The four largest Greek banks have failed and would also have defaulted had capital controls not been imposed at the outset of the week, due to deposit withdrawals and the ECB’s decision not to raise the Bank of Greece’s (BG) Emergency Liquidity Assistance (ELA) ceiling, says Fitch Ratings.
The Greek banking system’s liquidity and solvency positions are very weak and some banks may be nearing a point where resolution becomes a real possibility. The ECB is responsible for supervising and authorising the four major Greek banks, and the BG is resolution authority for the others.
Resolution of Greek banks, if required, is unlikely to be straightforward. We believe it would be politically unacceptable to impose losses on Greek creditors and that efforts would be made to find a solution which avoids this but still complies with EU legislation. Existing bank resolution laws in Greece are relatively mature, sharing many similarities with the EU’s Bank Recovery and Resolution Directive, although they exclude explicit bail-in of senior unsecured creditors.
In April, Panellinia Bank was liquidated under this framework, with selected assets and liabilities sold to Piraeus Bank. Panellinia’s small size may have facilitated speedy resolution. Potential resolution of any more systemically important banks would be far more complex.
Recapitalisation of Greek banks using domestic resources would be impossible due to the sovereign’s weak financial condition. The remaining EUR10.9bn European Financial Stability Facility notes available to cover potential bank recapitalisation or resolution in Greece were cancelled by the European Stability Mechanism (ESM) when Greece’s bailout programme expired on 30 June.
The Greek deposit insurance fund, which could be used to recapitalise banks contained only around EUR3bn at end-2013. No pan-EU deposit insurance fund yet exists but under the recast Deposit Guarantee Schemes Directive, EU banks can access other countries’ deposit insurance funds. Other EU member states would be highly unlikely to agree to share due to a lack of confidence in Greece and its banking system.
The ESM could still inject funds directly into the banks, but a precondition would be the bail-in of 8% of liabilities and own funds. This would most likely wipe out much or all equity in a failed bank. The equity/assets ratios of the four largest Greek banks were 8%-10% at end-1Q15, but losses incurred since are likely to have reduced this figure. Greek banks have issued limited debt, so ESM rules would theoretically make uninsured deposits vulnerable to bail-in if a bank were to suffer material erosion of own funds before any resolution action.
But any bail-in of uninsured deposits would be politically unacceptable for a Greek government and would also be unlikely to be palatable for Greece’s international creditors, as they overwhelmingly relate to “real economy” SMEs and retail customers. An alternative, more creative, solution would therefore probably be needed to resolve and/or recapitalise Greek banks. This would depend on political goodwill and the outcome of negotiations with creditors, which are still highly uncertain.
The liquidity position of Greek banks is much deteriorated without access to incremental ELA. The ECB only extends ELA to solvent banks and against acceptable collateral. The credit quality of Greek banks’ domestic loan books is exceptionally weak. We calculate that for the country’s four largest banks an aggregated total regulatory capital erosion equivalent to around 4.8% of risk-weighted assets (5% of domestic gross loans) would probably render them non-compliant with the EU minimum total capital requirement of 8%, assuming static risk-weighted assets.
At end-March, these banks reported 90-days-past-due loans equivalent to around 36% of total domestic loans, and arrears may since have risen significantly.
A swift lifting of capital controls is highly unlikely even if there is successful resumption of negotiations with the ECB, IMF and European Commission. Controls on Cypriot banks, lifted in May, lasted two years.