Page 23 - Azerbaijan State University of Economics

THE JOURNAL OF ECONOMIC SCIENCES: THEORY AND PRACTICE, V.82, # 1, 2025, pp. 19-35

Some researchers have sought to evaluate the effectiveness of the Beneish and Roxas
models by comparing their predictive results with other indicators that determine the
presence of fraudulent financial reporting. Thus, Svabova et al. (2020) compared the
outcomes of the Beneish model with real data from 1,900 Slovak companies identified
as manipulative during the period 2009-2018.
Their analysis revealed a match of 32.7% for fraudulent companies and 38.4% for
companies that did not commit fraud. Shakouri et al. (2021) examined the indicators of 161
companies listed on the Tehran Stock Exchange over the same period and found that the
Beneish model was able to distinguish between fraudulent and non-fraudulent companies
with 73% accuracy. Similarly, Golec (2019) in a study of 24 Polish firms listed on the
Warsaw Stock Exchange reported that the Beneish M-Score achieved an accuracy rate
ranging from 71% to 75%. Using the Beneish and Roxas model, Sylwestrzak (2022)
examined 63 non-financial Polish companies listed on the Warsaw Stock Exchange over
the period 2010-2021. Based on the analysis, the author developed and tested a hybrid
model that demonstrated the importance of incorporating both financial and non-financial
indicators, along with logistic regression techniques, to improve the accuracy of detecting
financial statement fraud. Thus, a review of the scholarly literature confirms that the
Beneish and Roxas models are sufficiently effective tools for identifying companies
engaged in financial statement falsification and manipulation. Their application is therefore
appropriate and valuable in enhancing analytical procedures within the audit process. The
purpose of this article is to identify signals of financial statement falsification among
Ukrainian marketing companies as part of analytical procedures in auditing, using of the
Beneish and Roxas models.

METHODS
The study is based on the use of Beneish (Beneish M-Score) (1999) and Roxas (Roxas
Score) (2011) models for atypical correlation between financial reporting indicators
and forecasting the risks of its falsification. Their general characteristics and
calculation procedure are given in Table 01.
The application of these models (Table 01) within the audit process enables
preliminary risk assessment, particularly by facilitating the prompt identification of
companies exhibiting potential signs of financial fraud. This is achieved through the
analysis of key financial indicators such as revenue trends, asset quality, and cash flow
performance.
Using the Beneish and Roxas models, an empirical analysis was conducted on 50
Ukrainian marketing companies for the period 2023–2024. The results allowed for the
identification of companies falling within the high-risk category - i.e., those with a
potential likelihood of financial statement manipulation, as well as those whose
financial reporting demonstrated a high level of reliability.

23