Financial distress

Manufacturing SMEs’ Financial Distress Prediction Model ตัวแบบในการพยากรณ์การเกิดการล้มละลายในอุตสาหกรรมการผลิตขนาดกลาง และขนาดย่อม โชติมา โรจน์สุรกิตติ บทคัดย่อ ประเทศไทยมีสัดส่วนการจ้างงานในอุตสาหกรรม SMEs ประมาณร้อยละ 69 ของจำนวนประชากร ทั้งหมด SME มีโครงสร้างทางการเงินโดยการดำเนินกิจการขึ้นอยู่กับการกู้ยืมเงินระยะสั้น ดังนั้นจึงควร ตระหนัก ถึงปัญหาทางการเงินของ SMEs โดยในการศึกษานี้ใช้แบบจำลองโลจิตเพื่อวิเคราะห์อัตราส่วนทางการเงินของ บริษัท SMEs จำนวน 385 บริษัท ผลจากการศึกษาพบว่า 37 บริษัท มีปัญหาทางการเงิน และ 348 บริษัทไม่มี ปัญหาทางการเงิน การศึกษานี้ประกอบด้วย 2 คำถามการวิจัย ได้แก่ (1) มีความแตกต่างอย่างมีนัยสำคัญใน สภาพคล่องทางการเงิน อัตราส่วนความสามารถในการจำแนกบริษัท SMEs ที่มีปัญหาทางการเงินและไม่มีปัญหา ทางการเงินในประเทศไทยหรือไม่ (2) แบบจำลองโลจิตเป็นแบบจำลองที่ดีในการวัดสภาพคล่องทางการเงิน ความสามารถในการทำกำไรและการยกระดับทางการเงินของบริษัทที่มีปัญหาทางการเงินในประเทศไทยหรือไม่ โดยทำการตรวจสอบหลักฐานเชิงประจักษ์จากอุตสาหกรรมการผลิตในประเทศไทยเพื่อระบุความแตกต่างระหว่างรูปแบบทางการเงินของ SMEs ที่มีปัญหาทางการเงินและไม่มีปัญหาทางการเงิน โดยการพัฒนาและทดสอบแบบ จำลองโลจิตเพื่อวิเคราะห์บริษัทที่มีปัญหาทางการเงินพบว่าสมมติฐานแรกได้รับการสนับสนุนซึ่งแสดงให้เห็นว่ามีความแตกต่างอย่างมีนัยสำคัญทางสถิติระหว่างอัตราส่วนทางการเงินของ SMEs ในประเทศไทยที่มีปัญหาทางการ เงินและไม่มีปัญหาทางการเงิน และสมมติฐานที่สองแสดงให้เห็นว่าแบบจำลองโลจิตสามารถจำแนกบริษัท SME ในประเทศไทยที่มีปัญหาทางการเงินและไม่มีปัญหาทางการเงินได้แม่นยำมากกว่าการจัดหมวดหมู่ตามโอกาสที่เป็นไปได้ การศึกษานี้สามารถช่วยผู้กำหนดนโยบาย เจ้าของ SMEs และที่ปรึกษาทางธุรกิจ ในการกำหนดกลยุทธ์ เพื่อพัฒนาอุตสาหกรรมการผลิตในประเทศไทยอย่างยั่งยืน นอกจากนี้แบบจำลองโลจิตในการศึกษานี้สามารถ นำไปประยุกต์ใช้ในอุตสาหกรรมอื่น ๆ เพื่อขยายการเจริญเติบโตของอุตสาหกรรมในประเทศไทย Abstract Thai SMEs employ about 69 per cent of the total population. However, SMEs structure in short term financial characteristics depends mostly on short term loan. Thus, researcher have to be aware of financial distress of SMEs. This study utilizes a Logit analysis model to examine financial ratio of 385 SMEs financial statements. The result shows that financial statement consists of 37 financial distress and 348 non-financial distress of enterprises. This study is conducted from 2 research questions which are (1) Are there significant differences in liquidity, leverage and profitability ratios of financial distressed and non-financial distress in Thai SMEs. (2) Is Logit model is appropriate for classifying Thai financial distress. The study has examined empirical evidence from Thai manufacturing industries to identify differences between financial profiles of financial distress and non-financial distress of SMEs before developing and testing the Logit analysis model in order to predict tendency of SMEs financial distress. The first hypothesis is accepted, which shows that there are statistically significant differences between financial ratios of financial distress and non-financial distress of SMEs in Thailand. The second hypothesis shows that the predictability of financial ratios in the Logit analysis model enables classifying Thai financial distress and non-financial distress of SMEs more accurately than an feasible classification. Finally, this study is expected to help policy-makers, SMEs owners and business consultants to determine strategies by providing sustainable development for Thai manufacturing SMEs to expand the growth of Thai industries. Introduction Small to medium sized enterprises (SMEs) in Thailand are defined as enterprises that comply with employment under 200 employees with investment capital under 100 million baht and fixed assets under 100 million baht. (OSMEP, 2005). In the competitive business world of today, in which flexibility, speed and adaptability are essential for survival and progress, small and medium sized enterprises (SMEs) play an extremely important role in any country’s economic development, especially for the 21 members of the Asia-Pacific Economic Cooperation. Thai SMEs have played a vital role in the Thai economy over three decades since the first National Five-Year Plan was instigated in 1960. As a part of this economic development, SMEs in Thailand have been successful in many ways, particularly with the SME’s share of GDP in Thailand has reached 39 per cent. If farming and agricultural processing income are also included, the share is supposed to rise at 50 per cent (OSMEP, 2003). During the same period, the SME share of manufactured goods’ exportation reaches 38.2 per cent of the total value of Thailand’s exports. When employment is taken into consideration, Thai SMEs employ about 69 per cent of the total population. However, SME has characteristics of short-term