related documents
RR67 - Corporate Financial Disclosure and Analyst Forecasting Activity - Preliminary Evidence for the UK
ACCA Research Report No. 67
The full report presents the results of a research project concerned with discretionary financial disclosures by UK quoted companies. It reports the findings of a survey of the quality of company accounts published in 1996 and investigates the extent to which the amount of discretionary information published by companies is associated with investment analyst following. Particular attention is paid to the information disclosed by UK quoted companies in their operating and financial reviews (or equivalent disclosures in statements such as the Chairman's Statement, the Chief Executive's Review, and the Finance Director's Review).
Most of the published research literature on the motives for and the effects of voluntary financial disclosure is based on data for US companies (see Healy and Palepu (2000) and Walker (1997) for recent reviews of this literature). In part this is due to the relative ease of access to high quality data that US researchers have enjoyed. In recent years the availability of UK data has improved, and one purpose of this report is to investigate the scope for combining machine-readable annual reporting data with data on analysts' forecasts and analyst following. In particular we show how the IBES International Data (IBES (1997)) can be used to study the determinants of analyst forecasting activity in the UK.
The main purpose of the research project was to examine the extent to which the publication of forward-looking information in the annual reports of companies is associated with analyst following. This research forms part of a broader project that is concerned with the causes and consequences of voluntary financial disclosures, and the quality of financial communication between companies and investment professionals.
There is a related body of work in this area that attempts to measure the effects of increased levels of disclosure on share prices, the cost of equity capital, and the timeliness of reported earnings (for leading examples see Botosan (1997) and Schleicher and Walker (1999) for the London and New York stock markets respectively). The current work is complementary to this body of research because it serves to confirm the existence of a rich channel of communication from companies to an influential group of investment specialists. The claims of an observable impact of increased disclosures on share prices are strengthened if one can trace a direct influence of financial disclosures on the research activities of professional investment advisors. The present report makes a modest advance in this direction.
In stock market based economies, such as the US and the UK, the informed pricing of corporate securities is vital for economic stability and the promotion of sustained levels of high quality investment by corporations. Until relatively recently, conventional financial indicators of performance, such as earnings and book value per share, were generally perceived as providing a useful starting point for company valuation. During the 1990's the usefulness of these conventional measures of financial performance as a basis for fundamental valuation have increasingly been called into question.
The diminishing relevance of conventional accounting indicators, has manifested itself in two, related, ways. First, we have witnessed an increased tendency for many firms to publish other forward-looking information within the annual report. Schleicher (1996), for example, has shown that annual reports increased materially in length over the period 1975 to 1996, and that most of the increase in the content of the annual report was in the un-audited, qualitative, and forward-looking sections of the report. Companies, it would appear, have recognised the limitations of the conventional indicators, and have produced alternative forms of information to address them. Second, a number of researchers have discovered that the ability of accounting earnings and book values, to explain company valuations, or changes in valuation over time, has materially declined.
This project presents evidence consistent with the view that one of the main considerations driving the financial disclosure choices of UK companies, is the need to supply forward-looking, value-relevant information to professional financial analysts. This evidence should be of interest to anyone who wishes to understand recent changes in the scope and content of annual reports. Accounting policy-makers, and others concerned with regulating the flow of information from firms to professional analysts may be interested in the richness of the communications that now routinely take place within the un-regulated portions of the annual report. Ordinary investors need to be aware of the range of indicators that move professional investment opinion. Some finance directors may find it interesting to learn that increased analyst following appears to carry a "price", in terms of the need to increase the quality of forward-looking information available to this class of users. Finally, some analysts may be interested in the way that improvements in financial disclosure appear to follow increases in analyst following.
In this project we construct an index of forward-looking financial disclosures based on a detailed analysis of the operating and financial reviews (or related disclosures) for a sample of 57 UK quoted companies. Our analysis of such disclosures reveals considerable variation in company practice. From a list of 154 line items disclosed by at least one company, we found that the number of line items disclosed ranged from a minimum value of 17 for the least forthcoming company, to a maximum value of 111 for the most forthcoming.
Further analysis of our disclosure scores revealed a high degree of robustness of our scoring methodology. In particular we found that the ranking of firms by disclosure stance was insensitive to whether the ranking was based only on the most frequently disclosed items, or whether it was based on the least frequently disclosed. We also found that a ranking of line items by frequency of disclosure was virtually identical between sub-samples of high disclosure firms and low disclosure firms. These sensitivity results suggest that our scoring technique possesses a high degree of internal consistency.
A practical problem with our scoring technique is that it is costly to implement. We estimate that each annual report takes about 0.5 days each to score. We therefore considered the extent to which cruder indicators of disclosure might be used in the context of large-scale statistical research requiring broad-brush numerical indicators of disclosure quality. In this respect we found a fairly high positive correlation between our disclosure index and the word count of the operating and financial review (or equivalent reviews). This suggests that word count might reasonably be used in large scale studies requiring a classification of firms into, say, High, Medium, and Low disclosure groups.
