Elsevier

Journal of Accounting and Public Policy

Volume 28, Issue 5, September–October 2009, Pages 401-418
Journal of Accounting and Public Policy

Differential effects of regulation FD on short- and long-term analyst forecasts

https://doi.org/10.1016/j.jaccpubpol.2009.07.003Get rights and content

Abstract

In this paper, we provide evidence that the quality of short-term analyst forecasts improved but the quality of long-term analyst forecasts deteriorated after the implementation of Regulation Fair Disclosure (FD). More specifically, our results show that the precision of the idiosyncratic information component of short-term forecasts improved in the post-FD period, whereas the precision of both the common and idiosyncratic information components of long-term forecasts declined. We attribute this result to the reduced disclosure of long-term, future-oriented information in the post-FD period. Thus our results support assertions by some analysts that FD has had a chilling effect on information that is relevant to long-term forecasts.

Introduction

Regulation Fair Disclosure (FD) (Securities and Exchange Commission, 2000) requires firms to disclose material information to all investors at the same time.3 Although several studies suggest that the implementation of Regulation FD has succeeded in providing equal access to the information disclosed by firms (e.g., Brown et al., 2004, Eleswarapu et al., 2004, Sunder, 2002, Mohanram and Sunder, 2006, Yang and Mensah, 2006), evidence of improvement or deterioration in the information environment remains inconclusive. Heflin et al. (2003) find no significant reduction in the availability of information after FD, whereas Ahmed and Schneible (2007) document decreases in the average information quality of both small and high-technology firms in the post-FD period. Mohanram and Sunder (2006), however, report that the precision of the idiosyncratic4 information component of short-term analyst forecasts (with a horizon of approximately one month) improved after regulation FD, although they detect no such improvement in the precision of the common information component of these forecasts. Given that the precision of idiosyncratic and common information components serves as a measure of the quality of the forecast, their results suggest an improvement in the quality of forecasts with no accompanying adverse effects of FD.

An analysis of the effect of FD on the quality of analyst forecasts, however, is not complete unless its effects on the qualities of both the short-term and the long-term forecasts are considered because both are shown by many studies to be useful to investors in estimating firm value (e.g., Brown et al., 1985, Mest and Plummer, 1999, Bandyopadhyay et al., 1995).5 In this paper, we extend Mohanram and Sunder (2006) research on the effect of FD to long-term forecasts. Given the improvement in the quality of short-term analyst forecasts as documented by them, we are particularly interested in whether FD has any adverse effect on the precision of long-term analyst forecasts. We also extend Ahmed and Schneible’s (2007) study, which documents a lower level of information quality for small and high-technology firms in the post-FD period. Because the long-term performance of small and high-technology firms is more likely to be uncertain and is more difficult to project than that of other firms, we include firm size and technology as control variables in our analysis to test their incremental effect on our forecast precision measures.

A number of surveys (SIA, 2001, AIMR, 2001b, PWC, 2001) show that financial analysts believe that firms have disclosed less valuable information after FD and that there has been a deterioration in the quality of oral communications. The managers of disclosing firms, in contrast, believe that they now disclose more information. This study provides a potential reconciliation of the differences between these two viewpoints.

In this study, we argue that although FD improved the quality of short-term forecasts, it impaired the quality of long-term forecasts. Short-term forecasts, especially those for a subsequent one-month period, are based on publicly available information, such as order backlogs (Rajgopal et al., 2003). For making these forecasts, it can be reasonably assumed that the operating structure of the firm (costs, capital, and ownership) remains constant. However, information such as current order backlogs is less useful in developing long-term forecasts. Moreover, the assumption of a constant operating structure is not as compelling for long-term forecasts as it is for short-term forecasts. A good long-term forecast requires information about possible changes in a firm’s operating environment and capital and ownership structures. Such information is typically based on disclosures of future strategies and operating plans by managers. Because of the risk and liability involved in publicly disclosing such information and because FD requires all disclosures to be made public, managers are less likely in the post-FD period to be as forthcoming with future-oriented strategy and operating plan disclosures compared to the pre-FD period. We argue that a reduction in such disclosures may have a negative impact on the quality of long-term analyst forecasts.

