Limitation of financial health prediction in companies from post-communist countries

Csikosova, Adriana; Janoskova, Maria; Culkova, Katarina Article Limitation of financial health prediction in companies from post-communist countries Journal of Risk and Financial Management Provided in Cooperation with: MDPI – Multidisciplinary Digital Publishing Institute, Basel Suggested Citation: Csikosova, Adriana; Janoskova, Maria; Culkova, Katarina (2019) : Limitation of financial health prediction in companies from post-communist countries, Journal of Risk and Financial Management, ISSN 1911-8074, MDPI, Basel, Vol. 12, Iss. 1, pp. 1-14, https://doi.org/10.3390/jrfm12010015 This Version is available at: https://hdl.handle.net/10419/238938 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Journal of Risk and Financial Management Article Limitation of Financial Health Prediction in Companies from Post-Communist Countries Adriana Csikosova *, Maria Janoskova and Katarina Culkova Department of Earth sources Management, Faculty BERG, Technical University of Košice, 042 00 Košice, Slovakia; [email protected] (M.J.); [email protected] (K.C.) *Correspondence: [email protected]; Tel.: +421-556-022-929 Received: 29 November 2018; Accepted: 15 January 2019; Published: 18 January 2019   Abstract: The financial health of a company can be seen as the ability to maintain a balance against changing conditions in the environment and at the same time in relation to everyone participating in the business. In the evaluation of financial health and prediction of financial problems of the companies, various indexes are used that can serve as input for expert estimation or creation of various models using, for example, multi-dimensional statistical methods. The practical application of the proper method for evaluation of financial health has been analysed in post-communist countries, since they have common historic experiences and economic interests. During the research we followed up the following indexes: Altman model, Taffler model, Springate model, and the index IN, based on multi-dimensional discrimination analysis. From the research results there is obvious a necessity to combine available methods in post-communist countries and at least to eliminate their disadvantages partially. Experiences from prediction models have proved their relatively high prediction ability, but only in perfect conditions, which cannot be affirmed in post-communist countries. The task remains to modify existing indexes to concrete situations and problems of the individual industries in the chosen countries, which have unique conditions for business making. Keywords: managing of financial health; risk of bankruptcy; prediction methods; post-communist countries 1. Introduction Idea of financial health can be seen as the ability of the company to maintain a balance against changing conditions of the environment and at the same time in relation to everyone participating in the business. A financial healthy company means a company that maintains its existence and is able to evaluate invested capital to the measure that is demanded by shareholders. Results of financial analysis are different in companies from various sectors, since companies have different property and financial structures and also different structures of economic results. Financial health demands achievement of sufficient profit, as well as long-term liquidity. Bankruptcy means a situation when a given organization does not have the possibility to overcome bad financial health and this situation is in accordance with available legislation in a given country. Such a situation is considered by experts as a corporate failure or business failure. In the evaluation of financial health and the prediction of financial problems of the companies, various indexes are used that can serve as an input for expert estimation or creation of various models by using, for example multi-dimensional statistical methods. The main disadvantage of such approach based on chosen statistical methods is its limited time availability and the complexity of model modification in changed conditions, conditioned by the availability of input data. The next problem is the improper structure of the input file. We must be very careful during selection of a proper method, construction of a correspondent model and also during interpretation, since this can J. Risk Financial Manag. 2019,12, 15; doi:10.3390/jrfm12010015 www.mdpi.com/journal/jrfm J. Risk Financial Manag. 