Forecasting Equity Premium in the Face of Climate Policy Uncertainty

This study examines the role of the US climate policy uncertainty (CPU) index in forecasting the equity premium, employing shrinkage methods such as LASSO and elastic net (ENet) to dynamically select predictors from a dataset spanning April 1987 to December 2022. Alongside CPU, other uncertainty predictors like economic policy uncertainty (EPU), geopolitical risk (GPR), and the volatility index (VIX) are considered to assess their complementary roles in out-of-sample (OOS) equity premium forecasting. The results reveal that while CPU alone cannot consistently predict the equity premium, it provides crucial complementary information when combined with other predictors, leading to a statistically significant OOS R2$$ {R}<^>2 $$ of 1.231%. The relationship between CPU and the equity premium is time varying, with a stronger influence observed during periods of economic downturn or heightened uncertainty, as demonstrated by wavelet coherence analysis. This study also identifies CPU's significant impact on industry-specific returns, particularly in climate-sensitive sectors, offering valuable insights for investment strategies and risk management in an era of increasing CPU.