Developing a Bankruptcy Prediction Model for Sustainable Operation of General Contractor in Korea

Transcription

1 Developing a Bankruptcy Prediction Model for Sustainable Operation of General Contractor in Korea SeungKyu Yoo 1, a, JungRo Park 1, b,sungkon Moon 1, c, JaeJun Kim 2, d 1 Dept. of Sustainable Architectural Engineering, Hanyang University, Seoul, Korea, Dept. of Architectural Engineering, Hanyang University, Seoul, Korea, a james_yoo@hotmail.com, b intercessory@naver.com, b zidanehy@hotmail.com, d jjkim@hanyang.ac.kr ABSTRACT The construction industry largely depends on the demand from clients and sensitively reacts to the environmental change such as change of the governmental policy, economic depression and reduction of consumer confidence, etc. This causes a result that affects profitability directly or indirectly, and as these phenomena are directly exposed to the financial statement showing the results of a company, the company goes through the procedure of deterioration of the financial index, insolvency, liquidation, workout and bankruptcy. Recently, the construction industry goes through the procedure of continuous depression of the housing market, lowering of profitability due to the international financial crisis, and deterioration of liquidity, so 130 general contractors was finally bankrupted for one year of The bankruptcy of general contractors generates chain bankruptcies of subcontractors and related companies and the large layoff situation, so it brings about social problems. Accordingly, the present research suggested a financial ratio that can previously grasp the dishonour of a company by utilizing the financial ratio of bankrupt companies and non-bankrupt companies, and finally developed a Bankruptcy prediction model with the target of the externally audited construction companies that have relatively more vulnerable financial soundness than the listed companies. The present research expects to achieve the improvement on a management index and sustainable operation of general contractors by managing prediction variables showing a bankruptcy signal continuously and taking a measure previously through self-diagnosis using the Bankruptcy prediction model. KEYWORDS: Bankruptcy Prediction Model, Multivariate Discriminant Analysis(MDA), Financial Ratio, Externally Audited Construction Company 1. INTRODUCTION The construction industry has the industrial structure of a Job-Order-Production type by a client, so it distinguishes from the manufacturing industry. Also, it is a reality that the construction industry has characteristics that execute many projects for a long period and excessively depend on the government policy and economic trend (Kangari, 1988). The demand of construction markets was largely shrunken due to the aftermath of the Global Financial Crisis in last 2008, so the unsold housing of 165,599 was recorded in December The increase of unsold housing is directly connected to the increase of accounts receivables and the rise of financing costs, it brings about a result that increases financial difficulty of the construction company and finally, makes sustainable operation difficult. This caused a result that makes 130 general contractors (Construction Association of Korea, 2009) and 273 specialty contractors (Subcontractor) go bankrupt for one year of The bankruptcy of the 405

2 medium-and large-sized general contractors triggers the chain-reaction bankruptcy and leads to a social problem. The financial statement is important information showing the financial status and management results and has been widely used a tool for evaluating a company. Accordingly, the purpose of the present research is to develop a Bankruptcy Prediction Model based on the financial statement and financial ratio of bankruptcy companies and non-bankruptcy companies, to discover its financial distress or default in an early stage and provide an opportunity that can by itself recover by forecasting a bankruptcy of insolvent construction companies in advance, and to minimize waste of expenses being consumed in liquidation, composition, workout and others through this General Contractor Specialty Contractor Increase and Decrease Rate of Unsold Hounsing (%) Figure 1. Yearly Bankruptcy Company of General Contractor and Specialty Contractor 2. BACKGROUND The corporate executives, investors and creditors often use the financial statement of a company as a decision-making tool, and further, also utilize the summarized information of its financial statement by using its financial ratio. However, they do not well know about what thing among numerous financial ratios effect a company directly or indirectly, and even if they try to analyze individually, it is a work to take a long time. The decision-making tools for corporate executives, investors and creditors are being mainly developed by Credit Rating Agency and Investment Strategy Agency. However, because its assessment procedure is complex and the expenses consumed is much, it becomes a burden to an executive of a small-sized company. Accordingly, they are continuing efforts to find out a method that can easily simplify and quantify its procedure. The Bankruptcy Prediction Model using the financial ratio was first suggested by a research of Beaver (1966). Afterward, the theoretical and empirical research on the corporate bankruptcy was widely executed in the financial management and accounting fields. Afterward, Altman (1968) developed Bankruptcy Prediction Model by improving the analysis method of Beaver and applying the Multivariate Discriminant Model (MDA). This model is often called as a Z-score model. 3. DEFINITIONS 3.1 Bankruptcy and Non-Bankruptcy Construction Company The bankruptcy is a legal term included in a category of failure and means a state that collides with insolvency due to non-sufficient fund in a checking account owned by a company or dishonor of a bill or a check or that checking transactions are stopped. It is duplicatedly being used as terms like Liquidation, Composition, Insolvency or Workout according to the legal procedure or the character of a research. The present research defined the companies received a declaration of insolvency from Korea Financial Telecommunications & Clearings Institute from January 1 to December 31, 2008 and the companies included in the bankruptcy company list announced for the same period by the 406

