The role of macroeconomic shocks in banking crises

The macroeconomic reasons for the recent banking crises in the Nordic countries are analysed using an econometric model estimated with panel data from the 1980s and 1990s. Two alternative dependent variables are used: the ratio of banks’ loan losses to lending and enterprise bankruptcies per capita. The explanatory variables are the lagged dependent variable, lagged percentage change in GDP, an income surprise variable combined with lagged aggregate indebtedness, a real interest rate surprise variable combined with lagged aggregate indebtedness, and a deregulation dummy. The innovation in this paper is the use of surprise variables based on macroeconomic forecasts. According to the results, high indebtedness combined with negative macroeconomic surprises contributed to the recent banking crises in Sweden, Norway and Finland. Also the effects of the preceding financial liberalization and lending boom on bankruptcies and loan losses can be traced in the results. The econometric testing did not indicate direct effects of the exchange rate or the terms of trade on the banking crises. Denmark did not suffer a banking crisis because the macroeconomic surprises were smaller there and the initial debt burden was lighter than in the other Nordic countries. This was the result of, among other things, earlier financial deregulation, which was conducted in a fairly balanced way, and a different economic policy regime, as Denmark belonged to the ERM.