Oil Price Uncertainty and Predictability of Multi-Scale Positive and Negative Bubbles in the BRICS: Evidence from a Nonparametric Causality-in-Quantiles Test

First, we employ the Multi-Scale Log-Periodic Power Law Singularity Confidence Indicator (MS-LPPLS-CI) approach to identify positive and negative bubbles in the short-, medium, and long-term for Brazil, Russia, China, India and South Africa, i.e., the BRICS bloc, using weekly data from November 2003 to December 2020. Second, we utilize a nonparametric causality-in-quantiles approach to analyse the predictive impact of Generalized Autoregressive Conditional Heteroskedasticity (GARCH) model-based measure of oil price uncertainty on the bubble indicators. Barring some of the uppermost conditional quantiles, we detect strong evidence of predictability across the entire conditional distribution of the MS-LPPLS-CIs, with higher impacts for the short time-scale and for negative bubbles in particular. Our findings are robust to a low-frequency model-free measure of oil price uncertainty, and for advanced stock markets. Our findings have significant implications investors and policymakers.