financial structure which depends mostly on short term loan. Thus, researcher have to be aware of financial distress of SME so that researcher have to look for statistic model that analyzes financial ratio to predict tendency of company’s financial distress. Prediction models of SMEs’ financial distress can help business managers to foresee the cautions of financial risk of SMEs before making credit decisions and to inform policymakers by highlighting key prior areas. Against this background, this study develops and tests a model to identify SMEs’ financial distress. This study’s major premise is about financial ratios in the isolation fail to provide sufficient basis for making informed judgment about SME failure. Accordingly, researcher develops and tests a multiple discriminant analysis model to distinguish between financial distress and non-financial distress of SMEs using three categories of financial ratios: liquidity, leverage and profitability. Thus, the following research hypotheses are pursued: The objectives of this study are: -To identify and to confirm the factors which are contributing to business performance or failure. -To construct the management of innovative performance as a conceptual model for SME(s) in Thailand. Many studies have conducted the importance of SME especially the growth of manufacturing industry in Thailand following the definition of Thai SMEs, manufacturing importance towards Thai economy and SME, financial distress definition, the benefit of financial ratio for identifying financial distress, development of statistical models to predict financial distress. Many firms allow their customers to delay payment for delivered goods by offering trade credit and enable their business partners to cope with liquidity problems. The results of empirical studies show that trade credit is a very important source of external finance in short term. To date, a number of empirical and theoretical studies analyze the demand and the provision of trade credit: With respect to the demand for trade credit, research findings suggest that bank credit constraining firms is more likely to resort of trade credit (Biais & Gollier, 1997; Petersen & Rajan, 1997). Suppliers are possibly willing to provide trade credit to their customers if they have better information about the business and the credit risk of their customers more than banks and if they have fewer problems to obtain external finance more than their customers (Schwartz, 1974). Moreover, firms may provide trade credit in order to get price discrimination since expansion of the credit period implies some reduction with the effective price (Chee K. NG, Smith, & Smith, 1999). Hence, suppliers are fesibly willing to offer more trade credit with the most price elastic segmentation of the market, e.g. credit rationed firms, or to provide further price discrimination because of the long-term interest to be a survival as the business partner (Petersen & Rajan, 1997). Manufacturing importance towards Thai Economy or SMEs has quite tremendous exist. comprising on average 25 per cent of each addition to GDP (incremental GDP), or 70 per cent of all industrial added value. Manufacture is characterized by a high reliance on agricultural products, including rubber products, textile products, food processing, beverages, and tobacco. Thailand's food and agriculture share added manufacturing value around 36 per cent. The second foremost manufacturing area is textiles, clothing, and leather products, that are produced mainly for exportation, with 23 per cent of manufacturing added value. Machinery and transport equipment, which consists most of repair and motor vehicles, are accounted for 11 per cent, and chemicals accounted for 7 per cent, and the other 23 per cent including processed minerals, wood, rubber, carpets, batteries, rope, gunnysacks, plastic goods, tires, footwear, and an expanding domestic small arms production. However, the genesis of financial distress corporate in term of “financial distress” is used in a negative connotation to describe the financial situation of a company confronted with a temporary lack of liquidity and encouraged some difficulties that ensue in fulfilling financial obligations on schedule with the full extent. (Gordon ,1971), (Davydenko, 2005). Financial distress is frequently determined in terms of failure, default, bankruptcy, or distressed restructure, that are overall dependent on underlying methodology and research objectives. As a consequence, theoretical and empirical models of financial distress exhibit to a certain extent as a one-sidedness in the context of question analyses which mainly concentrate on the momentary perspective. When the adverse process has reached its