Patrick Honohan: The future shape of banking
Address by Mr Patrick Honohan, Governor of the Central Bank of Ireland, at the IIEA Conference on "The future of banking in Europe", Dublin, 2 December 2013.
3 Dec 2013
Long before the single currency, the potential for a single market in European banking was created by the Second Banking Directive in 1989. A bank licensed anywhere in the European Union could thereby provide services throughout the entire Union. The most basic legislative ingredient for what is now thought of as a banking union was created. Curiously, this powerful piece of legislation did not have the impact that one might have expected. Pan-European banking supermarkets did not rapidly emerge. Instead, the stronger forces of financial integration were global ones. An Australian bank bought one of the six full service Irish banks; expanding Irish banks bought into (then non-EU) Poland and the US more than into continental Europe.
It is not always easy to predict where the future of banking lies, whether for Europe or for Ireland. More dependent for financial services on banks than on markets in debt and equity, Europe has had to work harder to repair its financial system following the banking collapse of 2008 and this task is complicated by the subsequent widespread debt overhang.
Much of this conference has been devoted to the question of banking supervision and resolution, matters of official policy in the proposed Banking Union. I would like to look further ahead to ask what type of banking system we may see - or wish to see - in Ireland within the Banking Union.
Let us recall that banks have to perform a multidimensional set of functions in the economy:
- ensuring that payments are reliably made for the transactions that underpin all economic activity;
- providing safety for the liquid assets of the general public: households and firms;
- evaluating creditworthiness of borrowers, monitoring their financial condition and ensuring that loans are repaid;
- acting as broker for a range of other financial services.
All this we expect banks to do. And that is part of why it is so difficult to resolve a crisis which - to a greater or lesser extent all over Europe - has weakened both banks' financial strength and their operational capacity to deliver satisfactorily on all of these functions.
What might change in Europe?
As the banking systems of Europe are repaired, how will their organisation differ? What position will they hold in European finance?
Two types of indicator of what might be in store could be sought by (i) comparing the current situation with that in other advanced economies in order to assess whether the gaps between the systems might tend to be closed and (ii) extrapolating recent trends and technological innovations.
As far as structural differences today between the EU and US financial systems are concerned, two striking differences are in integration and in the relative importance of bank and market finance.
The US banking market is along several dimensions more integrated than that of Europe, and the trend has been continuing. Just to give one related indicator, concentration, I noticed from a recent study that the share of the top 50 banks in the US grew by 10 percentage points during the first decade of the new millennium (continuing a long trend), whereas the share of "large banks" in the EU shrank by 5 percentage points between 2007 and 2013. Geographic segmentation along country lines has also increased in recent years in Europe, whereas the traditional segmentation along state lines for retail banking in the US has progressively weakened. Perhaps the increasing segmentation in Europe is an immediate post-crisis tendency - driven for example by EU competition requirements in relation to State-aid granted to failing banks, as well as by the pressures in the euro area - and will be reversed. I hope so.
The US has long had a larger nonbank financial sector, including organised securities markets, which provide a useful "spare tyre" for SME finance. In Europe, nonbank market finance has a much smaller share. Even if this gap partly reflects the wider range of financing activities traditionally undertaken in the universal banking model which has long characterised continental banking systems, it is surely possible that with the weakening of European banks, the time may be ripe for this gap to be closed.
The growth of nonbank finance in Europe to close the gap may be flavoured and facilitated by some recent financial technological trends, already observed in the US and elsewhere. For example, what is the long-term potential of innovations such as internet-based equity crowd-funding and peer-to-peer lending? How much evolutionary growth will there be in lending by collective investment schemes, trusts and unincorporated financial vehicles and how will they impact the role of banks? And I suppose one cannot mention technological change without at least alluding to Bitcoin. Some nascent developments will surely fizzle out, but others may not. All raise controversial and complex regulatory issues which need further study. But these and other trends could serve to shrink the share of banks in European finance.
Nevertheless, universal banking can be a very profitable business, not least because, when banks fail, their creditors are often bailed out by Government. Anticipating this, investors require a smaller premium to fund the bank. The value of this implicit government guarantee can be significant, especially during market turbulence.
In Ireland for the past five years banking has not been profitable. This is largely because of the difficulty banks have had in coping with the legacy assets on their books. A large proportion of these assets yield little by way of interest because of the low euro area money market rates to which their yield is contractually set. A large and - at least until very recently - growing, share of the assets is also not fully performing. The Irish banks have set aside, and will still need to set aside, sizable provisions to enable them to absorb the inevitable credit losses on some of this portfolio, as it becomes more evident what part of the loans cannot be paid in full, and since the property-related collaterals that backed the loans at the outset no longer suffice as a backstop.
