By Amy Wood
Each year Glass Lewis (GL) grades each company's pay-for-performance (PFP) alignment and such grades weigh heavily in the analysis underlying the GL vote recommendations for say on pay. The GL model grades based on 36 measurement points, including measurement points that are relative to GL-selected peer companies. Historical criticism of the GL model has included complaints about lack of transparency re the measurement points and complete lack of disclosure of the GL-selected peer group.
GL recently announced that it is making changes to its model following its partnership with Equilar that include:
- Transparency about GL selection criteria for peer companies
- A new market-based peer selection methodology (intended to move closer to a peer group that would be an appropriate peer group for company to use when determining compensation as opposed to the prior GL model that was not intended to do that, but rather to apply consistent criteria in effort to present an "apples to apples" comparison)
- Disclosure of the GL-selected peer groups, noting which of the GL-selected peers are in company's self-selected peer group and highlighting any outliers in the self-selected peer group that have much larger market caps, etc.
Here are a few additional highlights about the changes GL is making:
- New market-based peer groups: The biggest change under the updated model is the new peer group selection methodology. Instead of the old system where GL selected 4 peer groups based on industry sector, size, sub-industry and geographic location (with the 4 groups including between 50 and 125 companies), GL will now use Equilar's market-based peer selection system. This new system builds a pool by starting with the company's self-selected peers and then adding all of the peers of the self-selected peers. Equilar then maps out the "connections" between each company and the subject company and then ranks the pool based on the strength of those connections – the top 30 companies based on strength of connections are then chosen as the peer group for the GL PFP analysis and ultimate grade.
Note that Equilar offers a subscription that gives companies access to the same database that GL will be using to construct its new peer groups.
- Performance criteria: Historically, GL measured performance based on change in stock price, change in shareholder return, change in book value per share, change in operating cash flow, EPS growth, return on equity and return on assets. Going forward, GL will no longer measure change in stock price (because it is too similar to change in shareholder return) or change in book value per share.
- Compensation: Historically, GL used the one-year compensation for the CEO and top 5 executives. Going forward, GL will use a three-year weighted average of total compensation for the CEO and top 5 executives. The most recent year will be more heavily weighted in this calculation (but the exact formula was not disclosed).
- Methodology: The fundamental methodology is not changing – GL will rank a company's performance relative to the peer group and pay relative to the peer group and then measure the gap between performance rank and pay rank. The larger the gap, the worse the score.
- Letter grades: GL gives letter grades (A through F) as scores under its PFP model. Under the prior model, GL allocated grades based on a forced curve where 10% of companies would get an A, 20% would get a B, 40% would get a C, 20% would get a D and 10% would get an F. Under the new system, there will be no forced curve but rather grades based on the new scores that are relative to the new and more refined peer groups.
- Qualitative factors: GL will continue to take a holistic approach in its analysis underlying recommendations for say on pay – to the extent there are quantitative issues under the model, GL will continue to consider the same qualitative factors to identify the perceived level of compensation risk. Part of this analysis will include identifying any companies in a company's self-selected peer group that GL views as real outliers based on the Equilar analysis. Other examples of qualitative factors include lack of performance-based long-term incentives, tax gross-ups and poor disclosure.