TY - JOUR AU1 - Handika, Rangga AU2 - Putra, Iswahyudi Sondi AB - PurposeThis paper aims to indirectly evaluate the accuracy of various volatility models using a value-at-risk (VaR) approach and to investigate the relationship between the accuracy of volatility modelling and investments performance in the financialized commodity markets.Design/methodology/approachThis paper uses the VaR back-testing approach at six different commodities, seven different volatility models and five different time horizons.FindingsThis paper finds that the moving average (MA) VaR model tends to be the best for oil, copper, wheat and corn (long horizon) whereas the exponential generalized autoregressive conditional heteroscedastic (E-GARCH) VaR model tends to be the best for gold, silver and corn (short horizon). Our findings indicate that MA volatility model should be used for oil, copper, wheat and corn (for longer time horizons) commodities whereas E-GARCH volatility model should be used for gold, silver and corn (for short time horizons) commodities. We also find that there is a positive relationship between an accurate VaR performance and commodity return. This indicates that a good job in modelling volatility will be rewarded by higher returns in financialized commodity markets.Originality/valueThis paper indirectly evaluates the accuracy of volatility model via VaR measure and investigates the relationship between the accuracy of volatility and investments performance in financialized commodity markets. This paper contributes to the literature by offering VaR approach in evaluating volatility model performance and reporting the importance of performing accurate volatility modelling in financialized commodity markets. TI - Commodities returns’ volatility in financialization era JF - Studies in Economics and Finance DO - 10.1108/SEF-10-2015-0254 DA - 2017-08-07 UR - https://www.deepdyve.com/lp/emerald-publishing/commodities-returns-volatility-in-financialization-era-0NrNtIdVI3 SP - 344 EP - 362 VL - 34 IS - 3 DP - DeepDyve ER -