|
|
This study uses the experience of the restructuring of the 72 banks in the euro area that received capital support during the period 2007 to 2014 to provide a rough estimate of the funds required for bank resolution, should a crisis of a similar magnitude recur. The aided banks, representing approximately 45% of the total euro area bank assets, experienced losses totalling €313 billion. Most of these losses arose in significant banks.
About half of the banks that received state aid during the past crisis might have received public support under the new framework of the BRRD as well. Of the 62 banks for which there was data available, 17 banks did not report a single year in which the capital fell short, while for 14 banks a bail-in would have been sufficient to meet the minimum capital requirements during the sample period. The remaining 31 banks, accounting for only 6% or € 1.9 trillion of the total euro area bank assets, had losses requiring more funds than obtained through a minimum bail-in of 8% total liabilities, including own funds.
The SRF would have covered only a fraction of the losses given that its contribution is limited to 5% of the total liabilities. The reason is that there were many cases in which the total losses were relatively large (as a % of total liabilities). The banks that would have demanded funds from the SRF were responsible for about two-thirds of the total losses. Those banks could have asked the SRF to inject up to approximately €72 billion to cover resolution costs (this is an upper limit, there would have been no obligation for the SRF to do this). The contribution of the private sector through (excess) own funds and bail-in of creditors (totalling €153 billion) would have been much larger than the maximum contribution of the SRF (€72 billion).
Our estimates seem fairly robust. Changing the assumptions for the deleveraging of activities or alternative methods to calculate the losses only produces relatively small changes in total funds required. The key variable remains the level of capital demanded by the supervisor after re-capitalisation. If this level had been set systematically higher than 8%, say at 12% (of RWA) as in Cyprus, banks with moderate losses would have needed more funding (the SRF would anyway not have been able to put more funding into the banks where it would have hit the 5% of liabilities ceiling). In this case the SRF might have needed up to €101 billion.
The SRF will start collecting funds only from 2016. The target is 1.0% of insured deposits, which would yield a total of €55 billion, but that will be available only at the end of an eight-year transition. Moreover, the SRF could call on three years of advance payments, yielding a total potentially available to the SRF of about €76 billion, enough to cover, once fully contributed, the estimated needs for a ‘normal’ recapitalisation scenario.
Alternative funding (for restructuring), which in any case is difficult to quantify, would be borrowings from other resolution funds and third parties, such as potential contributions from deposit guarantee schemes.
Our estimate of the potential funding needs results from observations from a once-in-a-generation crisis and covers a period of eight years. This suggests that if another exceptional crisis occurred alternative funds might be called. The SRF is of a sufficient size to handle the failure of one or several small and medium sized banks, but it is insufficient to cope with the failure of the largest banks in the first two years of crisis. The SRF will probably not have sufficient funds to handle a general banking crisis, in such a situation bridge financing of up to €45 billion might be needed. Taking into account some safety margins a bridge financing facility of a similar size to the €60 billion of the European Stability Mechanism`s direct recapitalisation instrument seems appropriate.
Our broad conclusion is that a backstop facility, in addition to the existing alternative funding measures, might be required during the transition period, but only in extreme circumstances.
It needs to be borne in mind that the funding of the SRF would be an investment in the equity capital of the restructured banks. We have not been able to estimate the value of the equity the SRF would hold today if it had existed during the crisis. But the fact that most of the national rescue operations did not result in large losses suggests that a large part of the funding would have represented an investment, not a loss. The same might be the case for funds provided under a bridge financing facility.
Finally, in this paper it is assumed that the SRF will de facto only be used for solvency support, because its contribution is too restrictive and its size too small to deliver a credible contribution in the event of liquidity problems. For liquidity the SRF is likely to depend on the ECB, which may entail additional uncertainty in times of financial distress.