Detecting the Manipulation of Financial Information by Using Artificial Neural Network Models

Abstract

Despite their widespread usage, models of accrual based methods in detecting false financial statements have been subject to significant criticism. An alternative to the accruals approach is to use binary probit and logit models and some other multivariate statistical techniques where they combine accruals and some other financial ratios and/or indexes. The objective of this paper is to explain the historical evolution of the accrual based methods where they provide some evidence of earnings management practices and than extend to some other alternative methods in detecting manipulative practices in financial reporting. This paper also, introduces a new method that has been widely used in detecting financial distress companies. An Artificial Neural Network Model, which is based on the concept of using artificial neurons, to estimate the manipulative financial reporting practices of the companies listed in the Istanbul Stock Exchange (ISE). The results indicate that the proposed Artificial Neural Network Model outperforms the traditional statistical techniques used in earnings manipulation practices.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 11 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.