These results are not surprising since providing a consistent and “line-of-sight” performance goal is always the best way to provide incentive to management. The era of using stock options, which was more like a “lottery ticket” (as described by Warren Buffet), has come to an end.

“In all instances, we pursue rationality. Arrangements that pay off in capricious ways, unrelated to a manager’s personal accomplishments, may well be welcomed by certain managers. Who, after all, refuses a free lottery ticket? But such arrangements are wasteful to the company and cause the manager to lose focus on what should be his real areas of concern. Additionally, irrational behavior at the parent may well encourage imitative behavior at subsidiaries”

— Warren Buffett, commenting on stock options in The Warren Buffett CEO: Secrets from the Berkshire Hathaway Managers, by Robert P. Miles, published by John Wiley & Sons

“We seek to design our executive officer compensation programs to attract, retain and motivate key executives who drive our success and industry leadership.”  Versions of this statement can often be found as the lead-in for a public company’s Compensation Discussion and Analysis (“CD&A”) section of their Proxy Statement. Following closely behind this statement is a declaration that executive pay that is market competitive and reflects performance is key for accomplishing the goals of the compensation program while also being in alignment with shareholders.

An important question coming out of this is — what sort of compensation really motivates top executives today?  And, can companies be assured that whatever motivates top executives will also drive company performance?  If these two factors are not connected, motivating executives the wrong way could potentially harm company performance.

For example, if a company provides short-term and/or long-term incentives to executives with payouts tied primarily to revenue growth, it will likely motivate executives to increase the size of the company in ways that could reduce profit margins and perhaps stock price growth.  Understanding the interaction of incentive measures with each other, and with other company measures of performance, is critical in designing an effective incentive program.

For most S&P 500 companies, long-term incentives (“LTI”) comprise 60 to 70 percent of total direct compensation (salary, bonus, LTI grants). In our annual study of short- and long-term incentive design at the top 200 public companies (“Incentive Design Study”) we found that 86 percent of companies granted performance-vested equity. Of these companies, 59 percent used a relative measure or measures to determine vesting. Seventy-nine percent (79 percent) of these relative measures were Total Shareholder Return (“TSR”). So when you boil it all down, approximately 40 percent of the top 200 companies (“Top 200”) used a relative TSR measure in its LTI mix.

Performance shares have been around for many years, and now comprise the majority of the LTI award (on a grant date value basis) for the CEO of a large U.S. company. However, does the use of performance measures, particularly TSR measures, in LTI grants result in positive results in company performance as measured by TSR? And, does relative TSR motivate executives to perform at a high level? We believe there is reason to doubt a convincing link between performance-based grants and company performance. Moreover, we suspect TSR is not a measure that is motivational.

The results of the 2013 proxy season will be summarized to show the changes in executive compensation with explanation for such changes. Overall, there is little evidence of reductions in CEO and top executive pay. What has changed is that a higher percentage of pay is based on achieving performance targets and not just the stock price.

As we review CEO pay from 2013 into 2014, we see this shift to performance-based LTI continuing. We have also reviewed the relationship between performance-based LTI and company performance. We posit that performance measures that relate most closely to company performance imply these measures are both meaningful and motivational.

The influence of proxy advisory firms and shareholder activists has become stronger. Investors cannot review each company, so they rely on Institutional Shareholder Services (“ISS”) and Glass Lewis for voting guidance. ISS is becoming more accountable and responsive to corporate concerns. Under pressure from corporations, the SEC has compelled ISS to reach out to corporations for feedback on voting policy, which resulted in changes to peer group selection and definition of pay (e.g., realizable pay).

ISS also has an explicit guideline that a CEO’s long term incentive pay should be at least 50 percent performance-based. This has helped to drive the prevalence of performance-based grants for top executives of the Top 200 from 75 percent in 2009 to 88 percent in 2012. Moreover, 60 percent of Top 200 companies using performance-based grants like performance shares or performance stock units had weights of 50 percent or higher in their LTI mix. In 2009, only 48 percent of Top 200 companies had weights of 50 percent or higher.