2019,12, 15 2 of 14 lead to considerable bad estimation of its predictive ability. But the advantage is its very good quality of prediction during the existence of a qualitative data file. Approaches that are based on calculation of the total score of the company according to its values of financial indexes have advantages, but also disadvantages. The main advantage is the simplicity of realization and simple interpretation. The disadvantage is the determination of strict boundaries for interpretation. This can lead to a ranking of companies with almost identical values of financial indexes for various groups. A higher disadvantage of the approach is that it does not consider in most cases possible relations among indexes that are evaluated individually. This means that both approaches should be combined, while both approaches can be used and their disadvantages can be partially eliminated. Our findings show that prediction indexes in individual sectors in post-communist countries record different results. Due to the poor index of payment disability, which evaluates financial health of the companies in V4, all models should be modified by this index. The findings encourage an evaluation of the accuracy of bankruptcy prediction models by examining a large sample of companies and evaluating the real benefits obtained from the acquired information. Based on the finding we can state that before selection of the method for prediction of financial health, the economic conditions of an analysed subject must be considered. The findings could serve as a base for further research in other economic spaces. 2. Literature Review Classical statistical methods were used for a long time for development of single dimensional discrimination analysis. The most used statistic method is the multi-dimensional discrimination analysis, followed up by logit analysis (Altman and Saunders 1997). The next classification methods are risk index models, probit analysis and linear probability models. Beaver (1966) was one of the pioneers of models for bankruptcy prediction, based on financial rate indexes. He was the first author that applied a single dimensional model to results of financial rate indexes and compared individual results at prospering and non-prospering companies with a goal to define a model for bankruptcy prediction. The result of this method is often different from the practice, when a number of rate indexes has non-linear dependence on bankruptcy status (Keasey and Watson 1991). While using single dimensional model rate indexes, resulting from financial accounting, the evaluation of importance of one concrete index individually is very difficult, since a majority of indexes are connected. The single dimensional model is different from reality, when the financial situation of the company is seen as one unit, which cannot be evaluated only according one rate index. In reaction to Beaver (1966), Tamari (1966) understood that Beaver’s analysis is not definite, since according to one index a company could be evaluated as prospering and according to another index a company could be ranked among non-prospering companies. Therefore, Tamari created so-called risk index model that presents very simple point system, including various rate indexes, which are generally accepted as indexes of the financial health of the company. The advantage of the risk index model is its intuitiveness and easy applicability, but on the other hand it can be considered also as its main disadvantage, since it means an index with rather subjective characteristics, due to for example subjective determination of individual weights in the Tamari index. In 1968, Altman (1968) applied a technique of multi-dimensional statistical analysis (MDA) in connection with a prediction of bankruptcy and created a model, called the Z-score model. This method presents multi-dimensional discrimination analysis. Eisenbeis (1977) determined a final Z-score model, regarding also new standards of financial reporting for areas of business. Several authors emphasized the importance of first two limited assumptions and their possible mistakes, majority of studies about bankruptcy did not try to analyse if used data filled these assumptions. In practice, used data only rarely fill all three assumptions. MDA models are very often applied by improper way and conclusions are disputable. The first assumption of multivariate normally distributed (MND) models is often not observed (Deakin 1976). This can have the consequence of a bad picture of importance of tests and J. Risk Financial Manag. 