Etikonomi Volume 17 (1), 2018: 57 - 68 P-ISSN: 1412-8969;
Added on 2023-04-21
12 Pages6999 Words235 Views
|
|
|
Received: December 28, 2017; Revised: January 25, 2018; Accepted: February 5, 2018
1, 3 Universitas Malikussaleh. Muara Batu, North Aceh, Aceh, Indonesia
1,2 Universitas Syiah Kuala. Jl. Teuku Nyak Arief, Banda Aceh, Aceh, Indonesia
E-mail: 1ghazali.syamni@unimal.ac.id, 2mshabri@unsyiah.ac.id, 3wiedyanav@gmail.com
DOI: htttp://dx.doi.org/10.15408/etk.v17i1.6559
Etikonomi
Volume 17 (1), 2018: 57 - 68
P-ISSN: 1412-8969; E-ISSN: 2461-0771Ghazali Syamni1, M. Shabri Abd. Majid2, Widyana Verawaty Siregar3
Abstract. Various bankruptcy prediction models have been used to measure the movement
of stock prices, and thus the firms’ performance. This study is aimed at empirically exploring
the usefulness of the Olhson, Almant Modification, Grover, Springate, and Zmijewski models
for predicting bankruptcy of the 19 coal mining companies. It also attempts to measure
the effects of the scores of these bankruptcy prediction models on the stock prices of the coal
mining companies in Indonesia. The technique of analysis that used in this research is panel
regression. The results of the study showed that the bankruptcy prediction scores of the Ohlson
and Almant Modification were found to be the dominant prediction models that affected the
stock prices of the coal companies in Indonesia. This indicates that the bankruptcy prediction
model can be used as one of the approaches to measure the movement of stock prices and
performance of the coal mining companies in Indonesia.
Keywords: bankruptcy, stock prices, coal mining companies.
Abstrak. Berbagai model prediksi kebangkrutan telah digunakan untuk mengukur
pergerakan harga saham dan sekaligus kinerja perusahaan. Penelitian ini bertujuan
untuk mengeksplorasi secara empiris kegunaan model Olhson, Almant Modification,
Grover, Springate, dan Zmijewski dalam memprediksi kebangkrutan 19 perusahaan
pertambangan batubara. Penelitian ini juga menguji dampak model prediksi kebangkrutan
terhadap harga saham perusahaan pertambangan batubara di Indonesia. Teknik analisis
yang dipergunakan pada penelitian ini ialah teknik regresi panel. Hasil penelitian ini
menemukan bukti bahwa model prediksi Ohlson dan Modifikasi Almant merupakan
model prediksi dominan yang mempengaruhi harga saham perusahaan batubara di
Indonesia. Hal ini mengindikasikan bahwa model prediksi kebangkrutan dapat digunakan
untuk memprediksikan pergerakan harga saham dan sekaligus kinerja keuangan industri
batubara di Indonesia.
Kata kunci: kebangkrutan, harga saham, perusahaan pertambangan batubara
How to Cite:
Syamni, G., Majid, M.S.A., & Siregar, W.F. (2018). Bankruptcy Prediction Models and Stock Prices of The Coal Mining
Industry in Indonesia. Etikonomi: Jurnal Ekonomi. Vol. 17 (1): 57 – 68. doi: http//dx.doi.org/10.15408/etk.v17i1.6559.
Bankruptcy Prediction Models and Stock Prices
of the Coal Mining Industry in Indonesia
1, 3 Universitas Malikussaleh. Muara Batu, North Aceh, Aceh, Indonesia
1,2 Universitas Syiah Kuala. Jl. Teuku Nyak Arief, Banda Aceh, Aceh, Indonesia
E-mail: 1ghazali.syamni@unimal.ac.id, 2mshabri@unsyiah.ac.id, 3wiedyanav@gmail.com
DOI: htttp://dx.doi.org/10.15408/etk.v17i1.6559
Etikonomi
Volume 17 (1), 2018: 57 - 68
P-ISSN: 1412-8969; E-ISSN: 2461-0771Ghazali Syamni1, M. Shabri Abd. Majid2, Widyana Verawaty Siregar3
Abstract. Various bankruptcy prediction models have been used to measure the movement
of stock prices, and thus the firms’ performance. This study is aimed at empirically exploring
the usefulness of the Olhson, Almant Modification, Grover, Springate, and Zmijewski models
for predicting bankruptcy of the 19 coal mining companies. It also attempts to measure
the effects of the scores of these bankruptcy prediction models on the stock prices of the coal
mining companies in Indonesia. The technique of analysis that used in this research is panel
regression. The results of the study showed that the bankruptcy prediction scores of the Ohlson
and Almant Modification were found to be the dominant prediction models that affected the
stock prices of the coal companies in Indonesia. This indicates that the bankruptcy prediction
model can be used as one of the approaches to measure the movement of stock prices and
performance of the coal mining companies in Indonesia.
