The opening keynote at the Financial Stability Conference 2019 delivered Martin F. Hellwig, Director emeritus of the Max Planck Institute for Research on Collective Goods. The well-known academic expert recalled developments and outlined steps as reaction to the financial crisis, and offered insights and assessments on the interlinkages of monetary policies, politicial settings, European integration and banking.
Financial Stability Conference 2019
28 October 2019
Keynote – Opening Speech
Governments, Banks and European Monetary Union
Martin F. Hellwig, Director emeritus, Max Planck Institute for Research onCollective Goods
Martin Hellwig opened his speech with the claim that the most important challenge to European Monetary Union is political, rather than economic, and problems should not be treated as merely technical. Money is a source of power. Since central banking is about money, it is ipso facto political and needs political legitimacy. In Germany, the decades of the Bundesbank’s preaching about the importance of price stability served to support the political legitimacy of the Bundesbank, as a counterweight to greedy unions and profligate governments. In the European Economic and Monetary Union (EMU), the problem is that politics is either national or local, and therefore, it is difficult to establish political legitimacy for a supranational institution.
The most important challenge to European Monetary Union is political, rather than economic, and problems should not be treated as merely technical.
These issues can be illustrated in the context of the Banking Union (BU) with two conflicting propositions that firstly, EMU needs BU and secondly, BU lacks political legitimacy. Both propositions are based on the fact that banks are political and politics is national or local but not supranational.
In the first decade of the monetary union there was no problem. By comparison to preceding decades, in fact there was a de-politicization of monetary policy. Discussions were technical rather than political, addressing for example the question whether Otmar Issing‘s two-pillar policy was appropriate or whether the European Central Bank (ECB) should move to inflation targeting, without looking at monetary aggregates.
The financial crisis changed all this. Since then we have seen a re-politicization of monetary policy, in which the central bank‘s supranational character was a source of both strength and weakness. The fragmentation of politics along national lines allowed a dissociation of the ECB from politics, but the politics itself came to be hostile to the ECB. Some of the irritation was due to the crises, but some of it also reflects the deeper fault-lines between national politics and supranational monetary policy making.
The fragmentation of politics along national lines allowed a dissociation of the ECB from politics, but the politics itself came to be hostile to the ECB.
Banks enter this field because, on the one hand, they are an integral part of the monetary system, managing the payment system and playing a key role in monetary-policy transmission, and, on the other hand, they are also an integral part of national and regional politics, where important people, politicians in positions of power, as well as various members of local and national elites, always have wonderful ideas that just need funding from banks.
Paradoxically, the Treaty pays hardly any attention to banks. Banking regulation and supervision remained in the national domain. Articles 127(5) and 127(6) of the Treaty on the Functioning of the European Union (TFEU) say that the ECB should assist regulators and supervisors in what they are doing, but the relationship between banks and monetary policy is not addressed. Moreover, the Treaty says nothing about the Lender of Last Resort (LOLR) function of a central bank. In the early 2000s, there were many Memoranda of Understanding (MoU) on the relation between national and supranational institutions in dealing with banks; these MoUs stated that solvency problems of banks and recapitalizations would be dealt with by finance ministers, liquidity problems of individual banks by national central banks under the auspices of emergency liquidity assistance (ELA), and general, system-wide liquidity problems by the ECB’s intervening in markets. The notion that finance ministers might be unable or unwilling to deal with solvency problems and recapitalize the banks was not considered. Nor was the nature of liquidity problems that might arise.
In the first decade of monetary union, none of this mattered. There were no concerns about bank insolvencies or liquidity problems. As far as the transmission mechanism was concerned, interbank markets seemed to work without frictions. So the entire question of how precisely central bank money is distributed did not matter much, because the financial system used interbank markets to redistribute funds if the initial allocation had not been appropriate. These markets facilitated huge capital flows, especially from Germany to Ireland, Spain, Greece and Italy. To some extent, these capital flows also balanced current accounts imbalances, which is why in the first years of monetary union target balances were small.
All this changed in September 2008. Following the Lehman Brothers collapse there was a complete breakdown of interbank markets. Capital flows were reversed, interbank loans were recalled or not renewed, and securities were sold – much of it across national borders, i.e. financial systems, which had become more integrated in the run-up to the crisis, fragmented again along national lines.