Faced with these problems, some foreign-owned banks, whose parents also are weakened by the crisis, have decided to withdraw from the Irish market, either altogether or from large segments of it. Two significant Irish-controlled banks - Anglo and INBS - have been liquidated. Two others, AIB and EBS, have been merged. Aside from the group of export-oriented banking firms operating in the export-oriented IFSC context, the banking landscape is now much sparser than it was.
The ability of Irish banks to make their full contribution to financial services needed for the recovery and into the longer term depends on their being fully repaired after the far-reaching damage wreaked in the boom and its aftermath. Much has been done here already in terms of recapitalisation, deleveraging and reduction in the cost base. But the case-by-case resolution of impaired loans, both for the SME sector and the household sector, still has a long way to go. At last there are clear indications that the operational processes put in place to accomplish this goal are gaining traction and that the problem can finally be brought under control.
What will Irish banking look like at the end of the decade; what would we like it to look like? For the sake of argument I will assume that the legacy issues will by then have been put behind us. Official policies and the mandates and operational imperatives of the surviving banks are focused on ensuring this.
Banking in 2020 will have to be composed of entities that are financially solid and profitable, delivering the services needed for strong performance of the non-financial economy.
The prospective organisational shape of the banking system can be considered along a number of dimensions. Few or many? Big or small? Specialised or full service? Internationally diversified or locally focused? Regional or national? Commercial or mutualist? Under MNC or local control? With public or private ownership? Let me discuss these aspects one-by one, recognising of course that they are not fully independent but raise overlapping issues.
Few or many?
The question here is about ensuring a sufficient degree of competition. And there is an increasing focus on this. Making sure that banks compete is not just a question of numbers, however, or the preventing of formalised collusion. It may not be necessary to have more than a handful of banks to ensure that they will have the incentive to compete for business and offer keen pricing (without sacrificing prudence).
Many writers have cautioned against excessive competition for market share in banking, and there have even in recent years been distinguished advocates of, for example, deposit interest ceilings to inhibit reckless competition for funds. In the run-up to the crash in Ireland, banks had unlimited access to funds in the global money market, so they had less need to drive up deposit rates. They sourced funds, not from reckless foreign speculators (there were such, but they did not buy low-yielding Irish bank bonds!), but in a highly liquid international market for what were considered safe bank bonds and deposits. Instead, the destructive competition was for market share in loans to property developers, and to those who would buy what the property developers and construction firms would build.
Unless they sold out before the crash, shareholders of the banks lost out because of this seemingly reckless competition, as did the holders of subordinated bank debt. Of course senior bank management did a lot better on average, given their very sizable earnings during the boom. But it was the Government that picked up the tab for the remaining creditors - mostly depositors and a smaller sum in respect of bondholders. Excessive competition can mean excessive risk-taking, and the losses are often socialised (though less so under the proposed resolution directive).
At present, some interest rates and other charges are much higher than would happen in a healthy and fully competitive banking market, but some are below - especially the legacy tracker mortgages whose interest tracks the ECB policy rate which is 4 percentage points lower now than before the crash. Still, it is too easy to attribute all of the much higher spreads on SVR mortgages to the banks' need to cross-subsidise trackers and to cover loan-losses on the back-book. The rising spreads on UK standard rate mortgages also since the crisis suggests that some of the rise should be attributed to belated recognition of the true risks and costs of mortgage lending.
A fully competitive market would need not so much a plethora of many banks, but a sufficient number of sound banks free of crippling legacy issues. If weakness of their parents, and concern about the prospects of banking profitability, have discouraged some foreign owners, new foreign entrants and re-engagement by those now hibernating or marking time will eventually shave spreads down to new equilibria. The domestic banks will certainly need to be fully repaired by then.
Big or small?
Globally, the question of optimal banking size has been a matter of considerable dispute. Are there economies of scale and scope in banking? Or do bigger banks have the possibility to make greater profits because their creditors are reassured by the implicit government guarantee provided by their having grown "too big to fail"? Some policy makers have argued for breaking up global mega-banks until this implicit guarantee no longer has much force. Others have tended to the alternative approach of simplifying the business of big banks and modularising them so that if part of the whole gets into trouble, it can be isolated and the remainder resolved without the need to call on Government support to keep essential parts going.
None of the Irish banks ever got to the €3 trillion scale of the largest European megabanks, but even at €200 billion in total assets for each of the top two - well in excess of national GDP - and €100 billion for the third, each of the largest nationally-controlled Irish banks was on a scale that could and did create problems that threatened the solvency of the State itself. Since then, these figures have shrunk by more than a quarter, though the total assets of the two largest remain comparable to GDP in size.