Keywords: bankruptcy, stock prices, coal mining companies.
Abstrak. Berbagai model prediksi kebangkrutan telah digunakan untuk mengukur
pergerakan harga saham dan sekaligus kinerja perusahaan. Penelitian ini bertujuan
untuk mengeksplorasi secara empiris kegunaan model Olhson, Almant Modification,
Grover, Springate, dan Zmijewski dalam memprediksi kebangkrutan 19 perusahaan
pertambangan batubara. Penelitian ini juga menguji dampak model prediksi kebangkrutan
terhadap harga saham perusahaan pertambangan batubara di Indonesia. Teknik analisis
yang dipergunakan pada penelitian ini ialah teknik regresi panel. Hasil penelitian ini
menemukan bukti bahwa model prediksi Ohlson dan Modifikasi Almant merupakan
model prediksi dominan yang mempengaruhi harga saham perusahaan batubara di
Indonesia. Hal ini mengindikasikan bahwa model prediksi kebangkrutan dapat digunakan
untuk memprediksikan pergerakan harga saham dan sekaligus kinerja keuangan industri
batubara di Indonesia.
Kata kunci: kebangkrutan, harga saham, perusahaan pertambangan batubara
How to Cite:
Syamni, G., Majid, M.S.A., & Siregar, W.F. (2018). Bankruptcy Prediction Models and Stock Prices of The Coal Mining
Industry in Indonesia. Etikonomi: Jurnal Ekonomi. Vol. 17 (1): 57 – 68. doi: http//dx.doi.org/10.15408/etk.v17i1.6559.
Bankruptcy Prediction Models and Stock Prices
of the Coal Mining Industry in Indonesia
Ghazali Syamni. Bankruptcy Prediction Models and Stock Prices
http://journal.uinjkt.ac.id/index.php/etikonomi
DOI: htttp://dx.doi.org/10.15408/etk.v17i1.655958
Introduction
Various approaches have been adopted to measure the company financial performances;
one of them is using the bankruptcy prediction models. The bankruptcy prediction model has
been used to analyze the company performances of different industries. The first bankruptcy
prediction model was introduced by Altman (1968), known as Almant Z-Score. This model
has been widely used and still being relevant to predict a company whether it is bankrupt, in
grey area or healthy (Altman, et al, 2017). In 1995, Edward Almant later modified the model,
so that it can be used for predicting bankruptcy of manufacturing and non-manufacturing
companies.
After 1970s, several other models have been introduced to predict bankruptcy such
as Springate (1978); Ohlson (1980); Zmijewski (1983); and Grover and Lavin (2001).
The names of the models were given based on the name of the researchers who introduced
them for the first time. In his study, Jayasekera (2017) have identified four models of the
bankruptcy prediction, namely: the mathematic, neural network, statistic and the market
models. Meanwhile, Wu, et al (2010) categorized the bankruptcy models into the discriminant
model popularized by Altman in 1968, the logit model introduced by Ohlson in 1980, the
probit model developed by Zmijewski in 1984, the hazard model proposed by Shumway in
2001, and the Black-Scholes-Merton (BSM) probability model introduced by Hillegeist, et
al (2004).
In predicting the bankcruptcy, these models have different levels of accuracy based
on their measurements used (Purnajaya and Merkusiwati, 2014). For example, the Ohlson
model has added the company income variable and in totality, the model has seven variables.
