Business Failure Prediction Research Articles

Business Failure Prediction of Construction Contractors Using a LSTM RNN with Accounting, Construction Market, and Macroeconomic Variables

AbstractIn the construction industry, predicting business failure and providing early warnings are critical challenges in the prevention of business failure chain reactions. Most relevant studies h...

Decision tree based model of business failure prediction for Polish companies

Research background: The issue of predicting the financial situation of companies is a relatively young field of economic research. Its origin dates back to the 30's of the 20th century, but constant research in this area proves the currentness of this topic even today. The issue of predicting the financial situation of a company is up to date not only for the company itself, but also for all stakeholders.
 Purpose of the article: The main purpose of this study is to create new prediction models by using the method of decision trees, in achieving sufficient prediction power of the generated model with a large database of real data on Polish companies obtained from the Amadeus database.
 Methods: As a result of the development of artificial intelligence, new methods for predicting financial failure of the company have been introduced into financial prediction analysis. One of the most widely used data mining techniques in this field is the method of decision trees. In the paper, we applied the CART and CHAID approach to create a model of predicting the financial difficulties of Polish companies.
 Findings & Value added: For the creation of the prediction model, a total of 37 financial and economic indicators of Polish companies were used. The resulting decision trees based prediction models for Polish companies reach a prediction power of more than 98%. The success of the classification for non-prosperous companies is more than 83%. The created decision tree-based prediction models are useful mainly for predicting the financial difficulties of Polish companies, but can also be used for companies in another country.

Predicting Business Failure of Construction Contractors Using Long Short-Term Memory Recurrent Neural Network

AbstractPredicting business failure of construction contractors is critical for both contractors and other stakeholders such as project owners, surety underwriters, investors, and government entiti...

Small Business Success or Failure Prediction: A Comparative Study in Ghana and Israel

This paper tests the Lussier success versus failure prediction model in Ghana, an emerging country and Israel, a developed country. The logit results of Israel (N = 205; p = 0.000) and Ghana (N = 208; p = 0.000) support the model validity to predict the success or failure of a group of businesses. The accuracy rate of predicting a specific business as successful or failure was greater than 85 percent in both countries, with high R square. Testing the t-values of the individual variables in each country revealed some differences between the two countries; results and implications are presented.

Business failure prediction using cart-based model: A case of Slovak companies

Business failure prediction using cart-based model: A case of Slovak companies

Business Failure Prediction: A Tri-dimensional Approach

Investigations of corporate failure prediction research usually implement binary classification into one of the distinguished groups – Distress or non-Distress companies. This study looks at a tri-dimensional approach which cluster firms into three (3) distinct dimensions namely - non-distress, semi-distressed and distressed. The study used secondary data from 2011 to 2015 obtained from the Ghana Stock Exchange (GSE) spanning across six industries, namely, Banking & Finance, Distribution, Food & Beverage, Insurance, Manufacturing and Mining & Oil. The study initially adopted the Altman (1968) Z score bankruptcy model to classify companies into non-distress, semi-distressed and distressed. Further analysis was conducted using the Hierarchical agglomerative cluster analysis to cluster companies into non-distress, semi-distressed and distressed. A comparison was then made between the Hierarchical agglomerative clustering against the Altman (1968) Z score bankruptcy classification to obtain higher classification. The outcome of the analysis revealed that the Hierarchical agglomerative cluster analysis and the Altman (1968) Z score bankruptcy model can both be used to classify companies into nondistress, semi-distressed and distressed based on the tri-dimensional approach instead of the binary classification (distressed and non-distressed). The study recommends that future research can explore other clustering methods for bankruptcy prediction to achieve higher and better classification.

Hybridizing kernel‐based fuzzy c‐means with hierarchical selective neural network ensemble model for business failure prediction

AbstractMore and more ensemble models are used to forecast business failure. It is generally known that the performance of an ensemble relies heavily on the diversity between each base classifier. To achieve diversity, this study uses kernel‐based fuzzy c‐means (KFCM) to organize firm samples and designs a hierarchical selective ensemble model for business failure prediction (BFP). First, three KFCM methods—Gaussian KFCM (GFCM), polynomial KFCM (PFCM), and Hyper‐tangent KFCM (HFCM)—are employed to partition the financial data set into three data sets. A neural network (NN) is then adopted as a basis classifier for BFP, and three sets, which are derived from three KFCM methods, are used to build three classifier pools. Next, classifiers are fused by the two‐layer hierarchical selective ensemble method. In the first layer, classifiers are ranked based on their prediction accuracy. The stepwise forward selection method is employed to selectively integrate classifiers according to their accuracy. In the second layer, three selective ensembles in the first layer are integrated again to acquire the final verdict. This study employs financial data from Chinese listed companies to conduct empirical research, and makes a comparative analysis with other ensemble models and all its component models. It is the conclusion that the two‐layer hierarchical selective ensemble is good at forecasting business failure.