Can we risk having banks of this size for the long-run? Against this concern there are of course economies of scale to be considered. A full-service bank needs to have IT systems of a sophistication that undoubtedly needs a degree of scale to achieve acceptable unit costs. Personally, my reading of the evidence is that a €100 billion bank is not much larger than is needed to deliver the full range of banking services at the lowest unit cost.
Foreign or domestically controlled?
One way of achieving low unit costs is, of course, to rely on multi-national corporations in banking as we do to a large extent in the manufacturing and internationally traded non-financial services sector already in Ireland.
Having foreign ownership can help absorb problems. Even if some of the entrants of the early 2000s (Danske, Halifax-Bank of Scotland, Rabo) and earlier (Royal Bank of Scotland, Kredietbank) - all coming from within the European Union - stoked up the over-heated credit competition in the boom, they certainly also absorbed a sizable chunk of the losses in the bust. And foreign shareholders owned well over half of AIB and Bank of Ireland when the crisis broke. (At present, foreign shareholders hold more than 70 per cent of the equity in Bank of Ireland).
But what about in normal times? To what degree is it safe to rely on foreign controlled entities for delivering the full range of banking services? The international literature notes the tendency for foreign-owned banks to serve mainly large companies and wealthy individuals when they enter a banking market; but it also emphasises that their entry can drive the incumbent banks to refocusing on SMEs and middle-income households, with a beneficial effect on availability and pricing of services to this clientele.
Perhaps then the best mix is to rely on foreign-owned concerns for those services that require large scale, while not neglecting the importance of the local relationship banking that is necessary for supporting the SME sector.
Internationally oriented or domestic oriented?
Does that mean that those banks that are, in the future, locally controlled should not be looking to export business? This is a question to which one would tend to give a different answer after the crisis than before, though maybe the centralisation of banking supervision and resolution in the euro area will change it again.
Looking at the experience of the Icelandic banks, whose export-orientation had got to the point where over 70 per cent of their net income was said to be generated from international business, but which then spectacularly blew up destroying a sizable segment of the Icelandic economy even though their foreign creditors were not bailed out by the Icelandic government, it would appear that the potential for future international expansion of banks headquartered in small countries may be doomed.
Irish banks too became quite internationalised in the boom, not just through their foreign subsidiaries (or through their traditional branch-bank operations in Northern Ireland and Britain) but also through the financing of Irish developers' international property portfolios. Outside of the UK substantially all of that has been divested. Should or could it be recreated? In principle, one could say that the creation of the SSM and a common resolution mechanism offers the potential for some future Irish-controlled bank to become pan-European without implying a threat to the Irish sovereign, at least if its funding and lending is effectively ring-fenced from the national economy. After all, that is consistent with the logic of the single market. But I feel that such a prospect lacks the ring of authenticity. This is probably not a path to seek out in the immediate future. While recognising the UK links, the main focus of Irish banks should now be on Ireland.
Regional or sectoral?
Should one go further and imagine a banking system that is regionally specialised? For a country of Ireland's size and economic homogeneity - even if a distinction can be made between the greater Dublin area and the rest - there seems little reason to press for a downsizing and specialising of national banks to a regional focus.
Even less case I would think, to promote the idea of sectorally specialised banks (for agriculture, housing, commerce, etc.). After all, specialisation in property-related lending was the downfall of all the banks. The demise of both of the former state-owned sectoral banks ACC and ICC when they were turned by their acquirers into predominantly property lenders only adds an ironic twist to this conclusion. A degree of sectoral diversification not only insulates a bank from such risks, but helps build a more comprehensive information base on which to rebuild relationship lending. Besides, the sectoral mix of banking business is already changing, with service sectors looming much larger than before; their evolving financial product needs are quite different from those of manufacturing, construction or the primary industries. A bank that merely focuses on one sector misses the cross-sectoral interactions and business opportunities that require financial services.
To say that national banks need not be pressed to downsize or split into regionally or sectorally specialised sub-units is not to dispute the value of the community bank. A community bank may be able to offer the range of retail services that low or middle-income households and small and micro enterprises require even if operating on a scale well below €1 billion - and do it professionally and efficiently, and with sensitivity to local conditions.
Though we don't use the term, already in Ireland we have a large involvement in community banking with an astonishingly deep credit union sector (it has greater penetration measured by the ratio of accounts to population than in any other country apart from tiny islands in the West Indies). Its governance is mutualist rather than capitalist: this strengthens the bonds between the members, while limiting the potential for expansion. Though quite large in aggregate (reporting almost €14 billion in assets), the sector is very fragmented with almost 400 separate credit unions around the country.