Meanwhile, the modified Almant and Springate models have the similar four variables to
the Olson model, yet they have different types of variables, except the working capital of
the total assets. Finally, the models categorized the firm either into the healthy, grey area,
or bankrupted company with different scores. The detailed measurements of the models are
explained in the methodological section.
Many previous empirical studies in the developed countries have used different models
to predict the company performances. For example, in the United States of America; Charitou,
et al (2013) used the Black-Scholes-Merton (BSM) model to predict the bankruptcy of the
non-financial companies. In England, Tinoco and Wilson (2013) predicted the bankruptcy
by using the network and Almant Z Score models. Ko, et al (2017) predicted the bankruptcy
of the solar energy company in Taiwan using the Z score model. Xu and Zhang (2009)
predicted the bankruptcy of the listed companies in the Japanese stock exchange using the
Almant Z and the Ohlson scores and then regressed them with the financial performances of
the bank institution and Keiretsu firms as the dependent variable.
Similar studies on the bankruptcy prediction in the developing countries have
also used different types of the bankruptcy prediction models. Marcinkevičius and
Kanapickienė (2014) predicted the bankruptcy of the construction company in Lithuania
using the Altman, Springate, Taffler and the Tisshaw models. Karas and Režňáková (2015)
measured the bankruptcy prediction of manufacturing company in the Republic of Czech
http://journal.uinjkt.ac.id/index.php/etikonomi
DOI: htttp://dx.doi.org/10.15408/etk.v17i1.655958
Introduction
Various approaches have been adopted to measure the company financial performances;
one of them is using the bankruptcy prediction models. The bankruptcy prediction model has
been used to analyze the company performances of different industries. The first bankruptcy
prediction model was introduced by Altman (1968), known as Almant Z-Score. This model
has been widely used and still being relevant to predict a company whether it is bankrupt, in
grey area or healthy (Altman, et al, 2017). In 1995, Edward Almant later modified the model,
so that it can be used for predicting bankruptcy of manufacturing and non-manufacturing
companies.
After 1970s, several other models have been introduced to predict bankruptcy such
as Springate (1978); Ohlson (1980); Zmijewski (1983); and Grover and Lavin (2001).
The names of the models were given based on the name of the researchers who introduced
them for the first time. In his study, Jayasekera (2017) have identified four models of the
bankruptcy prediction, namely: the mathematic, neural network, statistic and the market
models. Meanwhile, Wu, et al (2010) categorized the bankruptcy models into the discriminant
model popularized by Altman in 1968, the logit model introduced by Ohlson in 1980, the
probit model developed by Zmijewski in 1984, the hazard model proposed by Shumway in
2001, and the Black-Scholes-Merton (BSM) probability model introduced by Hillegeist, et
al (2004).
In predicting the bankcruptcy, these models have different levels of accuracy based
on their measurements used (Purnajaya and Merkusiwati, 2014). For example, the Ohlson
model has added the company income variable and in totality, the model has seven variables.
Meanwhile, the modified Almant and Springate models have the similar four variables to
the Olson model, yet they have different types of variables, except the working capital of
the total assets. Finally, the models categorized the firm either into the healthy, grey area,
or bankrupted company with different scores. The detailed measurements of the models are
explained in the methodological section.
Many previous empirical studies in the developed countries have used different models
to predict the company performances. For example, in the United States of America; Charitou,
et al (2013) used the Black-Scholes-Merton (BSM) model to predict the bankruptcy of the
non-financial companies. In England, Tinoco and Wilson (2013) predicted the bankruptcy
by using the network and Almant Z Score models. Ko, et al (2017) predicted the bankruptcy
of the solar energy company in Taiwan using the Z score model. Xu and Zhang (2009)
predicted the bankruptcy of the listed companies in the Japanese stock exchange using the
Almant Z and the Ohlson scores and then regressed them with the financial performances of
the bank institution and Keiretsu firms as the dependent variable.