lowest point, the decision towards insolvency or distressed restructure has to be made (Gilson, 1989). However, picking single negative events for the analysis of financial distress as a whole possibly incorrect and it may leads to biases. Distortions feasibly arise because of the examination of the deepest point of financial distress, also known as default, which ignores the fact that the largest losses and increae of financial inflexibility happen several periods before this event occurs. (Ward and Foster, 1997), The first step to evolution of the failure prediction of quantitative firm model was taken by Beaver (1966), who developed a dichotomous classification test based on a simple t-test in a univariate framework. He used individual financial ratios from 79 failed and non-failed companies that were matched by industry and assets size in 1954 to 1964 and identified a single financial ratio. Beaver’s study was then followed by Altman (1968), who suggested a Multivariate Discriminant Analysis (MDA). By utilizing 33 bankrupt companies and 33 non-bankrupt companies between 1946–1964, five variables were selected with foremost relevant factors to predict bankruptcy before being analyzed with many model studies afterward. Development of statistical models fot predicting financial distress has played important role for good projection towards financial distress, for instance, principal component analysis of models and methodologies, cluster analysis, CHAID, the logistic model. In order to identify the “healthy” and “unhealthy” of Thai listed companies for the year 2008 researcher applied several models and methodologies, such as the principal component analysis, a hierarchical cluster, tree model of CHAID decision and the logit model. These models appropriately classify the listed companies quite good and provide relevant information regarding financial ratios that better predict financial distress. The PCA and cluster analysis indicate the following variables: profit margin, return on assets (ROA), return on equity (ROE), profit per employee, current ratio, debts on equity and growth rate on total assets, wheras the tree model of CHAID decision indicated profit margin, return on assets and turnover growth, and lastly, the logit model indicate profit margin and debts on equity. Research Questions 1. Are there any significant differences in liquidity, leverage and profitability ratios of financial distress and non-financial distress of Thai SMEs? 2. Is logit model is the appropriate model for measures of liquidity, profitability, and financial leverage to classify Thai financial distress? Hypotheses The research questions of this study give rise to the following hypotheses (H): H1: There are significant differences in liquidity, leverage and profitability ratios of financial distressed and non-financial distress of Thai SMEs. H2: A logit analysis model with measures of liquidity, profitability, and financial leverage classifies Thai financial distress Financially and non-financially distressed SMEs are more accurately than a possible classification. The first hypothesis enables researcher to test differences between individual financial ratios of financially distressed and non-financially distressed SMEs, whilst, the second hypothesis relates to developing and testing prediction of SMEs’ failure model by taking several ratios into account. Research Methodology Used Descriptive research in this research transform the raw data into the form that has a clear understanding and is easy to interpret data, which is presented in a meaningful way (Sekaran, 1993). Descriptive research is studied to determine the answer of who, what, when, where and how questions (Zikmund, 2003). The definition of SMEs in Thailand is explained in the previous as a basis for the purpose of identifying the sampling population before applying further operational approach was to follow making the best use of available data in Thailand because information regarding number of SMEs’ employees and fixed asset size are not available. Therefore, asset size is used as a criterion to classify the size of businesses in this study. This research is expected to apply by adopting the recommendation for Thai SMEs in order to enhance productivity and efficiency of SMEs’ products including quality management to strengthen Thai economics. For this research, the sample size will be estimated by the sample population proportion approach with confidence level at 95% and maximum allowance for random sampling error at 0.05. Thus, the sample size for this research can be calculated with the following formula (Zikmund, 2003) n = Z2pq E2 Or n = Z2p(1-p) E2 Where, n = number of items in sample Z2 = square of the confidence