Measuring Financial Cycles : Empirical Evidence for Germany, United Kingdom and United States of America

This study contributes to the literature by identifying the most appropriate factor to detect and measure Financial Cycles, similar to Gross Domestic Product (GDP) for Business Cycles. Four financial variables were included in the study: Credit, House Prices, Share Prices and Interest Rates. The filter used to estimate and extract the cycles from the original time series was the Christiano and Fitzgerald (2003)'s one. Then, three methods, namely the Concordance Index, the Granger Causality Test and the AUROC Test, were used to identify which of the four variables is the most accurate proxy to measure and estimate financial cycles. In all of them, the results pointed to the same variable: Share Prices. A comparison between Share Prices and GDP shows a higher capacity of the financial variable to predict financial and economic crises than GDP, which justifies the recent increasing interest of macroprudential policymakers on Financial Cycles. Our conclusions are robust to different time periods and alternative filtering procedures