Forecasting Corporate Business Failure with an Innovative Model: An Application on UK Construction Companies

Abstract

This research focuses on developing a bankruptcy prediction model in the UK construction industry. It fills the gap of previous researches which did not focus on the construction industry and in the UK. The aim of this research is to establish accurate model which successfully classify firms into their respective financial status of failed and non-failed. This research analyses financial variable extracted from FAME database by Bureau Van Dijk for the development of bankruptcy prediction model. Ranging from SIC trade codes of 41100 (Development of building projects), 41 (Construction of buildings), 42(Civil Engineering), 43 (Specialised construction activities), 68 (Real estate activities), and 71(Architectural and engineering activities and related technical consultancy). Then the model is tested against the collected sample and compared with two other Z-score model. These models have an average accuracy of 80%. The model with the strongest predictive power (SIC 42) correctly predicts 93.2% (non-bankrupt) and 90.9% (bankrupt) while the weakest model (SIC 68) classify 51.2% (non-bankrupt) and 70% (bankrupt) correctly into their respective group. The developed model is far superior in classifying firms into their financial status when compared to the other two prediction models. Furthermore, the established model is robust enough that it has an average classification accuracy over the years of 70%. A practical case study is conducted utilising the developed Z-score model and ratio analysis. Macroeconomic variable is then analysed where long term interest rate, inflation, 3 months’ treasury bills has a negative impact on the firms Z-score. However, construction output and gilt repo interest rate has a positive relationship to the Z-score of companies. The estimate of covariance is also seen to build up from the year 2006 which acts as a warning sign that a financial crisis is emerging. This result point towards the fact that both macroeconomics and microeconomics contributes significantly to business failure. The developed bankruptcy prediction model serves as a warning sign for management to pay attention to. As firms enter the “grey zone”, management should act to save firms from defaulting. Ultimately, this model is applicable widely for other industry of world economies.