2019,12, 15 3 of 14 estimated measure of mistakes (McLeay and Omar 2000). It is necessary to mention that at the normal situation MND demands univariate normally distributed models (UND). Due to the mentioned some researchers test values only in UND conditions. Some authors amend UND and try to make UND transformation of indexes before their including to the model. Taffler (1983) and Altman et al. (1977) adapted values that do not correspond with UND through the transformation. A second assumption that must be tested before model development, based on MDA, is the assumption of dispersion matrixes equality. In case the assumption is violated, it tests importance of differences in variables between declined and prospering group of companies will be influenced. Third assumption mentions that during optimal score selection, deciding about the group, probability of mistaken evaluation should be considered in combination with costs of mistake type I and II (Zavgren 1983). To date, when MDA where clear dominant method for models creation, this method has been replaced by less demanded statistic techniques, as for example logit analysis (LA), probit analysis (PA) and models of linear probability (LPM). By these methods were created evaluation models of conditioned probability (CPM) (Zavgren 1983). These models are constructed from a combination of variables that make the difference between group of prospering and nonprospering companies. Ohlson (1980) used LA to create his models. Zmijewski (1984) was on the other hand orientated to PA. Since then LA has been considered the most favourable method for bankruptcy prediction. Number of studies, using PA is less, since in comparing with LA it demands bigger number of calculations. CPM enables to estimate a probability of the unsuccessful situation in dependence on raw of company characteristics, mainly by non-linear estimation of maximal probability. Models are based on a certain assumption of probability distribution. Models based on LA, assume logarithmic distribution of probability (Maddala 1977). In models of linear probability, it is assumed that the relation between individual variables of the given model and probability of bankruptcy is linear. LA is in literature of company prediction the method of conditioned probability most used. 3. Materials and Methods The use of bankruptcy indexes had been analysed in post-communist countries, since those countries have not only common historic experiences with communism, but also mentality, cultural background, traditions, as well as strategic economic interests and complex reforms and transitions to the market economy. We orientated the research mainly to the V4 group—Czech Republic, Slovakia, Hungary and Poland—since the situation in V4 was very similar also after the transformation to the market economy. V4 countries have very successful economic situation; all countries are growing rapidly than average in EU. But there is threat that after achievement of certain level the growth will be stopped (Onaran 2011). The main object of searching was chosen companies in individual industrial sectors from V4 countries. 3.1. The Data The data for prediction of financial situation in chosen industrial companies was obtained as aggregate data available at the Ministry of Industry and Trade of the Czech Republic (Ministry of Industry and Trade CZ 2017) and Ministry of Economy in Slovakia (Ministry of Economy SR 2017), Hungarian Central Statistical Office (Hungarian Central Statistical Office 2017) and Ministry of Entrepreneurship and Technology Poland (Ministry of Entrepreneurship and Technology Poland 2017). Other necessary information to solve the research problem was data from the Register of financial statement (Register of Financial Statement SR 2017) and individual web sites of companies. Annual financial statements of 2007–2017 of investigated companies and industries had been analysed. Set of 30 financial statements had been acquired from companies that present representatives in individual industries. 3.2. Methods Due to the research of prediction indexes used in post-communist countries we followed up the following indexes: Altman model, Taffler model, Springate model and Index IN, based on J. Risk Financial Manag. 2019,12, 15 4 of 14 multi-dimensional discrimination analysis. We decided to follow up these indexes since the research of Kanapickiene and Kanapickiene and Marcinkevicius (2014) proved that the Taffler and Altman’s Z” Score Model for emerging countries models are the least accurate. In multi-dimensional discrimination analysis, the financial situation of the company is predicted by various combinations of simple characteristics, which means by certain files of various indexes, to which various weights are given. Their task is to predict the financial situation through achieved results and with correspondent reliability to rank the company among prospering or non-prospering companies. 