A European strategy for sour loans
EBA’s proposal to establish a European “bad bank” illustrates the determination of Europe’s authorities to reduce non-performing loans more swiftly, and to do so without sharing risk or deviating from the European Bank Resolution Directive. The idea is not so much to relieve bank balance sheets as to promote the emergence of a secondary market in non-performing loans.
Non-performing loans (NPLs) topped €1 trillion in the European Union (EU) in June 2016, according to the European Banking Authority (EBA). They accounted for more than 10% of loans in ten countries, sitting at 47% in Greece and Cyprus, over 19% in Portugal and Slovenia, 16% in Italy, and between 12% and 15% in Ireland and a number of south-eastern countries. Central and northern European countries, meanwhile, are posting ratios below the EU average. The build-up of sour loans on bank balance sheets poses a systemic problem by acting as a drag on sector profitability, creating latent recapitalisation requirements and interfering with the pass-through of monetary policy via the credit channel. Although situations vary across countries, Europe’s authorities are hoping to create a common framework for NPL resolution. In its capacity as supervisor, the European Central Bank (ECB) has addressed part of the problem by issuing guidance on NPL management by banks. The guidance calls on banks to set up dedicated teams and adopt quantitative targets for reducing NPL stock, typically over three years. The other pillar of the strategy is based around creating a secondary market in NPLs to outsource their management. With this in mind, in late January 2017 EBA suggested establishing a European bad bank that could potentially hold €250 billion worth of bad loans (a network of national bad banks would be an alternative – and less politically sensitive – solution). The plan’s designers point to several benefits, including harmonised information gathering on assets, reduced informational asymmetries between banks and investors, narrower bid/ask spreads, economies of scale in managing NPLs of smaller banks, and a knock-on effect on the remaining stock of arrears. Who foots the bill? EBA is proposing that bad bank structures be funded by issuing state-backed bonds under two conditions: (i) no risk-sharing between countries; (ii) compliance with European rules on state aid, meaning that guarantees must be remunerated. However, bad banks would bear no losses. After buying loans at their “real economic value” – and it remains to be seen how this is to be determined – bad banks would be allowed to dispose of them at a lower price if necessary within three years and pass the loss on to the initial lenders. Recapitalisation of lenders by national governments would then be possible subject to a bail-in involving junior creditors. For Italy, the return of the bad bank concept has particular resonance. Despite provisioning efforts and increased portfolio disposals in 2015 and 2016, Italy’s domestic banking sector is still saddled with a stock of impaired loans (sofferenze). Although they have become less systemic following the introduction late last December of a plan to provide state aid to struggling banks, these loans carry a definite cost for the economy (curtailing credit for businesses) and mean that a portion of the sector remains vulnerable (access to market financing, profit and even solvency eroded, liquidity risk). The gradual decrease in outstanding sofferenze that began last year thanks to the pick-up in growth will not be enough for a radical clean-up. Admittedly, the authorities did introduce a series of measures last year that: - help banks manage future credit risk more effectively (patto marciano, pegno non possessorio); - enable banks to securitise sofferenze more extensively (state guarantee for senior tranches, bigger tax breaks on provisions); - support banks that are unable to recapitalise on the markets (Atlante fund, which receives contributions from healthy banks); - launch a secondary market in sofferenze (Atlante II fund, which will invest in mezzanine and junior tranches of securitised sofferenze).
Yet all these measures will not be enough to step up the pace of consolidation. As innovative as they are, the new measures merely illustrate the authorities’ ongoing efforts to avoid the severe restrictions associated with the bail-in regime for just over a year, when assistance of the sort received by Spanish banks in 2012 would be much simpler. Remember that the stock of sofferenze began to be a source of concern after the new state aid discipline came into effect (2013), at a time when Italian government finances were already in no position to fund an ad hoc ring-fencing structure for the system.
In late December, following the failed market recapitalisation of Monte dei Pasch, Italy’s weakest bank, the government was forced to take a step towards setting up a bad bank for the banking system – its outlines and procedures are still far from being determined – as part of a plan to stop any potential risk of a wider banking crisis. The plan also extended the government’s support for sector liquidity for six months by providing a remunerated guarantee for new bond issues and set aside a budget of €20 billion for preventive recapitalisation. All in all, the idea of creating a bad bank, even if only at the Italian level, certainly deserves to be looked at again, given the progress made. In particular, the high provisioning ratio now reached across the sector as a whole (46% for NPLs generally and between 60% and 70% for sofferenze) could pave the way for wholesale deconsolidation of sofferenze at a lower cost than in the past, which we believe to be the only way for the Italian banking system to regain its position in the workings of the economy.
On a more general note, whether in Italy or Europe as a whole, a bad bank is not a miracle solution. For one thing, banks continue to bear the risk until the final disposal price is known, and that price will always be well below the net book value, i.e. even after taking previously released provisions into account. For another, country situations vary widely, and we learned from the last crisis that the effectiveness of bad bank structures was verified only for uniform assets (real estate) that are subject to simple restructuring and liquidation rules. Accordingly, the diversity of NPL types in some countries and the complexity of bankruptcy laws and legal systems raise questions about the comparative advantage of bad banks in countries currently reporting the highest NPL levels.
Delphine Cavalier et Anna Sienkiewicz