Rare disasters and multilayer spillovers between volatility and skewness in international stock markets over a century of data: The role of geopolitical risk

Abstract

Measuring risk and understanding risk spillover across markets lie at the core of the decision-making process of every financial market participant and monetary authority. However, the bulk of the literature treats risk as a function of the second moment (volatility) of the return distribution, based on the implicit unrealistic assumption that asset returns are normally distributed. In this paper, we examine risk spillovers involving robust estimates of both second and third moments of model-implied distributions of stock returns using a multilayer approach; then, we assess the ability of geopolitical risk, as a proxy for rare disaster risk, to forecast layer-based risk spillovers using machine-learning methods. Considering a century of data on the stock indices of the G7 and Switzerland from May 1917 to February 2023, the results show the following: Firstly, the risk spillover among stock markets exists within each layer (i.e. volatility and skewness), with a stronger effect in the volatility layer. Secondly, the risk spillover is significant across the two layers, highlighting how various aspects of risk information are transmitted across major stock markets. Thirdly, geopolitical risk affects both risk layer values, based on an out-of-sample forecasting exercise. Specifically, global measures of geopolitical risk can forecast risk spillovers at shorter horizons up to 6 months, whereas, at longer horizons, the forecasting exercise is dominated by market-specific characteristics.