The main purpose of the study was to examine the relation between the quality of forward-looking information disclosed in the annual report and analyst forecasting activity. In undertaking this task we needed to control for a number of other factors that influence analyst forecasting activity, especially firm size. The landmark paper of Bhushan (1989) guided our choice of relevant control variables. Our analysis reveals that the level of analyst forecasting activity is primarily influenced by the quality of disclosure. Firm size, and firm volatility also have an effect, but these effects are not as strong as the quality of disclosure effect.
When we carried out a companion analysis that treated disclosure quality as the dependent variable we found analyst forecasting activity to be the main determinant. In particular forecasting activity completely dominated firm size as a determinant of disclosure quality. Our finding of a strong positive association between analyst forecasting activity and the level of financial disclosure is broadly consistent with the findings of US researchers (notably Bhushan (1989) and Lang and Lundholm ((1993), and (1996)).
In the final chapter of full report we make a preliminary attempt to investigate the main direction of causation between disclosure quality and analyst forecasting activity. It is logically possible for analyst forecasting activity to be induced by higher levels or disclosure, or for higher levels of disclosure to be driven by higher levels of analyst following. Indeed it is logically possible for both types of effect to operate simultaneously.
We investigate the direction of causation issue using a sample of 230 companies for two consecutive financial years. To avoid the costs of scoring 460 annual reports we used the page length of the annual report as a crude proxy for disclosure quality. Since most firms' annual reports adopt a similar format from one year to another, the page length gives a fairly sensitive indication of firms that have changed their disclosure stance.
Using standard econometric modelling methods we tested whether increases in page length are either preceded by or followed by associated changes in analyst forecasting activity. The results indicated a statistically significant increase in page length following an increase in analyst forecasting activity. This result indicates that as analyst interest in a firm increases, so the firm is driven to increase the level of disclosure in its annual report.
We also found some evidence of the reverse effect. We found a, slight, tendency for analyst forecasting activity to increase following an increase in the page length of the annual report. This effect was not as strong as the "increased disclosure follows increased forecasting activity" effect. These results on the direction of causation are novel for the UK and run counter to the evidence reported by Lang and Lundholm (1996) for the New York stock market.
The scope for UK academic research in this subject area has been restricted by the lack of high quality, machine-readable data, especially machine-readable annual reports, and to some extent by the limitations of software for working with textual as opposed to statistical data. Appendix 1 of the full report discusses some of the data problems the researchers encountered as guidance to others contemplating research in this area. One particular issue, on which more work could be done, would be to investigate the possibility of reducing the costs of implementing the scoring methodology by qualitative/textual analysis computer software, along the lines of Previts et al (1994), and Rogers and Grant (1997).
Key References (The full report contains a guide to further reading).
1. Bhushan, R., (1989), "Firm Characteristics and Analyst Following", Journal of Accounting and Economics, Vol. 11, pp. 255-274.
2. Botosan, C.A., (1997), "Disclosure Level and the Cost of Equity Capital", The Accounting Review, Vol. 72, pp. 323-349.
3. Healy, P.M. and Palepu, K.G.(2000), "A Review of the Voluntary Disclosure Literature." Working paper presented at the Journal of Accounting and Economics Conference 2000.
4. IBES International Inc.(1997) International Detail History CD. IBES. New York.
5. Lang, M.H. and Lundholm, R.J, (1993), "Cross-Sectional Determinants of Analysts Ratings of Corporate Disclosures", Journal of Accounting Research, Vol. 31, pp. 246-271.
6. Lang, M.H and Lundholm, R.J., (1996), "Corporate Disclosure Policy and Analyst Behaviour", The Accounting Review, Vol.71, pp. 467-492.
7. Previts, G.J., Bricker, R.J., Robinson, T.R., and Young, S.J.(1994) "A Content Analysis of Sell-Side Analyst Company Reports", Accounting Horizons, Vol 8 No. 2 pp. 55-70.
8. Rogers, R.K., and Grant J. (1997) "Content Analysis of Information Cited in Reports of Sell-Side Financial Analysts", Journal of Financial Statement Analysis, Vol 3. No. 1 pp. 17-30.
9. Schleicher, T. (1996), "Corporate Financial Disclosure and Share Price Anticipation of Earnings", Master of Philosophy Dissertation, University of Manchester, England.
10. Schleicher, T. , and Walker, M. (1999 ) "Share Price Anticipation of Earnings and Management's Discussion of Operations and Financing", Accounting and Business Research (forthcoming).
11. Walker, M., (1997) "The Economics of Corporate Financial Communication", The Association of Chartered Certified Accountants, Occasional Research Paper No. 19.
