electricity, and by 2040 it is predicted to produce 22% of the world's electricity. Therefore, the development of

a coal company's stock price can reflect companies’ management performances in controlling risk which in turn

can affect the level of volatility of the company's stock price and become an indicator for investors in making

investment decisions in order to get a return. tall one. The formulation of strategic risk of coal subsector

companies with the application of the vector autoregressive (VAR) model becomes the basis of this research,

where strategic risk is proxied through the growth of stock prices and returns in each coal company that is the

sample of the study. The method that will be used in this research is descriptive quantitative through the

application of the VAR model to be able to describe the causality relationship between companies. The results

obtained are the VAR(2) model of each coal subsector company, which is used as an initial identification of its

strategic risk so that the coal subsector company can make mitigation steps in dealing with these strategic risks.

1 Introduction

Over time, the need for energy, especially coal will

increase so that many countries that do not have

sufficient natural energy resources to meet their

energy needs will import to meet their needs and

domestic stability, [1]. In addition, Indonesian coal

production is expected to continue to increase,

especially to meet domestic needs (power

generation and industry) and foreign demand

(exports), [2]. The development of coal production

Indonesian coal exports makes Indonesia one of the

largest coal exporters in the world besides Australia,

[4].

reached 115 million tons or less than the domestic

making investment decisions in order to get a high

return, [9]. The objective of this study then is to

examine the causal relationship of strategic risk of

stock returns of coal subsector companies in

Indonesia. The novelty of this study is in examining

the dynamic relationship of strategic risk caused by

the unstable rate of return between coal companies

in Indonesia.

offered in terms of strategic risk management

control that can be applied in the banking world,

which are a comprehensive definition of strategic

risk itself, and a framework that uses the cost of

equity component to estimate the amount of

economic capital needed to mitigate the risk.

strategic risk. On the one hand, the framework

simulates the bank's net income and uses a Value at

Risk (VaR) approach to measure economic capital

requirements. The framework can also be used to

stock prices, sharia stock indexes, and changes in

the rupiah exchange rate against foreign currencies.

The results of their research also revealed that each

variable is only influenced by its respective

historical data, and the results of the impulse

response function indicated that the response of all

variables is difficult to reach the zero point or

equilibrium point after a shock to other variables

occurs in the short term. The VAR(1) modelling is

then used as a model to predict the data for each

strategic risk analyzed using the VAR approach in

coal-based companies has not been widely carried

out. For this reason, the state-of-the-art research that

we are going to do is to provide a causal relationship

model. Figure 3 shows the results of the optimal lag

test for the estimation of the VAR model.

Furthermore, from the VAR Lag Order output, the

optimal lag length is then indicated by criteria

marked with an asterisk, and in this case, the

optimal lag length occurs in lag 2 (shown in Figure

3) because it has the highest number of criteria with

asterisks. Then, after getting the optimal lag length

in lag 2, the next step is to estimate the VAR lag 2

model, which is as Figure 4.

Fig. 4: Results of VAR Lag 2 Model Estimation

After getting the optimal lag on the VAR model, the

next step is to perform a cointegration test on the

VAR model estimation. The criteria for the VAR

model to pass the cointegration test can be seen in

the Prob** value, if it is less than 0.05 then the

VAR model passes the cointegration test. The

Fig. 5: Johansen Cointegration Test Output

variable PTBA becomes:

ADRO is:

concluded that PTBA stock returns are significantly

affected only by stock returns ADRO first lag.

Meanwhile, for the ADRO variable, the VAR(2)

model shows that ADRO stock returns are

significantly affected by PTBA stock returns in the

only influenced by itself and not by the historical

value of other variables. The following are the

results of the Granger causality test from the

VAR(2) model, as shown in Table 3.

Table 3. Granger Causality Test Statistics Model

Model VAR(2) above, it can be concluded that the

PTBA variable with a probability value of more

than 0.05 PTBA return value variable is not

[1] Suharyati, S. H. Pambudi, J. L. Wibowo, and

[4] M. Maulidia, P. Dargusch, P. Ashworth, and

[8] A. Gill and C. Shah, “Determinants of

[22] R. S. Tsay, Multivariate Time Series