Off‐site monitoring systems for predicting bank underperformance: a comparison of neural networks, discriminant analysis, and professional human judgment

Abstract

This study compares the ability of discriminant analysis, neural networks, and professional human judgment methodologies in predicting commercial bank underperformance. Experience from the banking crisis of the 1980s and early 1990s suggest that improved prediction models are needed for helping prevent bank failures and promoting economic stability. Our research seeks to address this issue by exploring new prediction model techniques and comparing them to existing approaches. When comparing the predictive ability of all three models, the neural network model shows slightly better predictive ability than that of the regulators. Both the neural network model and regulators significantly outperform the benchmark discriminant analysis model's accuracy. These findings suggest that neural networks show promise as an off‐site surveillance methodology. Factoring in the relative costs of the different types of misclassifications from each model also indicates that neural network models are better predictors, particularly when weighting Type I errors more heavily. Further research with neural networks in this field should yield workable models that greatly enhance the ability of regulators and bankers to identify and address weaknesses in banks before they approach failure. Copyright © 2001 John Wiley & Sons, Ltd.

Citizens as consumers: Profiling e-government services’ users in Egypt via data mining techniques

“…However, this limitation "should not be a serious drawback if one simply desires classification from the model" (Swicegood and Clark, 2001, p. 176).…”

Bank failure prediction models: for the developing and developed countries

“…Various models can predict bankruptcy prediction problem. Classical statistical techniques influenced the formation of these models, such as intelligent techniques (Kumar and Ravi 2007), discriminant analysis (Swicegood and Clark 2001), BPNN (Swicegood and Clark 2001;Bell 1997), logistic regression (Bell 1997;Kolari et al 2002), feed forward neural networks (Olmeda and Fernandez 1997), logistic regression (Olmeda and Fernandez 1997;Canbas et al 2005), multivariate adaptive splines (Olmeda and Fernandez 1997), Cox proportional hazards mode (Lane et al 1986), Principal Method (Canbas et al 2005), Linear Program (Cielen et al 2004), SOM network (Alam et al 2000), Genetic algorithm (Min et al 2006), multivariate discriminant analysis (Cielen et al 2004), neural network techniques (Boyacioglu et al 2009), support vector machines (Boyacioglu et al 2009). …”

“…Previous studies only analyzed early warning systems in a single country (Kolari et al 2002;Olmeda and Fernandez 1997;Ravi and Pramodh 2008;Canbas et al 2005;Chauhan et al 2009;Cielen et al 2004;Alam et al 2000;Swicegood and Clark 2001). The first part of this study, however, analyzes financial early warning systems in five regional groups because of the global nature of financial distress.…”

Artificial chemical reaction optimization of neural networks for efficient prediction of stock market indices

“…Calderon and Cheh [26] argued that the standard MLP network is subject to problems of local minima. Again there is no formal guideline how to develop a network for the MLP technique [27]. It is suggested that to overcome the local minima problem, more nodes may be added to the hidden layers.…”