by Cydney Posner
Some consultants say yes. In this article, posted on CFO.com, two consultants argue that the use of the three-year time horizon frequently associated with performance-based restricted stock grants may not really be long enough, especially where the performance measure is relative total shareholder return (TSR). In fact, they contend, perhaps with a touch of hyperbole, it has the “potential to be dangerous” because the “payouts to executives may reward short- to mid-term stock price volatility rather than sustained long-term TSR performance.”
To assess this concern, the authors structured a study based on the 449 firms in the S&P 500 between 2004 and 2014, looking at TSR performance over four rolling 10-year periods. Using a standard performance-based plan, they compared the distribution of payouts of restricted stock for overlapping three-year performance periods and for five-year performance periods over the same time period. They also compared those payouts with 10-year TSR.
The authors explain that, optimally, relative performance and payout level should be closely aligned; i.e., “low performers” should receive small payouts while “high performers” should receive payouts that are “well above target or close to maximum.” In their view, the “‘right’ answer exists when payouts to LTI recipients mimic payouts to long-term shareholders.” Theoretically, “a 10-year award would generally have 100% alignment, but such a long-term award is obviously not practical.”
What they found “is that shorter-termed awards led to a higher variability in payouts within individual firms.” In their analysis, if “there is too much variation in payouts over time, especially when overlapping award cycles exist at once, recipients are more likely to view their awards as lottery tickets. And, of course, the problem with a lottery ticket framework is its passivity — it’s nice if the award happens to pay out but outside one’s control if it does not.” In addition, analyzing the distribution of payouts across companies, they found “a higher variation in payouts on three-year awards versus five-year awards, as well as a “clustering between a 75% and 125% payout (where 0% to 200% is possible) on three-year awards.” That is, with the three-year measurement periods, “a large majority of companies pay out near target even though their relative long-run performance will be varied.”
To analyze the alignment to shareholder returns for the two-performance periods, the authors compared the payouts under hypothetical three-year and five-year TSR awards relative to a company’s cumulative 10-year TSR. They found that, for the five-year performance periods, the awards at 179 companies matched 10-year TSR, while for the three-year awards, only 147 companies matched 10-year TSR.
In conclusion, the authors contend that their analysis raised questions about whether 3-year performance periods are always appropriate for relative TSR plans. They suggest that companies model their own alternative scenarios, particularly since companies may use indices other than the S&P 500, which could lead to different results. Recognizing that the ideal of a “truly long-term performance period like 10 years or 7 years, which also more closely corresponds to the life cycle of business strategies” is impractical, they suggest that compensation committees “think about a performance period of five years or even four years, which may provide a better balance between executives’ reluctance to wait to receive compensation and shareholders’ concerns that equity awards should reward long-term value creation. Additionally, if there are tenure concerns with five years, it is possible to structure a five-year performance period on an award that only requires three years of continuous service.”
Of course, different compensation consultants will have diverse views on these issues (and, most likely, different studies to back them up). In addition, different facts and circumstances at companies will often lead to a variety of judgments about the pros and cons of longer (or shorter) performance periods.