Original scientific paper
PRIMJENA CA SCORE MODELA NA PREDUZEĆIMA BOSNE I HERCEGOVINE
Selma Vidimlić
; Assistant Proffesor, University „VITEZ“
*
Saudin Terzić
; Indirect Taxation Authority BiH
* Corresponding author.
Abstract
This study examines the predictive performance of the CA Score model on a sample of 392 small and medium-sized enterprises (SMEs) operating in Bosnia and Herzegovina. Based on financial statements for 2021 obtained from the Financial Information Agency (FIA), firms were classified into two groups: those that maintained business continuity during 2022–2023 and those that experienced financial distress, defined as transaction account blockage exceeding 30 days.
The results indicate that the CA Score model achieves an overall classification accuracy of 65.82%, suggesting moderate predictive capability. The model performs slightly better in identifying financially stable firms than distressed ones, with limited sensitivity in detecting high-risk enterprises.
The findings highlight that the model’s reduced effectiveness is largely driven by structural characteristics of transition economies, including weak payment discipline, distortions in financial reporting, and inconsistent enforcement of insolvency regulations. These factors significantly limit the reliability of financial ratio-based models developed in advanced market economies.
Keywords
financial indicators; failure prediction; financial statements; CA Score; SMEs; insolvency.
Hrčak ID:
350738
URI
Publication date:
8.9.2026.
Visits: 0 *