An empirical study of the quality of disclosure in corporate annual reports in developing countries : the case of Bangladesh

Abstract

Research on corporate financial reporting in developing countries is too general and largely descriptive. The few research that assessed the quality of corporate disclosure in those countries suffer from methodological weaknesses and provide only a partial picture of the status of corporate financial reporting practices because of narrow scope. For example, quality of disclosure is measured by looking at a single criterion like voluntary disclosure. Also, most of the studies used the standard Ordinary Least Squares (OLS) method of estimation which is not appropriate in modelling disclosure quality, measured as an index bounded by zero and one. The purpose of this thesis is to measure the disclosure quality of corporate annual reports in Bangladesh, both cross-sectionally and longitudinally, and to determine whether key company attributes have any impact on the quality of disclosure using a modelling framework suitable for the disclosure index. A two-step research method is adopted to test the six null hypotheses stated in the thesis. The first step involves a questionnaire survey of auditors, accountants, bank loan officers and financial analysts (broadly defined as "users") to determine whether these groups perceive the importance of the selected disclosure items differently, and to measure whether corporate annual reports meet their information needs. The second involves a cross-sectional analysis of the 1987 -88 annual reports of 63 listed companies, and a longitudinal analysis of the annual reports of each of the 30 listed companies for a period of five years beginning in 1983-84. The disclosure index method is used to define disclosure quality. Two indexes are used: the overall disclosure quality consisting of both statutory and voluntary items, and the statutory disclosure quality comprising of mandatory items only. The results show that the user-groups had a significant consensus concerning the importance of information to be disclosed in corporate annual reports, but some significant differences emerged when the combined perceived information needs of the user-groups were matched with the overall disclosure quality. Consistent with some prior studies, it is found that a significant relationship existed between the perceived information needs and the overall disclosure quality for auditors and accountants, but not for bank loan officers and financial analysts. There were also several instances of noncompliance with the existing mandatory disclosure regulations, such as the Companies Act,1913 and the Securities and Exchange Rules, 1987, which have reduced greatly the usefulness and credibility of corporate annual reports in Bangladesh. It is noted that the overall disclosure quality had improved significantly over the five-year period surveyed. The logistically transformed Ordinary Least Squares regressions reveal that the qualifications of the principal accounting officer of the reporting company, size of the company's audit firm, and whether or not the firm was a subsidiary of a multinational company had significant impact on the overall and mandatory disclosure quality. It is found that the estimated probability of disclosing both statutory and voluntary information was 64.6 per cent for a company which was a subsidiary of a multinational company, whose principal accounting officer was a qualified accountant, and whose accounts were audited by a large audit firm. If a company did not possess any of the above characteristics, the estimated probability of disclosure was only 46.7 per cent- a fall of 17.9 per cent. In the case of statutory information only, the degree of compliance was higher; the probability of complying was 88.5 per cent for companies with the above characteristics, but only 67 per cent for those without. Show more