Prior to FD, managers provided private information to a select group of trusted analysts on the (implicit) assurance that they would not disseminate it on a wider basis. This assurance shielded managers from unwanted scrutiny and prevented the impairment of their firm’s competitive strength, did not increase the potential for unintended additional competition, and particularly shielded the managers and the firm from potential legal risks. The scope of these conference calls has been broadened in the post-FD period to include all interested parties (Jorgensen and Wingender, 2004). Because of this shift to a broader set of participants, the information that is disclosed in these conference calls is now public information. Public disclosure of a firm’s future performance expectations is likely to increase its legal risk, benefit its competitors, and result in higher litigation costs if the expected results are not achieved (Arya et al., 2005). Moreover, the public disclosure of long-term strategic information may severely affect the competitiveness of the firm within its particular industry (Darrough, 1993, Hughes et al., 2002), or it may encourage more firms to enter the industry, which, in turn, would increase competition and thereby reduce future profitability (Darrough and Stoughton, 1990). Consequently, managers are likely to be more cautious in the post-FD period about publicly disclosing the kind of information that they would have provided privately to trusted analysts in the pre-FD period.

In the pre-FD period, questions asked by analysts during conference calls primarily focused on financial and forward-looking information’s. Tasker (1998) survey revealed that a significant percentage of these questions related to industry trends (34%), management plans and initiatives (20%) such as new product roll-outs, and future operating performance (19%). The result was that managers signaled private information about the long-term prospects of firms. In contrast, only a small percentage of questions (9%) related to information that is relevant to short-term forecasts, such as backlogs and the number of units shipped, because this type of information could be obtained with little effort from public sources. Overall, these survey results suggest that private information was passed on to a group of select analysts prior to FD, and these analysts used that information to develop long-term forecasts. By precluding selective communication, FD has also reduced the incentive for analysts to probe managers for information on specific management plans and initiatives and future operating performance. Such information formerly accounted for nearly 40% of the questions asked.6 In conjunction with the reduced amount of information disclosed in conference calls, the reduced incentive of analysts to elicit more specific information further exacerbates the potential impairment of long-term forecast quality.

This discussion suggests that the precision of the idiosyncratic component of long-term forecast information is likely to be lower in the post-FD period. In contrast, consistent with Mohanram and Sunder (2006), we expect a greater precision of the idiosyncratic component in short-term forecasts. Using 1818 short-term and 2036 long-term forecasts for the pre- and post-FD periods, we conduct a comparative analysis to test our expectations about the impact of FD on short- and long-term forecasts. In our study, one-month-ahead forecasts for the coming year are considered short-term and 13-month-ahead forecasts are considered long-term. We use the methodology developed by Barron et al. (1998) to decompose information into its common and idiosyncratic components for evaluating the precision of both in short- and long-term analyst forecasts.

Consistent with our expectations, we find that the precision of the idiosyncratic information component of short-term forecasts improved, but the precision of both the idiosyncratic and common information components of long-term forecasts deteriorated in the post-FD period. Our findings for short-term forecasts are consistent with those of Mohanram and Sunder (2006), who, after controlling for other factors, report a significant increase in the precision of the idiosyncratic and total information in the last available forecast of quarterly earnings. However, our results show that Mohanram and Sunder (2006) findings do not extend to long-term forecasts.

Our study makes the following contributions to the literature. First, our results show that the quality (measured by the precision of the idiosyncratic and common components of information) of short-term forecasts improves but that of long-term forecasts deteriorates in the post-FD period. Thus, our findings indicate that the change in the analyst information environment following the introduction of FD has not been conducive to the development of long-term forecasts.7 Second, our findings provide a potential reconciliation between the perceptions of increased disclosure held by company executives and those of reduced disclosure held by security analysts following the implementation of FD. Apparently, the perceptions of security analysts are influenced by the deterioration in the information environment that is related to long-term forecasts, whereas those of managers are influenced by the improvement in the information environment for short-term forecasts. Third, our findings provide useful information to regulators about the impact of FD on the information environment. Because long-term forecasts are important for the smooth functioning of financial markets and the economy, regulators may need to consider how to mitigate the chilling effect of FD on information that is relevant to long-term forecasts.

The remainder of the paper is organized as follows. In Section 2, we provide the background and prior related work. Our research design is discussed in Section 3. Our results, including sensitivity tests, are presented in Section 4. In Section 5, we give the concluding remarks.

