CC BY
37
ORIGINAL PAPER
DOI: 10.26794/2587-5671-2023-27-2-192-202 JEL G10, M40, M41
(CO ]
Application of the Altman Z'' Score Model in Forecasting the Financial Position of BIST Companies
i. E. Gokturk, H. S. Yalqinkaya
Necmettin Erbakan University, Konya, Turkiye
ABSTRACT
For measuring financial performances of companies and identifying financial failure, there are a lot of models in the literature. Among these models, Z Score model is of the most used in terms of its being an accounting-based model and simple applicability. The purpose of this paper is found out whether the Z" Score model, which was revised by Altman, could be useful in making financial decisions about long-term firm value. For this purpose, panel cointegration analyzes were carried out among the variables, with the firm values of the publicly traded companies listed on the Turkish BIST (Istanbul Stock Exchange) as the independent variable and the Z" Score values as the dependent variable. Although the research is specific to Turkey, the results of the research are considered to be applicable globally, as Altman states that the Z" Score model can also be used by developing country companies. It has been proven that Altman Z" Score Model, applied in public company, has a high prediction power directed to financial success of the firms. According to the results of the analysis, 1 unit increase in the Z" Score values of the companies cause an increase of 0.353 units in the logarithmic return calculated over the firm value. Z" Score Model can be a precious indicator for heads of companies, accounting and financial managers, auditors, creditors, investors to make accurate decisions directed to assessing financial structures of companies in advance. Keywords: Accounting based prediction model; financial distress; Altman Z-Score; BIST
For citation: Gokturk i. E., Yalqinkaya H. S. Application of the Altman Z'' score model in forecasting the financial position of BIST companies. Finance: Theory and Practice. 2023;27(2):192-202. DOI: 10.26794/2587-5671-2023-27-2-192-202
ОРИГИНАЛЬНАЯ СТАТЬЯ
Использование модели Альтмана Z" при прогнозировании финансового положения компаний, зарегистрированных на турецкой фондовой бирже
И. Е. Гёктюрк, Х. С. Ялчинкайя
Университет Некметтина Эрбакана, Конья, Турция
АННОТАЦИЯ
Для измерения финансовых показателей компаний и выявления их финансового кризиса в научной литературе существует множество моделей. Среди них модель Z" Score является одной из наиболее используемых, поскольку она основана на бухгалтерском учете и проста в применении. Цель данного исследования - выяснить, применима ли модель Z" Score, усовершенствованная Альтманом, для оценки стоимости фирмы в долгосрочной перспективе. Проведен панельный анализ коинтеграции между переменными, где в качестве независимой переменной выступала стоимость компании, акции которой котируются на турецкой фондовой бирже BIST (istanbuL stock exchange), а в качестве зависимой переменной - Z" Score. Несмотря на то, что исследование проводилось конкретно в Турции, его результаты считаются применимыми во всем мире, так как Альтман утверждает, что модель Z" Score может также использоваться компаниями из развивающихся стран. Доказано, что модель ALtman Z" Score, примененная в отношении публичной компании, имеет высокую способность прогнозировать финансовый успех фирм. Согласно результатам анализа увеличение на 1 единицу значения Z" баллов компаний приводит к увеличению на 0,353 единицы логарифмического дохода, рассчитанного по стоимости фирмы. Модель Z Score может быть полезна для руководителей компаний, бухгалтерских и финансовых менеджеров, аудиторов, кредиторов, инвесторов при принятии верных решений, связанных с предварительной оценкой финансовых показателей компаний.
Ключевые слова: модель прогнозирования на основе бухгалтерского учета; финансовый кризис; Z-балл Альтмана; BIST
Для цитирования: Göktürk i. E., Yalqinkaya H. S. Application of the Altman Z'' score model in forecasting the financial position of BIST companies. Финансы: теория и практика. 2023;27(2):192-202. DOI: 10.26794/2587-5671-2023-27-2-192-202
© Göktürk i. E., Yalçinkaya H. S., 2023
BY 4.0
introduction
The firms, which cannot perform the requirements of market conditions, due to the fact that economic and financial troubles they face of, experience difficulties in sustaining their lives. The studies show that these problems that we can refer to business failure arise from bad management of firms [1].
Although there are many reasons for business failure, these reasons were classified in the literature as economic and financial failure in two main areas [2]. Economic failure emerges in case those firms cannot obtain income as much as it can meet costs forming in activity process [3, 4]. Economic failure may not create any problem for the business, which has a strong structure and paying power for their debts. Also, after the income generating activities, business can reach its profit targets in time.
Financial failure emerges in case of not being able to pay for debts in their due dates [5]. Financial failure is the insufficiency of the acts made directed to identifying financial performances of the firms [6-8]. Financial performance can be expressed as activity abilities of business managers for asset management and control [9]. Being able to take action by the managers can be possible by predicting financial failure in advance.
