Examining small bank failures in the United States: an application of the random effects parametric survival model

Abstract

Purpose The purpose of the current study is to identify variables that, when integrated into the random effects parametric survival model, could be used to forecast the failure rate of small banks in the USA. A bank’s income production, efficiency and costs were taken into consideration when choosing the internal components. The breakout of the financial crisis, bank regulations that affect how the banking sector operates and the federal funds rate are the primary external variables. Design/methodology/approach This study uses the random effects parametric survival model to investigate the causes of small bank failures in the USA from 1996 to 2019. The study identifies several characteristics that failed banks frequently display. The main indications that may help to identify the elevated risk of small bank failures include the ROA, the cost of funds, the ratio of noninterest income to assets, the ratio of loan and lease losses to assets, noninterest expenses and core capital (leverage) ratio to assets. Economic disruptions, financial market distress and industry-based regulatory redress by the government exacerbate the financial distress borne by small banks. Findings The study revealed that a failed bank typically demonstrates a certain number of characteristics. The key factors that might assist identify which bank would be most likely to collapse include the cost of funding earning assets, the yield on earning assets, core Capital (leverage) ratio to assets, loan and lease loss provision to assets, noninterest expense and noninterest income to assets. Additionally, when a financial crisis occurs or the government changes regulations that could raise the cost of compliance for small banks, the likelihood that a bank will fail increases. Originality/value Models based on survival theories are more suitable when the authors examine bank failure as a unique event that happens gradually. The authors use a random effects parametric survival model to investigate the internal and external factors that may influence prospective small bank failure. This model has been developed and used in the medicinal research field. The authors do not choose the Cox proportional hazards model because it does not work well with panel data.

Examining small bank failures in the United States: an analysis using (coarsened exact matching) CEM

“…Survival models, such as the Cox proportional hazards model used by Cox et al (2017), have been used to predict bank failures. Another approach used by Foley et al (2023) is the Random Effects Parametric Survival Model, which involves selecting a specific number of years before the bank failure event. It is worth noting that these techniques cannot effectively analyze panel data due to their inherent limitations and constraints.…”