Macroeconomic attention and commodity market volatility

Fameliti Stavroula,Skintzi Vasiliki
DOI: https://doi.org/10.1007/s00181-024-02613-z
IF: 3.2
2024-05-23
Empirical Economics
Abstract:In this paper, we empirically examine the relationship between the novel macroeconomic attention indices (MAI) and commodity market volatility. In-sample analysis indicates that MAI contribute significantly to the volatility fluctuations in commodity markets. In addition, we employ dimension reduction techniques, shrinkage methods, and combination models in an out-of-sample exercise to assess the predictive ability of MAI, alongside a variety of economic predictors including uncertainty measures, and global as well as US economic indicators. Our empirical results demonstrate the superior predictive ability of the elastic net and LASSO models incorporating MAI together with macroeconomic and uncertainty indicators. This empirical finding is reinforced through a series of robustness checks. However, dimension reduction methods exhibit superior performance in longer forecast horizons. Finally, MAI are more informative for commodity volatility forecasting during economic expansions and non-crisis periods. Our study offers new insights on commodity volatility forecasting.
economics,social sciences, mathematical methods
What problem does this paper attempt to address?