This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognizing you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.
Dissecting GRId
In February, RiskMetrics Group, Inc. (“RMG”), the prominent risk management and corporate governance services firm, introduced its Governance Risk Indicators (“GRId”) system, a new tool for assessing the corporate governance-related risks, including compensation-oriented risks, of public companies. GRId replaces RMG’s previous process for evaluating the corporate governance practices of the U.S. companies that it tracks, the Corporate Governance Quotient (“CGQ”) system.
In recent weeks, RMG has published technical materials explaining its new system and describing how it will work. This article summarizes the relevant information about the compensation-related aspects of the GRId system based on the documentation issued by RMG as of May 1, 2010.
GRId Overview
RMG is using GRId as a way to compile a comprehensive database of the corporate governance attributes and practices of the companies that it follows (both domestically and globally), as well as to evaluate and “grade” the level of risk in four specific areas of investor concern:
- Audit;
- Board Structure;
- Compensation; and
- Shareholder Rights.
Each year, companies will be assigned a color-coded risk rating (green for low concern, yellow for medium concern, and red for high concern) in each category based on RMG’s analysis of a company’s relevant policies and practices. While RMG has indicated that it will not use these GRId analyses when formulating voting recommendations on shareholder proposals and director elections, they are displayed prominently in RMG’s research reports and presumably investors will use the results to assess the strength (or weakness) of a company’s governance and compensation practices (based on RMG’s determination of prevailing market “best practice”) and make comparisons between companies.
How GRId Works
The GRId methodology for U.S. companies involves a set of 63 questions, which are used to examine a company’s existing policies and practices with respect to its audit function, board of directors’ structure, approach to certain compensation matters, and positions on various shareholder rights issues. Each of these categories is divided into subcategories, with individual questions focused on specific issues associated with that subcategory.
Is GRId Just Another “Black Box”?
Although GRId is being touted as an improvement over the previous Corporate Governance Quotient system (and, make no mistake, it is), it appears to suffer from many of the same flaws. Its inherent complexity – the opaque weightings of the relative significance of different policies and practices, the five point rating system with its seemingly arbitrary assignment of values, and its attempt to apply a comprehensive methodology across different global markets – almost ensure that it will be poorly understood by market participants and, potentially, erroneously applied. Perhaps more importantly, in spite of its characterization as a “fully transparent” methodology for assessing corporate risk, to date the GRId rollout has been anything but transparent. Over the past few weeks, RMG has continued to revise the methodology (often without disclosing the nature of or reason for the changes). This has included modifications to the scoring as well as some of the practice standards themselves. For example, in its most recent update, the number of points awarded for having a “double-trigger” (rather than a “single trigger”) change-in-control arrangement was changed to “0” from “3.” For some companies, this single revision may be sufficient to increase their compensation-related score by nearly 8% – which could alter their risk rating in this key area. Moreover, this problem may not be limited to GRId’s initial “launch” phase. RMG has indicated that the methodology will be revised annually when it updates its benchmark proxy voting standards – potentially creating a fluid environment that will be repeated year after year.
Given these shortcomings, it would not be unreasonable for companies to consider GRId to be a perpetual “work in progress.” Accordingly, for the foreseeable future, you should be vigilant in verifying both the accuracy of the data that RMG uses to perform its analysis and the ratings produced by the analysis itself. Until we better understand the market’s reaction to, and use of, GRId, the prospect of inaccurate or incomplete ratings is too great to ignore.
For 2010, the number of questions in each category is as follows:
- Audit – 6 questions;
- Board Structure – 14 questions;
- Compensation – 28 questions; and
- Shareholder Rights – 15 questions.
Each question is assigned a weighting and has a range of potential responses, each with a separate score (number of points). A response that reflects a policy or practice that RMG believes promotes good governance receives a positive number of points (up to a maximum of 5). Conversely, a response reflecting a policy or practice that, in RMG’s view, does not promote good governance or, in some instances, the absence of a policy or practice, receives a negative number of points (up to a maximum of -5). The aggregate (weighted) scores for each category are “normalized” on a 100 point scale to simplify comparability.
Scores are tabulated in each of the four categories and a “level of concern” rating (low, medium, or high) is assigned.
The principal – and most critical – differences between GRId and the CGQ system are threefold:
- GRID contains a more detailed review of a company’s governance and compensation practices; a process that will continue to mirror RMG’s benchmark proxy voting standards that it uses to formulate its voting recommendations for its institutional clients
- GRId assessments will be performed on an absolute basis compared to RMG’s proxy voting policies (which will be updated annually), rather than on a relative basis compared to a company’s peers and industry.
- GRId is more transparent; the publication of its methodology, including the questions posed, their relative weightings, and the range of potential scores, is a marked departure from CGQ, which was largely a “black box.” This transparency does not come without its cost, however, as GRId is anything but simple.
GRId’s Compensation Analysis
Given its historical role at the center of corporate governance, it is no surprise that nearly half of the GRId questions involve compensation. The 28 compensation-related questions for U.S. companies are spread across eight subcategories:
As expected, for the most part, these questions are designed to identify whether, in RMG’s view, a company’s executive and equity compensation programs reflect “poor pay practices” or could potentially encourage excessive risk-taking (such as stock option repricings without shareholder approval, guaranteed bonuses, and lucrative severance packages), as well as program features (such as “clawback” policies or stock ownership and holding requirements) that may mitigate such risks.
