Yves here. It is frustrating to see an article give a good statement of a problem and then not understand what a solution would entail. The authorities in the US punted on the idea of living wills, which was having the major banks draw up how they would be wound up in a crisis. They were given some wet noodle lashing and then the idea disappeared from the radar (I was paying reasonably good attention back in those days and the mechanism by which this change happened seems to have been executed in a dark alley).
The problem with this idea is right there, in the headline: “globally resolvable”. That is a legal and regulatory impossibility. The US and UK, and following their lead, bank regulators all around the world, have allowed a system to develop whereby banks operate across borders in a freewheeling manner, but when they are put down, it is a nation-by-nation process, subject to national bankruptcy laws. Recall the Lehman bankruptcy filing was in New York, and there was a big legal jurisdictional row over the effort to unwind the large London operations in the UK.
Lehman was a securities and derivatives firm, not a bank. The fine points of resolution are outside my ken, in no small measure because officials move and earth not to have it happen, witness the shotgun marriage of Credit Suisse and UBS and the heavily subsidized acquisition of First Republic by JP Morgan. So it’s not as if outsiders have much chance to study resolutions. But a big international bank that has deposit-taking operations in foreign countries are likely to have them set up at theoretically independent subsidiaries with the various required national licenses and limits on their ability to upstream funds to the holding company. Canadian bank TD Bank has a US bank that is the eight largest here and I assume it operates more or less in that manner.
Without going into overmuch detail, the problem is not foreign banks serving retail customers. It is banks operating internationally. Consider the long established behavior of banks (and former investment banks) trading Treasuries. Any TBTF bank and some also rans have Treasury trading operation in the major time zones: UK/Europe, Asia, New York. The pass the book from one dealing center to another as daylight moves across the globe. It’s remarkably slipshod to let such large positions leave a regulated entity in one country and move into another, but that’s become accepted.1
A second example is that big banks have multinational customers that expect them to offer services in multiple jurisdictions. That will usually include cash management and multi-currency accounts.
That activity makes a bank unresolvable if it happens at any scale. As we wrote in 2010:
GTS is a big cash management/information service. It is also a bread and butter earner for Citi. Per the Journal:
Otis Otih, the treasurer of candy maker Mars Inc., uses GTS to handle most payments to employees and vendors of Mars operations in 68 countries. “Citibank is the only truly, truly global company for us — I don’t see any alternative,” he says.
As an example of what the unit allows multinationals to do, an Asian subsidiary of a European company can deposit funds with Citigroup locally and the money will instantly show up on the ledger of the parent a continent away. The system makes it easier for corporate treasurers to manage their finances, and many corporate and government clients outsource a wide range of other finance work to GTS….
Executives told officials with the Treasury Department and the Fed that GTS’s technology and presence in more than 100 countries made it too dangerous for the U.S. to let Citigroup collapse. The Treasury gave the bank a second big helping of $20 billion just six weeks after an initial $25 billion infusion from the Troubled Asset Relief Program, partly in recognition of GTS’s importance to the financial system, according to government and company officials….
While Citigroup is primarily known for its retail banking and credit-card businesses, the GTS unit is increasingly integral to the parent company’s functioning. Clients that move funds through GTS leave a lot of cash on deposit at the unit, which funnels the money to other parts of Citigroup for lending or other uses. GTS’s deposit-gathering muscle has grown more important since the financial crisis began, now providing about 40% of Citigroup’s $800 billion of deposits.
Yves here. GTS is a big piece of what makes Citi a difficult to disarm bomb. One of the swords of Damocles that the big bank had over the officialdom is that, prior to the crisis, it had $500 billion of uninsured foreign deposits. If Citi looked wobbly, sensible depositors would withdraw funds, and that could quickly morph into a run. Moreover, the any other international bank with meaningful cross border deposits could come under scrutiny (although it is odd more of this has not happened in the wake of the implosion of Iceland, which left a number of UK borrowers high and dry until concerted pressure on Iceland produced some restitution).
The wee problem is this sort of service is likely not hugely important profit-wise to banks. But it’s VERY stick. It’s perceived as extremely important to corporate Treasurers and switching costs would be high. And having your bank tied to you via this type of product creates oodles of opportunities for upselling.
So to have resolvable banks, they need to be either:
National only (or mainly), which does not at all work for multinational or even somewhat international companies, and it also makes securities dealing across borders monstrously cumbersome. Bye bye international capital flows!