The use of an accounting-based structure in measuring fiscal performance and analysis of fiscal tables obtained from accounting data contribute performance measurement. The ratios that will be obtained from financial tables will provide information about the existing situation and future of the business. The analyses to be made will provide critical information for many parts such as company managers, investors, creditors and government while identifying possible problems to emerge in financial structures of companies provides for the in-business users to take necessary precautions from financial point of view, in viewpoint of out-business parts, it will be a guide in arranging the relationships with company [10].
Not being able to identify the problems occurring in financial structures of businesses and not being able to take the necessary actions in time are generally ended with bankrupt, and this case causes the formation of serious costs negatively affecting both companies and all stakeholders related to the company [1, 11, 12]. In these stages, taking businesses to bankrupt, using certain methods for predicting failure will enable to take action with lower cost without facing to heavy cost of bankrupt [13, 14].
It is known that many prediction methods were developed for identifying the problems in financial structures of the businesses in advance [15-21].
One of the most encountered methods in the literature for testing financial structure and identifying bankrupt risk is Altman Z Score model. Besides that, Altman Z Score model is used in the areas such as merging and acquisition, credit risk analysis and return methods as danger measurement, this model has been begun to be used for the purpose of measuring performance [22]. The Altman Z Score model is a useful model not only for predicting bankruptcy, but also for measuring financial performance [23].
The most interesting study for identifying financial performances and successes via proportioning the variables taking place in financial tables was carried out by Beaver in 1966. In this study, used a univariate analysis as a traditional method [24]. After this study, Altman, using Multiple Discriminant Analysis (MDA), developed a new model in 1968. This model was broadly accepted besides that it had a high accuracy level, due to the fact that it could be used in assessing financial performances of firm managers, analyses of management accounting and presenting predictions in identification of fiscal structures as well as decision making processes of the various stakeholders such as investors, creditors, auditors, consultants [25, 26].
The study was divided into three sections, including introduction and conclusion sections. The first section theoretically describes Altman Z Score model and, the second section includes empirical examination of the literature about this method. The third section includes database and study method and, the last section presents a discussion about Turkey-specific results and their effects.
altman z score model
in literature
Z Score Model, which is the first multiple variable model for measuring performances of financial structures and identifying financial failure, was developed by Altman Edward in 1968 [27]. Easily application of the method through accounting data and its success on performance results directed to identifying financial failure enabled the model to be used in a wide area and to be acceptable [28, 29].
The first Altman Score Model is based on the assumption that there is a linear relationship between financial failure and the ratios obtained from financial tables [30]. In the study, using 22 financial ratios,
calculated through the values obtained from financial tables, were used. In the direction of the results obtained, 5 financial ratios, which is accepted that they identify financial failure, were reached [31].
Financial rates, accepted that they identify failure in Altman's original model, are [6]:
X1: Working capital/Total Assets;
X2: Retained earnings/Total assets;
X3: Earnings before interest and taxes/total assets;
X4: Market value of equity/Book value of total debt;
X5: Sales/Total assets.
Discriminant function, first developed by Altman and called Z Score, is as follows [15]:
Z = 0.012 (X1) + 0.014 (X2) + 0.033 (X3) + 0.006 X4 + + 0.999 (X5).
The possible results of Z Score are determined according to the flowing limit values [15]:
• Z ^ 1.80: High Risk (Distress Zone);
• 1.81 ^ Z ^ 2.99: Uncertain (Grey Zone);
• Z > 2.99: Low risk (Safe Zone).
It was identified by the researcher that the first model formed could predict accurately the financial successfulness of unsuccessfulness of the firms in the rate of 95% with the one year ago data and accurately in the rate of 72% with two years ago data [15].
The first Altman Z Score model was formed for public businesses. However, financial structures of non-public businesses differ from financial structures of public businesses. Especially, due to weakness in cash flows of non-public businesses, they face to more financial problems [32]. Altman noticing this problem updated model in the following years.
Discriminant function revised by Altman is as follows [33]:
Z': 0.717 * X1 + 0.847 * X2 + 3.107 * X3 + 0.420 * X4 + 0.998 * X5
Xl = working capital / total assets;
X2 = retained earnings/ total assets;
X3 = earnings before interest and taxes / total assets;
X4 = book value of equity/ book value of total liabilities;
X5 = sales/ total assets.
After this modification, Z Score classification areas were also again determined,
• Z' < 1.23: High Risk (Distress Zone);
• 1.23 ^ Z' ^ 2.90: Uncertain (Grey Zone);
• Z' > 2.90: Low risk (Safe Zone).
Altman, for Z Score model to be able to be used except manufacturing businesses, revised the model once more. Altman, eliminating the variable of X5 sales/total assets, formed four variable Z" score model.
Four variable Z" Score model, is as follows [17, 33]: Z" = 6.5 6 (XI) + 3.2 6 (X2) + 6.72 (X3) + 1.05 (X4). After this modification, Z Score classification areas were also again determined. According to this: Z" Score < 1,10: High Risk (Distress Zone) 1.10 ^ Z" Score ^ 2.60: Uncertain (Grey Zone) Z" Score > 2.60: Low risk (Safe Zone) Altman theory, for distinguishing whether or not quantitative models such as Multiple Discriminant Analysis (MDA) will show sufficient performance from financial point of view, is important in terms of that it shows that financial ratios can be used. Altman Z Score model, which has a scientific support and uses accounting data, is also continuously used at the present time by market markers both in academic studies and as an indicator of basic analysis.