3.3. Altman Model Altman improved the Beaver (1966) univariate method by establishment of a multi variant approach that reflects the financial situation of the company better. Altman analysed 66 companies and selected two groups of companies—one group before bankruptcy and a second group of excellent companies. By multivariable discrimination analysis he created weight of individual indexes and determined values for companies ranking to three groups. By this way he predicts future financial development of the company and the possibility of bankruptcy. Altman found out that following indexes reflect the best financial situation and its future development (Altman 2000). By this way constructed discrimination function for the company in the following equation: Z = 0.012X1+ 0.014X2+ 0.033X3+ 0.006X4+ 0.999X5(1) where: X1= working capital/total assets, X2= undivided profit/total assets, X3= earnings before interest and taxes (EBIT)/total assets, X4= market value of equity/debts, X5= sales/total assets. 3.4. Index IN Index IN presents next possibility of how to evaluate financial health of the company by bonity and bankruptcy models (Neumaierováand Neumaier 2002). By these indexes, we can determine with certain probability if acompany belongs among bonity or bankrupting companies, or if it is able to create value for its owners. Index IN is modified to several types according to time of their rising: index IN95, IN99, IN01, IN05, when Neumaier and Neumaierová(2005) are their authors. Index IN95 arose in 1995 according to data obtained in 1994. Data were obtained and elaborated for the industry as a whole, as well as for its individual sectors. Success of the index is over 70%. The index is created by six indexes with correspondent weights: IN95 = 0.22X1+ 0.11X2+ 8.33X3+ 0.52X4+ 0.10X5−16.8 X6(2) where: X1= assets/debts, X2= EBIT/interest costs, X3= EBIT/assets, X4= revenues/assets, X5= floating assets/(short term liabilities + short term bank credits), X6= unpaid liabilities/revenues. Intervals of index IN95: IN95 < 1 company in financial difficulty, J. Risk Financial Manag. 2019,12, 15 5 of 14 IN95 = 1–2 grey zone, IN95 > 2 company without financial problems. The given index is orientated to the ability of the company to pay its liabilities; it did not deal with the demands of owners for value creation. This regards the following index IN99. It reflects the demand of the owner; therefore, weights in IN95 are changed. Index IN99 is recommended to use in cases when it is not possible to state costs of equity for calculation of EVA index (economic value added). Success of the IN99 index is estimated at level over 85%. 3.5. Index Bonity B The financial health of a company can be evaluated also from the view of financial management quality. This task can be fulfilled by index B, evaluating the bonity of the company, provided by qualified financial management. B = 1.5X1+ 0.08X2+10X3+ 5X4+ 0.3X5+ 0.1X6(3) where: X1= cash flow/debts, X2= total capital/debts, X3= earnings before taxes (EBT)/total capital, X4= EBT/total revenues, X5= stocks/total assets, X6= total revenues/total capital. Evaluation scale: positive value means positive and healthy situation of the company. Negative values mean a negative and unhealthy situation, the lower the value, the worse the situation of the company. 4. Results Using prediction indexes in company from post-communist countries show different results, as illustrated by Table 1. Table 1. The example of prediction indexes using in company from post-communist country. Method Used Data Obtained Evaluation IN test 0.852 Rather not creating value Z-score 1.5456 Bankruptcy threat Index B 2.05348 Healthy situation Overall evaluation 1. Profitable company 2. Profitability is not sufficient 3. Necessity to decrease debts In spite of the positive bonity index, which means situation in the company is healthy, management of cash-flow can be evaluated positively, Z-score shows possible threat by bankruptcy, which demands a need for a detailed analysis, especially an analysis of debt management, etc. It is demanded also due to the unstable situation, given by IN test, which speak company does not create value. During the detailed analysis, we found the reason for such situation is high capital costs, which can be solved through improving of financial structure, using of debt management tools, decreasing of debts, etc. Prediction indexes of bonity in individual sectors in Slovakia recorded the following results given by Table 2. J. Risk Financial Manag. 