In August last year—12 years after the Dodd-Frank mandate— the SEC finally adopted a new rule that requires disclosure of information reflecting the relationship between executive compensation actually paid by a company and the company’s financial performance: the pay-versus-performance rules.   To a significant extent, the approach taken by the SEC in this rulemaking was prescriptive and some of the prescriptive aspects of the rules were quite complex; the SEC opted not to take a “wholly principles-based approach because, among other reasons, such a route would limit comparability across issuers and within issuers’ filings over time, as well as increasing the possibility that some issuers would choose to report only the most favorable information.”  But there was some flexibility built into the new rules. How would companies implement the more flexible disclosure requirements?   That was the question considered by Compensation Advisory Partners, which published a report on the  versions of pay-versus-performance disclosure from the earliest filers among the S&P 500. A similar study of a slightly larger group was conducted by equitymethods. The goal in each case was to try to get a sense of how companies were responding to the new disclosure requirements. What choices were companies making on peer groups, financial measures or “Company-Selected Measures”? How were companies describing the relationship between pay and performance? Just what did the new disclosure look like?

Tabular list of “most important” performance measures.  Outside of the PvP table, the rules require companies to provide an unranked tabular list of the three to seven most important financial performance measures, including the Company-Selected Measure, used to link executive comp actually paid during the fiscal year to company performance. Non-financial performance measures may also be included so long as the company has disclosed at least three most important financial performance measures. “Financial performance measures” are “measures that are determined and presented in accordance with the accounting principles used in preparing the issuer’s financial statements, any measures that are derived wholly or in part from such measures, and stock price and total shareholder return.” CAP found that the median number of measures in the tabular list was 4 and the mean average was 4.6. In addition, 64% of companies in the sample disclosed only financial measures, while 36% disclosed financial and non-financial measures. Among non-financial measures, 56% were combined operational/ ESG,  22% were ESG only, 11% were operational only and 11% were individual performance. CAP noted that two companies identified fewer than three financial measures and one company used different lists for the CEO and other NEOs.

Clear description of relationship between compensation actually paid and performance.  Using the information in the pay-versus-performance table, companies are required to provide “clear descriptions” of several relationships over the five-year period: the relationship between executive comp actually paid to the PEO and, on average, to the other PEOs and the company’s TSR; the relationship between the company’s TSR and peer group TSR; the relationship between executive comp actually paid and net income; and the relationship between executive compensation actually paid and the Company-Selected Measure. CAP found that 84% of the sample companies used bar charts and/or line graphs to display the required relationships; just 16% used a narrative only.  Some companies did not comment on the graphics, while others  added some explanatory narrative about, for example, pay/performance alignment.

Data in Summary Comp Table compared with comp actually paid.  Comp actually paid turned out to be one of the more complex aspects of the new rule—it took the SEC about 30 pages to describe how to determine comp actually paid.  “Executive compensation actually paid” under Item 402(v) is total comp as reported in the SCT, modified through complex formulations to adjust for pension benefits and equity awards.  CAP reported significant variation over the three-year period in the relationship between comp actually paid and the SCT data, as well as significant variation based on company and stock price performance and potential payouts of past awards.  CAP even identified three companies in the sample that reported negative comp actually paid in one year.  According to CAP, because the majority of pay for most CEOs and NEOs is tied to long-term incentives, “point-in-time stock price had a significant impact on the Compensation Actually Paid amounts.” CAP reports that, if the ratios of PEO comp actually paid and SCT data are added for all three years, the ratio tends to align to companies’ three-year TSR, which CAP suggests is probably because the majority of PEO compensation is in stock-based comp.

Location.    Of companies in the sample, none included the PvP disclosure in the CD&A; most placed the disclosure following the existing comp tables and/or near the CEO pay-ratio disclosure.  CAP viewed this placement to suggest that the required analyses under the new rules “were not part of the compensation decision-making process of the compensation committee. [CAP does] not expect the PvP outcomes to become a primary factor in analyzing future compensation decisions but it will likely be part of the discussion going forward.”

Length of disclosure/ voluntary disclosure.   Disclosures ranged from three to seven pages, with a median of four pages and a mean average of 4.3 pages. Most companies minimized any additional voluntary disclosure, focusing instead on just complying with the complex new rules. CAP noted that the more extensive disclosures often included more detailed footnotes and charts reconciling the calculations of comp actually paid. One company voluntarily provided a five-year table when only three years are required at this point. CAP found a few companies in its sample that voluntarily provided some additional measures reflecting the relationship between comp actually paid and the metrics used in their incentive plans.

equitymethods study

In its study, equitymethods looked at 36 companies that had filed as of March 3, 78% of which were large accelerated filers and 22% of which were smaller reporting companies.  Of this sample, 19% had market caps over $25 billion, 50% were between $1 billion and $25 billion, 19% were between $100 million and $1 billion and 11% had market caps below $100 million.   This study also looked at  various  decision points, such as the comparative peer group for TSR, disclosure location  and the number of metrics listed in the tabular disclosure.

