Causality test on gold prices and economic risk (VAR model application)

Abstract

The presence of gold as an investment asset has roots from the past to the present. The volatility of the value of economic risk can reflect the gold prices’ volatility that can affect investors' decisions in investing amid uncertainty. This study aims to predict the reciprocal relationship between gold price volatility and economic risk proxied by BI 7-days Repo Rate. The Method applied in this study is the Vector Autoregression (VAR) model to dynamically express such causality. The result found that VAR (1) model is the best-fit model to analyse the causality between variables. In addition, VAR (1) model also is tested by Granger Causality as well as used in their respective impulse response. Finally, VAR (1) model is applied to forecast the next 12-month data for both variables and has high accuracy forecasting estimation with low MSE, RMSE, and MAPE. The study then can be used as one of considerations for gold investors in making investment decisions.