The actual GRId questions, along with their weightings, are set forth in Exhibit A to this article.
Compensation – Our Observations
Thorough Disclosure is Critical.
Since the answers to questions are based on a company’s public filings (as well as its corporate web site), GRId places a premium on responsive disclosure. In fact, some of the compensation-related questions actually penalize a company if it does not provide any disclosure about a specific topic. These include the performance measures and related information for annual performance-based incentive compensation arrangements, holdings periods for shares acquired by executives through stock option exercises and the vesting of restricted stock awards, and the amount of severance payable to executives in the event of a change in control. In some of these areas (for example, equity holding periods), the absence of any disclosure produces a negative score while disclosure of the absence of a specific policy will result in a neutral (0) score.
Questions Affect Your Rating Differently.
The weightings of the questions vary between subcategories. Consequently, in some areas, a single response can dramatically shift a company’s rating. Coupled with its use of color-coded ratings, such a shift may have a noticeable impact on how a company’s governance practices are viewed by the investor community. For example, the nature of the trigger event under your change-in-control agreements (that is, whether the agreements are subject to a “single” or a “double” trigger) counts for nearly 23% of the entire rating in the Compensation category (generally, other questions account for between 2–5% of the rating). Consequently, the presence of a single trigger or a modified single trigger provision in an executive agreement can potentially move you from a “low” level of concern to a “medium” level of concern or from a “medium” level of concern to a “high” level of concern.
Ratings Don’t Track Proxy Voting Guidelines Consistently.
While it is clear that the questions in the Compensation category generally follow RMG’s proxy voting guidelines, in some instances the presence or absence of a specific compensation policy or practice is treated more harshly under GRId than the proxy voting guidelines. For example, while under its proxy voting guidelines RMG generally supports stock ownership guidelines, it tends to let companies determine the appropriate ownership requirement. Under GRId, however, RMG evaluates whether chief executive officer ownership guidelines are “robust” (six times base salary), “standard” (three to five times base salary), or “substandard” (less than three times base salary) and only rewards “robust” guidelines. Moreover, “substandard” guidelines are penalized three points. Given such discrepancies, it will be important to re-evaluate the GRId methodology each year, as RMG plans to revise it annually in conjunction with its proxy voting policy updates.
Ratings Are Likely to Influence Program Design.
RMG’s use of a standard set of questions to determine its ratings will give it significant additional leverage over the corporate governance practices of U.S. companies. For example, the inclusion of equity holding periods as a subcategory, and the potential penalty for not addressing this subject, will likely compel companies to consider this practice – even though the utility of such holding periods is still being debated. Thus, the GRId system may become analogous to the SEC’s use of “negative disclosure” as a way to drive behavior. As a result, companies will need to recognize that GRId may, indirectly, become another pressure point for conforming to RMG’s governance policies even where legitimate reasons to not follow a specific policy may exist.
Ratings Will Not Affect Proxy Voting Recommendations – but May Affect Votes.
RMG continues to make clear that it does not intend to use GRId ratings as a factor in formulating its proxy voting recommendations, including, presumably, recommendations on “Say on Pay” resolutions. Consequently, even if RMG judges a company’s corporate governance-related risks to be a “low” level of concern, it does not ensure a favorable vote recommendation for an employee stock plan proposal or the reelection of directors. On the other hand, the presence of a compensation policy or practice that generates a ‘high” level of concern under GRId will not necessarily preclude a company from receiving a favorable vote recommendation. As a practical matter, it remains to be seen how RMG’s institutional clients and other investors will make use of the GRId analyses. Even where accompanied by a favorable vote recommendation, a GRId score reflecting a “high” level of concern in the Compensation category may be sufficient to produce a negative vote – particularly on a close question. Until we’ve completed at least one proxy cycle under the GRId system, its influence on actual voting decisions is unclear at best.
Verifying GRId Data
Until the full impact of the GRId system is understood, companies should plan to verify the data that RMG compiles to conduct its analysis to ensure that this data is complete and accurate. It’s important to note that you may not have much time to verify the accuracy of your GRId data once it has been compiled. While RMG should automatically send your company an electronic message notifying you when your data is available for review, we recommend requesting a log-in number (for this, contact RMG at support-corporate@riskmetrics.com or call (301) 556-0570). Companies that already have access to RMG’s Governance Analytics may use their current password to access their data.
Because of the potentially limited amount of time between the availability of this information and your annual meeting of shareholders, we recommend that you check your profile as soon as it is posted to ensure that any errors have been identified (and fixed) and a corrected report is issued, and thereafter at least annually.
For more information, see the GRId data verification web page at www.riskmetrics.com/data_verification.
Need Assistance?
Compensia has had significant experience in helping companies to design and implement their executive compensation programs. If you have any questions on the subjects addressed in this Thoughtful Pay Alert or would like assistance in assessing their likely impact on your executive compensation plans and arrangements, please feel free to contact us.
Download a pdf of this article »