Location.  No companies in the sample included the PVP disclosure in CD&A, which made sense to the author because CD&A is intended to capture comp committee decision-making, and the author did not anticipate that companies would be structuring pay decisions around PvP calculations. In the sample, 75% of companies included it next to the CEO pay-ratio disclosure, and 25% included it near the other existing comp tables. The author attributed the decision to locate the PvP disclosure away from the comp tables to a rejection by most companies of the idea that PvP effectively explains “value evolution,” and the comp table seemed to be implicitly connected with CD&A—apparently no one wanted to go there.

Supplemental disclosures.   Companies may voluntarily provide supplemental measures of compensation or financial performance or other supplemental disclosures, so long as they are “clearly identified as supplemental, not misleading, and not presented with greater prominence than the required disclosure.”  The study found that only 19% of companies in the sample provided any voluntary supplemental disclosures. That means 81% decided against it. 

Peer group TSR.   The author notes that there has apparently been a fair amount of confusion here about whether a broad market index could be used (the rule points to Item 201(e)(1)(ii), which excludes any broad market index) and the use of peer groups from CD&A that were not explicitly used for “benchmarking.”  CDI 128D.05  confirms that the peer group does not need to have been formally used for “benchmarking.” To perform the TSR calculation, the company may use a peer group that is disclosed in its CD&A as a peer group “actually used” to help determine executive pay, even if the peer group is not used for “benchmarking” under Item 402(b)(2)(xiv), as that term is explained in CDI 118.05.  (See this PubCo post.) This study showed that 75% of companies in the sample used the 201(e) performance graph peer group, 21% used the peer group from CD&A and 4% used the CD&A relative TSR peer group.

Tabular list of performance measures. This study found that companies in the sample used an average of 3.7 financial measures and 0.7 non-financial measures. According to the author, the selection should be rather mechanical, drawn from metrics in the annual and long-term incentive plans.  The study found 11% of non-SRCs that  omitted the tabular list without explanation.

Company-Selected Measure.   Non-SRCs are required to include a Company-Selected Measure unless there were no financial performance measures used, not already in the table, to link executive compensation actually paid. Nevertheless, 7% of non-SRCs did not include a Company-Selected Measure. According to this study, 54% of companies used earnings as the Company-Selected Measure, 19% used cash flow, 15% used return and 12% used sales; none used stock price. The author was surprised not to see relative TSR, since that metric is widely used in incentive plans, but suggested that companies may have been confused about whether it was permissible. CDI 128D.09 clarified that issue by expressly permitting the use of relative TSR.

Relationship disclosures.  As in the CAP study, the authors found that most companies chose to use graphics to explain relationships: 58% used graphics only, 27% combined graphics and narrative, 12% used narrative only and 3% used a percent change table (which “illustrates the slopes or percentage changes across PvP table variables”). Over 90% in the sample used the dual-axis graph referenced in the PvP rules.  The author also studied the narrative disclosures, distinguishing between “shallow” analysis and “deep” analysis “based on whether the copy rehashes information already present (shallow analysis) versus offering up new information as to why the numbers in the PvP table are what they are (deep analysis).”  Guess what? The author concluded that 62% of companies provided only a shallow analysis.  In addition, the author “classified 100% of the narrative-only cases as shallow.”

Common mistakes.   The author also identified several common errors (in the hopes of eliciting improvement), especially in the prescriptive components of the rule, of which there are many.

• Granular equity calculations.   In CDI 128D.04, Corp Fin advises that companies should provide footnote disclosure of each of the amounts deducted and added under the complex pension and equity award adjustment calculations (Items 402(v)(2)(iii)(B)(1)(i) – (ii) and Items 402(v)(2)(iii)(C)(1)(i) – (vi)).  At March 1, the author found that 31% of companies provided aggregate calculations only and did not provide the requisite granular calculations.  (The CDIs were posted on February 10.)
• Valuation assumption disclosures.  For equity awards, the rules “first require the year-end reporting and valuation of awards granted during the fiscal year and then the year-over-year change in fair value of such awards until the vesting date (or the date the registrant determines the award will not vest).” The rules provide a complex formula for valuation of equity awards, and require footnote disclosure of any valuation assumptions that materially differ from those disclosed as of the grant date for those awards.  Only 6% of companies in the sample disclosed these assumptions.  Citing several pages, the author suggests that some of the commentary in the adopting release might indicate that more disclosure would be appropriate. For example, fn 224 of the release indicates that “there may be a material difference in assumptions if the registrant has made changes to key assumptions that would have materially changed the grant date fair value if the assumption(s) applied as of grant date.” However, given that there is no bright line, the author advises documenting internally the rationale for the approach taken and running comparison analytics to assess how sensitive the final values are to the actual assumptions.
• Other areas of risk or potential misconception.  The author highlights five other potential risk areas worth noting:

1. “Failing to provide Item 402(v)(5) relationship disclosures altogether
2. Omitting one of the required Item 402(v)(5) relationship disclosures (such as the company TSR to peer TSR comparison)
3. Adding supplemental disclosure that violates the requirement that any supplemental disclosure be clearly labeled as supplemental, not be made more prominent than the required disclosure, and not be misleading
4. Including a non-financial measure in the Tabular List prior to providing three financial measures (e.g., there are two financial measures and one non-financial measure)
5. Using an Item 201(e) peer group that is a broad market index and therefore falls within Item 201(e)(1)(i) but not the requirement that it fall within Item 201(e)(1)(ii), which excludes any broad market index”

[View source.]