3 Construction Guarantee as bankruptcy companies, and defined the companies not included in the above-mentioned category as non-bankruptcy companies. 3.2 Range of the Sample The construction industry of Korea is largely divided into a general contractor group and a specialty contractor group. The present research aimed to develop a bankruptcy prediction model suitable for a general contractor group rather than handling the total construction industry. The reason is because the general contractor and specialty contractor have a big difference in the size aspect of the company and have different major business portfolios. Also, when the total construction industry is selected as a sample, the characterization of a model is easy to deteriorate. And the present research selected an externally audited company group that is comparatively easy to secure samples and shows high reliability in the financial statement. The reason is because as the number of bankruptcy companies among listed construction companies is very few, there is difficulty in sample collection and analysis, and because as the non-externally audited companies does not receive external audit, there is a problem in reliability of the financial statement. The estimation samples finally decided in the present research consist of 25 bankruptcy construction companies and 25 non-bankruptcy construction companies, and the data used in this research are taken from the Total Solution 2000(TS2000) of Korea Listed Companies Association. 4. Methodology 4.1 Selection Method of the Prediction Variable The present research used the T-test technique to extract predictor variables between two groups (bankruptcy and non-bankruptcy). The T-test technique is a statistical method for testing whether the average values of between two samples are the same in a statistical meaning, so it can be said that as the T-score value gets larger, the average values of the bankruptcy company group and the nonbankruptcy group are different from each other. The present research finally deduced 12 prediction variables by analyzing 150 financial ratios extracted from the respective company s financial statement for 3 years. Table 1. List of Prediction Variables Variable Name of Financial Ratio Formula of Financial Ratio Growth rate of stockholder`s equity Total Equity at the end of the period /Total Equity at the end of the Past Period Growth Rate Of Sales Total sales at the end of the period / Total sales at the end of the Past Period Net Income To Sales Current net Income / Sales 100 Stockholders' Equity To Total Equity Stockholder`s equity / Total Equity 100 Current Ratio Current Assets / Current Debt 100 Quick Ratio Quick Assets / Current Debt 100 Receivables to Inventories Receivables / Inventories 100 Debt Ratio Total Liabilities / Stockholder`s equity 100 Total Borrowings And Bonds Payable To Total Assets (Long and Short term Liabilities+ Bonds Payable) / Total Equity 100 Borrowing Ratio Interest Earned Ratio Operating Profit / Interest Expense Stockholders' Equity Turnover Sales / Stockholders' Equity (Short-term Borrowings+ Current Bonds Payable+Current Long-term Borrowings+Bonds Payable+Long-term Borrowings) / Stockholder`s equity

4 Inventories Turnover Sales / Inventories Receivables Turnover Sales / Receivables 4.2 Multivariate Discriminant Analysis The analysis methods that are being most used in the bankruptcy prediction field until present are the Multivariate Discriminant Analysis (MDA), Logistic Regression Analysis (LGA) and Artificial Neural Network Analysis (ANNA). The present research used the MDA technique that is the most used in bankruptcy prediction theses. The MDA technique is an analysis method that when the dependent variables are category type variables like a bankruptcy or non-bankrupt type and the independent variables are continuous variables, it is used for estimating a function dividing the difference of two groups by using a series of independent variables. The discriminant analysis was achieved by using a statistical computer program (SPSS 17, Excel). The present research has prediction variables(financial ratio) of x, x, x x and sample companies of kinds, used the discriminant function like the equation (1) as an estimated equation when the dependent variable (Z) is divided as Z 1, in case of bankruptcy companies and is divided as Z 0 in case of non-bankruptcy companies. Z d d x d x d x d x (1) Where Z is the discriminant score of th company; and d is the intercept; and d is the discriminant weight of independent variable; and x is the independent variable. There are many cases that the normal distribution is not formed according to the property of the Financial Ratio, and there are many cases that the distribution chart also has skewness in a positive (+) direction. A method removing outlier is being mainly used to reduce a biased value, but it can have serious influence on the result value and show arbitrarily manipulated results. Accordingly, the financial ratio used in the present research was normalized as a natural logarithm function, and the variables were induced closely to the normal distribution. 5. RESULTS OF THE ANALYSIS As a result of carrying out a Multivariate discriminant analysis by applying a total of 50 estimation samples of construction companies and 14 predictor variables, it was analyzed that the True-Positive- Rate is 78% before one year of bankruptcy occurrence, 76% before 2 years of bankruptcy occurrence and 68% before 3 years of bankruptcy occurrence. It can be seen that as it gets remote from the bankruptcy time, the True-Positive-Rate gets to rapid lower, and this says that as it approaches the bankruptcy time, the companies have a high possibility to become delinquency or default. And as the analysis result of the present research, the discriminant weight of Debt Ratio( ), Total Borrowings And Bonds Payable To Total Assets( ) and Borrowing Ratio( ) was analyzed as the highest for 3 years, it can be judged that the financial ratios related to debt of construction companies most well represent the bankruptcy of a company. Table 3. Three Year Predictive Accuracy of the MDA Model Year to Prior to Bankruptcy True Positive Rate (Percent, %) 1 st Year 78 2 nd Year 76 3 rd Year 68 The model of the first year before bankruptcy: 408

