Savvy Financial Analysts Adapt to Regulation ‘Fair Disclosure’

The prediction was dire: if financial analysts ceased to get preferential access to information from the companies they followed, the quality of their forecasts would suffer and cause volatility in the markets.

That was the claim made by financial analysts in October 2000 when the Securities and Exchange Commission passed Regulation Fair Disclosure (Reg FD), restricting firms from disclosing nonpublic information to preferred analysts and institutional shareholders. Specifically, Reg FD required that firms conduct investor communications in such a way that all investors received material information at the same time.

After analysts lost their access to inside company contacts, however, their forecast accuracy did not worsen. In fact, as demonstrated in a recent study by Shyam Sunder (Kellogg School of Management) and Partha Mohanram (Columbia University), Reg FD encouraged analysts to obtain independent information, an important benefit for the investors who rely on their advice.

Sunder and Mohanram set out to discover how analysts changed their operations to maintain their forecast accuracy after Reg FD was passed. Using data from the Institutional Brokers Estimate System and COMPUSTAT, the researchers examined the following: whether analysts changed how they weighted the different kinds of information they obtained; whether they changed the number of firms they covered; and whether some individual analysts produced less accurate forecasts, even though average performance had stayed the same. Their sample covered October 1999 through December 2001 and included firms with at least four forecasts available both prior to and after the passage of Reg FD.

Using a model developed by Barron, Kim, Lim and Stevens (1998), Sunder and Mohanram found that Reg FD leveled the playing field among analysts. Overall, large brokerage houses lost their edge in producing superior financial forecasts. However, analysts from big brokerage firms that were classified as “all star” analysts by Institutional Investor magazine in 1999 continued to produce superior forecasts after Reg FD was enacted. When access to privileged information was restricted, the better analysts developed new information sources which helped them maintain their superior forecasting ability.

Sunder and Mohanram noted a decline in the number of firms covered by individual analysts. In the big brokerage houses, there was an overall drop of one firm per analyst. For “all star” analysts, however, there was essentially no decrease in the number of firms covered.

Further, the researchers determined that analysts placed greater emphasis on obtaining information independently, in part by reducing coverage of well-followed firms and increasing coverage of firms that had received less attention. The quality of “common information” (information disclosed by companies to all analysts) remained unchanged after the passage of Reg FD. The quality of “idiosyncratic information” (information that analysts developed independently through their own methods) actually improved (Figure 1).

Figure 1: Mean information precision around Reg FD

Sunder gave an example of how and why Reg FD has resulted in increased development of idiosyncratic information:

Sunder and Mohanram speculate that some analysts may prefer to invest their intellectual efforts on firms where they can distinguish themselves “instead of herding around extensively followed firms.” Also, Reg FD may have raised the cost of following the well-followed firms, making it worthwhile to invest effort in following a greater variety of companies.

Leveling the field for analysts has implications for individual investors, Sunder pointed out. “Quid pro quo was part of the old system,” he said. “If analysts get their information from the companies, they can become a marketing tool for the company, talking up their stocks. Reg FD has resulted in more diverse and better analyzed views.”

Further reading:

Barron, Orie E., Oliver Kim, Steve C. Lim and Douglas E. Stevens (1998). “Using analysts’ forecasts to measure properties of analysts’ information environment.” Accounting Review, 73(4): 421-433.