Recession fears and stock markets: An application of directional wavelet coherence and a machine learning-based economic agent-determined Google fear index

Abstract

Recession fears play a pivotal role in investment decision making and policy development aimed at reducing the likelihood of a recession and managing its impact. Using machine learning, we develop an economic agent-determined daily recession fear index using Google searches that isolates recession-related fears from overall stock market uncertainty. We study the evolving impact of recent recession fears on stock markets using directional wavelet coherence that distinguishes between positive and negative associations. Recession fears negatively impact world and G7 stock markets and trigger heightened volatility, with Japan being the most resilient. Monetary policy tightening in response to record inflation levels significantly contributes to persistent recession fears, suggesting that policymakers should consider co-ordinating responses to avoid an excessive global economic slowdown. Our methodology offers a high frequency monitoring tool that can be applied to analyse evolving relationships between variables and can be generalised to study the influence of specific events on financial markets by isolating topic-specific components from general proxies for uncertainty, attention or sentiment.