Financial and business cycles in the US: A non-parametric time–frequency investigation

In this study, we contrast U.S. financial and business cycles using turning point and wavelet analysis. These non-parametric methods enable us to identify the key characteristics of financial cycles and assess their relationship with business cycles without imposing assumptions about their cyclical or secular components. Contrary to the conventional view in the literature, we find that financial and business cycles are more similar than generally assumed. Wavelet analysis reveals that: i) since the 1990s, the dominant frequency range of both cycles has shifted towards lower frequencies; and ii) the observed increase in their average length is linked to a change in the relationship between financial and business cycles − from shorter business cycle frequencies (4–8 years) to higher medium-term frequencies (8–16 years). Focusing on the post-1990s period and using a measure of the financial cycle that includes equity prices, we find that the average lengths of business and financial cycles have become more aligned, at approximately 9 and 10 years, respectively. From a policy perspective, these findings cast doubt on the need for macroprudential policy as a distinct tool separate from traditional macroeconomic stabilization policy.