Exploration of random polynomial regression efficiency in internal audit systems: the case of greek enterprises

Abstract

Corporate governance is strongly affected by the implementation of certain standards and followed procedures in modern firms, rendering firm auditing as a crucial parameter for their sustainability. The corporate scandals depict the vigorous importance of transparent and solid corporate governance in todays, rapidly emerging financial environment. At the same time, the lack of operational rules and monitoring mechanisms allows for expanding further the already existing gap between the designed and the actually followed operational strategy in all operational sectors of a company. As a consequence, increased fraudulent incidents accompanied by highly-uncertain business risks may therefore hinder business’ robustness. In this business environment, Internal Audit is a key factor in aiding the administration to make right decisions. Decision making for internal auditing, in such a complex inter-affected ecosystem though may suggest for processing a great amount of data which have to be carefully selected and cleansed out of huge data- lakes and warehouses. In order to cope with a cumbersome human task, well established regression techniques are being utilized over a certainly defined problem on internal auditing and control application. A parameterized MATLAB script instance designed so as to enable multiple regression tests was developed for such a purpose. A linear in the parameters, polynomial regression model was utilized in order to test and estimate several different polynomial models of abstract-order using a common least-squares fitting over a surveyed dataset among chief executive auditors from Greek Stock Exchange (listed at the last quarter of 2018) companies. The analysis over the surveyed results highlight the strong interdependence between internal control quality, and internal auditing function execution.