For monetary union, these developments posed three challenges: First, the liquidity crunches in 2007/08 and 2011/12; these crunches were successfully countered with huge liquidity injections, for example, the Long-Term Refinancing Operation (LTRO) of 2011/2012. Second, the transmission problem of how to implement monetary policy when interbank markets are not functioning; this problem was solved by moving to a system of full allotment, where the banks’ applications for central bank loans were automatically granted – at the conditions fixed by the central bank. This measure plays a role in German political debate because, under full allotment, central-bank funding substituted for interbank funding, enabling a flow of funds back from the periphery countries to for instance Germany, which then appears in the Target Balances of national central banks in the Eurosystem. In Germany, the populist interpretation has been monetary union using Bundesbank money to bail out Irish or Greek banks, without regard to the fact that these “bailouts” really benefited the borrowing banks‘ creditors, often German and French banks and that, under the Treaty, there is not such thing as Bundesbank money, it is all ECB or Eurosystem money.
Third, many of the banks to which the Eurosystem lent liquidity support were weak, on the brink of insolvency, and there was no way the ECB could weed out those that should not be supported. This weakness raised questions of principle about supporting “zombies”, as well as questions about the effectiveness of the transmission mechanism. In the case of the LTRO of 2011/12, many funds went to weak banks, which had a preference for lending to their own governments, rather than the real economy. This behavior hampered the transmission of monetary policy to the real economy, but in contrast to the other two challenges, it was not in the ECB’s power to do much about it. Dealing with weak banks was a task for national authorities rather than the ECB. National authorities however were not eager to address the issue.
Many banks to which the Eurosystem lent liquidity, were weak, near insolvency, and there was no way the ECB could weed out those that should not be supported.
In recognizing and addressing problems, banks and supervisors always have a certain leeway because the valuation of assets that are not traded in organized markets involves an element of arbitrariness. If a borrower’s debt service is in arrears, how much do you write down the value of the loan? If you believe that the borrower’s problems are temporary, perhaps not at all because you “are sure” that he will eventually pay up. If you suspect that he will never repay, you may still resist a write-down because a write-down makes you look bad. And who is there to challenge your assessment that the borrower will “surely” end up paying? The supervisors may have the same incentive to procrastinate because when the problems are laid open, they also look bad, especially if they have no ready remedy for dealing with them.
Governments may also like procrastination because it enables them to avoid using public funds for recapitalizing the banks in question. If public budgets are squeezed, they may not even be able to provide for the recapitalizations. In contrast, if weak banks get funding from the central bank and pass the money on to the government, the government gets an indirect access to the printing press, which it may like. The central banks themselves may not be averse to procrastination, because they also like to avoid turmoil.
If weak banks get funding from the central bank and pass the money on to the government, it gets an indirect access to the printing press, which it may like.
All such delays are costly however, because eventual cleanups are much more difficult after long procrastination. There also is a cost to not having enough exit from the industry, i.e. to maintain excess capacity in the industry, inducing aggressively competitive behavior, squeezing margins, and causing significant systemic risk.
BU was created to deal with the problems and the incentive distortions involved in weak banks’ being fully left to national authorities. The Single Supervisory Mechanism (SSM) took supervision out of the national domain, introduced the ECB as a single supervisory institution and gave independence to national supervisors cooperating with the ECB in the implementation of European law. Given the scope for conflict between the different participants, I see the experience of the SSM as positive. It has contributed significantly to loss recognition and recapitalization in European banking.
I see the experience of the SSM as positive. It has contributed significantly to loss recognition and recapitalization. But resolution remains a problem.
But resolution remains a problem. The new rules of the Bank Recovery and Resolution Directive (BRRD) and the Single Resolution Mechanism (SRM) are not working well. The SRM was applied only once, in the case of Banco Popular Español (BPE). In this case, the failing bank was sold to another bank. The sale took place overnight and there were no alternatives because BPE was subject to a run by large depositors and the SRM had no way to replace the funding that was disappearing. The legal norms, the BRRD and the SRM Regulation, simply do not provide for any funding that the bank might need while the resolution authority is choosing its strategy. Perhaps the drafters of these norms thought that funding in resolution might work like funding in an insolvency procedure, where new creditors are often willing to come in because they are given priority over all previous creditors.
For a bank, however, the problem is more serious because so much of a bank’s debt is short-term, provided by institutions that have short-term funding themselves, e.g. money market funds. If this short-term debt is frozen in resolution, the lenders themselves may be in trouble, for example, because their own funding breaks away, as in the case of Reserve Primary and other US money market funds after the Lehman Brothers bankruptcy. If this short-term debt is not frozen, prioritization of new creditors does not work well because there are too many of them relative to the presumed remaining value of the bank’s assets, and, moreover, if they are themselves run upon, they do not have a choice but must withdraw anyway. Without solid funding, however, there is no time for the resolution authority to choose its strategy.
