Abstract:For decades, the prediction of bank failure has been a popular topic in credit risk and banking studies. Statistical and machine learning methods have been working well in predicting the probability of bankruptcy for different time horizons prior to the failure. In recent years, bank efficiency has attracted much interest from academic circles, where low productivity or efficiency in banks has been regarded as a potential reason for failure. It is generally believed that low efficiency implies low-quality management of the organisation, which may lead to bad performance in the competitive financial markets. Previous papers linking efficiency measures calculated by Data Envelopment Analysis (DEA) to bank failure prediction have been limited to cross sectional analyses. A dynamic analysis with the updated samples is therefore recommended for bankruptcy prediction. This paper proposes a nonparametric method, Malmquist DEA with Worst Practice Frontier, to dynamically assess the bankruptcy risk of banks over multiple periods. A total sample of 4426 US banks over a period of 15 years (2002-2016), covering the subprime financial crisis, is used to empirically test the model. A static model is used as the benchmark, and we introduce more extensions for comparisons of predictive performance. Results of the comparisons and robustness tests show that Malmquist DEA is a useful tool not only for estimating productivity growth but also to give early warnings of the potential collapse of banks. The extended DEA models with various reference sets and orientations also show strong predictive power.

An inverse semi-oriented radial data envelopment analysis measure for dealing with negative data

A directional distance based super-efficiency DEA model handling negative data

Constructing a robust data envelopment analysis model to evaluate the profitability efficiency of banks with financial loss

An inverse data envelopment analysis model to consider ratio data and preferences of decision-makers

A Revised Inverse Data Envelopment Analysis Model Based on Radial Models

An inverse optimization model for imprecise data envelopment analysis

Additive integer-valued data envelopment analysis with missing data: A multi-criteria evaluation approach

A general inverse DEA model for non-radial DEA

Alternative Secondary Goals in Multiplicative Two-Stage Data Envelopment Analysis

Assessing sustainability indicators using inverse integer-valued data envelopment analysis with undesirable outputs

An Interval Efficiency Measurement in DEA When considering Undesirable Outputs

Inverse data envelopment analysis model to preserve relative efficiency values: The case of variable returns to scale

Supplier selection: comparison of DEA models with additive and reciprocal data

DATA ENVELOPMENT ANALYSIS WITH MISSING DATA: A MULTIPLE LINEAR REGRESSION ANALYSIS APPROACH

Uncertain Data Envelopment Analysis with Imprecisely Observed Inputs and Outputs

A new approach for fair efficiency decomposition in two-stage structure system

A novel inverse DEA-R model with application in hospital efficiency

Restructuring units in the simultaneous presence of desired and undesired factors

Performance evaluation of Chinese commercial banks by an improved slacks-based DEA model

Bank efficiency and failure prediction: a nonparametric and dynamic model based on data envelopment analysis

A new inverse DEA cost efficiency model for estimating potential merger gains: a case of Canadian banks