Section snippets

Background and related work

Regulation FD prohibits the issuers of securities or the persons acting on their behalf from making selective disclosures of potentially material non-public information about a firm to a small group of investment professionals, such as security analysts. In effect, FD eliminates the practice of providing private information directly to select analysts.

The passage of Regulation FD has been controversial from its inception because of ambiguity in its potential effect on the investor information

Study period

We define the pre-FD period as from October 1, 1998 to September 30, 1999 and the post-FD period as from January 1 to December 31, 2002.

Data collection and sample selection

We obtain our data on analyst forecasts from the IBES consensus database. The data on the other variables are collected from Standard and Poor’s Compustat database. Price data are collected from the CRSP database.

In our analyses of FD’s effects on short- and long-term analyst forecast accuracy and precision, we only consider forecasts for firms that are followed by at least three analysts because the precision variables, S and H, are not meaningful if the number of analysts is less than three.

Concluding remarks

This paper provides evidence that FD reduces the precision of both common and idiosyncratic components of long-term (13-month-ahead) earnings forecasts, while it improves the precision of idiosyncratic component of short-term forecasts, consistent with Mohanram and Sunder (2006). We attribute this decline in the quality of long-term forecasts to reluctance by managers to publicly disclose predominantly subjective, strategic, and future-oriented information because of potential litigation risk

Acknowledgements

We are grateful to the editor, Prof. Martin Loeb, two anonymous reviewers and to the workshop participants in City University of Hong Kong, The Hong Kong Polytechnic University, State University of New York at Buffalo, Rutgers University, University of Technology at Sydney and the 2006 AFAANZ Conference for their comments and suggestions. We also thank Profs. Shyam Sunder and Partha Mohanram for their valuable comments on an earlier version of this paper.

References (36)

  • Association for Investment Management and Research (AIMR), 2001b. AIMR surveys, March 26, 2001...
  • W. Bailey et al.

    Regulation fair disclosure and earnings information: market, analyst, and corporate responses

    The Journal of Finance

    (2003)
  • R.D. Banker et al.

    Sensitivity, precision, and linear aggregation of signals for performance evaluation

    Journal of Accounting Research

    (1989)
  • O. Barron et al.

    Using analysts’ forecasts to measure properties of analysts’ information environment

    The Accounting Review

    (1998)
  • P. Brown et al.

    Security analysts’ multi-year earnings forecasts and the capital market

    (1985)
  • B.J. Bushee et al.

    Managerial and investor responses to disclosure regulation: the case of Reg FD and conference calls

    The Accounting Review

    (2004)
  • M.N. Darrough

    Disclosure policy and competition: Cournot vs. Bertrand

    The Accounting Review

    (1993)
  • V. Eleswarapu et al.

    The impact of regulation fair disclosure: trading costs and information asymmetry

    Journal of Financial and Quantitative Analysis

    (2004)
  • Cited by (18)

    • Textual disclosure complexity and analysts’ weighting of information

      2024, Journal of Contemporary Accounting and Economics
    • Anti-selective disclosure regulation and analyst forecast accuracy and usefulness

      2020, Journal of Corporate Finance
      Citation Excerpt :

      Heflin et al. (2003) find no change in accuracy whereas Gomes et al. (2007) report improved accuracy for large-cap stocks but not overall. Srinidhi et al. (2009) report increased accuracy of short-term forecasts (which are most common), but reduced accuracy of long-term forecasts. Bailey et al. (2003) report unchanged accuracy of forecasts for the nearest quarter but reduced accuracy of forecasts for future years.

    • Analysts' strategic inducement of management guidance

      2016, Journal of Accounting and Public Policy
    • Regulation Fair Disclosure and analysts' reliance on earnings announcements

      2013, Journal of Accounting and Public Policy
      Citation Excerpt :

      These studies conclude that analysts may reduce coverage because they need more efforts to analyze firms in post-RFD. Using the private information measure developed by Barron et al. (1998), Mohanram and Sunder (2006) and Srinidhi et al. (2009) find that analysts produce more private information in order to issue forecasts in the post-RFD period. Janakiraman et al. (2007) find that analysts delay the issuance of their first forecast in post-RFD, which implies that analysts take more time to analyze the information.

    View all citing articles on Scopus
    1

    Tel.: +852 2788 7924.

    2

    Tel.: + 732 445 3539.

    View full text