Lien utile
- Publications de la Banque centrale européenne
The leverage cycle in Luxembourg's banking sector
| Numéro | 66 |
| Date | October 2011 |
| Auteur | Gaston Giordana and Ingmar Schumacher |
| Résumé | In this article we investigate the leverage cycle in Luxembourg’s banking sector using individual bank-level data for the period 2003 Q1 to 2010 Q1. We discuss the mechanics behind the leverage cycle in Luxembourg’s banks and show that these banks predominantly adjust leverage by changing both loans and deposits. One of our findings is that Luxembourg’s banks have a procyclical leverage. This procyclicality is not due to marking-to-market but because Luxembourg’s banks are liquidity providers to the EU banking sector. This also explains the different evolution of leverage compared to the US commercial banks (Adrian and Shin [1]) that, even though their balance sheet structure is similar to that of the Luxembourgish banks, target a constant leverage. To further understand what drives leverage in Luxembourg’s banks we empirically investigate the role of bank characteristics as well as real, financial and expectation variables that proxy for macroeconomic conditions in the pre-crisis and crisis period. We find that off-balance sheet exposures have different effects in the pre-crisis and crisis period, and that the share of liquid assets in the portfolio only affects the amount of security holdings. In terms of macroeconomic variables, we find that the Euribor-OIS spread is a significant driver of the build-up in leverage in the pre-crisis period. The reason is that most banks in Luxembourg are either branches or subsidiaries. This, firstly, makes leverage a less relevant indicator of riskiness for investors. Secondly, it implies that in times of liquidity shortages, mother companies or groups demand further liquidity from their branch or subsidiary. The downturn in leverage during the crisis can be accredited to reductions in expectations, which we proxy by an economic sentiment indicator. It can also be explained by increasing bond prices which induce depositors to shift their funds from bank deposits into bonds. We find no important role for GDP growth.
JEL classification: E51, E52, E58, G21, G28. Keywords: leverge dynamics, banking sector, GMM estimation, crisis effect. |
| Téléchargement | Cahier d'étude 66 (pdf, 1 MByte) |