Performance metrics and LTI mix continues to be a primary topic of interest, with continued emphasis on the pay-for-performance linkage and the percentage of total pay tied to performance continuing to rise. But is it possible that performance-based grants have no significant relationship to company performance?  And if this is the case, is this all about optics and tax deductibility, or is it about real performance?  This paper presents an analysis of Long-Term Incentive Plans (“LTIPs”) in place at Top 200 companies.

Pay-for-Performance and ISS

Performance is usually measured in terms of annual stock price performance (increases) or, in the case of a dividend-paying company, TSR.

Performance is also often described or related to financial performance. Any positive factors that support a pay-for-performance result are usually presented in the CD&A portion of the annual proxy in an effort to induce a convincingly positive Say-on-Pay vote. Along with the increase in the use of performance-based grants, the use of TSR as a measure has increased, especially relative TSR, which is company TSR as compared with a peer or industry group.

Fifty-one percent (51 percent) of Top 200 companies making performance-based grants used a TSR measure, up from 43 percent in 2008. Since ISS measures company performance using TSR, it would seem logical to include TSR as an LTI measure. In addition, with relative TSR there is no need to set goals three years into the future, which is often a difficult and uncertain exercise.

The increased use of performance-based LTI is partially in response to the ISS methodology for determining their Say-on-Pay recommendation. If the initial screening pay-for-performance test fails, the qualitative test needs to be acceptable to ISS in order to avoid and “Against” recommendation from ISS. See Figure 2 below for an illustration of the ISS Say-on-Pay methodology.

As a result of these ISS evaluation factors, which are reviewed by most major investors, executive pay has undergone significant changes in the past five years. They include:

  • A distinct shift from time-vested stock options, not considered performance-based by ISS, to performance-vested stock grants linked to financial performance, TSR, or both
  • Pay increases that come primarily from increased equity grants (i.e., long-term incentives)
  • Increased use of TSR as a long-term incentive measure

Long-Term Incentives—Performance Plans

As further evidence of the shift toward meaningful performance plans, our firm’s analysis of the Top 200 public companies in the U.S. shows a steady increase in the number of companies using either performance shares (“PS”) or performance share units (“PSU”) with threshold, target and maximum payout opportunities.

We believe proxy advisors have influenced this shift toward performance-vested grants through their policies assessing the structure of CEO compensation. The introduction of Say-on-Pay votes in 2011 has also had an effect. And since the proxy advisor recommendations regarding Say-on-Pay are based in part on these considerations, the Say-on-Pay mechanism has helped to shape these pay changes through the unplanned empowerment of the proxy advisors.

Given these changes and the specific focus on performance-vested grants, it seems likely many companies would be able to improve their performance as compared to competitors by establishing goals that motivate executives.

Broad Analysis of Incentives and Performance

Our firm has conducted its annual Incentive Design Study for the past six years. After publishing our 2012 Incentive Design Study in December of 2013, we began to review more closely the relationship between select LTI performance measures and TSR performance. Specifically, we reviewed the LTI performance measures used over the past five years in relation to each company’s five-year TSR. Companies that choose measures that are motivational and connected to business performance should be expected to underpin TSR growth.

For this analysis, 195 of our current Top 200 list had five full years of data for TSR and LTI information. Two companies went from private to public ownership within the past five years (2008 to 2012) and three companies went from public to private ownership. In these cases, we do not have a full five-year TSR value.

Over the past five years, this group of companies used an average of 1.8 measures for those years performance-based LTI was granted. Sixty-four percent (64 percent) of these companies granted performance-vested LTI all five years. Eight percent (8 percent) did not grant performance-vested LTI in any of the five years.

Total Shareholder Return (TSR)

Fifty-three percent (53 percent) of companies, 103 in all, used a TSR measure at least once in the 2008 to 2012 period with an average five-year TSR of -0.18 percent. This is lower than the overall average of 1.15 percent. We found that companies with LTIPs not using TSR had an average return of 2.67 percent. Companies that did not grant performance-vested LTI for any of the five years had average returns of 2.54 percent (negative 0.47 percent, excluding Apple and Amazon). Companies using TSR for just one to four years performed poorly. This suggest that TSR as a performance measure might be limited in motivating executives in a way that translates actions and decisions into results that benefit shareholders.