level in standard error units P = estimated proportion of success q = 1-p, or estimated proportion of failures E2 = square of the maximum allowance of error between the true proportion and the sample proportion. The allowable error is 0.05 or 5%. Therefore, the total of the sample size to be researched is n = Z2p(1-p) E2 = (1.96)2(0.5)(1-05) (0.05)2 = 348.16 348 samples The result of the sample size is equal to 348 samples after formularcalculation. In order to recheck approach accurately, researcher gatheres some online data and information from the online information which is developped by Department of Business Development (DBD) which obtain from the Ministry of Commerce, Thailand (http://www.dbd.go.th) (Department of Business Development, 2008). Secondary sources are journals, Internet, newspaper, magazine articles, textbook and previous studies. The purpose of going thoroughly with secondary material also was to find support and guidance for the research that has been undertaken. Statistical Treatment of Data In this study, researcher uses financial ratio to calculate logit model in order to determine the financially distressed company. The logistic model is a conditional probability model that uses maximum likelihood estimation to provide the conditional probability of a firm belonging to a certain group given the values of the independent variables for that firm. It is a single-period classification model (Shumway, 2001) which can be decribed by the following function: An important issue in using binary state prediction models such as logit analysis is the selection of the cutoff probability, which determines the classification accuracy. In order to classify an observation into two groups; the estimated probability from the logit model and a pre-determined cutoff probability. If the estimated probability is lower than the cutoff, the observation will be classified as an inferior performer and if the estimated probability is above the cutoff, it is placed in the superior performer group. A total of 385 financial SMEs statements are used comprising those 37 financially and 348 non-financially distressed enterprises. The list of the distressed firms is obtained from the website of the Legal Execution Department, Ministry of Justice, Thailand (http://www.led.go.th) (Legal Execution Department, 2008). Thai SMEs that apply to the Thai Bankruptcy Court, the Central Bankruptcy Court and the Civil Court until 2002 was used to select financially distressed enterprises with assets lower than THB 2,000 million. SMEs in the sample may or may not have ceased operations following the bankruptcy because the future of these firms will depend on factors such as the progress of their loan restructure and plans for improving their performance. Sixty-eight sets of financial statements, i.e., balance sheets and income statements, of financially distressed SMEs were completed and usable. After identifying SMEs using the explained criteria earlier, 198 financial statements of non-distressed SMEs are considered completely and usably for the study. To avoid a possible sampling bias and to be consistent with the approach, researcher used for selecting financially distressed SMEs after developing the model and a new sample with three different sets is used for testing the model’s reliability. Variable Definition The nine independent variables, are most commonly used by previous studies, which can be classified into liquidity, leverage, and profitability. These ratios are outlined below: 1. Liquidity quickly refers to how an asset can be converted into cash, i.e., the ability of current assets to meet current liabilities. Liquidity can be divided into 3 sections; current assets to total assets ratio (CATA), current liability to total assets ratio (CLTA), and working capital to total assets ratio (WCTA). Those 3 measures can be explained as follows; 1.1. Current assets to total assets ratio (CATA) is amount of cash, account receivables, bills, inventory and other current assets of total assets in percentage. 1.2. Current liability to total assets ratio (CLTA) refers to amount of account payables, and other short-term liability of total assets in percentage. 1.3. Working capital to total assets ratio (WCTA) means current assets or less current liability of total assets in percentage. 2. Leverage or gearing refers to the use of debt to supplement investment, or the degree to which business is utilizing borrowed money. Three measures of leverage can be categorized as long-term liability to total assets ratio (LLTA), total liability to total assets ratio (TLTA), and debt to equity ratio (DE) as follows; 2.1. Long-term liability to total assets ratio (LLTA) specifies amount of long-term liabilities of total assets in percentage. 