5 Z (2) Table 3. Result of multivariate discriminant analysis (first year) Actual Group First year before bankruptcy Frequency The model of the second year before bankruptcy: 0(Non-Bankruptcy) Predicted Group 1(Bankruptcy) 0(Non-Bankruptcy) (Bankruptcy) 7 18 Z (3) Table 4. Result of multivariate discriminant analysis (second year) Actual Group Second year before bankruptcy Frequency The model of the third year before bankruptcy: 0(Non-Bankruptcy) Predicted Group 1(Bankruptcy) 0(Non-Bankruptcy) (Bankruptcy) 8 17 Z (4) Table 5. Result of multivariate discriminant analysis (third year) Actual Group Second year before bankruptcy Frequency 0(Non-Bankruptcy) Predicted Group 1(Bankruptcy) 0(Non-Bankruptcy) (Bankruptcy) CONCLUSIONS According to the property of the construction industry, as most financing is flowed in from nonmonetary institutions, it cannot help using debt essentially. However, the construction industry can prevent delinquency or default capable of being generated afterward by maintaining a proper debt ratio through self-diagnosis of the construction company using the Bankruptcy Prediction Model and comparison with competitors with the same business type. The weight of ratios related to debt was analyzed highly in the results of the present research, but the respective variables should be considered together. As mentioned above, the bankruptcy of construction companies causes big damage socioeconomically beyond a dimension of an executive or workers belonging to the company and cannot avoid chain-bankruptcy of related companies. In case of the construction industry of Korea, the other 409

6 companies except large companies have a small scale and don t have a systematic management system, so several hundreds of construction companies get to face bankruptcy, but if the executive of a company could recognize the possibility of the bankruptcy in advance, he or she can minimize the loss due to this. The Bankruptcy Prediction Model developed in the present research informed that the ratio related to debt is the most important part in sustainable operation of general contractors, and makes a company previously recognize to approach bankruptcy by itself, so it can be said that the meaning of the model is large. But there are many cases that the blind dependence on the prediction model cannot reach intuition or a common sense of human beings because it is just a numeric value appeared through complex calculation. When the moral hazard or opaque trade of the management is carried out in spite that the financial statement of a company is healthy, we cannot estimate its result with the developed model. Accordingly, the prediction model should become an assistant means of experience and intuition of the management in decision-making. The insufficient portion in the present research is to experience difficulty in analysis because the number of samples of bankruptcy general contractors is insufficient, and is not to perfectly remove Multi-Collinearity between predictor variables in the MDA analysis process. The reason is because although the meaning that the financial ratio contains is different, there are many portions that the values included in the ratio are duplicated. And as all the samples are limited to externally audited companies, the prediction model has a limitation that cannot be applied to KOSDAQ and listed companies. It is judged that the development on the prediction model of general contractors and specialty contractors as well as the prediction model including qualitative variables of companies and Macro-Economic indicators having a close relationship with the construction industry is necessary through a diversification process of the future prediction model. REFERENCES Almula K. and David A., Predicting Construction Company Decline Journal of Construction Engineering and Management, ASCE, 130(6), pp Altman E., Financial ratio, discriminant analysis and the prediction of corporate bankruptcy. Journal of finance, American finance association, pp Altman E. and Hotchkiss E., Corporate Financial Distress and Bankruptcy: Predict and Avoid Bankruptcy, Analyze and Invest in Distressed Debt, Third Edition WILEY, USA Beaver W., Financial ratios as predictors of failure. Empirical research in accounting: selected studies, Supplement of Accounting Research, Institute of Professional Accounting, pp Construction Guarantee, Annual Bankruptcy Construction Company List Construction Guarantee in Korea. Construction Association of Korea, Yearly Bankruptcy Company of General Contractor. Construction Association of Korea. Kangari R., Business Failure in Construction Industry Journal of Construction Engineering and Management, ASCE, 114(2), pp Koop G., ANALYSIS OF FINANCIAL DATA WILEY, USA, pp , Korea Specialty Contractors Association, Yearly Bankruptcy Company of Specialty Contractor. Korea Specialty Contractors Association. Korea Financial Telecommunications & Clearings Institute, Declaration Company List of Insolvency(1/Jan/2008~31/Dec/2008) Korea Financial Telecommunications & Clearings Institute. Ohlson J., Financial Ratios and the Probabilistic Prediction of Bankruptcy. Journal of Accounting Research, 18(1), pp Ross S., Westerfield R. and Jordan B., Essentials of Corporate Finance, 6 th Edition McGraw- Hill, USA 410