Smallish. But big companies don’t like small bank, and trading/asset management business have high barriers to entry and very large economies of scale. You need to have a full complement of products in at least all the major time zones, with the needed IT, back office, and sales/customer support staff. So you have a very high required level of operations just to be a player. The economies of scale result from greater efficiency of running large (v. not so large) volumes over the same cost base, and the information advantage of “seeing” more trades from more customers
Subject to an international resolution regime. Good luck with that.
We’ve said for many years that banks enjoy more government support than any nominally private sector business. They cannot be regarded as private companies. They are the world’s biggest welfare queens.
Since it’s contrary to prevailing ideology to nationalize banks if they are too big and/or complicated to wind up, the next best approximation is to regulate them like utilities. But the supposed elites are too deferential to business executives, and also confuse specialized skills with “talent”. In this arena, regulators are particularly afraid of trying to mind trading operations.
But the fact that the recent crop of US bank tsuris were due to Banking 101 failures, as in asset/liability management and ignoring liquidity risks, says bank regulators are even weaker than they were in the runup to the 2008 crisis. Not a happy finding.
By Aymo Brunetti, Professor, Department of Economics University of Bern. Originally published at VoxEU
The subsidised emergency takeover of Credit Suisse by UBS brings the current global ‘too big to fail’ regime into question. This column argues that an in-depth analysis of the global resolution framework by both regulators and academics is needed. The main question is whether a resolution of a global systemically important bank is indeed feasible in plausible scenarios. An affirmation would clearly be the best possible result of this analysis. However, if such a resolution proves not to be realistic, then there should be no hesitation to drastically reduce the global risks of such institutions via regulation of their business models.
Sunday, 19 March 2023 was a historic day for Swiss economic policy – and not in a positive sense. It was the day Swiss authorities announced a subsidised shotgun marriage between the two globally systemic Swiss banks. Credit Suisse had suffered a dramatic loss in confidence in its business model that triggered a spectacular bank run, which was ultimately resolved by a de facto takeover by UBS. This event is not only relevant for Switzerland but for the entire regulatory framework of global financial markets. It puts a huge question mark on whether global ‘too big to fail’ (TBTF) provisions for such cases will ever work as planned. Instead of a resolution according to the gone concern framework promoted internationally by the Financial Stability Board (FSB 2021), the Credit Suisse case was handled as if we were back in 2008. The government stepped in and used taxpayer money to avoid a potentially catastrophic breakdown of an obviously still ‘too big to fail’ institution. And all international observers applauded this swift action that clearly went against the provisions painstakingly designed, implemented, and internationally coordinated in the past decade.
Swiss authorities will certainly hurry to produce an in-depth analysis of the case and its regulatory consequences. The case, however, goes far beyond this and deserves a thorough international inquiry as it is relevant for the regulation in all financial centres. In my view, the follow-up to the Credit Suisse case must now finally and unequivocally achieve the goal declared in 2008: there must be no company that is too big to fail! And ultimately there are only the approaches mentioned in the title of this column. Either a global resolution without state support is possible in plausible crisis scenarios, or the big banks must become ‘smaller’; the quotation marks mean less risky, in the sense that it may no longer be allowed to do business that endangers global financial stability.
Resolvability Remains the Silver Bullet
The concept that a business idea can fail and thus a company can go bankrupt is one of the fundamentals of a functioning market economy. If this possibility does not exist, it leads to seriously distorted incentives and to dysfunction. Large banks that are ‘too big to fail’ are therefore unacceptable in a market economy, and correspondingly intensive efforts were made to find a solution after the Global Crisis. The goal of global regulatory efforts became clear very quickly: a large bank must become resolvable and thus lose its ‘too big to fail’ status. This means that in the event of a crisis, the authorities must be able to order a restructuring or an orderly bankruptcy. If this succeeds, it is clearly superior to all other regulatory approaches. Indeed, any alternative means regulating the banks’ business activities, which has the disadvantage of any planning approach. Not only would the authorities have to know the existing businesses and their risks and regulate them, but they would also have to continuously assess any innovation and re-regulate accordingly. In view of these bureaucratic challenges, it is much more effective to leave the big banks their entrepreneurial freedom on the condition that they can be wound up at any time.
