The Impact of NAS: Evidence From Value Relevance of Earnings in Bangladesh

Abstract

This study examines capital market evidence on the impact of Non-audit service (NAS) on the value relevance of reported earnings in a developing country context-using Bangladesh as the example. The main research question is addressed by an econometric analysis - using panel data methods - of data on securities traded on the Dhaka Stock Exchange in Bangladesh, over the period 1995-99. The study includes a comprehensive test of weak-form efficiency of the Bangladesh stock market. The test results support weak form efficiency and therefore lend support to the assumption of semi-strong form efficiency made in interpreting the value relevance results. Analysis of selected regulatory frameworks, with respect to NAS and auditor independence, shows that the provision of NAS can threaten the integrity of both independence in perception and independence in fact. There are some NAS for which no safeguard seems to be adequate and which are, therefore, appropriately subject to prohibition. On the other hand, for some non-audit services the threats are not so clear-cut, and auditors are then required to apply professional judgment so that the seriousness of potential threats is balanced against the effectiveness of specified safeguards. It is clear from the analysis outlined here that unlike regulatory frameworks of developed countries (such as in the US, UK and NZ) which relate to auditor independence and NAS, there is no detailed regulation relating to non-audit services in the ICAB's Manual of Professional Ethics. Therefore, if the ICAB wants to promote the ethical standards of its members, and increase public confidence in them, it will need to either revise and upgrade its own code of ethics to those of developed countries, or simply adopt the IFAC code of ethics. The study examines the impact of NAS on returns-earnings relationship and finds that the expected relationship is weaker for companies that receive a NAS from their external auditor. The regression results also indicate that the adverse effects of NAS on returns-earnings relationship are more serious when NAS is provided by non-big auditors. On the other hand, the adverse effect of NAS is weaker for Big auditors, indicating that the choice of a large auditor does seem to alleviate any negative impact of NAS. The analysis also reveals that a qualified opinion signals bad news about a company's performance which might ultimately impact on the company's returns-earnings relationship, but the results concerning the effect of audit qualification are insignificant. Finally, it is evident from the present analysis that in Bangladesh, audit quality may not necessarily be identified with auditor size, since auditing by a big firm in itself doesn't appear to make audited earnings more informative. A possible implication of the result is that companies need not worry about having their financial statements audited by a lower-cost non-big auditor, provided that no non-audit services are obtained from the same auditor. On the other hand, companies purchasing a relatively large amount of non-audit services from their auditor are likely to find it worthwhile, for their reputation, to hire a big audit firm, albeit with a fee premium.