2.2. Total liability to total assets ratio (TLTA) is amount of short-term and long-term liabilities of total assets in percentage. 2.3. Debt to equity ratio (DE) identifies the amount of debt divided by equity. 3. Profitability refers to the ability of a firm to generate net income. In the three categories, ratios that are applicable to all selected companies in the sample were chosen as follows; 3.1. Total income to total assets ratio (TITA) centers on amount of total core and other income of total assets in percentage. 3.2. Earnings before interest and tax expenses to total assets ratio (EBITTA) are all earnings before interest and tax expenses as a percentage of total assets. 3.3. Earnings after interest and tax expenses to total assets ratio (EAITTA) means all earnings after interest and tax expenses as a percentage of total assets. Results Hypothesis 1: Test of differences in financial ratios. Comparison of descriptive statistics concerning financial ratios of financially and non-financially distressed SMEs is presented in Table 1. The variables of interest are the ratios that relate to current liabilities, long-term debts, and profitability. Comparison of financial ratios’ mean for the two groups of SMEs shows that financially distressed SMEs have lower liquidity, higher leverage and lower profitability than non- financially distressed. This is consistent with the theoretical expectation that non-financially distressed companies exhibit higher liquidity, greater profitability, and lower levels of debt. The distressed firms have a great deal of liabilities, which are greater than their assets value. Table 1 shows total liability and long-term liability to total ratio assets are over 100 per cent for financially distressed SMEs, which resulted in distressed firms having negative equity (DE = -4). In ideal circumstances, liabilities would be kept under total assets, and excessed equity of debt. The study testes the statistically significant differences between financially and non-financially distressed SMEs. Parametric t-tests are conducted to cover specific nine variables in order to identify statistically significant differences between the financial ratios for the two groups of SMEs in the sample. The tests show results that are fit to our expectations (Table 2). In that the variables exhibit statistically significant differences for both parametric and nonparametric tests at a 0.1% level of significance. The financially distressed SMEs exhibit lower liquidity, higher leverage and lower profitability than non-financially distressed SMEs. Thus, the test of differences shows that there is significant difference of liquidity, leverage and profitability ratios of financially and non-financially distressed Thai SMEs. TABLE 1: THE FINANCIALLY DISTRESSED (FD) AND NON-FINANCIALLY DISTRESSED (NFD) SME’S COMPARATIVE DESCRIPTIVE STATISTICS FD-SMEs NFD-SMEs Variable Mean SD. Mean SD. Liquidity 1) CATA 40.864 25.291 71.597 20.892 2) CLTA 172.208 44.510 41.381 23.553 3) WCTA -107.412 75.491 40.271 20.674 Leverage 4) LLTA 107.451 82.658 19.240 16.657 5) TLTA 297.550 92.517 46.648 29.650 6) DE -278.846 152.295 38.614 28.684 Profitability 7) TITA 57.326 85.031 155.594 18.461 8) EBITTA -15.416 39.942 19.254 19.019 9) EAITTA -19.489 58.523 12.162 15.681 TABLE 2: COMPARATIVE PARAMETRIC (t -TEST) RESULTS OF FD-SMES AND NFD-SMES Parametric t-test t value Sig. (1-tailed) Result Liquidity of FD-SMEs is less than that of NFD-SMEs 1) CATA of FD-SME < that of NFD-SMEs -7.591 0.000 Sig. difference 2) CLTA of FD-SMEs > that of NFD-SMEs 5.218 0.000 Sig. difference 3) WCTA of FD-SMEs < that of NFD-SMEs -6.090 0.000 Sig. difference Leverage of FD-SMEs is greater than that of NFD-SMEs 4) LLTA of FD-SMEs > that of NFD-SMEs 7.890 0.000 Sig. difference 5) TLTA of FD-SMEs > that of NFD-SMEs 8.506 0.000 Sig. difference 6) DE of FD-SMEs > that of NFD-SMEs -4.704 0.000 Sig. difference Profitability of FD-SMEs is less than that of NFD-SMEs 7) TI of FD-SMEs < that of NFD-SMEs -10.107 0.000 Sig. difference 8) EBIT of FD-SMEs < that of NFD-SMEs -6.298 0.000 Sig. difference 9) EAIT of FD-SMEs < that of NFD-SMEs -5.672 0.000 Sig. difference Remark: Sig. (2-tailed) divided by 2; * Significant at 0.05 level; ** Significant at 0.01 level Hypothesis 2: SME failure prediction model development and testing. Having established that the differences in financial profiles between the two groups of SMEs are statistically significant, a distress prediction model for Thai SMEs was developed and its accuracy assessed. A multiple legit analysis model was developed for Thai SMEs in the sample with a view to classifying the firms into financially distressed (FD)and non-financially