The role of oscillatory modes in US business cycles

Andreas Groth , M. Ghil (1, 2, 3) , Stéphane Hallegatte (4) , Patrice Dumas (4)

1 IGPP - Institute of Geophysics and Planetary Physics [Los Angeles]
2 AOS - Department of Atmospheric and Oceanic Sciences [Los Angeles]
3 LMD - Laboratoire de Météorologie Dynamique (UMR 8539)
4 CIRED - centre international de recherche sur l'environnement et le développement

We apply multivariate singular spectrum analysis to the study of US business cycle dynamics. This method provides a robust way to identify and reconstruct oscillations, whether intermittent or modulated. We show such oscillations to be associated with comovements across the entire economy. The problem of spurious cycles generated by the use of detrending filters is addressed and we present a Monte Carlo test to extract significant oscillations. The behavior of the US economy is shown to change significantly from one phase of the business cycle to another: the recession phase is dominated by a five-year mode, while the expansion phase exhibits more complex dynamics, with higher-frequency modes coming into play. We show that the variations so identified cannot be generated by random shocks alone, as assumed in "real" business-cycle models, and that endogenous, deterministically generated variability has to be involved. © OECD 2015.

Mots clés: Advanced spectral methods; Comovements; Frequency domain; Monte Carlo testing; Time domain; JEL classification: C15; C60; E32

Domaines: Sciences de l'environnement; Economies et finances

Andreas Groth, M. Ghil, Stéphane Hallegatte, Patrice Dumas. The role of oscillatory modes in US business cycles. OECD Journal: Journal of Business Cycle Measurement and Analysis, 2015, 2015 (1), pp.63-81. ⟨10.1787/jbcma-2015-5jrs0lv715wl⟩. ⟨hal-01239779⟩