<![CDATA[Background: The Indonesian Rural Bank (BPR) sector faces increasing liquidation risk, with 81 BPRs/BPRS having been liquidated between 2016 and 2025, driven by stricter capital regulations, including Peraturan Otoritas Jasa Keuangan (POJK) No. 5/POJK.03/2015. However, most existing prediction models remain static and fail to capture the dynamic process of financial deterioration, creating a gap in the development of early warning systems. Objective: This study developed a liquidation prediction model for Indonesian Rural Banks (BPRs) using a multi-period logit approach covering the period from 2013 to 2024. Methods: A matched-pair design was applied, consisting of 67 liquidated banks and 134 healthy banks, resulting in 201 banks and producing 603 observations across three periods (t-1, t-2, and t-3). Results: The model demonstrated strong explanatory power, with a Nagelkerke R² value of 79.1%. Simultaneously, financial ratios significantly influenced liquidation risk. Partially, Equity to Total Assets (ETA), Cash Ratio (CR), Loan to Asset Ratio (LAR), and Net Interest Margin (NIM) reduced the probability of liquidation, whereas Loan to Deposit Ratio (LDR), Cost of Funds (COF), and Non-Performing Loan (NPL) increased it. NPL emerged as the strongest predictor of failure (Wald = 71.463), triggering a chain reaction that reduced liquidity, increased funding costs, compressed profit margins, and weakened capital adequacy. Conclusion: The model achieved an overall accuracy of 90.7%, with 95.5% sensitivity and 81.1% specificity, demonstrating strong performance as an early warning system for predicting BPR liquidation and supporting regulatory risk monitoring.]]>
