Modeling and forecasting the volatility of carbon emission market: The role of outliers, time-varying jumps and oil price risk

The objectives of this study are three-fold. First, we aim to assess whether outliers or extreme observations occur in the European Union Allowance (EUA) data. Second, we examine if time-varying jumps are present in the carbon emission market. Third, we use the crude oil volatility index (OVX) to investigate the effect of oil market uncertainty on the emission price volatility. In order to detect the possible outliers in the EUA market, we employ a standard methodology proposed by Ane et al. (2008) and identify several outliers in the emission data. After spotting and then removing those extreme points, we apply the GARCH-jump models to both original and outlier free observations. The results of the jump model show that while time-varying jumps do exist in the uncorrected data, most of the jump parameters, however, become insignificant in case of outlier-free observations. Next, the application of an extended EGARCH model, in which OVX is introduced in the GARCH specification, demonstrates that emission prices are highly sensitive to oil market implied volatility and that the impact of OVX on the EUA market appears to be asymmetric. Additionally, the use of forecast encompassing test documents that considering the outlier-free data and using the information content of OVX would improve the volatility forecasts for the carbon emission market. The results of our research carry important implications for both investors and policymakers. (C) 2017 Elsevier Ltd. All rights reserved.