Risk analysis on agricultural commodity portfolio using Value at Risk (VaR) and Expected Shortfall (ES) based on ARIMA-GARCH

Abstract

Investment in commodities has become an alternative investment that is increasingly in demand by the public in the last fifteen years. In commodity investment, there are two things that investors consider, namely return and risk. One way to calculate risk is to use Value at Risk (VaR) and Expected Shortfall (ES). The main reason of this research is to determine the value of Value at Risk (VaR) and Expected Shortfall (ES) of selected agriculture commodities which are Wheat, Cocoa and Cotton using the time series model approach. The data used in this research is the daily closing price of selected commodities from January 3, 2017 to December 31, 2020. In the time series modeling process, the models used for predicting commodities price movements are Autoregressive Integrated Moving Average (ARIMA) for the mean model, and Generalized Autoregressive Conditional Heteroscedasticity (GARCH) for the volatility model. The values of mean and variance acquired from the model are then used to calculate the Value at Risk (VaR) and Expected Shortfall (ES) of each selected commodity. Based on the analysis, obtained that from the selected commodities, the estimated risk for selected commodities varies, where based on Value at Risk, Cotton has the lowest risk with a Value at Risk of 0.02189155, and Cocoa has the highest risk with a Value at Risk 0.02435271. However, Expected Shortfall gives a different conclusion, where Cocoa has the lowest risk with an Expected Shortfall value of 0.02435271 and Cotton has the highest risk with an Expected Shortfall value of 0.03114681.