Glass Lewis & Co. Updates Its Pay for Performance Model
Earlier this month Glass Lewis & Co. announced updates to its proprietary pay for performance model used to develop voting recommendations on say on pay proposals. The changes went into effect for annual meetings taking place on or after July 1, 2012.
Similar to the prior model, Glass Lewis will perform both quantitative and qualitative reviews when evaluating pay for performance. The changes announced only affect the quantitative review; the qualitative review remains unchanged.
Glass Lewis made the following four major revisions to its quantitative pay for performance model:
- Peer Group Selection – Adopted “market-based” approach to building peer groups using a company’s self-selected peers and peers of self-selected peers
- Compensation Analysis – Named executive officer compensation evaluated over three years rather than one year
- Performance Analysis – Reduced number of performance metrics reviewed from seven to five
- Pay for Performance Grades – Eliminated the forced bell curve in favor of letter grades based on actual degree of pay for performance alignment relative to peer group
Further analysis of the changes can be found below.
These changes are a positive step for both investors and companies. We have long been strong advocates of the use of self-selected groups. In our experience, the self-selected peer groups are the product of considerable thought and represent companies that are competitors for sales, talent or investment dollars. Additionally, it is these peers that companies use when conducting their own assessment of compensation practices and making compensation-related decisions.
We are also in favor of expanding the compensation analysis from a one-year to a three-year analysis. We believe this change appropriately places greater emphasis on evaluating and rewarding long-term performance. This change is also appropriate considering Glass Lewis’ performance analysis looks at three years of data. Lastly, the change will help eliminate one-year anomalies such as one-time retention grants which can cause a company’s score to drop.
The elimination of the forced bell curve grading system is another positive development. Anecdotally, we’re aware of several companies who have consistently scored well in the previous model only to be graded a “B” or “C” due to the forced curve. In our opinion, that never seemed fair. If a company has performed well and compensated executives accordingly, the company should receive a high score regardless of how other companies fared.
Finally, it appears Glass Lewis is making an effort to be more transparent. Make no mistake, this new model is still a “black box,” but they appear to be making an effort to communicate more with companies, investors and governance practitioners. This will help companies anticipate and, where appropriate, avoid making decisions which jeopardize favorable vote recommendations from Glass Lewis.
Peer Group Selection
Glass Lewis has adopted Equilar’s “market-based” approach to building peer groups. Under this approach, Equilar’s algorithm analyzes publicly-disclosed relationships among thousands of companies in order to create a network of the strongest peers. The resulting peer group will include up to 30 companies. The network of companies used in this analysis includes a company’s self-selected peer group, peers of the self-selected peers, companies that use the subject company as a peer and companies that appear with the subject company in a peer group. Factors that determine connection strength include:
- Direction of peer relationships
- Similarity between peer groups
- Number of times two companies “connect” paths in the network
- Distance of connecting paths between two companies
Under the past approach, Glass Lewis used four peer groups, resulting in an average of 100 peers per company, to analyze a company’s compensation. Groups were based on GICS codes, enterprise value and geographic location, with GICS codes and enterprise value serving as the primary determinants.
Under the new model, compensation will be analyzed on a three-year weighted average basis. This is a change from the previous model that analyzed only one year of compensation for named executive officers. Total compensation continues to be calculated as the sum of:
- Non-Equity Incentive Plans
- Grant Date Fair Market Value of Equity Awards
- All Other Compensation
The only change to performance analysis has been the removal of two metrics: change in stock price and change in book value per share. Glass Lewis will continue to evaluate performance metrics on a three-year weighted average basis. The five performance metrics remaining in the model are:
- Total Shareholder Return
- Change in Operating Cash Flow
- EPS Growth
- Return on Equity
- Return on Assets
Pay for Performance Grades
Glass Lewis has abandoned their forced curve grading system and will now assign grades based on a company’s relative percentile position of compensation and performance rankings compared to its peers. Grades will be assigned as follows:
- A = Performance > Compensation by 60 to 100%
- B = Performance > Compensation by 30 to 59%
- C = Performance > Compensation or Compensation > Performance by 0 to 29%
- D = Compensation > Performance by 30 to 59%
- F = Compensation > Performance by 60 to 100%