Sovereign Risk and the Bank Lending Channel: Differences across Countries and the Effects of the Financial Crisis

Abstract

This article analyses how sovereign risk affects the bank lending channel of monetary policy, and tests whether these effects differed before, during, and after the onset of the financial crisis. This issue was analysed only in the eurozone during the sovereign debt crisis. However, these results are difficult to extrapolate to other countries. First, Europe is the only developed region that has experienced sovereign risk concerns. Second, it has a centralised monetary regime controlled by the European Central Bank, so it is more difficult to adapt monetary decisions to the specific level of sovereign risk in each country. To overcome these limitations, our analysis is based on two country scenarios: (1) developed countries (eurozone vs. noneurozone countries); and (2) developing countries. We find that the role of sovereign risk in the transmission of monetary policy is very complex, and its significance not only varied before, during, and after the global financial crisis, but also in developed and developing countries.

The Bank lending channel and sovereign risk: Effects on significant and less significant banks after the implementation of the Banking Union

“…However, they discovered little evidence of a relationship between sovereign risk and the loan supply after expansive monetary policies. Indeed, Cantero-Saiz et al [ 39 ] established that sovereign risk only affects the bank lending channel in developed countries, especially before the financial crisis. Moreover, during the 2008 financial crisis, the expansive monetary policies carried out did not lead to an increase in bank loans.…”

“…Our initial sample consists of all the banks from the eurozone included both in the S&P Capital IQ database between 2014 and 2020 and in the ECB's official list of significant and less significant banks (755 banks). 6 Moreover, according to Cantero‐Saiz et al [ 39 ], we remove: 1) banks with growth rates of loans and/or deposits greater than 300%; and 2) banks with loans 100 times greater than deposits. After these adjustments we end up with 736 banks.…”

“…Δ ln (GDP) is the real GDP growth rate. This variable controls for the business cycle [ 39 ]. As other studies suggest, there is a causal link that runs from economic growth to bank credit supply [ 47 , 48 ].…”

“…Δi is the change in the short-term money market rate and represents monetary policy changes. This is the variable most frequently used in the literature to test the existence of the bank lending channel with bank-level data [ 39 , 49 ]. To capture the effects of significant and less significant banks on the bank lending channel, we interact Δi with the variables SIG and LessSIG (significant banks and less significant banks, respectively).…”

“…Like previous studies, we also include interactions between the monetary policy variable ( Δi ), and SIZE, LIQ and CAP. As we are interacting continuous variables, similar to previous studies, to facilitate the interpretation, we normalize in Equation (1) the banks’ variables ( SIZE , LIQ , and CAP) with respect to their mean across all banks in the sample as it is shown in Equations (2) , (3) , (4) [ 23 , 39 ]. 10 A i,t is the total assets, L i,t is the level of treasury and temporary financial investments, C i,t is the total capital, N t is the number of banks and T is the number of years.…”