The High volatility in mining sector shares leads this sector as one of the sector with a high level of risk. It makes the risk of investment in mining sector shares become interesting to be studied. Measuring the risk of loss on investment in mining sector shares can be analyzed using the method approach of Value at Risk (VaR). Meanwhile, the estimation is analyzed using Volatility Autoregressive Conditional Heteroscedasticity (ARCH) and Generalized Autoregressive Conditional Heteroscedasticity (GARCH). This study aims to form the optimum model of the ARCH-GARCH model for the Mining sector shares so that the value of investment risk can be estimated using the Value at Risk method approach. Value at Risk is currently run by financial managers as an important tool in the entire risk management process. The data of the study was obtained from Indonesia Stock Exchange (www.idx.co.id) and Yahoo Finance in the form of daily stock price per January 1, 2014 until December 31, 2018. This study revealed that the return data of the mining sector shares was the stationary data that does not have normal distribution. Besides, the measurement of Value at Risk (VaR) through the process of estimating volatility with the ARCH-GARCH model using a 95% confidence level and a holding period provides information on the maximum potential loss on each share return value. The conclusion was that the longer the holding period, the higher the level of loss. This study is expected to provide benefits for investors in considering the investment decision making. Besides, this study is also expected to bring the benefit to the company in the form of improving the performance of company management. This study can provide empirical evidence on the theory of risk analysis with the Value at Risk method.

Keywords : The Mining Sector Shares, Return Stock, Value at Risk (VaR), ARCH/GARCH Model, Stationarity.