Digital Bankruptcy Early Warning System Berbasis XBRL Pada Perusahaan Textile Bursa Efek Indonesia

Abstract

Abstract: The Indonesian textile and textile products (TPT) industry is facing acute structural pressure, with over 126,000 workers laid off between 2023 and 2025 due to the closure of 59 companies. This crisis underscores the critical need for early warning systems capable of detecting financial distress before it reaches irreversible stages. Previous studies on financial distress in Indonesian textile companies have relied on manual extraction of financial data from PDF reports—an approach that is time-consuming, error-prone, and inherently not scalable for real-time monitoring. This study addresses that gap by integrating eXtensible Business Reporting Language (XBRL) instance documents, mandated by the Indonesia Stock Exchange (IDX) for all listed companies, with the Altman Z-Score model within a single automated workflow. The research employs an applied quantitative approach. Using purposive sampling, eight textile companies listed on the IDX with complete XBRL data for 2024–2025 were selected as the sample. A web application system was developed using JavaScript and React JS via CDN, without build tools such as Webpack or Node.js, to extract financial data automatically from XBRL ZIP files and compute the Altman Z-Score. System evaluation results show 100% extraction accuracy across all standard idx-dei and idx-cor taxonomy tags, with 16 XBRL files processed in 11.2 seconds and Z-Score calculations completed with 100% accuracy after data conversion to JSON format. Financial health analysis reveals three distinct groupings: ARGO, TFCO, and CNTX are classified in the safe zone (Z > 2.99); BELL and SSTM are in the grey zone (1.81 ≤ Z ≤ 2.99); while INOV, MYTX, INDR, and ESTI fall into the distress zone (Z < 1.81). A notable anomaly was identified in POLY, whose extremely high Z-Score (21.30–24.43) is distorted by an abnormal retained earnings-to-total assets ratio rather than reflecting genuine financial health. This study contributes methodologically by demonstrating a scalable, open-data approach to financial distress analysis, and provides empirical evidence that XBRL infrastructure in Indonesia is sufficiently mature for automated systematic financial surveillance.