distressed (NFD) categories. Two approaches are used in selecting variables for the model: 1) using all variables; and 2) selecting variables based on correlation results. Using the first approach, i.e., incorporating all the nine variables into the model, long-term liability to total assets (LLTA) and working capital to total assets (WCTA) ratios did not pass the tolerance criteria (i.e., the minimum tolerance level of 0.001, see Hair et al. (1998)). This indicates that some variables are likely to exhibit non-normal distributions and also multi-collinearity. Therefore, the second approach is employed to closely examine correlation results and select variables for the model with a view to excluding some of the highly correlated variables (Table 3). TABLE 3: CORRELATIONS MATRIX Variables CATA CLTA WCTA LLTA TLTA DE TITA EBITTA EAITTA CATA 1.000 CLTA -0.151 1.000 WCTA 0.581** -0.611** 1.000 LLTA -0.157 0.438** -0.495** 1.000 TLTA -0.319** 0.567** -0.598** 0.610** 1.000 DE 0.241** -0.190 0.315** -0.303** -0.298** 1.000 TITA 0.426* -0.181 0.245* -0.317** -0.287* 0.207* 1.000 EBITTA 0.226* -0.359** 0.601* -0.315* -0.352* 0.213 0.265** 1.000 EAITTA 0.351* -0.407** 0.443** -0.381* -0.612** 0.249* 0.345** 0.512** 1.000 * Correlation is significant at the 0.05 level (2-tailed) ** Correlation is significant at the 0.01 level (2-tailed) TABLE 4: STEPWISE LOGISTIC REGRESSION: ANALYSIS OFMAXIMUM LIKELIHOOD ESTIMATES Variables Coefficient Std.error Wald df Sig.(2-tailed) Exp (B) Constant -1.806 0.523 6.244 1 0.013 CATA 0.308 0.021 5.149* 1 0.027 1.352 WCTA 0.454 0.435 4.698* 1 0.030 2.569 LLTA -0.221 0.125 3.147* 1 0.046 0.802 DE 0.318 0.203 8.529** 1 0.004 1.182 TITA 0.215 0.052 7.211** 1 0.005 1.155 EBITTA 0.195 0.170 5.309* 1 0.025 1.215 * Correlation is significant at the 0.05 level (2-tailed) ** Correlation is significant at the 0.01 level (2-tailed) From Table 4, It has been founded that Logit Regression Model has 6 independent variable: Current assets to total assets ratio (CATA), working capital to total assets ratio (WCTA), long-term liability to total assets ratio (LLTA), debt to equity ratio (DE), total income to total assets ratio (TITA) and earnings after interest and tax expenses to total assets ratio (EAITTA). These 4 variables are analysed with confidence level at 95% (Use * symbol) which are current assets to total assets ratio, working capital to total assets ratio, long-term liability to total assets ratio, earnings after interest and tax expenses to total assets ratio. While another 2 variables; debt to equity ratio (DE) and total income to total assets ratio (TITA) have 95% of confidant level. In the same time, current assets to total assets ratio (CATA), working capital to total assets ratio (WCTA), debt to equity ratio (DE), total income to total assets ratio (TITA), and earnings before interest and tax expenses to total assets ratio (EBITTA) have positive coefficient with Exp (B) equal to 1.352, 2.569, 1.182, 1.155, 1.215 respectively. If Exp(B)>1 means independent variable will stimulate the possibility of non-financial distress. It can show that if current assets to total assets ratio (CATA), working capital to total assets ratio (WCTA), debt to equity ratio (DE), total income to total assets ratio (TITA) or earnings after interest and tax expenses to total assets ratio (EAITTA) has increased 1 %, it can stimulate the possibility of non-financially distress that equals to 1.352, 2.569, 1.182, 1.155, 1.215 times respectively. Only one variables which is long-term liability to total assets ratio (LLTA) that has negative coefficient. Furthermore, Exp(B) =0.802 (Exp(B)<1) means that this independent variables will stimulate the possibility of non-financial distress. If long-term liability to total assets ratio (LLTA) has increased 1 %, it will decrease the possibility of being non-financial distress 0.802 times. The result of regression has found that Chi-square statistics equal to 20.399, which has significant level of 0.01. It means that independent variables in model are appropriate or it has some independent variables that have significant effect to financial or non-financial distress of Thai manufacturing SMEs (TABLE 5 (a)). TABLE 5(b) has discovered that 2 Log likelihood statistics equal to 223.257, which is lower than 2 log likelihood that has only fixed variable (243.65652). It means that this logit regression model has resulted in the same way as empirical study and it defines that R2 can be properly predicted statistics which equal to 45.2 (R2 of Cox& Snell) and 51 per cent for Nagelkerke R2. TABLE 5(c) has shown the effectiveness of logit regression, which can predict financially distressed Thai SMEs. Manufactures are corrected 84.08 per cent and