Abstract:
We use a nonparametric quantiles-based model to analyse the predictability of long-spans (nearly or over one century) of annual volatility of Canada, France, Germany, Italy, Japan, Switzerland, the United Kingdom (UK) and the United States (US), based on information contained in domestic (banking, currency, inflation, sovereign debt, and stock market) and global crises. We find that, in general, global crises tends to have a stronger causal impact on market volatility than domestic crises, but domestic stock market crashes also play an important role in explaining equity market volatility of Germany, the UK and the US. Interestingly, extreme ends of the conditional distribution of market volatility cannot be predicted, irrespective of whether domestic or global crises are used as predictors.