Detecting Earning Manipulation in a Developing Economy: An Empirical Study Using Beneish M-score Model

Manipulation of financial statements entails the intentional and often ill-motivated manoeuvring of financial
records towards a pre-determined target. In such cases, motivations include achieving budgetary targets and
rewarding senior managers with generous rewards – a classic instance of conflict of interests. Such
manipulations have lately become increasingly frequent and severe in Bangladesh. In this back drop, the
concerned board of directors is looking for improved surveillance techniques to better prevent and/or, detect
and investigate possible financial frauds. In their quest for proactive approach against manipulation of financial
statements, the board of directors look for warning signs and the present empirical study provides a profile of
a company that is likely to manipulate its financial statements. In this study, data from 2016-2017 financial
reports were utilized that correspond to 105 companies, excluding banks, non-banking financial institutions,
insurance companies, and mutual funds listed at the Dhaka Stock Exchange (DSE) and the likelihood of
accounting manipulation was quantified applying Beneish M-score model. It revealed that the maximum
M-score was 7.06 and the minimum was -8.98, where higher scores indicate increased likelihood of accounting
manipulation. Using a cut-off point of -1.78, twenty five companies were found to be suspected of accounting
manipulation, while using a cut-off point of -2.22, fifty seven companies were found to be likely manipulator.
Later a logistic model was developed to relate the likelihood of accounting manipulation to several company
specific variables that were not explicitly considered in the Beneish M-score model. The findings are likely to
benefit analysis of profiles of companies prone to accounting manipulation and thus could contribute to better
corporate governance practices in emerging economies such as Bangladesh.