Amid great controversy, the SEC implemented Regulation Fair Disclosure (Reg FD) on October 23, 2000. The new rules prohibit companies from disclosing material information such as earnings guidance to a subset of analysts without also announcing it publicly. Over three years later, we can begin to investigate the impact of these new rules. Some may find the results surprising.

At the time, the adoption of Reg FD created heated discussion among market participants. Critics were fearful that the new rules would hinder the release of information, leading to greater uncertainty and greater analyst forecast error. Certainly, many investors rely on analyst earnings estimates to make investment decisions. Consensus estimates are a critical element in forward valuation models and can serve as a useful benchmark by which investors can compare their own forecasts.

Our research suggests that both the critics and supporters of Reg FD vastly overstated the initial impact of these new rules. In fact, Reg FD seems to have had little noticeable effect on the accuracy and dispersion of analyst earnings forecasts. This finding indicates that, for the time being, consensus estimates still provide a useful barometer for investors in their analysis of a company's earnings prospects.

Analyst Accuracy
The accuracy of analyst forecasts has changed dramatically over the last decade. The chart tells the story of this virtual evolution and shows that the impact of Reg FD is merely a footnote to the larger story. Using detailed analyst data from IBES, we calculated the average realized error in the consensus estimates for stocks in the Russell 1000 universe going back to 1984. The error is expressed as a percentage of the absolute value of the realized earnings per share (EPS). In order to eliminate the impact of outliers, we omitted stocks where the denominator was less than $0.03.

            Source: SSgA, based on IBES data

There are several striking features in the data. First, analysts consistently overestimated EPS during the 1980s and the early 1990s. This suggests that corporations were less aggressive at guiding analysts during that time period. It also suggests that analysts were biased in favor of the stocks that they followed and assigned a higher weight to positive information than to negative information.

Over time, analyst accuracy noticeably improved. The chart demonstrates this phenomenon, as the percentage error declines dramatically during the latter half of the 1990s. What was behind this improvement? It is likely that corporations became more sensitive to missing their earnings estimates and began to guide analysts more explicitly, hoping to temper the volatility of their stock price.

Over the past several years, consensus estimates have systematically underestimated earnings. This finding runs counter to the common perception that analysts are overly optimistic. The chart shows that the average consensus estimate has been too low every quarter since 1999.

The recent earnings pessimism among analysts seems unrelated to the advent of Reg FD. We now have 11 full quarters of data since the passage of Reg FD and the average error has been 2.55% during those quarters versus 3.01% in the preceding 11 quarters. This finding suggests that Reg FD had no meaningful impact on analyst accuracy. In other words, analyst estimates still provide a useful guide for investors, particularly for those savvy enough to adjust for the slight, systematic downward biases in these estimates.

Dispersion
Dispersion is often used to measure the degree of uncertainty surrounding earnings estimates. Those in favor of Reg FD felt that dispersion would decrease in the aftermath of the new rules, arguing that since all analysts would have access to the same information, uncertainty and, therefore, dispersion would decrease. Those opposed to Reg FD believed that management would reveal less information under the new rules, leading to decreased certainty and greater estimate dispersion.

We measured the historical estimate dispersion each quarter from Q4 1984 through Q2 2003, by dividing the standard deviation of forecasts by the absolute value of the mean forecast. We found no discernable change in dispersion post-Reg FD. In fact, we found that dispersion declined systematically during the 1980s and the early ‘90s, in lockstep with increasing analyst forecast accuracy. Dispersion fell by 50% during this period and has remained fairly constant since 1994. In other words, Reg FD did not alter the degree of forecast uncertainty.

Conclusion
Analyst earnings estimates still provide a useful guide for investors post-Reg FD. There has not been a significant decrease in forecast accuracy since the issuance of the new rules. In addition, recent consensus forecasts are dramatically better than the forecasts of ten and 15 years ago.

Moreover, the degree of uncertainty surrounding consensus estimates has remained constant since the mid 1990s.

Regulation FD created both great hope and great fear that it would dramatically change the fundamental relationship between analysts and corporate management. Our findings indicate that the new rules were neither a panacea for nor the death knell of analyst forecast accuracy. In hindsight, it appears that Regulation FD did not merit the heated discussion that it created.

This material is for your private information. The views expressed are the views of Ric Thomas and Jennifer Sjostedt only through the period ended February 25, 2004 and are subject to change based on market and other conditions. The opinions expressed may differ from those with different investment philosophies. The information we provide does not constitute investment advice and it should not be relied on as such. It should not be considered a solicitation to buy or an offer to sell a security. It does not take into account any investor's particular investment objectives, strategies, tax status or investment horizon. We encourage you to consult your tax or financial advisor. All material has been obtained from sources believed to be reliable, but its accuracy is not guaranteed. There is no representation nor warranty as to the current accuracy of, nor liability for, decisions based on such information. Past performance is no guarantee of future results.

Posted On: March 04, 2004