Walker and Tsulta, 2001
Executive summary
The full report presents the results of a research project concerned with discretionary financial disclosures by UK quoted companies. It reports the findings of a survey of the quality of company accounts published in 1996 and investigates the extent to which the amount of discretionary information published by companies is associated with investment analyst following. Particular attention is paid to the information disclosed by UK quoted companies in their operating and financial reviews (or equivalent disclosures in statements such as the Chairman's Statement, the Chief Executive's Review, and the Finance Director's Review).
Most of the published research literature on the motives for and the effects of voluntary financial disclosure is based on data for US companies (see Healy and Palepu (2000) and Walker (1997) for recent reviews of this literature). In part this is due to the relative ease of access to high quality data that US researchers have enjoyed. In recent years the availability of UK data has improved, and one purpose of this report is to investigate the scope for combining machine-readable annual reporting data with data on analysts' forecasts and analyst following. In particular we show how the IBES International Data (IBES (1997)) can be used to study the determinants of analyst forecasting activity in the UK.
The main purpose of the research project was to examine the extent to which the publication of forward-looking information in the annual reports of companies is associated with analyst following. This research forms part of a broader project that is concerned with the causes and consequences of voluntary financial disclosures, and the quality of financial communication between companies and investment professionals.
There is a related body of work in this area that attempts to measure the effects of increased levels of disclosure on share prices, the cost of equity capital, and the timeliness of reported earnings (for leading examples see Botosan (1997) and Schleicher and Walker (1999) for the London and New York stock markets respectively). The current work is complementary to this body of research because it serves to confirm the existence of a rich channel of communication from companies to an influential group of investment specialists. The claims of an observable impact of increased disclosures on share prices are strengthened if one can trace a direct influence of financial disclosures on the research activities of professional investment advisors. The present report makes a modest advance in this direction.
In stock market based economies, such as the US and the UK, the informed pricing of corporate securities is vital for economic stability and the promotion of sustained levels of high quality investment by corporations. Until relatively recently, conventional financial indicators of performance, such as earnings and book value per share, were generally perceived as providing a useful starting point for company valuation. During the 1990's the usefulness of these conventional measures of financial performance as a basis for fundamental valuation have increasingly been called into question.
The diminishing relevance of conventional accounting indicators, has manifested itself in two, related, ways. First, we have witnessed an increased tendency for many firms to publish other forward-looking information within the annual report. Schleicher (1996), for example, has shown that annual reports increased materially in length over the period 1975 to 1996, and that most of the increase in the content of the annual report was in the un-audited, qualitative, and forward-looking sections of the report. Companies, it would appear, have recognised the limitations of the conventional indicators, and have produced alternative forms of information to address them. Second, a number of researchers have discovered that the ability of accounting earnings and book values, to explain company valuations, or changes in valuation over time, has materially declined.
This project presents evidence consistent with the view that one of the main considerations driving the financial disclosure choices of UK companies, is the need to supply forward-looking, value-relevant information to professional financial analysts. This evidence should be of interest to anyone who wishes to understand recent changes in the scope and content of annual reports. Accounting policy-makers, and others concerned with regulating the flow of information from firms to professional analysts may be interested in the richness of the communications that now routinely take place within the un-regulated portions of the annual report. Ordinary investors need to be aware of the range of indicators that move professional investment opinion. Some finance directors may find it interesting to learn that increased analyst following appears to carry a "price", in terms of the need to increase the quality of forward-looking information available to this class of users. Finally, some analysts may be interested in the way that improvements in financial disclosure appear to follow increases in analyst following.
In this project we construct an index of forward-looking financial disclosures based on a detailed analysis of the operating and financial reviews (or related disclosures) for a sample of 57 UK quoted companies. Our analysis of such disclosures reveals considerable variation in company practice. From a list of 154 line items disclosed by at least one company, we found that the number of line items disclosed ranged from a minimum value of 17 for the least forthcoming company, to a maximum value of 111 for the most forthcoming.
Further analysis of our disclosure scores revealed a high degree of robustness of our scoring methodology. In particular we found that the ranking of firms by disclosure stance was insensitive to whether the ranking was based only on the most frequently disclosed items, or whether it was based on the least frequently disclosed. We also found that a ranking of line items by frequency of disclosure was virtually identical between sub-samples of high disclosure firms and low disclosure firms. These sensitivity results suggest that our scoring technique possesses a high degree of internal consistency.
A practical problem with our scoring technique is that it is costly to implement. We estimate that each annual report takes about 0.5 days each to score. We therefore considered the extent to which cruder indicators of disclosure might be used in the context of large-scale statistical research requiring broad-brush numerical indicators of disclosure quality. In this respect we found a fairly high positive correlation between our disclosure index and the word count of the operating and financial review (or equivalent reviews). This suggests that word count might reasonably be used in large scale studies requiring a classification of firms into, say, High, Medium, and Low disclosure groups.
The main purpose of the study was to examine the relation between the quality of forward-looking information disclosed in the annual report and analyst forecasting activity. In undertaking this task we needed to control for a number of other factors that influence analyst forecasting activity, especially firm size. The landmark paper of Bhushan (1989) guided our choice of relevant control variables. Our analysis reveals that the level of analyst forecasting activity is primarily influenced by the quality of disclosure. Firm size, and firm volatility also have an effect, but these effects are not as strong as the quality of disclosure effect.
When we carried out a companion analysis that treated disclosure quality as the dependent variable we found analyst forecasting activity to be the main determinant. In particular forecasting activity completely dominated firm size as a determinant of disclosure quality. Our finding of a strong positive association between analyst forecasting activity and the level of financial disclosure is broadly consistent with the findings of US researchers (notably Bhushan (1989) and Lang and Lundholm ((1993), and (1996)).
In the final chapter of full report we make a preliminary attempt to investigate the main direction of causation between disclosure quality and analyst forecasting activity. It is logically possible for analyst forecasting activity to be induced by higher levels or disclosure, or for higher levels of disclosure to be driven by higher levels of analyst following. Indeed it is logically possible for both types of effect to operate simultaneously.
We investigate the direction of causation issue using a sample of 230 companies for two consecutive financial years. To avoid the costs of scoring 460 annual reports we used the page length of the annual report as a crude proxy for disclosure quality. Since most firms' annual reports adopt a similar format from one year to another, the page length gives a fairly sensitive indication of firms that have changed their disclosure stance.
Using standard econometric modelling methods we tested whether increases in page length are either preceded by or followed by associated changes in analyst forecasting activity. The results indicated a statistically significant increase in page length following an increase in analyst forecasting activity. This result indicates that as analyst interest in a firm increases, so the firm is driven to increase the level of disclosure in its annual report.
We also found some evidence of the reverse effect. We found a, slight, tendency for analyst forecasting activity to increase following an increase in the page length of the annual report. This effect was not as strong as the "increased disclosure follows increased forecasting activity" effect. These results on the direction of causation are novel for the UK and run counter to the evidence reported by Lang and Lundholm (1996) for the New York stock market.
The scope for UK academic research in this subject area has been restricted by the lack of high quality, machine-readable data, especially machine-readable annual reports, and to some extent by the limitations of software for working with textual as opposed to statistical data. Appendix 1 of the full report discusses some of the data problems the researchers encountered as guidance to others contemplating research in this area. One particular issue, on which more work could be done, would be to investigate the possibility of reducing the costs of implementing the scoring methodology by qualitative/textual analysis computer software, along the lines of Previts et al (1994), and Rogers and Grant (1997).
Key References (The full report contains a guide to further reading).
1. Bhushan, R., (1989), "Firm Characteristics and Analyst Following", Journal of Accounting and Economics, Vol. 11, pp. 255-274.
2. Botosan, C.A., (1997), "Disclosure Level and the Cost of Equity Capital", The Accounting Review, Vol. 72, pp. 323-349.
3. Healy, P.M. and Palepu, K.G.(2000), "A Review of the Voluntary Disclosure Literature." Working paper presented at the Journal of Accounting and Economics Conference 2000.
4. IBES International Inc.(1997) International Detail History CD. IBES. New York.
5. Lang, M.H. and Lundholm, R.J, (1993), "Cross-Sectional Determinants of Analysts Ratings of Corporate Disclosures", Journal of Accounting Research, Vol. 31, pp. 246-271.
6. Lang, M.H and Lundholm, R.J., (1996), "Corporate Disclosure Policy and Analyst Behaviour", The Accounting Review, Vol.71, pp. 467-492.
7. Previts, G.J., Bricker, R.J., Robinson, T.R., and Young, S.J.(1994) "A Content Analysis of Sell-Side Analyst Company Reports", Accounting Horizons, Vol 8 No. 2 pp. 55-70.
8. Rogers, R.K., and Grant J. (1997) "Content Analysis of Information Cited in Reports of Sell-Side Financial Analysts", Journal of Financial Statement Analysis, Vol 3. No. 1 pp. 17-30.
9. Schleicher, T. (1996), "Corporate Financial Disclosure and Share Price Anticipation of Earnings", Master of Philosophy Dissertation, University of Manchester, England.
10. Schleicher, T. , and Walker, M. (1999 ) "Share Price Anticipation of Earnings and Management's Discussion of Operations and Financing", Accounting and Business Research (forthcoming).
11. Walker, M., (1997) "The Economics of Corporate Financial Communication", The Association of Chartered Certified Accountants, Occasional Research Paper No. 19.