A1093
Title: Abrupt variance shifts and volatility forecasting in the renewable energy markets: A comprehensive analysis
Authors: Akram Hasanov - Monash University (Malaysia) [presenting]
Abstract: Research on volatility modelling in the renewable energy markets has significantly increased over the past decade owing to the crucial role played by renewable energy sources in addressing climate change concerns. Using daily data on various renewable energy stock indices, various econometric models are comprehensively examined with and without accommodating structural shifts in the variance, and their forecasting performance is assessed at multiple horizons through a sliding window scheme. One of the novel aspects is the consideration of the forecasting capabilities of GARCH-class models, augmenting the regime dummies in the variance models. First, single regime GARCH-class and stochastic volatility models are used with different window sizes and distributional assumptions. Furthermore, MS-GARCH models are considered under several distribution assumptions. Second, a subset of competing models employs the break detection algorithm to determine the estimation windows considering the existence of structural breaks. The main results show that incorporating the endogenously detected structural breaks in the model estimations leads to considerable forecasting accuracy gains. The findings enrich the ongoing debate on the most effective approach to accounting for structural shift information to enhance the precision of volatility forecasts in the out-of-sample analysis of renewable energy stocks.
