A tale of two tails: 130 years of growth-at-risk

TL;DR

This paper analyzes 130 years of growth-at-risk data using a time-varying model, revealing how risks and their determinants have evolved, with implications for policymakers and financial stability understanding.

Contribution

It extends growth-at-risk analysis over a long period, highlighting the time-varying nature of risks and determinants, and linking financial stress and other indicators to GDP growth risks.

Findings

Risk levels and determinants vary significantly over time.

Upside risks were more stable during the Great Moderation period.

GDP growth distribution has narrowed since the Bretton Woods era.

Abstract

We extend the existing growth-at-risk (GaR) literature by examining a long time period of 130 years in a time-varying parameter regression model. We identify several important insights for policymakers. First, both the level as well as the determinants of GaR vary significantly over time. Second, the stability of upside risks to GDP growth reported in earlier research is specific to the period known as the Great Moderation, with the distribution of risks being more balanced before the 1970s. Third, the distribution of GDP growth has significantly narrowed since the end of the Bretton Woods system. Fourth, financial stress is always linked to higher downside risks, but it does not affect upside risks. Finally, other risk indicators, such as credit growth and house prices, not only drive downside risks, but also contribute to increased upside risks during boom periods. In this context, the paper also adds to the financial cycle literature by completing the picture of drivers (and risks) for both booms and recessions over time.