Measuring credit procyclicality: A new database

Abstract

Today, the data available to estimate the credit cycle involve a trade-off between country coverage and frequency. In addition, there are pending methodological issues to estimate credit trend and cyclical components. To address these limits, we build a new database on credit metrics that relaxes the trade-off and includes credit procyclicality measures along three alternative methods- HP filter, the modified HP filter and basic SSA. The credit gaps in our database are statistically consistent with bank credit gaps estimated with BIS data and they have the advantage of being available for 163 countries instead of 43 countries. Armed with this new and expanded data, we investigate classic empirical questions in the literature with a comparative perspective across countries by income level.