ªï Ë 7 Ë 14 - ÿπ π 2547 «.«25 REHABCO and recovery signal : a retrospective analysis Worasith Jackmetha* Abstract An investigation of the REHABCOûs financial position and performance using the Altman model as a retrospective analysis is conducted. Original Altmanûs cutoff points are set out to quantify a recovery signal of the ailing firms. Results indicate that, on average, those ailing firms in the REHABCO still have poor performance and financial position. There are only two companies with the Z-score greater than the higher cutoff point, that means they are, theoretically, in the safe area. However, at the end of the year 2003 the two healthier firms are still in the REHABCO. This can be interpreting that the result of the Altman model is only a guideline of the recovery investigation, and the companies still need special treatments in order to reach strong financial position and performance on continuous basis. Keywords: Financial statement analysis; Altman Z-score; Bankruptcy prediction; Thai stock market. Introduction The most severe economic crisis in Thailand since the end of World War II started in 1996. Consequence of this crisis came with a large number of companies that faced operating and financial problems, and some of these are listed in the Stock Exchange of Thailand. In early 1998, a new sector çcompanies under rehabilitationé (hereafter referred to as çrehabcoé) was set up in the Stock Exchange of Thailand for those ailing firms. The criteria of classifying companies to be listed in the REHABCO have been announced by the Stock Exchange of Thailand. Major criteria that most of the ailing companies faced were shareholderûs equity been in red, and foreign debt problem that might be due to overaggressive policies of debt and expansion. These companies had to submit a plan called çrehabilitation plané to the Stock Exchange of Thailand. After approved by companyûs shareholders, the plan has been implemented, and * Department of Finance, Faculty of Business Administration, Huachiew Chalermprakiet University.
«.«26 ªï Ë 7 Ë 14 - ÿπ π 2547 the primary aim is to be returned to former sector. However, those companies have to demonstrate strong financial position and performance on a continuous basis. Corporate bankruptcy prediction models are routinely included in finance textbooks to illustrate benefits of predicting the possibility of bankruptcy (Grice.2002). Experienced financial analysts are often able to predict the likelihood of financial distress, which leads to bankruptcy, by monitoring firmûs performance closely over a long period (Hughes.1993). Prediction models that relate to firmûs financial statement are discussed in the next section. The purposes of this study is two-fold. First, to access financial health of listed companies in the REHABCO. Second, to test the Altman model in the Thai context, especially with the ailing enterprises to determine whether there is a Z-score that stays above bankruptcy zone. Contributions of this study may lead to more investigations in business practice, not only for testing a recovery sign but also as a tool for preventing of financial distress. Bankruptcy prediction development Studies of corporate bankruptcy were evident in 1930s, and attempt to develop bankruptcy prediction models has been a popular area among scholars since the late 1960s. Most of them relied heavily on accounting measures (financial ratios), such as profitability, liquidity, and solvency. Beaver (1967, cited in Altman.2000) one of the earliest researchers used univariate analysis of a number of financial ratios to discriminate between bankrupt and non-bankrupt firms. This analysis can be conducted in both on a cross-sectional basis (comparisons with firm in the same industry) and on a time-series basis (comparisons within the same firm over time). However, the univariate financial ratio analysis may not appropriate for some firms, especially those highly diversified on both products and geographical basis (Hughes.1993) because each ratio is determined separately from others Consequently, the analysis cannot conclude whether a particular ratio is favorable or not. Hence, different ratios may not be in line with others. The most influential in the early studies is multivariate discriminate analysis model (MDA), which was developed by Edward Altman in 1968 (Keating, et al.2003). The multivariate discriminate analysis, known as Z-score model that used multiple discriminant analysis to derive discriminant function based on five categories of financial ratios including liquidity, profitability, leverage, solvency, and activity (Grice.2000). The model was based on a sample composed of 66 manufacturing companies, half of which were filed a bankruptcy petition. The implication of the Altmanûs Z-score is to examine the indicated ratios in time period t to make prediction in the following period (t+1). The following generation of bankruptcy studies including Santomero and Vinso in 1977, Ohlson in 1980, and Zmijewski in 1984 (Keating,
ªï Ë 7 Ë 14 - ÿπ π 2547 «.«27 et al.2003), was conducted under multinomial choice techniques, such as maximum likelihood logit and probit analysis. Ohlson (1980) used logit analysis to derive default prediction model by using nine measures: firmûs size, leverage, liquidity, and performance with 105 bankrupt and 2,058 non-bankrupt industrial firms, while Zmijewski (1984) conducted his model by using financial ratios that measured firm performance, leverage, and liquidity and estimated the model using probit analysis. His probit analysis weighted the log-likelihood function by the ratio of the population frequency rate to the sample frequency rate of the distress and non-distress groups. Data This study is designed to test a financial model that has been developed since 1960s by using secondary data in Thai stock market, the REHABCO. Data used in this study comprises latest annual financial statement (balance sheet and income statement) of 51* ailing firms that provide financial position and operating result, and closing price of each stock. In case of no closing price at year-end of the latest accounting period due to a temporary trading suspension, prior closing price