Premium
Audit Report in Payment Default Prediction: A Contingency Approach
Author(s) -
Laitinen Erkki K.,
Laitinen Teija
Publication year - 2009
Publication title -
international journal of auditing
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.583
H-Index - 21
eISSN - 1099-1123
pISSN - 1090-6738
DOI - 10.1111/j.1099-1123.2009.00396.x
Subject(s) - audit , business , payment , actuarial science , logistic regression , accounting , econometrics , variables , walk through test , computer science , finance , external auditor , economics , internal audit , machine learning
This study analyzes the contribution of audit reports on default prediction. It is proposed that modification of an audit report contributes additional information on financial variables and moderates information given by these variables in payment default predictions. However, the effect of audit report modification on non‐financial variable information is assumed to depend on the set of selected predictors. The effect of audit information is evaluated in an estimation sample of 1,062 default and 1,162 non‐default Finnish firms. Test data from 420 default and 48,166 non‐default firms are used to validate the classification results. The default prediction models are estimated by logistic regression analysis. Empirical evidence supports expectations on the additive and moderating effects of audit report modification on financial information. It also has an effect on non‐financial and combined information. Audit information only marginally improves classification accuracy of the financial information model. However, it significantly improves that of a non‐financial information model.