Similar studies on the bankruptcy prediction in the developing countries have
also used different types of the bankruptcy prediction models. Marcinkevičius and
Kanapickienė (2014) predicted the bankruptcy of the construction company in Lithuania
using the Altman, Springate, Taffler and the Tisshaw models. Karas and Režňáková (2015)
measured the bankruptcy prediction of manufacturing company in the Republic of Czech
DEVDEEP
2/18/2019, 2:20:12 PM
http://journal.uinjkt.ac.id/index.php/etikonomi
DOI: htttp://dx.doi.org/10.15408/etk.v17i1.655959Etikonomi
Volume 17 (1), 2018: 57 - 68
using the combining models of the discriminant and the Box Cox. In Saudi Arabia, Al-
Kassar and Soileau (2014) predicted the bankruptcy of different companies (i.e., the mill,
transportation, heritage and museum, commercial companies, and the replenishment of
oil companies) using the Z-score model. Similarly, using the Z score model, Hussain,
et al (2014) measured bankruptcy of the textiles industry in Pakistan, while Al-Rawi, et
al (2011) investigated the bankruptcy of the glassware company in Jordan. Karamzadeh
(2013) predicted the bankruptcy of 90 stock exchange companies in Teheran using the
Almant Z score and the Ohlson models. In Thailand, Pongsatat, et al (2004) using the
Almant and Ohlson models to predict the performances of 60 bankrupted companies and
60 non-bankrupted companies.
In the context of Indonesia, in predicting the bankruptcy of the companies, many
previous studies have widely used a single Almant Z Score (Sudiyatno and Puspitasari,
2010), while only few other studies used different types of the bankruptcy prediction
models for companies of different sectors. For example, Sembiring (2016) used the Ohlson
model to predict the bankrupted companies. In predicting the bankruptcy, Rachmawati
(2016) applied the Almant model for the insurance company, Boedi and Tiara (2016)
for the telecommunication company, and Yunan and Rahmasari (2015) for Shariah Stock
Performance.
On the other hand, the previous studies that used more than one models for predicting
the bankruptcy of the company in Indonesia. Putera, et al (2017) predicted the bankruptcy
of the mining companies by using the Altman, Springate and Ohlson models. Gunawan, et
al (2016) applied the Altman, Grover and Zmijewski models for the manufactures company,
while Rahayu, et al (2016) used the Altman Z-Score, Springate, and the Zmijewski models
for the telecommunication companies in Indonesia.
Furthermore, Effendi, et al (2016) used the bankruptcy prediction model of the
Springate to estimate the stock prices of the telecommunication companies, while Andriawan
and Salean (2016) used the Almant model and analyzed its impact to the pharmacy company’s
stock prices. Adrian and Khoiruddin (2014) applied the Almant model and analyzed its
impact to the manufacturing company’s stock prices. They documented that the Almant
model affected the stock prices of the manufacturing companies (Adrian and Khoiruddin,
2014), the pharmacy companies (Andriawan and Salean, 2016), and the transportation
companies (Amaliawiati and Lestari, 2014) in Indonesia. Effendi et al. (2016) found that the
Springate model affected the stock prices, while Wulandari and Norita (2014) found that the
Ohlson model (O-score) affected the stock returns of the textile and garment companies in
Indonesia over the period 2010-2014. In short, these studies only used a single bankruptcy
prediction model to estimate the stock prices in Indonesia.
The above-reviewed studies showed that many previous empirical studies have only
adopted a single model to predict the performance of the firm, and yet its connection to the
stock prices was rarely done using various bankcruptcy prediction models, thus provided
insufficient empirical findings. Bankruptcy prediction models might affect the stock prices,
indicating that when a company goes into bankruptcy, its stock price goes down or unchanged.
DOI: htttp://dx.doi.org/10.15408/etk.v17i1.655959Etikonomi
Volume 17 (1), 2018: 57 - 68
using the combining models of the discriminant and the Box Cox. In Saudi Arabia, Al-
Kassar and Soileau (2014) predicted the bankruptcy of different companies (i.e., the mill,
transportation, heritage and museum, commercial companies, and the replenishment of
oil companies) using the Z-score model. Similarly, using the Z score model, Hussain,
et al (2014) measured bankruptcy of the textiles industry in Pakistan, while Al-Rawi, et
al (2011) investigated the bankruptcy of the glassware company in Jordan. Karamzadeh
(2013) predicted the bankruptcy of 90 stock exchange companies in Teheran using the
Almant Z score and the Ohlson models. In Thailand, Pongsatat, et al (2004) using the
Almant and Ohlson models to predict the performances of 60 bankrupted companies and
60 non-bankrupted companies.
