ISS previously issued its 2012 policy update.  ISS has now delivered the promised additional guidance on the 2012 Pay-for-Performance (P4P) methodology in a technical document.

By way of back ground, ISS ‘ revised analysis will consider the following factors:

• Peer group alignment.
  • The degree of alignment between the company ‘s Total Shareholder Return, or TSR, rank and the CEO ‘s total pay rank within a peer group, as measured over one-year and three-year periods (weighted 40 percent/60 percent) (“relative degree of alignment”);
  • The multiple of the CEO’s total pay relative to the peer group median (“multiple of median”).
• Absolute alignment. The absolute alignment between the trend in CEO pay and company TSR over the prior five fiscal years – i.e., the difference between the trend in annual pay changes and the trend in annualized TSR during the period (“pay-TSR alignment”).

In cases where alignment appears to be weak, further in-depth analysis will determine causal or mitigating factors, such as the mix of performance- and non-performance-based pay, grant practices, the impact of a newly hired CEO, and the rigor of performance programs.

The new guidance has some utility in further understanding the ISS methodology but to many the ISS methodology will still be a black box.  Some highlights from the ISS guidance are set forth below.

What Does ISS Measure?

 ISS stresses it measures total compensation as set forth in the summary compensation table.  ISS feels this is a better measure than “realized” compensation for a number of reasons.  One reason is that is how companies benchmark their pay practices.  ISS focuses on the CEO’s pay because that package sets the “compensation pace” at most companies; also the compensation committee and board are most directly involved in and accountable for the decisions that generate the CEO’s pay.  ISS utilizes a standard set of assumptions to value equity-based grants, but the guidance does not suggest what those standard assumptions might be.

As for performance, ISS uses the TSR benchmark.  While there might be many ways to do this, ISS believes this is what shareholders want and it is objective and transparent.

Peer Group Alignment

Two measures are used, the “relative degree of alignment” and the “multiple of median.”

Relative Degree of Alignment (RDA)

The relative degree of alignment measure addresses the question: Is the pay opportunity delivered to the CEO commensurate with the performance achieved by shareholders, relative to a comparable group of companies? The measure compares the percentile ranks of a company’s CEO pay and TSR performance, relative to a comparison group of 14-24 companies.

Peer groups are generally constructed with reference to the company’s industry (based on GICS classification), revenue (or assets with respect to financial companies), and market value.  The comparison companies selected by ISS are not intended to serve the function that a “peer group” does for a board of directors when it benchmarks executive pay. Boards use peer groups to help determine an appropriate pay package for attracting and retaining executive talent.  The ISS comparison groups are intended, rather, to help evaluate the alignment of executive pay and company performance that results from a board of directors’ pay decisions over time.  Further details on the construction of peer groups are set forth in the appendix to the guidance.

To determine relative degree of alignment, the subject company’s percentile ranks for pay and performance are calculated for one- and three-year periods.  Because of the sensitivity of TSR to overall market performance, annualized TSR performance for all companies (subject company and comparison companies) will be measured for the same period: that is, the one- and three-year periods ending on the last day of the month closest to the fiscal-year end of the subject company. To illustrate: if a company’s fiscal year ends on November 29, 2011, then all TSRs will be measured over the periods December 1, 2010-November 30, 2011 (for one-year) and December 1, 2008-November 30, 2011 (for three-year).

Combined percentile ranks for pay and for performance are calculated, based on a 40% weighting for the one-year and a 60% weighting for the three-year ranks. The relative degree of alignment is equal to the difference between the ranks: the combined performance rank minus the combined pay rank.  If three years of data are not available for the subject company, the combined measure will reflect only the one-year rankings.

Values for the relative degree of alignment measure range between -100 and +100, with -100 representing the high pay for low performance (i.e., 100th percentile pay combined with 0th percentile performance), zero representing a high degree of alignment (the pay rank is equal to the performance rank), and positive values representing high performance for low pay.

Multiple of Median (MOM)

This multiple of median measure addresses the question: Is the overall level of CEO pay significantly higher than amounts typical for its comparison group? Is the company significantly more than comparable companies, even for strong performance?

Calculating this measure is straightforward: the company’s one-year CEO pay is divided by the median pay for the comparison group.  Values can therefore range from zero (if the subject company paid its CEO nothing) to infinity. In ISS’ back-testing analysis, the highest observed value was just over 25 times peer median.

Pay-TSR Alignment (PTA)

ISS’ new measure of long-term absolute alignment is intended address the question: have shareholders’ and executives’ experiences followed the same long-term trend? It is important to note that PTA is not designed to measure the sensitivity of CEO pay to performance – whether pay and performance go up and down together on a year-over-year basis. It is a long-term measure of directional alignment.

If you have managed to follow the calculations so far, this is where it gets really mind numbing:  At a high level, the measure is calculated as the difference between the slopes of weighted linear regressions for pay and for shareholder returns over a five-year period. This difference indicates the degree to which CEO pay has changed more or less rapidly than shareholder returns over that period.

The trend lines calculated by these regressions are analogous to a 5-year “trend rate” for pay and performance, weighted to reflect recent history. The final pay-TSR alignment measure is simply equal to the difference: performance slope minus the pay slope. Potential values for PTA are theoretically unbounded, but in practice they range from just over -100% to just over 100%,

Qualitative Evaluation

If misalignment in pay-for-performance quantitative measures is detected, ISS will perform an in-depth qualitative assessment to determine either the likely cause or mitigating factors.  The ISS guidance indicates a number of factors that ISS may consider.

For example, if a company exhibits long-term disconnect between pay and performance, ISS closely examines its disclosed benchmarking approach to determine whether that may be a contributing factor. For example, a preponderance of self-selected peers that are larger than the subject company may drive up compensation without regard to performance. Above-median targeting may have the same effect.

Special circumstances are also considered.  The qualitative analysis may also consider exceptional situations, such as recruitment of a new CEO in the prior fiscal year or unusual equity grant practices (e.g., bi- or triennial awards) that may distort a quantitative analysis.  Recruiting a new CEO is not a free pass however.  While shareholders may welcome a new CEO in light of lagging performance, they may nevertheless be concerned about a board that has been forced to pay dearly for outside talent but fails to appropriately link the new CEO’s pay to performance improvement.

Check dodd-frank.com frequently for updates on the Dodd-Frank Act and other important securities law matters.