are able to predict non-financially distressed Thai SMEs. Manufactures are corrected 87.93 per cent. Overall, this logit regression model can correctly predict 87.27 per cent. TABLE 5: REGRESSION ANALYSIS OF PREDICTIVE 6 INDEPENDENT VARIABLES FOR FINANCIAL DISTRESS AND NON-FINANCIAL DISTRESS OF THAI SMEs MANUFACTURING (a) APPROPRIATE IDEPENDENT VARIABLE IN MODEL (OMNIBUS TESTS OF MODEL COEFFICIENTS) Statistic Value Chi-square df Sig (2-tailed) Step 20.399** 6 0.002 Block 20.399** 6 0.002 Model 20.399** 6 0.002 * Correlation is significant at the 0.05 level (2-tailed) ** Correlation is significant at the 0.01 level (2-tailed) (b) CONFIRMITY CHECK OF MODEL (MODEL SUMMARY) Step -2 Log likelihood Cox & Snell R Square Nagelkerkerke R Square 1 223.257 0.452 0.510 - 2 Log likelihood has fixed variable = 243.65652 (c) CORRECTIVE PREDICTION OF FINANCIAL DISTRESS AND NON-FINANCIAL DISTRESS Predicted Observed FD-SMEs NFD-SMEs Percent Correct FD-SMEs (0) 30 7 81.08 NFD-SMEs (1) 42 306 87.93 Overall 87.27 In conclusion, Thai financial and non-financial distress of manufacturing SMEs can be classified into liquidity ratio with current assets to total assets ratio (CATA) and working capital to total assets ratio (WCTA), financial leverage ratio with long-term liability to total assets ratio (LLTA) and debt to equity ratio (DE), and profitability ratio with total income to total assets ratio (TITA) and earnings before interest and tax expenses to total assets ratio (EBITTA). Consequently, Thai financially distressed SMEs and non-financially distressed SMEs more accurately than a possible classification by chance. Discussion This study has empirically examined differences between financial profiles of financially and non-financially distressed SMEs in Thailand before developping and testing a logit model analysis to predict SMEs that have some financial difficulties and thus involve in high financial risk so that the first hypothesis is accepted. The results show that distressed firms had lower liquidity, higher leverage and lower profitability ratios. The financial ratios of distressed firms are taken for analysing to develop the prediction model. The second hypothesis is also accepted. This hypothesis predicts that Thai SMEs failure is amenable to prediction to a statistically significant extent using a logit model analysis. The predictive power of the model has a room for improvement, whereas, non-financial variables such as age of business, level of education of business owners or managers, change of auditors, and other qualitative details of business managers, number of years established may also enable researchers to more effectively detect the signs of a financial distress (Altman et al., 2008). However, the main focus of this study is to enhance the usefulness of accounting information by articulating individual ratios into a model. This is a useful approach as financial information is usually the only publicly available information about small firms (Deegan, 2009; Godfrey et al., 2010). Furthermore, the sample is illustrated from various industries, which make the model still amenable to improve by focusing on specific industries. Developed models using financial data from some industries may not be highly accurate in predicting distress for firms in other industries since financial characteristics of firms cannot be expected to exhibit similarity across several industries. Developing models for particular sectors could improve the predictive power of the model for business failure tends to vary by type of business. For instance, in the United States, the retail sector was the second largest category of corporate business failure between 1992 and 1997 (Dun and Bradstreet, 1998). Conclusion The study has examined empirical evidence from Thai manufacturing industries to identify differences between financial profiles of financially and non-financially distressed SMEs by developping and testing a Logit model analysis for predicting SMEs’ financially distress. The first hypothesis is accepted, which shows that there are statistically significant differences between financial ratios of financially and non-financially distressed SMEs in Thailand. The results also exhibited that financially distressed SMEs tend to exhibit lower liquidity than non-distressed manufacturing SMEs, which arise from the greater use of short-term liabilities. Financially distressed manufacturing SMEs exhibit higher leverage than non-distressed manufacturing SMEs and less profitability because of the higher amount of operating costs and involved interest expenses. The second hypothesis, that predicts financial ratios in a logit model analysis, enables classifying Thai financially and non-financially