Also, in this study carried out, it was focused on Altman Z" Score model, revised by Altman, that is highly acceptable in the literature [2, 15-17, 20, 24, 25, 33, 35-43].
methodology
The Aim of the Study and Database
The last revised Altman Z" score is a model that can be applied in developing countries [44]. In the study, Altman Z" score success has been tested by using the financial statement data of publicly traded companies in Turkey, which is the developing country model.
When Altman Z Score coefficients are examined, Z" < 1.1 shows high bankrupt risk; 1.1 <= Z" < 2.60, grey zone and Z" >= 2.60, good condition of the firm [45, 46]. If there is a relationship between Z'' Score values and financial performance of firms, this relationship will have to effect market value of firms. In order to test hypothesis formed, Z" Score results of the firms recorded in Istanbul Stock Exchange (BIST) and market values of these firms were used in the study. In the hypothesis established, all companies registered in BIST were evaluated to obtain the most accurate result, and financial sector representatives from these companies were excluded because they had different balance sheet structures. By examining the data continuity of the companies registered in BIST, the longest possible period for the research was determined. According to this examination, it was decided that the most suitable study period for the study was 38 periods between 2nd quarter (six months) of 2012 and 3rd quarter (9 months) of 2021. As a result of all of these examinations, 111 firms and quarterly financial tables of 38 periods of these firms and market values of the firms were included as dataset in the study
Table 1
BIST Firms Being the Study Subject
ADEL ARMDA BTCIM DOKTA GUBRF KRSTL PETKM TKFEN
AEFES ARSAN BUCIM ECILC HATEK KRTEK PETUN TOASO
AFYON ASELS CCOLA EGGUB HEKTS KUTPO PINSU TTKOM
AKCNS ATEKS CELHA EGPRO INDES LINK PKART TTRAK
AKENR AVTUR CEMTS EGSER IPEKE LKMNH PNSUT TUPRS
AKSA AYGAZ CIMSA ENKAI KAREL MNDRS PRKAB ULKER
AKSEN BAGFS CLEBI ERBOS KARTN MRSHL SARKY USAK
ALARK BAKAB CMENT EREGL KENT NETAS SELEC UTPYA
ALCAR BIMAS DERIM ESCOM KLMSN NUHCM SISE VESBE
ALCTL BIZIM DESA ETYAT KNFRT OLMK TATGD VESTL
ALKIM BLCYT DESPC FROTO KONYA ORGE TBORG VKING
ALMAD BRISA DGATE GENTS KOZAA OTKAR TCELL YAPRK
ANELE BRSAN DOAS GOLTS KOZAL OYAKC TGSAS YATAS
ARCLK BSOKE DOBUR GOODY KRONT PARSN THYAO
Source: Compiled by the author.
carried out. The firms being subject of the study were shown in Table 1.
Z" Score values of the companies were calculated as shown in the formula (1).
Z" Score = 6.56X1 + 3.26X2 + 6.72X3 + 1.05X4 (1)
Firm value data, which is the independent variable of the research, is calculated from the market values of the companies shown in Table 1, reported in BIST. In the calculation of the firm value variable, the logarithmic return was calculated over the 3-month market values of the companies included in the analysis.
Ln (return) = ln (P+/P) (2)
The purpose of using logarithmic return value instead of using firm value, is to avoid the stationary problem that may arise in the time series. As a result, the increase or decrease in the logarithmic return will not harm the basic question of the research, since it directly depends on the increase or decrease in the value of the firm.
In the dataset formed, due to the fact that it contains both time and cross-sectional vales, dataset turned into panel data set in the scale of 38*111.Although that analyses are made on the
panel datasets is partly similar to time series, it also contains many differences. The most important one of these differences and element that is necessary to be studied is horizontal cross-sectional dependence. Other than cross-sectional dependence, additionally, studying homogeneity on dataset is highly important for panel data analysis. Depending on analyzing these two elements, the analyses to be made will differ.
horizontal cross-sectional dependence
Cross-sectional data exhibit a behavior in the direction of commonly moving, the correlation appears between cross-sectional data, and this state refers to horizontal cross-sectional dependence. Between cross-sectional data, the results of the analyses to be made without considering horizontal cross-sectional study are relatively different and can be misleading for the researcher [47-49].
For studying horizontal cross-sectional dependence, a lot of models were developed. These models are LM test [47], CD test [48] and NLM test [50]. The method to be used to select the most suitable one among these tests is basically is: if time series (T) is bigger than cross-sectional series (N), LM test should be chosen; if N > T, CD test; and if N and T reach big values, NLM test; CD calculates correlation between the residuals obtained as a result
Table 2
The Results of Horizontal Cross-Sectional Dependence Test
Test Ln (return) Z'' Score
Statistic d.f. Prob. Statistic d.f. Prob.