was used to represent the fair market value. The annual financial statement and the closing prices are public information, which can be obtained from the Stock Exchange of Thailandûs database. The Stock Exchange of Thailand provides financial statement of listed companies regularly through its internet website. Historical stock prices are available in the intranet system of the Stock Exchange of Thailand. The model The Z-score model, widely used by both academics and practitioners (Keating, et al.2003) and proven to be a reliable tool for bankruptcy forecasting in a wide variety of context and market (Gregory.1995), is selected to conduct this study. The model compiled a list of twentytwo financial ratios and classified them into one of five categories: liquidity, profitability, leverage, solvency, and activity. Together with the ratios, a statistic technique, known as multivariate discriminate analysis was selected to derive a linear combination of variables that best separates between bankrupt and non-bankrupt groups. The original Altmanûs work uses cutoff score of 2.675 and 1.81. The Z-score above 2.675 indicates non-bankrupt firms and the score below 1.81 signals ailing firms, while the area between the two cutoff score is ignorance zone. The model was 95 percent accurate in predicting one year prior to bankruptcy and 72 percent for the data two to five years prior to bankruptcy (Hughes.1993). In this study, a retrospective analysis is conducted to investigate financial health of those ailing entities in the REHABCO, and at * At the beginning of 2002 there were 52 companies in the REHABCO, and during the year two of them were merged together.
«.«28 ªï Ë 7 Ë 14 - ÿπ π 2547 the same time to reveal whether there is any recovery signal. The linear function that discriminates between bankrupt and non-bankrupt firms is (Altman.2000): Z = 0.012X 1 + 0.014X 2 + 0.033X 3 + 0.006X 4 + 0.999X 5 (1)* where X 1 = working capital/total assets, X 2 = retained earnings/total assets, X 3 = earnings before interest and taxes/total assets, X 4 = market value of equity/book value of total liabilities, X 5 = sales/total assets, and Z = overall index. Result and discussion To follow up financial health of companies listed in the REHABCO, the Z-score is calculated as a retrospective analysis based on the original work of Altman (1983). The first variable, working capital/total assets ratio measures net liquid assets to total capitalization. Working capital is the difference between current assets and current liabilities. The result from Table 1 shows the mean of the first variable of -1.01. This mean, on average, the firms listed in the REHABCO have negative working capital, which implies a liquidity problem. In addition, there are only 20 firms that shown positive net liquid assets. From the investigation, liquid assets are not only lower than current liabilities but the figure also reveals a relatively high portion of negative liquid assets when compared to total assets. This figures comply with Altman (1983), who mentioned that a consistent operating loss firm will reveal lower current assets size when compared to total assets. Table 1 Component ratios of the REHABCO Variable Max Min Mean S.D. X 1 0.39-22.35-1.01 3.34 X 2 0.39-31.51-2.93 6.04 X 3 2.06-0.66 0.02 0.35 X 4 4.66-26.40-1.35 4.15 X 5 2.12 0.00 0.61 0.60 * Due to format arrangement of Altmanûs original work, variables X 1 through X 4 must be calculated as absolute percentage value but variable X 5 has to be expressed in a different manner, a number of times.
ªï Ë 7 Ë 14 - ÿπ π 2547 «.«29 The second variable is retained earnings/total assets ratio (RE/TA). Retained earnings is an accumulated firm performance over its entire life. In general a relative young firm will probably have lower RE/TA ratio due to less time spent for accumulating profits. The RE/TA ratio, on the other aspect, measures firmûs financial leverage (Altman.1983). Firms with lower RE/TA ratio tend to use external financing than that of higher ratio. Calculation shows only two companies can turn their retained earnings to be positive. Table 1 indicates that average number of retained earnings to total assets ratio is as low as -2.93 times. This means average cumulative loss of the REHABCO is almost 300 percent of amount of total assets, and the consequence of this figure reveals that these ailing firms finance its assets with external funds. On the earning power side, earnings before interest and taxes/total assets ratio measure firmûs pure operating performance in relation to itûs assets. This means no tax or leverage factors are involved. Calculation of the third variable reveals that more than half of the companies (28 firms) have positive earnings before interest and taxes. However, on average there are only 2 percent earning power of their total assets, and hence the variation is ranging from minus 66 percent to more than 200 percent. The forth variable, market value of equity/book value of total liabilities is more efficient than commonly used ratio, net worth/total debt (book value). The ratio adds market value of all shares of stock, preferred and common. Table 1 shows a high variation of the ratio, starting at lowest from as low as minus 26.4 times to 4.66 times, but average figure is minus 1.35 times. There are only eight out of fifty-one firms that have positive value of equity at market price. These can be interpreted as a bad prospect about future of the REHABCO. The result confirms Altman (2000) that the X 4 measurement of bankruptcy firm shows lower asset value when compared with liabilities. The last composition of Z-score is asset turnover ratio. Asset turnover ratio is commonly used in measuring firmûs performance. Average figure of this ratio is only 0.61 times, means that on average these ailing firms generate sales only 61 Baht from their asset of 100 Baht. The ratio not only measures the ability of firmûs asset to generate sales, but also reveals competency of management in dealing with industryûs competition. However, firms listed in the REHABCO came from various sectors (industries) and due to the scope of the study, we cannot conclude about the competitiveness of individual firms.
