Measuring Financial Cycle Length and Assessing Synchronization using Wavelets

Identifying financial cycle dynamics is an important endeavor for researchers and policymakers alike, due to the new architecture of the macroprudential framework in which mitigating excessive credit growth, through countercyclical capital buffers, plays a central role in limiting the likelihood of future crises. The present paper focuses on measuring financial cycle length for a series of developed as well as emerging economies by applying Continuous Wavelet Transform (CWT) techniques, which have the ability to decompose time series on a wide range of frequencies and identify statistically significant cyclical behavior. Using credit-to-GDP data collected by the Bank for International Settlements (BIS) for 13 countries, our results confirm the established hypothesis that financial cycles are significantly longer than business cycles, in the case of developed economies, underpinning the European framework for setting the countercyclical buffer rates. Using Wavelet Coherence measures, we find statistically significant co-movement in time-frequency between several EU members, as well as tighter relationships between euro area members. The main conclusion is that policymakers from emerging economies should monitor financial cycle dynamics more closely, using the additional assumption of shorter cycles, in order to identify the build-up of systemic risk through excessive credit growth in a timely manner.