LM 229992.0 6216 0.00 242424.0 6216 0.00
CD 4.795,748 0.00 4.923.657 0.00
NLM 2.005,975 0.00 2.117.474 0.00
Source: Compiled by the author.
of ADF (Augmented Dickey-Fuller) in horizontal cross-sectional dependence test [51].
For being able to be calculated Pesaran CD test,
CD =
2T
N-1 N
\
P] = pn =
N (N -1)
r<
¿-It=1
IIP,
i]
\ i=1 ]■=i+1 J
T
e«efi
i \V i \ 1
(IL f¡ )" № ) '
(3)
(4)
e., and e.. are the residuals obtained as a result of
it ¡1
regression, represent correlation and p(y i and j. The hypotheses formed for the test made are [48]: H0: There is no correlation between cross-sections; Hj: There is a correlation between cross-sections. In this study carried out, in the size of N = 111 and T = 38, there are two variables as Z" and Ln (return). The results of Pesaran CD test and the other horizontal cross-sectional dependence test are presented in Table 2.
When the test results of three horizontal cross-sectional dependence, made on the variables of Z" Score and Ln (return) that are the subject of the study, are examined, depending on the result of p < 0.05 for each test and variable, H1 is accepted, while H0 is rejected. According to these acceptations, it is reached the conclusion that the variables have correlation on their own cross-sections and cross-sectional dependence cannot be rejected.
homogeneity test
If a variation in one of cross-sectional variables also shows similar effects on the other cross-sectional variables, we say that panel data structure is homogenous, otherwise, that it is heterogeneous [52]. In addition, whether or not panel data is
homogenous play's important role in the preference of the analyses that will be made later.
In this study made, using Hsiao C. [53] homogeneity test, homogeneity of panel-data set was tested. In Hsiao Homogeneity Test, three hypotheses are formed as H1, H2 and H3 are formed. While H1 and H2 hypotheses accept that model coefficients are homogeneous, H3 hypothesis accepts that these coefficients are partly homogeneous [54]. The results of Hsiao Homogeneity Test are shown in Table 3.
In the results of Hsiao Homogeneity Test, since p < 0.05, homogeneity of panel-data set cannot be rejected. According to the results of horizontal cross-sectional dependence and homogeneity tests, it will be decided which test can be used for unit root tests, the next step.
In case that there is no horizontal sectional dependence, while 1st generation unit root tests can be applied, in case that there is horizontal cross-sectional dependence, 1st generation unit root tests give misleading results. In case that there is horizontal cross-sectional dependence, MADF [55], SURADF [56] and CADF (CIPS) [57] tests are recommended, which are among 2nd generation unit root tests. However, from among these tests, in the cases of N > T and T < N, CIPS test gives the most reliable results [58]. Also, in this study carried out, CPS unit root test was chosen, depending on specified reasons.
CADF test, developed by Pesaran, was developed in terms of cross-section and while both T > N and e N > T, it gives reliable results in panel data analyses under horizontal sectional dependence. The other feature of this test is that it is a heterogeneous test. Pesaran CADF is based on the model given below [58]. In case that there is no autocorrelation, panel data model is the same as that shown in formula (5).
Table 3
The Results of Hsiao Homogeneity Test
H1 = Null Hypothesis: panel is homogeneous Alternative Hyp.: H2
H2 = Null Hypothesis: panel is heterogeneous Alternative Hyp.: H3
H3 = Null Hypothesis: panel is homogeneous Alternative Hyp.: panel is partially homogeneous
Hypothesis F-Stat P-Value
H1 0.00 1.00
H2 0.00 1.00
H3 0.00 1.00
Source: Compiled by the author.
Table 4
Pesaran CIPS Unit Root Test Results
Z" CIPS* = -2.219 N,T = (111.38) Ln (return) CIPS* = -4.730 N,T = (111,38)
10% 5% 1% 10% 5% 1%
Critical values at -2.01 -2.06 -2.14 -2.01 -2.06 -2.14
Source: Compiled by the author.
Yt = (l-0;>,+0,Y,t-i + i
(5)
when f is accepted as unobserved factor and if uit has a structure of single factor, uit is expressed as shown in formula (6) [51]:
uu = Y i ft + eu ,
(6)
when uit is placed in the model, the new model will turn into the shape shown in the formula (7).
AYit =a +PiYi ,t-1 +Yift + et (7)
kYit =Y u -Y u-, a =(l-0>;- and P =(1-0,) (8)
Pesaran, taking arithmetic means of each series, H0 and H1 hypotheses through CIPS values.
CIPS (N ,T ) =
H0: There is unit root; H : There is no unit root.
I J. N T )
N
(9)
Depending on the theoretical explanations, the results of unit root realized are shown in Table 4.
As shown in Table 4, since CIPS value for both variables is bigger than critical value of 1%, H0 is rejected, and it is accepted that there is no unit root in the variables.