Resolution under the BRRD and SRM Regulation is also impeded by political resistance. The cases of Monte dei Paschi di Siena, Banco Popolare di Vicenza and Veneto Banca, as well as HSH Nordbank and Nord LB all show national authorities going to great lengths to avoid having “their” banks go into resolution and investors in “their” banks subjected to bail-in, i.e. to a sharing of losses if there is not enough equity to absorb them. They prefer to use taxpayer money for “precautionary” recapitalizations rather than to apply the rules and make investors share the losses. They prefer even more not to be forced to acknowledge losses so that recapitalization needs to not become apparent. This is why last year there was so much resistance to the SSM’s proposing to tighten the rules for provisioning against risks from non-performing loans. Since the beginning of Banking Union, volumes of non-performing loans have declined somewhat, but they are still high, somewhat below 600 billion euro versus one trillion euro in 2014; in crisis-affected Cyprus and Greece and in Italy, they still pose significant risks to the entire financial system. Remarkably, throughout these years, the notion of using the resolution mechanism to deal with the problem does not seem to have been considered. Instead there were proposals to put problem loans into government-funded asset management companies, “bad banks”, with clawback provisions making the originating banks pay for eventually remaining losses – if they are able to do so.
Cases so far show national authorities going to great lengths to avoid having “their” banks go into resolution and investors in “their” banks subjected to bail-in.
From the perspective of EMU, improvements in resolution are essential to making Banking Union work. And Banking Union, with a well-functioning supervisory mechanism and a well-functioning resolution mechanism, is essential if monetary transmission is to get back to normal, with a workable symbiosis of the central bank and commercial banks in the creation of money. While many call for the creation of a European Deposit Insurance Scheme, the reform of resolution is actually much more important because deposit insurance does not matter much if the institutions do not even enter into resolution.
From EMU perspective, improvements in resolution are essential to making BU work. And BU is essential if monetary transmission is to get back to normal.
However, Banking Union lacks political legitimacy. Political legitimacy is a result of public political discourse. In the EU, such discourse usually takes place at the local level and at the national level but not at the supranational level. This is particularly true of banks. Banks are an important part of local and national political systems for several reasons. First, they are a source of money. Everybody, citizens and politicians alike, has an idea of what the banks’ should be used for, for example a house for me or an investment in wind energy. A German Landesbank provides the head of its regional government with the power to move a few million without asking a parliamentary committee; this power is worth every euro of taxpayer money, so the regional government will hardly agree to put the bank into resolution!
Second, banking itself is an area of industrial policy. Cyprus, Ireland, the UK, Iceland, and to some extent Switzerland promoted the financial sector as a way of enhancing economic growth. Third, investors in banks may be politically important. In the case of the German bank Hypo Real Estate, which was bailed out in 2008/09, investors included not only Deutsche Bank and Allianz, which might be thought of as systemic risks, they also included the established churches, the public television system, many municipalities and pension institutions, so the bailout forestalled a discussion of quite a different kind of systemic risk. In the case of BPE, the large depositors were mainly local and regional governments, so a bail-in would also have created significant political problems. In the case of the Venetian banks, the bail-in of equity and subordinated debt involved many small entrepreneurs who had invested their savings in preferred stock of the banks on the grounds that such provision of capital was allowing the banks to maintain or even increase their lending to them. The bail-in imposed serious losses on these entrepreneurs and contributed to the revulsion against EMU and Banking Union that shaped the result of the 2018 election in Italy. From the perspective of national politics, the BRRD and the SRM are politically illegitimate.
From the perspective of national politics, the BRRD and the SRM are politically illegitimate.
For someone used to Sunday school teaching, all this is really smelly. The notion that banks should provide funding for the government’s pet projects is a subversion of parliamentary budget authority. To be sure, there is no direct involvement of government money, but there are risks for taxpayers. The risks are kept hidden until they are realized, and then the taxpayer has to pay the bill. By then, it was just bad luck even though, with proper governance and proper foresight about risks, the “bad luck” could have been avoided. The same criticism applies to mercantilist industrial policy using “light touch” regulation to promote a country’s financial industry at a risk of serious damage to the country if there is a crisis. And the mis-selling of hybrid liabilities – and its toleration by the authorities – is altogether scandalous. The bailouts violate a fundamental principle of a market economy, namely that everyone should be responsible for the consequences of their own actions. This principle is a necessary correlate of the freedom to act as one likes as long as one abides by the rules of the law.
The notion that banks should provide funding for the government’s pet projects is a subversion of parliamentary budget authority.