2019,12, 15 6 of 14 Table 2. The Index B in Slovakian industrial sectors. Negative Score Positive Score Production of office equipment +0.978 Mining of iron ores −1.8 Production of construction materials +0.733 Ship construction −0.689 Beverage production +0.467 Steel production −0.372 Other mining +0.575 Automotive production −0.005 Production of dairy products +0.558 Due to the determined differences in bankruptcy indexes using in post-communist countries we found a necessity to provide during prediction following: •To use for evaluation more than one index and to compare the results. •To follow up indexes in time development. • Further important fact that is necessary to consider during using of the indexes is working with data that are calculated in time horizon—one year, which means expression of performance per year. •After finding a possible bankruptcy, to undertake a deeper analysis. • To consider the undeveloped capital market in post-communist countries due to its short history existence. • To count legislative restrictions, which can be overcome by common IFRS (international financial reporting standards), which due to the transition of several post-communist countries to the European Union (EU) (as for example Slovakia) are considered. • To overcome insufficient preparation of financial managers to new models for financial health prediction. • To make a comparison between the micro-economic and macro-economic environment of the company, since a possible bankruptcy is caused by internal, as well as external factors of the company. • To count on shortages of existing methods, for example in post-communist countries due to the undeveloped capital market there is sometimes difficult to state capital values of calculated indexes and values are considered only as accounting values or values are determined by estimation. • To consider new methods and their modification to concrete conditions, since most indexes for bankruptcy prediction are not created for conditions of companies from post-communist countries and must serve only as approximate orientation for prediction. Indexes must serve only as inspiration and financial analytics should make own indexes for post-communist countries. The research of previous studies showed number of post-communist countries deals with the problem of payment ability in industries. The indexes were therefore modified in 1995, 1999 and 2001 to such problems, mainly to indexes, mentioned in following Figure 1. J. Risk Financial Manag. 2019, 12, x FOR PEER REVIEW 6 of 14 Mining of iron ores −1.8 Production of construction materials +0.733 Ship construction −0.689 Beverage production +0.467 Steel production −0.372 Other mining +0.575 Automotive production −0.005 Production of dairy products +0.558 Due to the determined differences in bankruptcy indexes using in post-communist countries we found a necessity to provide during prediction following: • To use for evaluation more than one index and to compare the results. • To follow up indexes in time development. • Further important fact that is necessary to consider during using of the indexes is working with data that are calculated in time horizon—one year, which means expression of performance per year. • After finding a possible bankruptcy, to undertake a deeper analysis. • To consider the undeveloped capital market in post-communist countries due to its short history existence. • To count legislative restrictions, which can be overcome by common IFRS (international financial reporting standards), which due to the transition of several post-communist countries to the European Union (EU) (as for example Slovakia) are considered. • To overcome insufficient preparation of financial managers to new models for financial health prediction. • To make a comparison between the micro-economic and macro-economic environment of the company, since a possible bankruptcy is caused by internal, as well as external factors of the company. • To count on shortages of existing methods, for example in post-communist countries due to the undeveloped capital market there is sometimes difficult to state capital values of calculated indexes and values are considered only as accounting values or values are determined by estimation. • To consider new methods and their modification to concrete conditions, since most indexes for bankruptcy prediction are not created for conditions of companies from post-communist countries and must serve only as approximate orientation for prediction. Indexes must serve only as inspiration and financial analytics should make own indexes for post-communist countries. The research of previous studies showed number of post-communist countries deals with the problem of payment ability in industries. The indexes were therefore modified in 1995, 1999 and 2001 to such problems, mainly to indexes, mentioned in following Figure 1. Figure 1. Modification of IN index in post-communist countries. In 2000, the authors of IN95 and IN99 decided to construct the index that could connect the characteristics of both previous indexes and in this way evaluated the ability of the company to paid debts and at the same time to create value for owners. By discrimination analysis authors came to IN01, applied for industrial companies. While the IN test is from the view of the owner, the IN95 index for the creditor evaluates mainly a rating of the company. Figure 1. Modification of IN index in post-communist countries. In 2000, the authors of IN95 and IN99 decided to construct the index that could connect the characteristics of both previous indexes and in this way evaluated the ability of the company to paid J. Risk Financial Manag. 