Since it was identified that both variables were stationary at the level, the presence of the relationship between both variables will be studied by Panel LS test.
panel ls test
The presence of correlations between horizontal cross-sections of the variables was identified in the previous sections. In addition, it was also identified that the coefficients of the variables exhibited a homogenous distribution. Depending on these identifications, on homogenous panel data that has horizontal cross-sectional dependence, PDOLS (Panel Dynamic Least Square) Model among second generation analyses can be used [51]. PDOLS analyses come to our face as long-term predictions and in order to eliminate the effect of intersectional correlation, add the premise and lagged value of the variables to model and internal feedback to independent variable disappears [59]. PDOLS analysis model are shown below [60];
Yt =a+xtt ß+I cij ^u+1
(10)
j=-q
cri, added to the model formed is the coefficient of premise and lagged values of explanatory variable, taken first difference.
For being able to be realized Panel LS test, it is necessary to study that model has to which of fixed
Table 5
Effect Hypothesis
Hausman Test Chow(F) Test Breush-Pagan LM Test
H0: random effect H0: pooled effect H0: pooled effect
H1: fixed effect H1: fixed effect H1: random effect
Source: Compiled by the author.
Table 6
Effect Test Results
P degeri H0 Hi Etki
Hausman Test 0.9972 Ok Reject random effect
F Test 0.9998 Ok Reject pooled effect
B-P LM Test 1.0000 Ok Reject pooled effect
Source: Compiled by the author.
Table 7
PDOLS Test Results
Dependet Variable: Ln(return)
Variable Coefficient Std. Er. t-Stat. Prob.
Z'' 0.353028" 0.004486 78.68831 0.00
R-squared 0.837398 Adjusted R-squared 0.759935
"p < .01
Source: Compiled by the author.
effects, random effects or pooled effect. In order to be able to identify the existing effect in the model, it is necessary to make Hausman, Chow (F) and Breush-Pagan LM test. Without entering to the theoretical explanation of these tests, hypotheses of any test were shown in Table 5.
The hypotheses formed were tested on both horizontal cross-l sectional and time plane and test results were shown in Table 6.
When the results of the tests are examined, Hausman test accepts that the variables have the random effect, while F Test enables to be made preference between pooled effect and fixed effect for the variables. According to the results of F Test made, the variables have pooled effect. Breush-Pagan LM test enables to be made preference between random effect and pooled effect and, according to the result of the test, identified that the variables have random effect.
As a result of all tests made, the following model is suggested, in which Ln (return) is dependent variable
and Altman Z" Score value is independent variable.
q
Ln(return )tt = a+Zit p + XCjAZ«t+vt (11)
j=-q
According to PDOLS analysis results (Table 7) since P < 0.05, Z'' affects the variable of Ln(return), hence, hypothesis is not rejected. In addition, in the analysis, when R 2value is examined, significance level of the model formed was identified as 0.837 and this significance level can be accepted as relatively high. Lastly, it is understood that Altman's Z" model affects firm value in the rate of 0.35. Depending on all results obtained, although Altman Z" model is formed to measure financial failure of the firms, it is demonstrated that it can be also used to predict the increase or decrease in firm value in the long term.
conclusion and suggestions
In Turkey, identifying the financial conditions of the companies recorded to BIST and, thanks to this, according to the results obtained in this study carried out to measure the success of the model in terms of decision makers, it was identified that there was a relationship between Altman Z" score value and firm value at high significance level.
According to the PDOLS model established between the dependent variable Ln (Return) and the independent variable Z" value; A 1-unit increase in the Z" score value causes an increase of 0.353 units in the logarithmic return calculated over the firm value.
When the literature is examined, all of the studies on the Z" Score model have investigated the financial failures of the companies. In this study, it has been revealed that the Z Score model can be used not only to measure financial failure, but also to make firm value estimations in the future. From this point of view, this study has made an important contribution to the finance and accounting literature. Depending on this identification:
• Altman Z" score model is valid the other firms, recorded in Istanbul Stock Market, other than financial firms
• Increases in the value of Altman Z" score contribute to the increases of firm values.
• Following Altman Z" Score values will make contribution to long term investment.
• Since Altman Z Score values are calculated at the end of accounting records, especially in the period, in which financial tables are explained, that they also explain Z Score values will provide more prediction infrastructure for investors.
• Altman proposes the Z" Score model as a model that can be used by developing countries globally. Depending on this proposition, it is thought that the results of this study conducted in Turkey may be suitable for other developing country companies as well.
Depending on these results, in the next study, it will be suitable to prepare an index for investors through Altman Z Score value. In addition, the next studies can be focused on testing prediction accuracy of the models in sectorial basis. This study is focused on the companies, which are dealt in BIST. When considered that many small and medium -sized enterprises (SMSEs) in Turkey default due to cash flow problems, it will also be interesting accuracy of models to test on non- quotation SMSEs and service enterprises in stock market in Turkey.
references
1. Ahn B. S., Cho S. S., Kim C. Y. The integrated methodology of rough set theory and artificial neural network for business failure prediction. Expert Systems with Applications. 2000;18(2):65-74. DOI: 10.1016/S 0957-4174(99)00053-6
2. Moyer R. C. Forecasting financial failure: A re-examination. Financial Management. 1977;6(1):11-17. DOI: 10.2307/3665489
3. Petty J. W., Keown J. A., Scott D. F., Martin J. D. Basic financial management. 6th ed. Englewood Cliffs, NJ: Prentice-Hall; 1993. 684 p.
