Predicting Corporate Failure Using a Neural Network Approach

This paper investigates the performance of Artificial Neural Networks for the classification and subsequent prediction of business entities into failed and non-failed classes. Two techniques, back-propagation and Optimal Estimation Theory (OET), are used to train the neural networks to predict bankruptcy filings. The data are drawn from Compustat data tapes representing a cross-section of industries. The results obtained with the neural networks are compared with other well-known bankruptcy prediction techniques such as discriminant analysis, probit and logit, as well as against benchmarks provided by directly applying the bankruptcy prediction models developed by Altman (1968) and Ohlson (1980) to our data set. We control the degree of ‘disproportionate sampling’ by creating ‘training’ and ‘testing’ populations with proportions of bankrupt firms ranging from 1% to 50%. For each population, we apply each technique 50 times to determine stable accuracy rates in terms of Type I, Type II and Total Error. We show that the performance of various classification techniques, in terms of their classification errors, depends on the proportions of bankrupt firms in the training and testing data sets, the variables used in the models, and assumptions about the relative costs of Type I and Type II errors. The neural network solutions do not achieve the ‘magical’ results that literature in this field often promises, although there are notable 'pockets' of superior performance by the neural networks, depending on particular combinations of proportions of bankrupt firms in training and testing data sets and assumptions about the relative costs of Type I and Type II errors. However, since we tested only one architecture for the neural network, it will be necessary to investigate potential improvements in neural network performance through systematic changes in neural network architecture.     

Public Choice and Urban Water Quality*

Abstract . It is the purpose of this paper to examine critically the choice of “zero discharge” as me ultimate goal of our federal water pollution policy and to demonstrate that a goal of 90 percent removal of pollutants from effluents of urban areas will be more economically efficient. Our results indicate that in reaching “Zero discharge,” there is a minimal return in water quality for urban areas which are located on large bodies of water. The final conclusion is that the substantial resources, which would be required to surpass the level of 90 percent removal of pollutants, could be employed more effectively in other social programs.     

Comparing variations in labor costs for two vs. one full-time nurse manager

Labor costs for two versus one full-time nurse manager were compared. Results revealed reductions in nursing costs per patient day. These results suggest that innovative models may hold promise as organizations struggle to retain nurses and decrease costs.     

Protecting the privacy of personal information in employee benefit plans

The United States is not the only country dealing with privacy of health information. In the absence of clear legislative direction, this paper focuses on the common privacy principles that underlie Canadian privacy legislation and recommends a plan of action. The objective is to give plan administrators and trustees sufficient transition time to develop and implement the appropriate privacy strategies, thereby minimizing both the potential burden of compliance and the potential risks of noncompliance.     

Saving in Developing Countries: An Overview

This article reviews the current state of knowledge on the determinantsof saving rates, presenting the main findings and contributionsof the recently completed World Bank research project, ‘SavingAcross the World.’ The article discusses the basic designof the research project and its core database, the World SavingDatabase. It then summarizes the main project results and placesthem in the context of the literature on saving, identifyingthe key policy and nonpolicy determinants of private savingrates. Special attention is paid to the relationship betweengrowth and saving and the impact of specific policies on savingrates. The article concludes by introducing the studies includedin this special issue.     

Trade and Tradability: Exports, Imports, and Factor Markets in the Salter-Swan Model

We extend the Salter-Swan model to include both factor markets and semi-traded goods. In our model, changes in relative factor prices depend on changes in world commodity prices, factor endowments, and the trade balance. In contrast, only changes in world commodity prices can affect factor prices in the neoclassical trade model. The inclusion of semi-traded goods weakens the magnification effect in both the Stolper-Samuelson and Rybczynski theorems. When imports and domestic goods are poor substitutes, a characteristic of some commodities in developing countries, the sign of the Stolper-Samuelson theorem is reversed.     

Testing the endogenous growth model: public expenditure, taxation, and growth over the long run

Endogenous growth models, such as Barro (1990), predict that government expenditure and taxation will have both temporary and permanent effects on growth. We test this prediction using panels of annual and period‐averaged data for OECD countries during 1970–95, isolating long‐run from short‐run fiscal effects. Our results strongly support the endogenous growth model and suggest that long‐run fiscal effects are not fully captured by period averaging and static panel methods. Unlike previous investigations, our estimates are free from biases associated with incomplete specification of the government budget constraint and do not appear to result from endogeneity of fiscal or investment variables. JEL Classification: H30, O40 Validation du modèle de croissance endogène: dépenses publiques, fiscalité et croissance à long terme. Des modèles de croissance endogène comme celui de Barro (1990) prédisent que dépenses gouvernementales et fiscalité vont avoir des effets temporaires et permanents sur la croissance. On met cette prévision au test à l'aide de données annuelles et pour certaines moyennes couvrant des sous‐périodes pour les pays de l'OCDE (1970–95) dans le but de départager les effets à court et à long terme. Les résultats valident fortement le modèle de croissance endogène et suggèrent que les effets fiscaux à long terme ne sont pas pleinement capturés par des méthodes utilisant des moyennes ou des méthodes statiques. Contrairement aux résultats d'enquêtes antérieures, les résultats proposés ne souffrent pas de distorsions attribuables à une spécification incomplète de la contrainte budgétaire du gouvernement, et ne semblent pas être l'effet d'écho de l'endogénéité des variables fiscales et de l'investissement.