A systematic overview of the prediction of business failure

A systematic overview of the prediction of business failure

Predikcija poslovnog neuspeha preduzeća iz oblasti duvanske industrije Republike Srbije

Predikcija poslovnog neuspeha preduzeća iz oblasti duvanske industrije Republike Srbije

A systematic overview of the prediction of business failure

This paper proposes a systematic analysis and revision of the literature to predict business failure. The systematic literature review (SLR) is done using Scopus database that is one of the biggest scientific databases. The study analyses the state-of-the-art of the subject between 1968 and 2017. At first are defining the keywords relevant to the subject, and are synthesising the publications across years identifying a growing trend in the literature. Also are detailing the most relevant journals and cited papers. This kind of study is valid to evaluate the scientific production across the years, to find the most important contributions and to have a global vision of the subject. Furthermore this information is useful to researchers to begin the study of a subject and to know the literature tendency.

Prediction of Business Failure with Fuzzy Models

This paper extends the theory of fuzzy diseases predictions in order to detect the causes of business failure. This extension is justified through the advantages of the reference model and its originality. Moreover, the fuzzy model is completed by this proposal and some parts of it have been published in isolated articles. For this purpose, the fuzzy theory is combined with the OWA operators to identify the factors that generate problems in firms. Also, a goodness index to validate its functionality and prediction capacity is introduced. The model estimates a matrix of economic- financial knowledge based on matrices of causes and symptoms. Knowing the symptoms makes it possible to estimate the causes, and managing them properly, allows monitoring and improving the company’s financial situation and forecasting its future. Also with this extension, the model can be useful to develop suitable computer systems for monitoring companies’ problems, warning of failures and facilitating decision-making.

What Do Post-Communist Countries Have in Common When Predicting Financial Distress?

Business failure prediction is an important issue in corporate finance. Different prediction models are proposed by financial theory and are often used in practice. Their application is effortless, selecting only few key inputs with the greatest informative power from the large list of possible indicators. Our paper identifies the financial distress predictors for 5 post-communist countries (Bulgaria, Croatia, the Czech Republic, Hungary and Romania) based on information collected from the Amadeus database for the period 2011-2013 using CHAID decision trees and neural networks. We propose a short list of indicators, which can offer a synthetic perspective on corporate distress risk, adapted for these countries. The best prediction models are substantially different from country to country: in the Czech Republic, Hungary and Romania the flow-approach indicators perform better, while in Bulgaria and Croatia - the stock-approach indicators. The results suggest that the extrapolation of such models from one country to another should be made cautiously. One interesting finding is the presence of the ratios per employee as predictors of financial distress.

The use of accounting anomalies indicators to predict business failure

Most of the studies that try to predict business failure assume that accounts give a true and fair view of the financial position of a company, without considering that managers can discretionarily apply accounting rules or even perform accounting fraud. This paper takes a set of financial ratios especially designed to detect accounting anomalies as bankruptcy predictors. These ratios are not very common in bankruptcy prediction studies, but they come from creative accounting studies. The ratios try to identify abnormal depreciation figures, exaggerated receivables or deteriorating financial conditions preceding aggressive accounting practices. The empirical study has been performed from a sample of 51,337 public and private European companies, during the period 2012–2016. The analysis techniques applied were logistic regression and decision trees, allowing to obtain rules to predict the status of failed or non-failed. It is found that several indicators proposed in the literature as earnings management indicators present statistically significant differences between failed and non-failed firms, but they do not have enough predictive power to incorporate them into prediction models. However, an index developed to measure accounting anomalies exhibits high discriminatory power, similar to that of the classical financial ratios. The construction of the index and its application to private firm sample provide the main contribution of the paper, as the results suggest slightly better forecast accuracy only for the private firm sample. The inclusion of indicators to detect accounting anomalies should be considered when developing new models to predict bankruptcy, especially in private companies.

A territorial perspective of SME’s default prediction models

PurposeThe purpose of this paper is to test whether the qualitative variables regarding the territory and the firm–territory relationship can improve the accuracy rates of small business default prediction models.Design/methodology/approachThe authors apply a logistic regression to a sample of 141 small Italian enterprises located in the Marche region, and the authors build two different default prediction models: one using only financial ratios and one using jointly financial ratios and variables related to the relationship between firm and territory.FindingsIncluding variables regarding the relationships between firms and their territory, the accuracy rates of the default prediction model are significantly improved.Research limitations/implicationsThe qualitative variables data collected are affected by subjective judgments of respondents of the firms studied. In addition, neither other qualitative variables (such as those regarding competitive strategies, or managerial skills) are included nor those variables regarding the relationships between firms and financial institutions are included.Practical implicationsThe study suggests that financial institutions should include territory qualitative variables, and, above all, qualitative variables regarding the firm–territory relationship, when constructing business default prediction models. Including this type of variables, it could be able to reduce the tendency to place unnecessary restrictions on credit.Originality/valueThe field of business failure prediction modeling using variables regarding the relationship between firm–territory is a unexplored area as it count of a very few studies.