4. Brigham E. F., Gapenski L. C. Financial management theory and practice. 8th ed. Chicago, IL: The Dryden Press; 1997. 875 p.
5. Hunter J, Isachenkova N. Failure risk: A comparative study of UK and Russian firms. Journal of Policy Modeling. 2001;23(5):511-521. DOI: 10.1016/S 0161-8938(01)00064-3
6. Poston K. M., Harmon K., Gramlich J. D. A test of financial ratios as predictors of turnaround versus failure among financially distressed firms. Journal of Applied Business Research. 1994;10(1):41-56. DOI: 10.19030/jabr. v10i1.5962
7. McKee T. E. Rough sets bankruptcy prediction models versus auditor signalling rates. Journal of Forecasting. 2003;22(8):569-586. DOI: 10.1002/for.875
8. Meeks G., Meeks J. G. Self-fulfilling prophecies of failure: The endogenous balance sheets of distressed companies. Abacus. 2009;45(1):22-43. DOI: 10.1111/j.1467-6281.2009.00276.x
9. Chiarello T. C., Pletsch C. S., Da Silva A., Da Silva T. P. Financial performance, intangible assets and value creation in brazilian and chilean information technology companies. Revista Galega de Economia = Economic Review of Galicia. 2014;23(4):73-88. DOI: 10.15304/rge.23.4.2787
10. Lifschutz S., Jacobi A. Predicting bankruptcy: Evidence from Israel. International Journal of Business and Management. 2010;5(4):133-141. DOI: 10.5539/ijbm.v5n4p133
11. Andrade G., Kaplan S. N. How costly is financial (not economic) distress? Evidence from highly leveraged transactions that became distressed. The Journal of Finance. 1998;53(5):1443-1493. DOI: 10.1111/0022-1082.00062
12. Outecheva N. Corporate financial distress: An empirical analysis of distress risk. Doctor oeconomiae dissertation. St. Gallen: University of St. Gallen; 2007. 200 p. URL: https://www.e-helvetica.nb.admin.ch/ api/download/urn%3Anbn%3Ach%3Abel-130003%3Adis3430.pdf/dis3430.pdf
13. Schall L. D., Haley C. W. Introduction to financial management. 3rd ed. New York, NY: McGraw-Hill Book Company; 1983. 912 p.
14. Moyer R. C., McGuigan J.R., Kretlow W. J. Contemporary financial management. 4th ed. St. Paul, MN: West Publishing Company; 1992. 862 p.
15. Altman E. I. Financial ratios, discriminant analysis and the prediction of corporate bankruptcy. The Journal of Finance. 1968;23(4):589-609. DOI: 10.1111/j.1540-6261.1968.tb00843.x
16. Altman E. I. Corporate financial distress: A complete guide to predicting, avoiding, and dealing with bankruptcy. New York, NY: John Wiley & Sons, Inc;. 1983. 368 p. (Frontiers in Finance Series).
17. Altman E. I. Corporate financial distress and bankruptcy. 2nd ed. New York, NY: John Wiley & Sons, Inc.; 1993. 384 p.
18. Zmijewski M. E. Methodological issues related to the estimation of financial distress prediction models. Journal of Accounting Research. 1984;22:59-82. DOI: 10.2307/2490859
19. Springate G. L. Predicting the possibility of failure in a Canadian firm. Unpublished doctoral dissrtation. Burnaby, BC: Simon Fraser University; 1978.
20. Ohlson J. A. Financial ratios and the probabilistic prediction of bankruptcy. Journal of Accounting Research. 1980;18(1):109-131. DOI: 10.2307/2490395
21. Fulmer J. G., Moon J. E., Gavin T. A., Erwin M. J. A bankruptcy classification model for small firms. Journal of Commercial Bank Lending. 1984;66(11):25-37.
22. Calandro J. Considering the utility of Altman's Z"-score as a strategic assessment and performance management tool. Strategy & Leadership. 2007;35(5):37-43. DOI: 10.1108/ 10878570710819206
23. Hayes S. K., Hodge K. A., Hughes L. W. A study on the efficiency of Altman's Z to predict of specialty retail firms doing business in contemporary times. Economics and Business Journal: Inquiries & Perspectives. 2010;3(1):122-134.
24. Beaver W. H. Financial ratios as predictors of failure. Journal of Accounting Research. 1966;4:71-111. DOI: 10.2307/2490171
25. Grice J. S., Ingram R. W. Tests of the generalizability of Altman's bankruptcy prediction model. Journal of Business Research. 2001;54(1):53-61. DOI: 10.1016/S 0148-2963(00)00126-0
26. Platt H. D., Platt M. B. Understanding differences between financial distress and bankruptcy. Review of Applied Economics. 2006;2(2):141-157.