It is not to be wondered that the credit union sector was weakened by the crisis, though not to the same extent as the mainstream banking sector. The damage was limited by the conservative liquidity requirements that have long been imposed on the sector, supplemented by regulatory restrictions on new lending by weaker credit unions.
The Government has established and provided cash for two funds, a resolution fund for dealing with credit unions which are failing but for which an assisted transfer to another institution is warranted (such transfers may, for example, work out cheaper to the resolution fund than liquidation to the deposit guarantee fund), and a restructuring fund designed to strengthen credit unions not in danger of failing. This framework offers the opportunity now to ensure that community banking can be put back on a solid basis.
My own view, which is shared by some - though not all - experts and insiders, is that a degree of consolidation in the credit union sector is essential if it is to perform effectively for its members in the future. This would mean some dilution of the very local control enjoyed by the members of some very small town-based credit unions, but it would be more than compensated by greater operational efficiency, range of services, and solidity. Operating a modern credit union requires a greater degree of organisational skill than in the past, and the inevitable intensification of regulatory burden has also contributed to the need to reduce overall operational costs by consolidation, while still retaining a sufficient local presence. That is the way community banking has trended all over the world, especially where it is still thriving.
If I have conveyed a picture in which a handful of large multinational players actively caters to the larger firms and to those requiring more sophisticated financial services, alongside a handful of nationally controlled banks focusing mainly on delivering to the needs of the middle-market, while several dozen mutually-owned community banks are operating at county or large enterprise level, then this was my intention.
It is subtly different from the present situation where the two big players are nationally controlled, with a handful of other banks - one nationally controlled, the others MNC subsidiaries but nationally supervised banks - present but somewhat reticent. And a very fragmented credit union movement.
Two of the banks are wholly state-owned - though the Government is operating under relationship agreements (agreed with the Troika) which limit the extent to which Government intervenes in day-to-day operations of these entities.
What role will state ownership have in the future? Most likely, Government will have no long-term interest in holding a controlling stake in any of the existing banks. Indeed, it is likely to want to reduce its debt by disposing of its bank shares as quickly as is realistic. (Ideally, the capital needed to keep these banks going would come from a wider European fund, but that is a political debate at European level, to which I can make little useful contribution at this point.)
What about the vision of a Government-sponsored business or development bank, which recently has come back into vogue? There is a long and chequered history of such entities, and a sizable literature evaluating their successes and failures. Before the crisis the failures seemed to outweigh the successes by so much that few observers saw much value in reintroducing national development banks. (As mentioned, our own two, ACC and ICC, were sold off to foreign concerns at the turn of the millennium.) But the crisis and the sweeping failure of privately-owned banks has reignited the debate. One of the success stories is the German Government's KfW. Perhaps through the current interaction between the Irish Government and KfW we will learn, in a hands-on way, more of what it takes to make a Government-owned development bank a success.
Back to the future?
Was there a golden age of Irish banking that we should hanker after? Probably not. Yes there was a century and a quarter after the Munster Bank episode of 1885 with scarcely a failure. But, though the banks were safe, they were even less adventurous than those in Britain and lent mainly to commerce and the church (and later to agriculture), depositing the balance in the London money market for much of that interval. I have pointed out elsewhere that the convergence of Ireland to the frontier of European living standards in the last decade of the old millennium owed little to financing from the Irish banks.
Perhaps this reflection means that our ambitions should be modest. If the banks can manage to accelerate their processing of the large backlog of troubled legacy debt into sustainable solutions, much of what is needed to restore their health will have been achieved. If in addition the funding of their low-yielding tracker portfolio could be enhanced (something for which technical solutions have been devised, but require wider official European agreement that does not seem likely to be forthcoming in the short-run), then, when combined with the credit union restructuring framework being implemented by the Credit Union Restructuring Board (REBO), the foundations would be well laid for the long-term rehabilitation and repositioning of the banking sector.
Technological changes will of course influence the operations and image of banking in the future. I have spoken only briefly about such aspects. Such factors may not be as strong in a period of reconstruction than they were in the run-up to the crisis, where over-reliance was placed on risk management models and automated credit approval processes. But anyway that was not the flaw that caused the Irish banking collapse. Its converse, traditional relationship banking based on the lender understanding both the needs and the creditworthiness of customers, is in effect what was mishandled in Ireland during the exuberant years but is inevitably what we have to re-learn to ensure the recovery of the Irish economy.