EVALUATING FINANCIAL HEALTH OF GUJARAT STATE FERTILIZERS THROUGH ‘Z’ SCORE MODEL

The evaluation of financial health is very useful for financial managers, investors and other users. In this study it is tried to know whether Z score model is able to evaluate financial health of Chambal Fertilizers and Chemicals ltd for past 10 years (2007-08 to 2016-17). Analysis of this paper shows that the model is useful to know the financial soundness of Chambal Fertilizers. In this paper overall results of Z score model indicating that the financial position of the corporation is on alert to exercise the caution. These result shows that Altman model can give good analysis for fertilizers sector in India. Hence it can be concluded that user of financial statements like financial managers, analysts, investors etc can predict business failure or financial soundness of companies through Altman Z score model.

Business failure, efficiency, and volatility: Evidence from the European insurance industry

We analyze the operations and performance of insurance companies that left the insurance market from 2006 to 2013. Using a large European sample (2060 insurers from 16 countries and 250 failure events), we find that technical efficiency negatively and business volatility positively correlate with the probability of failure. Moreover, we document that firm growth has a U-shaped, nonlinear impact on failure probability. The research is original, in that we demonstrate the persistency of technical efficiency and business volatility in terms of early surveillance, extending valuable response time and thus broadening the available measures to prevent failures. We also demonstrate our identified failure indicators' applicability across financial crises and different regulatory philosophies. We apply the state-of-the-art data envelopment analysis, rare event logistic regression, hazard model of time to failure, and supporting vector machine approaches to the business failure prediction.

Development of Business Failure Prediction Model Using SmartPLS on Listed Construction Companies on Bursa Malaysia

Development of Business Failure Prediction Model Using SmartPLS on Listed Construction Companies on Bursa Malaysia

A framework for big data analytics approach to failure prediction of construction firms

This study explored use of big data analytics (BDA) to analyse data of a large number of construction firms to develop a construction business failure prediction model (CB-FPM). Careful analysis of literature revealed financial ratios as the best form of variable for this problem. Because of MapReduce’s unsuitability for iteration problems involved in developing CB-FPMs, various BDA initiatives for iteration problems were identified. A BDA framework for developing CB-FPM was proposed. It was validated by using 150,000 datacells of 30,000 construction firms, artificial neural network, AmazonElastic ComputeCloud, Apache Spark and the R software. The BDA CB-FPM was developed in eight seconds while the same process without BDA was aborted after nine hours without success. This shows the issue of not wanting to use large dataset to develop CB-FPM due to tedious duration is resolvable by applying BDA technique. The BDA CB-FPM largely outperformed an ordinary CB-FPM developed with a dataset of 200 construction firms, proving that use of larger sample size with the aid of BDA, leads to better performing CB-FPMs. The high financial and social cost associated with misclassifications (i.e. model error) thus makes adoption of BDA CB-FPMs very important for, among others, financiers, clients and policy makers.

A Review of Business Failure Prediction Models Used on Financial Distressed Companies

A Review of Business Failure Prediction Models Used on Financial Distressed Companies

The best of two worlds: Balancing model strength and comprehensibility in business failure prediction using spline-rule ensembles

Numerous organizations and companies rely upon business failure prediction to assess and minimize the risk of initiating business relationships with partners, clients, debtors or suppliers. Advances in research on business failure prediction have been largely dominated by algorithmic development and comparisons led by a focus on improvements in model accuracy. In this context, ensemble learning has recently emerged as a class of particularly well-performing methods, albeit often at the expense of increased model complexity. However, in practice, model choice is rarely based on predictive performance alone. Models should be comprehensible and justifiable to assess their compliance with common sense and business logic, and guarantee their acceptance throughout the organization. A promising ensemble classification algorithm that has been shown to reconcile performance and comprehensibility are rule ensembles. In this study, an extension entitled spline-rule ensembles is introduced and validated in the domain of business failure prediction. Spline-rule ensemble complement rules and linear terms found in conventional rule ensembles with smooth functions with the aim of better accommodating nonlinear simple effects of individual features on business failure. Experiments on a large selection of 21 datasets of European companies in various sectors and countries (i) demonstrate superior predictive performance of spline-rule ensembles over a set of well-established yet powerful benchmark methods, (ii) show the superiority of spline-rule ensembles over conventional rule ensembles and thus demonstrate the value of the incorporation of smoothing splines, (iii) investigate the impact of alternative term regularization procedures and (iv) illustrate the comprehensibility of the resulting models through a case study. In particular, the ability of the technique to reveal the extent and the way in which predictors impact business failure, and if and how variables interact, are exemplified.