Stock market confidence and copula-based Markov models

Abstract

This paper presents a descriptive model of stock market confidence conditional on stock market uncertainty in a first-order copula-based Markov approach. By using monthly closing prices of the VIX as a stock market uncertainty proxy for the United States and the copula of Fang et al. (2000) a stable nonlinear relation between confidence and uncertainty is derived. Based on the existence of a specific dependence structure uncertainty-reducing policies by US institutions which are intended to recover stock market confidence are evaluated with respect to its efficiency. The model implication for high uncertainty regimes is an aggressive uncertainty-reducing policy in order to avoid sticking in an uncertainty trap. Furthermore, uncertainty driving profit expectations force an uncertainty level which does not correspond to high confidence and calls for regulatory actions. Additionally, the methodological approach is appropriate for conditional quantile forecasts and a potential tool in risk management.