Search

The purpose of the paper is to propose two new procedures that deal with overfitting problem using neural techniques for variable selection and business failure prediction. The first procedure, called HVS-AUC, is based simultaneously on (i) the backward search, (ii) the HVS measure (Heuristic for Variable Selection), and (iii) the AUC criterion (Area Under Curve). The second procedure, called forward-AUC, is based on (i) the forward search and (ii) the AUC criterion. Using a sample of bankrupt and non-bankrupt firms in France, the implementation of the procedures using neural networks shows that the profitability, the repayment capacity, the taxation, and the importance of investment have a strong explanatory power in bankruptcy prediction. These procedures also provide more parsimonious and more efficient models compared to Linear Discriminant Analysis. [ABSTRACT FROM AUTHOR]

This paper illustrates the importance of referring to a dynamic approach when forecasting firms bankruptcies, paying a particular attention to French SMEs. Based on Shummay’s (J Bus 74:101–124, 2001), we build a duration model and extend it by incorporating unobservable heterogeneity. Moreover, we resort to a dynamic dichotomous specification in which “right side” censored data are taken into account. We emphasize the complexity of the calculations of integrals that must be implemented and show how to overcome this challenge by applying the Geweke, Hajivassiliou and Keane algorithm which involves the technique of the simulated maximum likelihood. The findings prove that our dynamic approach, which integrates macroeconomic variables and takes account of both random effects and exogenous shocks, provides credible results. Besides, our method provides the predictive content of macroeconomic variables and the unobservable heterogeneity, which is helpful in forecasting firms bankruptcies. [ABSTRACT FROM AUTHOR]

In order to mitigate the effects of the contraction in demand during economic crises, firms face the need to reduce the number of facilities in their networks. This reduction must be conducted taking into consideration both the possible actions of their rival firms and the reaction of the affected customers, so that the loss of market share is minimised. In this article, we analyse the facility closing problem of two firms operating in a duopolistic market. The problem is modelled as a non-cooperative game over a binary integer programming formulation of the firms' delocation problem. The possible outcome of the game is analysed for three different competitive scenarios: Myopic behaviour, Cournot conjectures, and Stackelberg strategies. These scenarios are analysed under the assumption that customers show certain level of loyalty to the firm that they initially resorted to. This assumption entitles us to establish the existence of Nash equilibria in the delocation game by means of the introduction of a social planner. Moreover, this social planner provides a mechanism for computing the Nash equilibrium in the Cournot delocation game. Additionally, we develop an algorithmic approach that provides a simple mechanism for finding equilibria under Stackelberg strategies. Experimental results indicate that for small reductions in the network size the solution under myopic behaviour is often an equilibrium. For large reductions, evidence has been found that there is a first mover advantage in the Stackelberg delocation game. [ABSTRACT FROM AUTHOR]