This approach remains the silver bullet of big bank regulation and has guided international ‘too big to fail’ regulatory efforts over the past decade. This was also the case in Switzerland where reports of two expert groups that included regulators, academics, and private sector representatives provided the blueprint for the current regulation (Swiss TBTF Commission 2010, Swiss Strategy Commission 2014). Based on these requirements, the big banks rebuilt their organisation at great expense so that the systemically important parts could be spun off in the event of a crisis. At the same time, bail-in capital and international agreements were to ensure that the rest of the bank could be restructured or wound up in an orderly manner.
Urgent Feasibility Analysis
The recent decision by the Swiss authorities not to wind down Credit Suisse according to this procedure now raises fundamental questions. Did they simply find and implement a less far-reaching alternative here, or would proceeding according to plan not have worked, or triggered a severe global financial crisis? Was a global resolution of Credit Suisse feasible, and if not, why not? These questions are of fundamental importance not only for Swiss financial market policy, but also for international ones, since ‘too big to fail’ regulation works according to these rules in all locations of global systemically important banks. Switzerland can and should undertake this analysis for itself as soon as possible but can also press with some legitimacy for a rapid and in-depth international evaluation of the event and its consequences; this is of interest to all major financial centres.
This in-depth analysis should cover the specific case of the demise of Credit Suisse in spring 2023, but also think through alternative plausible crisis scenarios. It can lead to three conceivable outcomes:
- A global resolution would be possible and defensible with acceptable risks.
- A global resolution would be too risky today because there are still substantial gaps in the concept and/or a bankruptcy of a major bank in clearly identifiable businesses would trigger massive turbulence.
- A global resolution is fundamentally not feasible under plausible scenarios.
Only based on this analysis – which should be available as soon as possible in view of the urgency of the problem and the political pressure – will it then be advisable to adjust ‘too big to fail’ regulation.
Of course, ideal would be to arrive at result 1, i.e. that resolution according to plan is possible. The conclusion would then be that the fears of the authorities in the case of Credit Suisse were exaggerated and that – if necessary, with a few cosmetic adjustments in the international agreements – in the future a major bank can be confidently sent into forced restructuring or bankruptcy in the event of a crisis. Unfortunately, based on the current state of knowledge, this outcome is rather unlikely. One would have to be very certain and have a consensus on this within the Financial Stability Board to come to the conclusion that no regulatory adjustments are needed.
In my assessment, the most likely outcome is 2: global resolution is not fundamentally impossible, but it will take some and substantial adjustments before this concept really works. It seems clear that a number of the reforms needed to achieve this would only be possible in an international context involving all relevant financial centres. Accordingly, it is likely to take a long time. Individual countries that are especially affected, such as Switzerland, should, therefore, seriously consider taking earlier actions such as significantly higher capital requirements for transactions that endanger global resolvability.
It cannot be ruled out that the conclusion 3 will be reached, i.e. that global resolution is fundamentally not possible and even reforms cannot save the system. This would be a very far-reaching finding, which would also have to lead to a fundamental realignment of big bank regulation internationally. From the perspective of Switzerland, which is particularly exposed, this would mean that the ‘too big to fail’ problem could only be solved by a drastic downsizing of UBS’s global systemically important activities. In fact, Switzerland could probably no longer be the host country of such a bank. Of course, it is good for a country with a sizeable financial centre if a major global bank is based here. But that is only on the condition that its business model can fail, just like any that of any other company. If that is not the case, then a small country especially should not hesitate to accept the withdrawal of the headquarters of such a bank; the potential costs of a failure are far too great. There should be no industrial policy style subsidies in favour of large banks.
Overall, it is important not to hastily discard the ‘too big to fail’ concept of global resolutions of large banks in view of the recent experience. Considerable efforts have been made by the banks and the authorities in recent years to make this possible. Resolvability remains the most efficient way to tackle the problem of the implicit ‘too big to fail’ subsidy for big banks. However, if the analysis concludes that the concept cannot be implemented even with additional regulatory efforts, then economic policy must aim at making the banks smaller and/or globally less risky, even with far-reaching interventions. This is especially the case for a small country with a large financial centre. Given the internationally record-breaking size of Switzerland’s largest bank in relation to GDP, the country simply cannot afford an UBS with a state guarantee. Larger countries might have the fiscal means to do this but subsidising global banks is an especially inefficient and unfair policy.
See original post for references
1 I would assume they are still treated as held by a single depositary in the US while this is happening…but having the economic reality diverge so greatly from the legal treatment opens big cans of worms. Regulators noted with some shock that the ways banks operated often had perilous little relationship to legal structures and who theoretically owned the activity, but they did nothing to clean that up.