Capital Efficiency

This category includes return ratios like return on invested capital, return on equity, return on assets and so forth. We also include economic value-added (economic profit) measures. Forty-four percent (44 percent) of companies, or 86 in all, used a capital efficiency measure at least once in the 2008 to 2012 period and had average five year TSR of 0.68 percent. This is lower than the overall average of 1.15 percent. We found that companies not using capital efficiency measures had a return of 1.36 percent, slightly higher than the average return.

Of the companies using a capital efficiency measure only those companies using a capital efficiency measure for three or more years beat the average (2.79 percent vs. 1.13 percent).

Based on these findings, companies using capital efficiency measures in their LTI design performed better than companies using TSR, yet the average TSR results for companies using capital efficiency measures were not impressive.

Earnings per Share (“EPS”)

Thirty-seven percent (37 percent) of companies (73 in all) used an EPS measure at least once in the 2008 to 2012 period, with an average five-year TSR of 1.37 percent. This is slightly higher than the overall average of 1.15 percent. We found that companies not using EPS had a slightly lower average return of 0.81 percent. Companies not granting performance-vested LTI had average returns of 2.54 percent (negative 0.47 percent excluding Apple and Amazon).

Of the companies using an EPS measure those companies using EPS for two or more years, the average returns were much higher than the overall average (3.53 percent vs. 1.15 percent). EPS appears to be effective in motivating executives and a more transparent, operational level than is the case with TSR. Moreover, the use of EPS resulted in average TSR returns that were higher than any other measure reviewed in our study.

From this analysis of three popular types of LTI measures, it suggests that using EPS is more closely related to strong TSR performance than use of relative TSR or capital efficiency measures. Companies typically provide investors with guidance on company EPS performance, so it is a measure that is consistently in focus for top executives. Accordingly, since executives prepare to present EPS results and guidance to Wall Street each quarter, it isn’t a stretch to assume CEOs across the U.S. are motivated to increase EPS on a trend basis.

This broad overview does not account for a number of important factors such as industry mix and prevailing industry practices, the rigor of the measures being employed, and the combination of measures being used. However, it appears EPS has a genuine effect on company performance and in motivating executives to focus on the right internal factors affecting performance.

Other Highlights

Number of years with same measure:

Referring to Figure 3 above, we distilled these values into weighted values for each of the number of years these measures were used to get the net effect. The TSR values were weighted by the number of companies in each cell and the number of years used. Clearly we see that using a consistently good measure for five or more years is a winner.

Our important conclusion is that incentive measures are most effective if the measures are used consistently from year-to-year, independent of the measure or measures being used. This can be observed in Figure 3 above where, in most cases, on average, the longer a measure is used, the higher the five-year TSR.

Looking across the sample of all measures for all companies, the table below indicates companies using at least one measure consistently for five or more years show higher TSR performance results (Figure 4). We believe having consistency in LTI design and in the use of measures provides a consistent focus for management and perhaps results in better performance. Changing LTI measures from year-to-year, recognizing that most performance periods are three years in length, can result in significant confusion and a loss of focus and motivation.

Number of years measures changed:  We can see similar results by looking at the number of measure changes made over a five-year period (2008-2012). If there were no measure changes over the five-year period, the average TSR was much higher than the overall average (5.07 percent vs. 1.13 percent).


TSR, particularly on a relative basis, is a poor LTI measure. For the majority of companies granting performance-based grants, relative TSR is a commonly used performance measure. However, using TSR as a measure has a negative influence on company performance, except in the case where it has been used as a performance measure for each of the past five years. Moreover, TSR, particularly when measured against a group of companies (including a stock index), does not motivate executives, and is similar a stock option in its nature (another form of lottery ticket), as it does not provide the line of sight necessary for oversight.

Stability of LTI design has real value. The stability of the measures used is as important as the types of measures used. As we saw for all three types of measures reviewed, TSR, EPS and capital efficiency, companies that had consistently used these measures for five or more years outperformed the industry average.

EPS is the best indicator of stock price performance. EPS was most closely related to TSR performance.  This is particularly the case for large, well-established companies that represent the Top-200 U.S companies that are in this study. EPS captures all revenue, expenses, write-downs and other actions typical for a publicly traded corporation. Line of sight has value and is more likely to motivate executives to focus on levers that affect performance the most.