distressed SMEs more accurately than a possible classification based on occasion is also accepted. Therefore, It is possible that Thai manufacturing SME could create risk of debt. This implies that policymakers need to provide financial help to manufacturing SME to create sustainable development. This study is contributed to develop and test a model for manufacturing SME in Thailand, which can also be applied in other emerging industries. Furthermore, the study has validated the model by using a new sample to test the model’s practical significance which makes model more practical than validates the model with acquiring samples. Policymakers’ ability to identify financial distress also assists the Government agencies to predict and prevent distress by providing protential assistance for distressed firms and issue policy to help non-distressed firm to run their business wisely and to avoid possible distress. SME owners need to set their business strategies to be not distressed. This study gives us the understanding towards their characteristics of financial ratio that have possibility to be distressed in order to assist in finding timely solutions for the problems. Business consultants has responsibility to give their clients some advices on how to develop viable financial strategies. From the further study’s point of view, it should be noted that a wide range of variables including non-financial data such as duration of business, educational level of business owners or managers, change of auditors, and qualitative details of business managers may also enable researcher more effective detecttion towards the signs of financial distress. Finally, the sample can illustrate other industries, which makes the model still amenable to further improvement by focusing on specific industries. Future research could be done centering on other industries and consider another non-financial variables into the analysis. References Altman, E I. (1968) "Financial Ratios, Discriminant Analysis and the Prediction of Corporate Bankruptcy", Journal of Finance, 23, pp. 589-609. Altman et al. (2008): How realistic are the prospects for “closing the gaps” in socioeconomic outcomes for Indigenous Australians’. CAEPR Discussion Paper No. 287/2008. Canberra: Centre for Aboriginal Economic Policy Research, Australian National University. Beaver, W. (1966): Financial Ratios as Predictors of Failure. In: Journal of Accounting Research, 5, 71-111. Biais, B., & Gollier, C. (1997). Trade credit and credit rationing. Review of Financial Studies, 10 , 905-937. Chee K. NG, C. K., Smith, J. K., & Smith, R. L. (1999). Evidence on the determinants of credit terms used in interfirm trade. The Journal of Finance, Vol. 54 No. 3 , 1109-1129. Davydenko, S. (2005): When Do Firms Default? A Study of the Default Boundary. Working Paper, London Business School. Deegan, C. (2009): Australian Financial Accounting, 6th edition, McGraw Hill Book Company, Sydney. The Dun & Bradstreet Corporation - Cost Basis (1998). The Dun & Bradstreet Corporation - Investor Relations. Retrieved June 8, 2013, from http://investor.dnb.com/costbasis.cfm?pagesect=june Gilson, S. (1989): Management Turnover and Financial Distress. In: Journal of Financial Economics, 25, 241-262. Godfrey C, Pickett KE, Parrot S. (2010): Estimating the costs to the NHS of smoking in pregnancy for pregnant women and infants. Department of Health Sciences, The University of York. Gordon, M. J. (1971): Towards a Theory of Financial Distress. In: The Journal of Finance, 26(2), 347-356. Hair, J. F. Jr., Anderson, R., Tatham, R., & W. C. (1998) Multivariate Data Analysis (5th ed.). Upper Saddle River, NJ: Prentice Hall. Petersen, M. A., & Rajan, R. G. (1997). Trade credit: Theory and evidence. The Review of Financial Studies, Vol. 10 No. 3 , 661-691. Office of Small and Medium Enterprises Promotion (OSMEP) (2003). The white paper onSmall and Medium Enterprises of Thailand in 2003 and Trend 2004. Bangkok: OSMEP. Office of Small and Medium Enterprises Promotion (OSMEP) (2005). Small and Medium Enterprises of Thailand in 2004 and Trends 2005. Bangkok: OSMEP. Sekaran, U. (1993). Research methods for business - A skill building approach (2nd ed.). New York: John Wiley &Sons. Schwartz, R. A. (1974). An economic model of trade credit. Journal of Financial and Quantitative Analysis, September , 643-657. Shumway, T. (2001) "Forecasting bankruptcy more accurately: A simple hazard model", Journal of Business, 74 (1), pp. 101-124. Ward, T., Foster, B. (1997): A Note on Selecting a Response Measure for Financial Distress. In: Journal of Business Finance and Accounting, 24(6), 869-879. Zikmund, W. G. (2003): Business Research Methods, 7th ed., (Ohio-USA, Thomson, South-Western). 3