«.«30 ªï Ë 7 Ë 14 - ÿπ π 2547 Table 2 Allocation of overall score of the REHABCO Classification Number of companies Safe area 2 Grey area 0 Bankruptcy Zone 49 The overall index, as appeared in Table 2, show that only two companies are in the safe area (Z-score is higher than 2.675), and the others are still in bankruptcy zone. Theoretically, one may consider the two firms whose Z-score above the higher cutoff point will continue as a going concern basis. While, the others forty-nine companies, which have Z-score far below the higher cutoff point have a minimal chance to survive in the business. Limitations and conclusion This study assesses the Z-score model to evaluate financial health of listed companies in the REHABCO. From the last section, we can see that most of the firms in the REHABCO have poor performance and very weak financial position. The results also reveal that there are only two companies, out of fifty-one that have Z-score above the higher cutoff point that means they are in the safe zone while most of them, which have very low score are still struggling. However, during the year 2003 thirteen companies were returned to the former sectors. These healthier companies (in opinion of the Stock Exchange of Thailand) have met criteria* set by the Stock Exchange of Thailand. Moreover, the two companies in the safe area as mentioned by Z-score are still in the REHABCO as at the end of the year 2003. Many scholars (for example Hughes 1993 and Gregory 1995) have found that the Z- score is only guideline for bankruptcy prediction and recovery investigation and does not give a clear result. The main point is the model is based on historical data and may fail to recognize changes in management structure and competitive situation. Normally, when firms face financial distress, they have to do something that help improve their financial position and performance such as: dispose of property; merge with other firms; reduce research and development funds; issue new shares; negotiate with creditors; liquidate; and lay off (Aiyabei.2002). Another limitation of using the model is the calculation of the forth (X 4 ) variable when market value of equity is in red. In case of a company has a huge figure of debt, the fourth variable will increase. Consequently this will lead to higher overall score. On the other hand, when the company has lower leverage ratio, the X 4 * There are five criteria: First, the company must have a positive shareholdersûsquity. Second, have net profit from main business in three consecutive quarters. Third, complete over 75 percent of debt restructure. Fourth, have positive cash flow after having booked interest expense. Finally, show strong financial position and performance on continuous basis.
ªï Ë 7 Ë 14 - ÿπ π 2547 «.«31 will decrease and eventually lower the Z-score, which contradicts the objective of this model. Direction for further research Further research may be conducted widely in Thai capital market, not only companies listed in The Stock Exchange of Thailand but also in the Market for Alternative Investment, and companies that issued bonds and listed in the Thai Bond Dealing Centre. The purpose of the studies is not aimed at detecting recovery sign but for early warning and prevention of financial distress and itûs consequence, bankruptcy. References Altman, E. (1983) Corporate Financial Distress : A Complete Guide to Prediting, Avoiding, and Dealing with Bankruptcy. New York : John Wiley..(1993) Corporate Financial Distress and Bankruptcy : A Complete Guide to Predicting & Avoiding Distress and Profiting from Bankruptcy. New York : John Wiley..(2000) çpredicting Financial Distress of Companies : Revisitng The Z-score and ZETA Model,é Working Paper. New York : Stern School of Business, New York University. Aiyabei, J. (2002) çfinancial Distress: Theory, Measurement & Consequence,é East African Journal of Humanities & Science. 11(1), 1-6. Gregory, E. (1995) çz score: A Guide to Failure Prediction,é CPA Online Journal. [Online]. Available at : http://www.nysscpa.org/cpajournal. (Accessed in November 2003) Grice, J. (2000) çinvestor Return and Predictions of Corporate Failure,é TSU Business Review. Summer, 17-21..(2002) çreestimations of The Zmijewski and Ohlson Bankruptcy Prediction Models, TSU Business Symposium,é Present Paper. January 31, February 1-2, 2002, Troy State University. Hughes, S. (1993) çbankruptcy Prediction Models,é Credit Control. 14(11): 16-19. Keating, E. et al. (2003) çassessing Models of Financial Vulnerability for the Nonprofit Sector,é Working Paper, Kennady School of Government, Harvard University. Stock Exchange of Thailand. (2003) Procedures and Guidelines for Listed Companies Facing Possible Delisting and Being Subject to Preparing Rehabilitation Plans Dated March 28, 2003. Stock Exchange of Thailand. [Online]. Available at : http://www.set.or.th (Accessed in December 2003)