GSJBW concludes that corporate governance is a factor that can be used "in a structured and quantitative way" in stock analysis.

The CGI rating system used by GSJBW examines governance across four categories: accounting quality, audit quality, board structure, and compensation. Companies are rated with an overall score ranging from 0 to 5 (with 5 reflecting the best governance practices), which is further broken down into subscores for the aforementioned categories (also ranging from 0 to 5). GSJBW constructed two model portfolios for each rating category: the "long portfolio" (consisting of companies with high ratings of 3, 4, or 5) and the "short portfolio" (containing companies with low ratings of 0, 1, or 2).

The performance of long only and long/short investment strategies was examined over the period Sept. 16, 2005, through May 31, 2006. Each long investment strategy outperformed the sample average with the most profound results found for the Overall Governance, Board Skills Subscore, Overall Board, and Remuneration categories. Likewise, each short strategy underperformed the average as expected. In line with these results, the combined long/short strategies all resulted in positive alpha.

GSJBW then employed the same strategies using CGI data back to August 2001. On a long only cumulative basis since that time, the strategy with the best returns was Audit (10% outperformance above benchmark), though companies with high Remuneration (8%), Board Skills (7%), and Overall Board (5%) scores also outperformed. Concentrating on a long/short basis, the best returning strategy was Remuneration (26% alpha), with Audit (22%), Board Skills (16%) and Overall Board (12%) all outperforming the overall average.

GSJBW also investigated whether corporate governance ratings could be indicators of the reliability of reported profit results. The logic behind this hypothesis is quoted here:

*"Well managed companies should have better operational momentum and therefore a higher probability to produce profit results better than market expectations."
*"Companies with higher quality reported information (e.g., as measured by 'Audit' or 'Accounting' ratings) are less likely to suffer accounting or reporting errors that require subsequent (typically negative) revision."
*"Well managed companies understand the need, and have the ability to, communicate and manage market expectations and are therefore likely to provide more reliable earnings outcomes relative to expectations."

In examining the June year-end reporting season, GSJBW looked at the reported surprise outcomes for each governance rating category for the June 2005 reporting season. The average reported surprise was reliably more positive for companies with higher governance ratings than for companies with lower governance ratings. Companies with better scores in the Overall Governance, Overall Board, and Board Skills categories and subcategories had more positive earnings surprises and fewer negative surprises. This trend was also observed for the Board Skills category back to August 2001. Top-rated Board Skills companies had a positive earnings surprise of 4.3 percent, compared to only 0.7 percent for low-rated companies.

Corporate governance rating factors are often used in risk analysis -- to screen out companies like Enron that might pose significant risks to investment returns and/or cause serious reputational damage. GSJBW used a "knockout" approach to weed out the lowest rated companies in its portfolio; they then assessed whether or not the "knocked out" companies actually did underperform. About 80 percent of the stocks with the most governance risk (those rated 0), which were screened out, actually did underperform. When the knockout rule was made less stringent --companies rated with a 0 or 1 were screened out -- about 70 percent of the knocked out stocks underperformed. And when stocks rated with a 0, 1, or 2 were knocked out, approximately 65 percent of the knocked out stocks underperformed. One can safely conclude that the use of a risk screen can increase the odds of avoiding underperforming equities.

GSJBW also investigated how the knockout screens affected portfolio returns and found that the use of these screens significantly improved the portfolio return above benchmark. By screening out stocks with 0, 1, or 2 ratings, the average return for the Board Skills, Audit, and Remuneration categories was 8.1 percent, compared to the benchmark of 4.2 percent.

In conclusion, the results found by GSJBW demonstrate a positive investment signal from the different investment strategies used in this analysis and further strengthen the case that governance factors can be used in a "structured and quantitative way."