An Empirical Analysis of the Impact of KAM Disclosure Quantity and Quality on Earnings Management: Evidence from U.K Listed Companies

Abstract

The agency theory asserts that timely disclosure of more high-quality information promotes greater transparency in corporate governance and reporting systems (Ribstein, 2005). This, in turn, reduces the degree of information asymmetry between top management (agent) and external stakeholders (principal), thereby limiting opportunities for managers to dishonestly manipulate financial statements.

Therefore, the aim of the current study is to empirically investigate whether the disclosure of key audit matters (KAM) in auditors' reports affects managers' reporting behaviour. More specifically, it uses a quantile regression method to analyse the relationship between earnings management (EM) and the quantity and quality of KAM disclosed by 201 non-financial firms which persistently listed on the U.K stock exchange FTSE350 from 2013 to 2018 with a total of 1206 observations.

EM is measured using the accrual-based approach and real earnings management. The KAM quantity is measured by the total number of KAMs disclosed in the auditing report while KAM quality is characterised by two variables which capture the specificity of accounting and entity-level risk factors. Account-level KAMs is represented by the total number of KAMs arising from accounting issues while entity-level KAMs, is measured by the total number of KAMs relating to firm-specific risk factors.

The results show that both KAMs' quantity and quality reduce managers' proclivity to misrepresent financial statements. This study found that disclosure of matters associated with entity-level risks favours the reduction of discretionary accruals than the revelation of accounting level audit matters at the percentiles from 0.15 to 0.50 and at the top of percentiles from 0.80 to 0.95. Also, it was found that account-level KAMs risk disclosures only influence the real activity earnings management and not accrual-based earnings management measurements. By contrast uncovering entity-level KAMs issues affects accruals-based EM but not real activity EM performance outcomes. In addition, it was discovered that most U.K firms' managers exhibit a preference for real activity earnings management than accruals-based EM to mispresent their earnings, presumably because of the lower possibility and consequences of getting caught, as real activity earnings management activities are harder to pursue and scrutinise for outsiders.

In general, the quantile regression results suggest that disclosing accounting and firm-level KAMs, significantly improve managerial financial reporting behaviour. This research will contribute to knowledge in various ways. Firstly, it will fill the gap and extend the literature on this subject, given that the study is novel by using quantile regression to analyse the impact of
KAM disclosure on EM. Secondly, this is the first study that examines the impact of KAM quantity and quality on two different practices of EM, namely accruals and real-based activity EM.