2019,12, 15 7 of 14 debts and at the same time to create value for owners. By discrimination analysis authors came to IN01, applied for industrial companies. While the IN test is from the view of the owner, the IN95 index for the creditor evaluates mainly a rating of the company. IN95 = 0.22X1+ 0.11X2+ 8.33X3+ 0.52X4+ 0.10X5−16.8X6(4) where: X1= assets/total capital, X2= EBIT/interest costs, X3= EBIT/assets, X4= revenues/assets, X5= current assets/(current liabilities + current bank credits), X6= liabilities after maturity/revenues. Evaluation scale for IN95 is following: IN95 > 2 financial healthy company, company is able to pay its liabilities, IN95 < 1 company has financial hardship, IN95 < 1–2 > financial health cannot be clearly evaluated. 4.1. Modification to Index IN99 With the aim of evaluating the financial ability of the company, not only the creation of value for the creditor but also for the owner, IN95 had been modified to IN99 by original authors, constructed according to data from 1698 post-communist companies. The calculation of the index is made by the following equation: IN99 = −0.017X1+ 4.573X2+ 0.481X3+ 0.015X4(5) Interpretation of IN99 is following: IN99 > 2.070 company creates value (achieves net profit), IN99 < 0.684 company creates negative value of net economic profit, IN99 < 0.684–2.070 > creation of value cannot be clearly determined, 4.2. Modification to Index IN01 Next variant that connect both previous indexes (IN95 and IN99) presents IN01, determined for industrial sectors. Its calculation is as follows: IN01 = 0.13X1+ 0.04X2+ 3.92X3+ 0.21X4+ 0.09X5(6) where: X1= total capital/debts, X2= EBIT/interest expenses, X3= EBIT/total capital, X4= revenues/total capital, X5= current assets/short term liabilities (in broader sense). Consequently, the interpretation of model results is given by following scale: IN01 > 1.77 company creates the value, IN01 < 0.75 company tends to bankruptcy, IN01 < 0.75–1.77 > future of the company is uncertain. J. Risk Financial Manag. 2019,12, 15 8 of 14 Considering the payment disability of companies in post-communist countries is made also in Altman, the Z-score was modified by Neumaier and Neumaierová(1995) to the following equation: ZMOD = 1.2X1+ 1.4X2+ 3.3X3+ 0.6X4+ 1.0X5+ 1.0X6(7) when: X6= overdue liabilities/revenues. At the same time, special conditions for business in companies from post-communist countries demand consideration of different weights during the indexes calculation. Table 3gives an illustration of the weights for chosen industries in Slovakia and the Czech Republic. Table 3. Weights of indexes in industries. Industrial Sector V1V3V4V6 Agriculture 0.24 21.35 0.76 −14.57 Fishery 0.05 10.76 0.90 −84.11 Raw material mining 0.14 17.74 0.72 −16.89 Mining of energy sources 0.14 21.83 0.74 −16.31 Mining of other sources 0.16 5.39 0.56 −25.39 Processing industry 0.24 7.61 0.48 −11.92 Grocery industry 0.26 4.99 0.33 −17.36 Textile and clothing industry 0.23 6.08 0.43 −8.79 Leatherworking industry 0.24 7.95 0.43 −8.79 Wood industry 0.24 18.73 0.41 −11.57 Paper and printing industry 0.23 6.07 0.44 −16.99 Coke ovens and refineries 0.19 4.09 0.32 −20.26 Production of chemical products 0.21 4.81 0.57 −93.0 Rubber industry and plastics production 0.22 5.87 0.38 −17.06 Construction materials 0.20 5.28 0.55 −43.01 Production of metals 0.24 10.55 0.46 −9.74 Machinery 0.28 13.07 0.64 −6.36 Electro technique and electronics 0.27 9.50 0.51 −8.27 Production of transport vehicles 0.23 29.29 0.71 −7.46 Other industries 0.26 3.91 0.38 −17.62 Electricity, water, gas 0.15 4.61 0.72 −55.89 Construction 0.33 9.70 0.28 −28.32 Business and repair of automotive 0.33 9.70 0.28 −28.32 Catering and accommodation 0.35 12.57 0.88 −15.97 Transport and communication 0.07 14.35 0.75 −60.61 Economy of Slovakia 0.22 8.33 0.52 −16.80 Next, modification of the index IN presents the index IN05 that enables us to reach complex conclusions about performance of the company. The index consists of five indicators. Two of them characterize the ability of the company to create profit, and two of them characterize EBIT calculation. Weights, added to individual indexes, had been given by authors Neumaierová(2005) by discrimination analysis. IN05 = 0.13X1+ 0.04X2+ 3.97X3+ 0.21X4+ 0.09X5(8) where: X1= assets/debts, X2= EBIT/interest costs, X3= EBIT/assets, X4= revenues/assets, X5= floating assets/short term debts.