The 95 percent habit: comp committees are being told to stop defaulting to PSUs

The 2005 accounting change did not make performance share units a better instrument. It made options costlier to report, and PSUs were the vehicle proxy advisers preferred ^[2][3]. The memo from Semler Brossy consultant Sam Gutenmacher and former colleague Michelle Metros is unusually direct about the reasoning at the time: flexibility in leveraging pay delivery, and *presumed* pay-for-performance alignment ^[1][3]. Presumed is the operative word, and two decades later some companies have gone further and swapped options out entirely ^[4].

The goal-setting problem the memo describes is worth reading as a pair rather than a list. Targets set too aggressively leave executives with low holdings and realized pay, which the authors link to dampened motivation and unwanted attrition ^[14]. Targets set too conservatively produce over payouts that draw criticism from proxy advisers and shareholders ^[15]. Neither outcome tells you whether the executive performed. Both tell you whether a forecast made three years earlier held up. That is the mechanism a committee is actually buying when it grants a PSU, and the memo notes investors themselves have flagged multi-year goal accuracy as a perceived weakness given sustained macroeconomic and geopolitical volatility ^[12].

Some companies cannot win that forecasting exercise regardless of conditions. High-growth and cutting-edge businesses struggle to set accurate three-year financial targets; so do companies in heavy investment phases, or those expecting their own numbers to be revised down ^[13]. The memo pushes further into metric fit: revenue and earnings may suit an established company in a mature industry while holding back a younger one whose strategic priority is market share or customer acquisition ^[17].

Then there is the arithmetic of leaving. PSUs went to 95 percent of S&P 500 CEOs last year and now account for 60 percent of the average LTI mix, against 44 percent in 2012 ^[5][6]. That is 16 percentage points of the mix, a relative gain of about 36 percent ^[7][8]. Roughly one in twenty S&P 500 CEOs received no PSU grant at all ^[9]. Compensation benchmarking runs on peer practice, so a committee acting on the latitude the memo identifies in recent ISS guidance ^[11] does so as a visible exception with a proxy statement to write about it.

The memo is also careful, and the caution should be read as part of the argument rather than around it. Its authors say the practices under discussion should remain prominent, and that performance-based vehicles work exceptionally well when they point executives at long-term shareholder value ^[16]. It comes from a compensation consultancy that sells the review it recommends ^[1], and it says alternative models are emerging without the prevalence data to show anyone has adopted them at scale ^[20]. What it does establish is that the proxy adviser apparatus which made PSUs standard is now the thing giving committees permission to question them ^[3][11], and that some influential investors have already started ^[10].