Subjects
53 Ciencias económicas
5303 Contabilidad económica
530301 Contabilidad financiera
Predicción de quiebras en el sector de la restauración en España
Abstract
En este trabajo se trata de predecir la quiebra de las empresas recogidas en el sector restaurantes y puestos de comida según la CNAE-2009, dentro del sector de la restauración en España. El objeto del análisis es predecir la quiebra con un año de antelación, con el propósito de servir de inspiración a empresas dedicadas a las fusiones y adquisiciones a implementar métodos novedosos para ayudar a tomar decisiones en torno a qué entidades adquirir. Para lograr este objetivo, se utilizan dos métodos de modelado: la regresión logística y las redes neuronales, implementadas en el entorno Spyder de Python. Los resultados muestran que los modelos pueden predecir la quiebra con un año de antelación, pero aún tienen margen de mejora. En cuanto a la información empleada para el análisis, el estudio se basa en información contable del último año disponible en la base de datos SABI, por lo que sería interesante considerar la incorporación de información no contable en los modelos para comprobar su impacto en la precisión predictiva.
The aim of this paper is to predict the bankruptcy of the companies included in the restaurant and food stalls sector according to CNAE-2009, within the catering sector in Spain. The aim of the analysis is to predict bankruptcy one year in advance, with the purpose of inspiring companies involved in mergers and acquisitions to implement novel methods to help them make decisions about which entities to acquire. To achieve this goal, two modelling methods are used: logistic regression and neural networks, implemented in Python's Spyder environment. The results show that the models can predict bankruptcy one year in advance, but there is still room for improvement. As for the information used for the analysis, the study is based on accounting information for the last year available in the SABI database, so it would be interesting to consider incorporating non-accounting information in the models to check its impact on predictive accuracy.
The aim of this paper is to predict the bankruptcy of the companies included in the restaurant and food stalls sector according to CNAE-2009, within the catering sector in Spain. The aim of the analysis is to predict bankruptcy one year in advance, with the purpose of inspiring companies involved in mergers and acquisitions to implement novel methods to help them make decisions about which entities to acquire. To achieve this goal, two modelling methods are used: logistic regression and neural networks, implemented in Python's Spyder environment. The results show that the models can predict bankruptcy one year in advance, but there is still room for improvement. As for the information used for the analysis, the study is based on accounting information for the last year available in the SABI database, so it would be interesting to consider incorporating non-accounting information in the models to check its impact on predictive accuracy.
Información detallada
Materias, derechos, colecciones e identificadores
Keywords
KBA, Predicción de quiebras, Regresión logística, Redes neuronales, Información contable, Ratios financieros, Bankruptcy prediction, Logistic regression, Neural networks, Accounting information, Financial ratios
URI / Handle
http://hdl.handle.net/11531/69548Rights
Attribution-NonCommercial-NoDerivs 3.0 United States