As you may know, RiskMetrics Group ("RMG") recently announced that it will be replacing its Corporate Governance Quotient ("CGQ") with a new governance ratings methodology – Governance Risk Indicators ("GRIds"). We have reviewed the GRIds materials and talked with people at RMG about the new system.

There will be many changes, of which the following are only the most significant:

  1. Replacing the prior CGQ data points will be 63 factors on which RMG will rate a company's corporate governance. Rating will be based solely on public filings.
  2. These rating factors will be grouped into four categories, or "dimensions" – Audit, Board Structure, Compensation and Shareholder Rights. Each factor will be given a maximum number of points that it will be possible to achieve for that factor and, within that maximum number, points will be awarded for the extent to which the company complies with that factor. Depending on the specific factor, it may call for an answer that is binary (yes or no), a number (e.g., the number of financial experts on the audit committee), a percentage (e.g., the percentage of independent directors) or some other reply. These points will then be aggregated to determine a "level of concern" – high, medium or low – for each of the four dimensions.
  3. RiskMetrics has now provided its initial levels of concern for each dimension as follows: (i) Audit, high risk is less than 75%, medium risk is 75 to 90% and low risk is above 90%; (ii) Board Structure, high risk is less than 60%, medium risk is 60 to 70% and low risk is above 70%; (iii) Compensation, high risk is less than 55%, medium risk is 55 to 70% and low risk is above 70%; and (iv) Shareholder Rights, high risk is less than 35%, medium risk is 35 to 60% and low risk is above 60%.
  4. Unlike the CGQ, GRIds scoring will not be relative to all other companies or to other companies in the industry, but will reflect only that company's percentage of the maximum possible points it received in each dimension. The old CGQ scores will be frozen as of early March and retired completely by June 30, 2010.
  5. The weightings given to many of the GRIds factors have been changed from the weightings for the CGQ data points.
  6. We have been told that the GRIds Profile for each company "will not directly impact" the existing (and continuing) proxy voting analyses that RiskMetrics has performed for many years. However, it will be included "up front" in the proxy voting analysis for each company.
  7. When a company's proxy statement is filed, RiskMetrics will calculate the company's GRIds Profile and will publish it shortly before the shareholders' meeting. There will be a data verification site made available to all covered companies but we have been told that RiskMetrics "cannot guarantee" that the site will be made available this year before the proxy voting analysis is made public.
  8. A company's GRIds profile will be updated three to four times per year and will be reviewed whenever a company makes an update request through the data verification website. The overall methodology for GRIds will be evaluated and potentially updated annually.

In reviewing the GRIds factors, we have found that many of them are carried over from the CGQ data points. (One salutary omission: Attendance at a RiskMetrics-approved director education program – a collateral RiskMetrics business.)

There continue to be anomalies, however. For example, RiskMetrics conceded to us that not having a particular governance practice, e.g., a long-term cash incentive plan, could prevent a company from getting points for having the right terms (according to RiskMetrics) for that practice. We were told that this problem would be considered. As another example, in considering "multiple voting rights," RiskMetrics says it will consider the "par value per share," which has no connection to voting rights. Also, RiskMetrics continues to view blank check stock, which is conventional among public companies, as a negative, although it will accept a "firm commitment" (not a policy and not a commitment contingent on independent directors' approval) not to use issue blank check stock "as an anti-takeover device."

As with the CGQ, the new GRIds system reflects RiskMetrics' own views of various corporate governance matters, despite the contrary views of many serious participants in the evolution of corporate governance and despite the varying benefit of particular governance practices from company to company. Nevertheless, RiskMetrics has decided, for example, that classified boards, executive board chairmen, shareholder rights plans and plurality voting are bad at any company.

RiskMetrics told us last week that the averages for the first 40 companies that it has pre-scored are Audit – 100%, Board Structure – 84%, Compensation – 71% and Shareholder Rights – 65%. As RiskMetrics publishes these averages for the 6,400 U.S. companies it covers, they should provide some indication of where a company places within the RiskMetrics universe.

According to RiskMetrics, "technical documents," including the points per factor, will be published this week and will disclose, among other information, the methodology that it will use to determine the assessment of risk.

Finally, RiskMetrics claims that "Governance Risk Indicators are one of many investment management tools to help institutions and other financial market participants measure and flag investment risk. We hope GRIds will . . . aid in the timely mitigation of portfolio risk . . . ." Note the new emphasis on "risk." Yet, the connection, if any, between various corporate governance practices and economic performance and/or risk is not clear. Indeed, several years ago, Institutional Shareholder Services ("ISS," now RiskMetrics) published a study, with Georgia State University, finding that shareholder rights plans and other takeover defenses correlated positively with higher shareholder returns (over three, five and ten years), stronger profitability measures (return on equity, return on assets, return on investment and net profit margin), higher dividend payouts and yields and higher interest coverage and operating-cashflow- to-liability ratios. ISS, reflecting its predisposition, called these results a "surprise;" but they were no surprise to business people and their advisers with experience in the often destructive results of hostile takeovers. Similarly, more recent academic studies have found no consistent relation between corporate governance indices and corporate performance and, specifically, no predictive validity for the now expiring CGQ.

Nevertheless, RiskMetrics has become a major force in influencing the voting of institutional shareholders and its positions cannot be ignored. Many of its views have become mainstream. The ultimate goal of any for-profit enterprise, however, is wealth maximization, not a high corporate governance score. Under Maryland law, a director's duty is to act in a manner that the director reasonably believes to be in the corporation's best interests, which may or may not be the same as what a proxy adviser, even one as influential as RiskMetrics, or some other external group thinks is good corporate governance.

We would be happy to review and discuss your GRIds profile with you as we have found that there are often mistakes, opportunities for partial credit, "low-hanging fruit" and other possible ways to improve a CGQ and now GRIds ratings without significantly affecting company operations.