The Application of Multiple-Output Quantile Regression on the US Financial Cycle

Michal Franta

The paper demonstrates the benefits of multiple-output quantile regression for macroeconomic analysis. The domestic financial cycle, which is characterized by the co-movement of credit and property prices, is a natural subject of such methodology. More precisely, I examine the tails of the joint distribution of US house price growth and household credit growth since the late 1970s to shed some light on the evolution of systemic risk and its links to various economic and financial factors. The analysis finds that the crucial indicators include the banking sector’s exposure to household credit, household leverage, house price misalignment and financial market volatility. This contrasts with the negligible role of real-economy factors. In addition, it is shown that the multiple-output quantile regression framework is a useful tool for forecasting and tracking systemic risk over time. The sustainable growth of house prices and credit can be distinguished from their growth accompanied by the rise in systemic risk to guide policymakers on an appropriate response.

JEL codes: C32, E44, G10

Keywords: Domestic financial cycle, multiple-output quantile regression, systemic risk

Issued: March 2023

Download: CNB WP No. 2/2023 (pdf, 815 kB)