27. Bemmann M. Improving the comparability of insolvency predictions. Dresden Discussion Paper Series in Economics. 2005;(8). DOI: 10.2139/ssrn.731644
28. Dichev I. D. Is the risk of bankruptcy a systematic risk? The Journal of Finance. 1998;53(3):1131-1147. DOI: 10.1111/0022-1082.00046
29. Francis J. Corporate compliance with debt covenants. Journal of Accounting Research. 1990;28(2):326-347. DOI: 10.2307/2491153
30. Miller W. Comparing models of corporate bankruptcy prediction: Distance to default vs. Z-score. SSRN Electronic Journal. 2009. DOI: 10.2139/ssrn.1461704
31. Takahashi M., Taques F., Basso L. Altman's bankruptcy prediction model: Test on a wide out of business private companies sample. iBusiness. 2018;10(1):21-39. DOI: 10.4236/ib.2018.101002
32. Charalambous C., Charuitou A., Kaourou F. Comparative analysis of artificial neural network models: Application in bankruptcy prediction. Annals of Operations Research. 2000;99(1):403-425. DOI: 10.1023/A:1019292321322
33. Hanson R. O. A study of Altman's revised four-variable Z"-score bankruptcy prediction model as it applies to the service industry. Doctor of business administration dissertation. Fort Lauderdale, FL: Nova Southeastern University; 2003. 153 p. URL: https://www.proquest.com/openview/d787fc069fb7c0c4b931 22e3faba6174/1?pq-origsite=gscholar&cbl=18750&diss=y
34. Tamari M. Financial ratios as a means of forecasting bankruptcy. Management International Review. 1966;6(4):15-21.
35. Ramaratnam M. S., Jayaraman R. A study on measuring the financial soundness of select firms with special reference to Indian steel industry — An empirical view with Z score. Asian Journal of Management Research. 2010;1(1):724-735.
36. Ijaz M. S., Hunjra A. I., Hameed Z., Maqbool A., Azam R. Assessing the financial failure using Z-score and current ratio: A case of sugar sector listed companies of Karachi stock exchange. World Applied Sciences Journal. 2013;23(6):863-870. DOI: 10.5829/idosi.wasj.2013.23.06. 2243
37. Tyagi V. A study to measures the financial health of selected firms with special reference to Indian logistic industry: An application of Altman's Z score. Industrial Engineering Letters. 2014;4(4):43-52. URL: https://iiste.org/Journals/index.php/IEL/article/view/12246/12599
38. Sansesh C. The analytical study of Altman Z score on NIFTY 50 companies. International Journal of Management & Social Sciences. 2016;3(3):433-438. DOI: 10.21013/jmss.v3.n3.p6
39. Shariq M. Bankruptcy prediction by using the Altman Z-score model in Oman: A case study of Raysut Cement Company SAOG and its subsidiaries. Australasian Accounting, Business and Finance Journal. 2016;10(4):70-80. DOI: 10.14453/aabfj.v10i4.6
40. Ko Y.-C., Fujita H., Li T. An evidential analysis of Alman's Z-score for financial prediction: Case study on solar energy companies. Applied Soft Computing. 2017;52:748-759. DOI: 10.1016/j.asoc.2016.09.050
41. Milasinovic M., Knezevic S., Mitrovic A. Bankruptcy forecasting of hotel companies in the Republic of Serbia using Altman's Z-score model. Hotel and Tourism Management. 2019;7(2):87-95. DOI: 10.5937/ menhottur1902087M
42. Muzir E., ^aglar N. The accuracy of financial distress prediction models in Turkey: A comparative investigation with simple model proposals. Anadolu Üniversitesi Sosyal Bilimler Dergisi = Anadolu University Journal of Social Sciences. 2009;9(2):15-48. URL: https://core.ac.uk/download/pdf/6827016.pdf
43. Yilmaz H., Yildiran M. Borsada i§lem gören i^letmelerde mali ba^arisizlik tahmini: Altman modeli'nin BIST uygulamasi. Aksaray Üniversitesi iktisadi ve idari Bilimler Fakültesi Dergisi = Journal of Aksaray University Faculty of Economics and Administrative Sciences. 2015;7(3):43-49.
44. Özdemir F. S. Turkish uniform accounting system and applicability of Altman Z score models in the context of public and private companies. Ege Akademik Baki§ = Ege Academic Review. 2014;14(1):147-161. (In Turk.).
45. Altman E. I. Predicting financial distress of companies: Revisiting the Z-score and ZETA models. 2000. URL: https://pages.stern.nyu.edu/~ealtman/Zscores.pdf
46. Yildiz A. Corporate ratings based on the corporate governance index and Altman z score applying with the logistic regression method. Süleyman Demirel Üniversitesi iktisadi ve idari Bilimler Fakültesi Dergisi = Suleyman Demirel University. The Journal of Faculty of Economics and Administrative Sciences. 2014;19(3):71-89. (In Turk.).
