A new paper by Bhagat and Bolton questions some of the generally accepted conclusions of the prior art on corporate governance indices, as well as offering some new conclusions.

Corporate governance indices have become a preferred method of capturing corporate governance quality. Gompers, Ishii, & Metrick (GIM) proposed the first index in 2003—their “G-index,” which assessed firms’ corporate governance quality based on their adoption or non-adoption of twenty-four governance provisions tracked by the IRRC. Characterizing firms as tending toward either democracy or dictatorship, GIM found a negative relation between dictatorship and firm value, as measured by Tobin’s Q. In addition, constructing portfolios of dictator and democracy firms, they found that democracy portfolios outperformed dictator portfolios by a statistically significant margin.

Following GIM, Bebchuk, Cohen, and Ferrell (BCF) devised their “E-index” (“E” for entrenchment). They took only six of the factors used by GIM, arguing that these six did all the entrenching work, and that an index composed of these six—pills, staggered boards, limits on charter amendments, limits on bylaw amendments, supermajority requirements for mergers, and golden parachutes—would have better predictive value than the G-index. Like GIM, BCF found that entrenchment was negatively correlated with firm Tobin’s Q, and that portfolios with low entrenchment had better stock returns that portfolios with high entrenchment.

Now come Bhagat and Bolton, with Corporate Governance Indices. They look at seven different governance measures, including not only the G-index and E-index, but also board stock ownership, CEO-chair separation, board independence, Brown and Caylor’s Gov-Score index, and an index created by The Corporate Library. They rely primarily on accounting measures of firm performance, eschewing stock market measures on the theory that they are subject to investor anticipation: long-term returns might not show a significant correlation with governance even if one exists. Among their interesting findings:

a. While better governance as measured by the G-index, the E-index, and other metrics is positively correlated with contemporaneous and subsequent operating performance (using accounting measures like ROA), none correlate with future stock performance, contrary to GIM and BCF.

b. Given poor firm performance, better governed firms as measured by the G-index and E-index are less likely to experience disciplinary management turnover despite poor performance.

An interesting study.

Corporate Governance, Empirical Legal Studies | Bookmark

TrackBack URL for this entry:
https://www.typepad.com/services/trackback/6a00d8345157d569e200d8352fcffa53ef

Links to weblogs that reference Empirical Corporate Governance: