The incremental information content of the annual report and accounts
1. This study examines the value to financial analysts and investors generally in the UK of firms' accounting disclosures and other information contained in the annual report and accounts for share valuation purposes. 2. Using daily share price data and a large sample of actively traded UK firms the relative information content of four major information releases, the preliminary announcement, annual report and accounts, annual general meeting and the interim report, are examined. Three different models are employed, two of which involved the calculation of the market risk measure; the firm's beta. To increase comparability with previous studies, some of the tests were replicated using weekly data. 3. Previous studies have shown a bias when calculating abnormal returns due to the size composition of the sample. Re-estimating the parameters using Ordinary Least Squares but including a size variable, showed virtually no effect on the magnitude of the estimated parameters. Neither the constant `alpha' nor the coefficient 11 of the size variable were statistically significant. The former is consistent with prior research. 4. Infrequency of trading is largely associated with small companies and gives rise to a downward bias when estimating betas. Although the sample comprised actively traded mainly large companies, the betas were re-estimated using a method which takes into account thin trading. The results indicated that parameter estimates have to take into consideration thin trading even when using predominantly large actively traded stocks. The degree of stability in the betas over successive periods was low though the difference in the average betas of the two periods was negligible. The pooled betas were therefore used. 5. The initial test was to ascertain the extent of the information content of the four events. The three different model formulations employed produced almost identical results suggesting a naive model with beta=1 may well be adequate in many such event study situations. Therefore, all further tests were conducted using only the market model with adjusted betas. Of the four events, the preliminary announcement and interim report, largely representing earnings and dividends announcements, had the highest information content, whilst little information in aggregate is conveyed to the market by the annual general meeting and the annual report and accounts. Similar results were produced using a different information measure and weekly data. No unusual share price activity was observed prior to 111 the event day or after event day plus one, which is consistent with the semi-strong form of the Efficient Market Hypothesis.6. Evidence was found in tests of the market's reaction to `good' and `bad' news that, on average, they are given equal value except for the interim report where there is a sharper reaction to `bad' news. 7. This study supports previous research showing an inverse relationship between company size and the abnormal return on an event day. When variance of returns is included in the regression, the size coefficient albeit statistically significant becomes negligible in magnitude which suggests, like prior work, that size is probable a surrogate for absent firm specific variables. 8. Using a control group of companies with low annual report and accounts abnormal returns but otherwise matching a sample of outlier companies with high annual report and accounts residuals, an analysis was made of both groups. Little difference was noted in the amount of statutory or voluntary information disclosed in the preliminary announcements of the groups. There was, however, significant evidence of greater price-sensitive information being provided in the annual reports and accounts of the outlier group. 9. There was little evidence to suggest that one group was more closely followed by analysts than the other. Capital gearing and income cover were, on average, little different. Apparently what was driving the returns of the outlier sample was company specific information. The information seemed to be largely contained in two sections of the annual report , the balance sheet and the chairman's statement, which previous studies have shown to be valued by users. 10. Tests of association between information content of the events produced conflicting results but analysis of the press comments seemed to furnish some evidence of an informational relationship between the preliminary announcement and the annual report and accounts of the outlier group not observed in the control group. Press comments on the annual reports of the outlier group bear some relationship to the press comments on their preliminary announcements. 11. No previous study has directly examined the informational value of the annual report and accounts for individual firms and used analysis of press comments to identify those parts of the annual report which seem to have information content. The value of this study lies in the new evidence that it provides suggesting the annual report and accounts does have some information value for the stockmarket for particular firms and in identifying those parts of the annual report found useful by investors. 12. The implications of this study are that whilst, in aggregate, there seems to be an apparent lack of incremental information in the annual report and accounts ,this is not a true reflection of its value to all market participants. Unless the stockmarket has access to this financial statement potential adverse valuation consequences may be missed. This study suggests summary accounts are unsuitable for active stockmarket participants.