47. Breusch T. S., Pagan A. R. The Lagrange multiplier test and its applications to model specification tests in econometrics. The Review of Economic Studies. 1980;47(1):239-253. DOI: 10.2307/2297111
48. Pesaran M. H. General diagnostic tests for cross section dependence in panels. Empirical Economics. 2021;60(1):13-50. DOI: 10.1007/s00181-020-01875-7
49. Altinta§ H., Mercan M. The relationship between research and development (R&D) expenditures: Panel cointegration analysis under cross sectional dependency on OECD countries. Ankara Üniversitesi SBFDergisi = Ankara University SBF Journal. 2015;70(2):345-376. (In Turk.).
50. Pesaran M. H., Ullah A., Yamagata T. A bias-adjusted LM test of error cross-section independence. The Econometrics Journal. 2008;11(1):105-127. DOI: 10.1111/j.1368-423X.2007.00227.x
51. Tatoglu Yerdelen F. Panel Zaman Serileri Analizi Stata Uygulamali. 3rd ed. istanbul: Beta Yayincilik; 2017. 359 p.
52. Kar M., Agir H., Türkmen S. Econometric estimation of the relationship between electricity consumption and economic growth in developing countries. In: Yardimcioglu F., Be§el F., Inan V., eds. Economic studies. Pesa Publications. 2018;2:305-321. (In Turk.).
53. Hsiao C. Analysis of panel data. New York, NY: Cambridge University Press; 2015. 562 p. (Econometric Society Monographs. No. 54).
54. Turgut E., U^an O. Investigation of the relationship between corruption and tax rate in the sample of OECD countries. Nigde Ömer Halisdemir Üniversitesi Sosyal Bilimler Enstitüsü Dergisi = Nigde Ömer Halis-demir University. The Journal of Graduate School of Social Sciences. 2019;1(3):1-17. (In Turk.).
55. Taylor M. P., Sarno L. The behavior of real exchange rates during the post-Bretton Woods period. Journal of International Economics. 1998;46(2):281-312. DOI: 10.1016/S 0022-1996(97)00054-8
56. Breuer J. B., McNown R., Wallace M. Series-specific unit root tests with panel data. Oxford Bulletin of Economics and Statistics. 2002;64(5):527-546. DOI: 10.1111/1468-0084.00276
57. Pesaran M. H. A simple panel unit root test in the presence of cross-section dependence. Journal of Applied Econometrics. 2007;22(2):265-312. DOI: 10.1002/jae.951
58. Ku^ukaksoy i., Akalin G. Testing of the Fisher hypothesis with a dynamic panel data analysis: An application on OECD countries. Hacettepe Universitesi iktisadi ve idari Bilimler FakUltesi Dergisi = Hacettepe University Journal of Economics and Administrative Sciences. 2017;35(1):19-40. (In Turk.). DOI: 10.17065/ huniibf.303303
59. ^etin M., Dogan i., I§ik H. The impact of energy consumption on environmental pollution: A panel data analysis. International Anatolia Academic Online Journal. 2014;2(1):15-29. (In Turk.).
60. Kao C., Chiang M.-H. On the estimation and inference of a cointegrated regression in panel data. In: Bal-tagi B. H., Fomby T. B., Carter Hill R., eds. Nonstationary panels, panel cointegration, and dynamic panels. Bingley: Emerald Group Publishing Limited; 2001:179-222. (Advances in Econometrics. Vol. 15). DOI: 10.1016/S 0731-9053(00)15007-8
about the authors / информация об авторах
ibrahim Emre Gokturk — Assist. Prof., Faculty of Health Science, Department of Social Service, Necmettin Erbakan University, Konya, Turkiye
Ибрагим Эмре Гёктюрк — старший преподаватель, факультет здравоохранения, кафедра социального обслуживания, Университет Некметтина Эрбакана, Конья, Турция https://orcid.org/0000-0003-2881-1684 Corresponding author / Автор для корреспонденции: iegokturk@gmail.com
Huseyin Serdar Yalginkaya — Assist. Prof., Department of Accounting and Tax, Eregli Vocational School, Necmettin Erbakan University, Konya, Turkiye
Хюсейн Сердар Ялчынкая — старший преподаватель, факультет бухгалтерского учета и налогообложения, профессиональная школа Эрегли, Университет Некметтина Эр-бакана, Конья, Турция https://orcid.org/0000-0002-5064-5144
Conflicts of Interest Statement: The authors have no conflicts of interest to declare. Конфликт интересов: авторы заявляют об отсутствии конфликта интересов.
The article was submitted on 03.11.2022; revised on 12.12.2022 and accepted for publication on 27.02.2023. The authors read and approved the final version of the manuscript.
Статья поступила в редакцию 03.11.2022; после рецензирования 12.12.2022; принята к публикации 27.02.2023.
Авторы прочитали и одобрили окончательный вариант рукописи.
