Abstract:Existing studies rely on exogenous drivers to improve the accuracy of the volatility forecasting of the least polluting fossil fuels, natural gas. However, the academic literature lacks a comprehensive analysis on the factors driving the volatility of clean energy stock indices, although the transition to renewable energy sources has resulted in the emergence of the clean energy sector as a fast-growing sector very appealing to environmentally responsible investors. This paper examines the ability of five uncertainty indices and seven global economic conditions to predict the realized volatility of clean energy stock and natural gas markets. Using daily return data of four exchange-traded funds to track the performance of the global clean energy stock market and natural gas prices, we construct the monthly realized volatility and apply various methods, including shrinkage methods. We find that both uncertainty indices and global economic conditions successfully predict clean energy realized volatility. The shrinkage methods consistently outperform dimensionality reduction methods and combination forecast methods for clean energy and natural gas. Whether clean energy ETFs or natural gas are used, the predictive information extracted from global economic conditions outperforms uncertainty indices using shrinkage methods. This suggests that real economic activities rather than text-based measures of uncertainty should be considered when investors and policy-makers analyze the volatilities of clean energy and natural gas. We further investigate the forecasting power of uncertainty indices and economic conditions in terms of variable selection, market conditions, and seasonality effects.

Natural gas volatility prediction via a novel combination of GARCH-MIDAS and one-class SVM

Natural gas volatility prediction: Fresh evidence from extreme weather and extended GARCH-MIDAS-ES model

Predicting natural gas futures’ volatility using climate risks

Predicting volatility in natural gas under a cloud of uncertainties

Prediction of Volatility Based on Realized-Garch-kernel-type Models: Evidence from China and the U.S.

Volatility prediction for the energy sector with economic determinants: Evidence from a hybrid model

Modelling and Forecasting Energy Market Volatility Using GARCH and Machine Learning Approach

Data Driven Natural Gas Spot Price Prediction Models Using Machine Learning Methods

Forecasting volatility with support vector machine-based GARCH model

Forecasting Volatility of Energy Commodities: Comparison of GARCH Models with Support Vector Regression

Predicting Natural Gas Prices Based on a Novel Hybrid Model with Variational Mode Decomposition

A hybrid econometrics and machine learning based modeling of realized volatility of natural gas

Research on Short-Term Gas Load Forecasting Based on Support Vector Machine Model

Natural gas spot price prediction research under the background of Russia-Ukraine conflict - based on FS-GA-SVR hybrid model

A hybrid deep learning approach by integrating extreme gradient boosting‐long short‐term memory with generalized autoregressive conditional heteroscedasticity family models for natural gas load volatility prediction

Variable Weights Combination MIDAS Model Based on ELM for Natural Gas Price Forecasting

Henry Hub monthly natural gas price forecasting using CEEMDAN–Bagging–HHO–SVR

Multivariate Modeling of Natural Gas Spot Trading Hubs Incorporating Futures Market Realized Volatility

Daily natural gas price forecasting by a weighted hybrid data-driven model

Volatility of clean energy and natural gas, uncertainty indices, and global economic conditions

Forecasting Natural Gas Spot Prices with Machine Learning