In the context of Indonesia, in predicting the bankruptcy of the companies, many
previous studies have widely used a single Almant Z Score (Sudiyatno and Puspitasari,
2010), while only few other studies used different types of the bankruptcy prediction
models for companies of different sectors. For example, Sembiring (2016) used the Ohlson
model to predict the bankrupted companies. In predicting the bankruptcy, Rachmawati
(2016) applied the Almant model for the insurance company, Boedi and Tiara (2016)
for the telecommunication company, and Yunan and Rahmasari (2015) for Shariah Stock
Performance.
On the other hand, the previous studies that used more than one models for predicting
the bankruptcy of the company in Indonesia. Putera, et al (2017) predicted the bankruptcy
of the mining companies by using the Altman, Springate and Ohlson models. Gunawan, et
al (2016) applied the Altman, Grover and Zmijewski models for the manufactures company,
while Rahayu, et al (2016) used the Altman Z-Score, Springate, and the Zmijewski models
for the telecommunication companies in Indonesia.
Furthermore, Effendi, et al (2016) used the bankruptcy prediction model of the
Springate to estimate the stock prices of the telecommunication companies, while Andriawan
and Salean (2016) used the Almant model and analyzed its impact to the pharmacy company’s
stock prices. Adrian and Khoiruddin (2014) applied the Almant model and analyzed its
impact to the manufacturing company’s stock prices. They documented that the Almant
model affected the stock prices of the manufacturing companies (Adrian and Khoiruddin,
2014), the pharmacy companies (Andriawan and Salean, 2016), and the transportation
companies (Amaliawiati and Lestari, 2014) in Indonesia. Effendi et al. (2016) found that the
Springate model affected the stock prices, while Wulandari and Norita (2014) found that the
Ohlson model (O-score) affected the stock returns of the textile and garment companies in
Indonesia over the period 2010-2014. In short, these studies only used a single bankruptcy
prediction model to estimate the stock prices in Indonesia.
The above-reviewed studies showed that many previous empirical studies have only
adopted a single model to predict the performance of the firm, and yet its connection to the
stock prices was rarely done using various bankcruptcy prediction models, thus provided
insufficient empirical findings. Bankruptcy prediction models might affect the stock prices,
indicating that when a company goes into bankruptcy, its stock price goes down or unchanged.
Ghazali Syamni. Bankruptcy Prediction Models and Stock Prices
http://journal.uinjkt.ac.id/index.php/etikonomi
DOI: htttp://dx.doi.org/10.15408/etk.v17i1.655960
Relying only on a single model to predict the bankcruptcy might lead to an inaccurate
estimation, thus it, in turns, lead to the improper policy recommendation. Anticipating this,
the present study used various models to predict the bankruptcy of the firms and analyzed
its impact to the firms’ stock prices in Indonesia. Thus, this study is among the first to use
various bankruptcy models of the Olhson, Almant Modification, Grover, Springate, and the
Zmijewski to estimate the stock prices in the Indonesian stock markets. This is the first
novelty of this study that is in its comparison among the bankruptcy prediction models.
Estimating comparatively among these models would show the most suitable and accurate
model to be adopted to predict the movement of stock prices in Indonesia. Secondly, this
study focuses its analysis on the coal mining companies, which has no previous empirical
studies on this important industry in the Indonesian economy. Considering the shortcomings
and mixed or inconclusive empirical findings of the previous studies on the relationships
between the bankruptcy prediction models and stock prices in Indonesia, thus this study is
hoped to provide a more comprehensive and enriching empirical evidences on this issue by
comparing various bankruptcy prediction models (i.e., the Olhson, Almant Modification,
Grover, Springate, and the Zmijewski models) and then analysing its impact to the stock
prices of the coal mining industry in Indonesia.