However, political legitimacy is not the same as legitimacy in the eyes of Sunday school teachers. Political legitimacy is established in political discourse. Political discourse at local and national levels however is driven by some of the very parties that benefit from conflicting with Sunday school teaching. From their perspective, Frankfurt and Brussels are illegitimate intruders. The European Commission, the SRB, the SSM and the ECB unfortunately are not present in local and national policy discourse, where blaming outsiders is a prominent strategy. At this level, there is hardly anyone who defends the need for the rules of the Banking Union and the interference that they mandate.
In local and national policy discourses there is hardly anyone who defends the need for the rules of the Banking Union and the interference that they mandate.
The sense of illegitimacy has been even stronger in those cases where the ECB has used its power over banking systems in order to take sides in distributive conflicts. In the case of Ireland, the letter written by the president of the ECB to the Irish government in the fall of 2010 demanding that there should be no bail-in of senior unsecured creditors (mainly German banks and the ECB itself), did a lot to de-legitimize the ECB with the Irish population. In other cases, such as the letter to the Italian prime minister in August 2011 or the ECB‘s role in the Greek crisis in 2015, there also was a view that the ECB was too much aligned with the creditors. To be sure, the typical German assessment was just the opposite, namely the ECB was (and is) seen as a mechanism of distribution of resources from Germany to the debtors in the peripheral countries. But the very difference between the different national discourses shows that the interplay between the supranational institutions in the monetary union and national policies is highly dysfunctional.
The very difference between the different national discourses in the EU shows that the interplay between the supranational institutions and national policies is highly dysfunctional.
This conflict is ultimately irresolvable. One often hears that the bank-sovereign nexus must be cut. This formulation was used at the Summit of June 2012 that decided in favour of Banking Union, and it has been used in discussions about an European Deposit Insurance Scheme (EDIS) ever since. But there always is a bank-sovereign nexus. Member States are sovereign and have sovereign powers over “their” banks, not just over supervision and resolution but also over employment law, tax codes and even ownership. (Remember that Mr. Varoufakis thought of nationalizing Greek banks!) Transnational or trans-European banks, as promoted by some, would alleviate the problem. But then the government of the country where such a bank has its seat still has sovereign power over it, and the governments of other countries where it has subsidiaries may want to impose their sovereign power as well.
What can be done? The answer is “muddling through”, perhaps with some small improvements. First, we need some strengthening of political legitimacy at the supranational level. At the supranational level, we need more public political discussion, which presumes more powers of the legislature and an executive that provides a counterweight to the regulatory administrations. In this context, we should think about a European finance minister and a European budget in terms of political processes and political legitimacy, rather than technocratic management. Second, the BRRD and the SRM need to be reformed to take account of the fact that bailouts are sometimes needed and, moreover, that the Commission’s state aid control may not be well suited to controlling such bailouts. State aid control is aimed at competition policy, rather than financial stability. In the case of the Italian banks, negotiations with the Commission about state aid control have contributed significantly to procrastination, which is harmful for financial stability. Reliance of state aid control on the private investor rule is inappropriate if some of the private investor participation is based on bail-in, as in the case of Banca Monte dei Paschi di Siena. Moreover, the judgement that exit makes bailouts acceptable because there is no more threat to competition is inappropriate if investors take the government’s bailout of some investors in the current case as a signal of its stance in future cases.
What can be done? “Muddling through” with some small improvements. First, we need some strengthening of political legitimacy at the supranational level.
Third, we need to take account of the loss of information capital of banks that are closed, especially when a crisis covers an entire region and many banks are affected. The question then is how to deal with an entire regional banking system, when the region and its banks are subject to an asymmetric shock, e.g. from changes in world trading conditions, as experienced by some of the industries in northern Italy whose competitiveness in European and world markets was strongly affected by the accession of Eastern European countries to the EU and by the expansion of Chinese exports. Given their experience in 1992, Swedes would say: nationalize, re-organize, and re-privatize, leaving the government to bear residual losses, but the BRRD does not leave room for such a strategy. In summary, the entire system of legal norms for bank resolution needs to be rethought.
Martin Hellwig concluded that ever since its beginnings in the 1950s, academic economists had reacted to European integration the way an engineer would react to a bumblebee. Just as engineers “know” that bumblebees cannot fly because they are too heavy and their wings are too small, so economists have always known that European integration cannot fly. But somehow bumblebees can fly despite the engineers’ assessments, and European integration has muddled through despite the economists’ skepticism. We must hope that Europe will also find ways to muddle through in the matter of Banking Union.
More on FSC Conference 2019