Predicting a Going-Concern Auditor’s Opinion: ANN Approach

Hassan, Manar; Wahdan, Mohamed Aly; Elsayed, Mohamed Saber; Mostafa, Mahmoud Abdel-Wahab

doi:10.21608/sjcs.2024.304077.1073

Predicting a Going-Concern Auditor’s Opinion: ANN Approach

نوع المستند : المقالة الأصلية

المؤلفون

¹ Assistant Lecturer in Accounting Dept. Menoufia University

² Professor of Auditing, Dean of the Faculty of Commerce, Menoufia University

³ Assistant Professor of Accounting, Faculty of Commerce, Menoufia University

⁴ Emeritus Professor in Accounting Dept, Menoufia

10.21608/sjcs.2024.304077.1073

المستخلص

The objective of this research is to Employ neural network techniques that can screen
out the most important variables when predicting GCOs. These factors include financial and non-
financial variables related to both auditor and auditee.
Design/methodology: The study was conducted on a sample consisting of 61 firms listed in the
Egyptian Stock Exchange (ESE) belonging to seven sectors during the period from 2018-
2021 with a total of (244) observations. The study adopts two stages to construct going concern
prediction models. In the first stage, ANN is used to screen out the most important variables. A
total of 9 variables are selected based on their importance value (importance value ≥ 0.05),
including CATA, Predictive ability of earnings, Return on investment, Current liabilities/ total
assets, Audit lag, Profit ratio, Sales revenue growth rate, and Managerial ownership. In the second
stage, the proposed model is constructed for predicting going concern uncertainties.
The results reject the study hypothesis and prove that ANN can be used in predicting
GCO with high accuracy. The research depended on contingency table to testing the accuracy of
ANN model through comparing the predicted results of 61 observations with the actual values.
The significant value is 0.000 which is less than (.05). Moreover, behind that, the Wilcoxon test is used to ensure
that there is no difference between the actual and predicted opinion by ANN. Result indicate that
the p-value equal (0.157) which is more than 0.05, which point out that the two groups aren’t
different.

الكلمات الرئيسية