The findings of this study are hoped to shed some lights for the inventors when
selecting which companies to invest the monies, for the managers to promote the
performance of the companies, and for the regulator to design policy in enhancing the
stock market in the biggest Muslim populous country in the world, Indonesia. The rest
of this study is structured as follows. Section 2 provides the empirical framework and
followed by the discussion of the findings and their implications in Section 3. Finally,
Section 4 concludes the study.
Method
Data of this study is gathered from the financial report of 19 companies in the coal
mining sector that are listed on the Indonesian Stock Exchange (IDX) over the period
2013 to 2015. These data are accessed through the website www.idx.co.id and firms are
selected using the purposive sampling technique. The firms investigated in this study are
the coal mining company that published their audited financial reports during the period
2013-2015. As for the stock prices, the closing stock prices of 19 coal-mining companies
are used.
To predict the bankruptcy of the companies, the bankruptcy prediction models of the
Olhson, Almant Modification, Grover, Springate, and the Zmijewski are used. The formula,
description, and score categorization for each model are presented in Table 1.
After measuring the scores for each bankruptcy prediction model, in the next step, the
panel regression model is estimated to explore the impacts of bankruptcy prediction models
to the stock prices of the coal mining companies in Indonesia. The scores of five bankruptcy
prediction models investigated in this study are then treated as the independent variables to
predict the stock prices.
http://journal.uinjkt.ac.id/index.php/etikonomi
DOI: htttp://dx.doi.org/10.15408/etk.v17i1.655960
Relying only on a single model to predict the bankcruptcy might lead to an inaccurate
estimation, thus it, in turns, lead to the improper policy recommendation. Anticipating this,
the present study used various models to predict the bankruptcy of the firms and analyzed
its impact to the firms’ stock prices in Indonesia. Thus, this study is among the first to use
various bankruptcy models of the Olhson, Almant Modification, Grover, Springate, and the
Zmijewski to estimate the stock prices in the Indonesian stock markets. This is the first
novelty of this study that is in its comparison among the bankruptcy prediction models.
Estimating comparatively among these models would show the most suitable and accurate
model to be adopted to predict the movement of stock prices in Indonesia. Secondly, this
study focuses its analysis on the coal mining companies, which has no previous empirical
studies on this important industry in the Indonesian economy. Considering the shortcomings
and mixed or inconclusive empirical findings of the previous studies on the relationships
between the bankruptcy prediction models and stock prices in Indonesia, thus this study is
hoped to provide a more comprehensive and enriching empirical evidences on this issue by
comparing various bankruptcy prediction models (i.e., the Olhson, Almant Modification,
Grover, Springate, and the Zmijewski models) and then analysing its impact to the stock
prices of the coal mining industry in Indonesia.
The findings of this study are hoped to shed some lights for the inventors when
selecting which companies to invest the monies, for the managers to promote the
performance of the companies, and for the regulator to design policy in enhancing the
stock market in the biggest Muslim populous country in the world, Indonesia. The rest
of this study is structured as follows. Section 2 provides the empirical framework and
followed by the discussion of the findings and their implications in Section 3. Finally,
Section 4 concludes the study.
Method
Data of this study is gathered from the financial report of 19 companies in the coal
mining sector that are listed on the Indonesian Stock Exchange (IDX) over the period
2013 to 2015. These data are accessed through the website www.idx.co.id and firms are
selected using the purposive sampling technique. The firms investigated in this study are
the coal mining company that published their audited financial reports during the period
2013-2015. As for the stock prices, the closing stock prices of 19 coal-mining companies
are used.
To predict the bankruptcy of the companies, the bankruptcy prediction models of the
Olhson, Almant Modification, Grover, Springate, and the Zmijewski are used. The formula,
description, and score categorization for each model are presented in Table 1.
After measuring the scores for each bankruptcy prediction model, in the next step, the
panel regression model is estimated to explore the impacts of bankruptcy prediction models
to the stock prices of the coal mining companies in Indonesia. The scores of five bankruptcy
prediction models investigated in this study are then treated as the independent variables to
predict the stock prices.
End of preview
Want to access all the pages? Upload your documents or become a member.