A Theory of the Credit-to-GDP Gap: Using Credit Gaps to Predict Financial Crises

The credit-to-GDP gap (credit gap) is thought to be a promising leading indicator of financial crises, but the trend of the credit-output ratio must be appropriately estimated before credit gaps can be used for this purpose. To appropriately estimate this ratio, it is necessary to know the mechanism by which it is generated, but this mechanism has not necessarily been sufficiently explained. In this paper, I examine the mechanisms underlying the creation of credit-output ratios and credit gaps, and show that the credit-output ratio has an exponentially increasing upward trend. A log-linear trend of the credit-output ratio therefore should be estimated when using the credit gap to predict financial crises. Estimates of log-linear trends of the credit-output ratio and credit gap of the U.S. indicate that the financial crises in 2008 and 1990 can be predicted if the credit gaps are well estimated and considered as an important leading indicator of financial crisis, even in “real time” (i.e., from the perspective of those years, not the present).