Policymakers are bungling their response to failing banks

Have we then learnt nothing?

From the way regulators on both sides of the Atlantic have handled bank failures over the past week, it feels to me like the answer is “no”. What is most striking is that the systems put in place after the last financial crisis were not used as intended — instead policymakers in the US and Switzerland hurried to come up with ad hoc solutions for Silicon Valley Bank and Credit Suisse. They may have thought that this was best to stem panic and safeguard financial stability. Many seem to agree that it is too idealistic to follow plans made for these crises in quieter times, and that “pragmatism” must rule. After all, no battle plan survives the first contact with the enemy. But that thought, I fear, is precisely the problem.

To see why, think about financial crises in the deepest sense.

Financial crises happen when the value of assets people think the financial system holds does not add up to the sum of all the claims they think they have on it, meaning that not all those claims will be honoured in full. It does not matter that what people expected to get may never have been realistic or rational. It does not even matter — at least in the first instance — whether their beliefs about the value of the assets in the system are overly pessimistic. What matters is that expectations are adjusted — and in that process a lot of damage can be done.

When suspicion sets in that there is not enough value to go around, people rush in a devil-take-the-hindmost kind of way to get their money out of danger, leaving others more exposed to the assets that are no longer trusted to be what they were thought to be. Because policy decisions will determine who finally bears any losses, an almightily political battle breaks out over how to allocate them. And while all this goes on, financial institutions will find it too risky to extend or make new loans and investments and investors will find it too risky to recapitalise banks and other entities.

All these processes — the first rapidly, the others grindingly — put a brake on economic activity, because of the uncertainty that poisons any choice to place new capital to work. That makes the losses even worse and gives another spin to the downward spiral. For that reason, you minimise the fallout by getting expectations realigned as fast as possible so that everyone can move on.

If we understand this basic anatomy of financial crises, some policy lessons are evident. The first is that you need to allocate the losses. That is to say, regulatory and legal mechanisms must crystallise losses and define where the writedowns must take place. The second is that you must do so as quickly as possible. The third is that you must do so in ways that create a sufficiently capitalised banking system after all losses are written down. The fourth has to do with politics: you must muster political backing for the inevitable redistributive effects, compensate those who legitimately need to be compensated, and tell those who don’t (and those who shoulder compensation costs) that this is the name of the game.

None of this is easy or pleasant. All of it is easier to do if you agree it in advance. That is why governments put in place laws and methodologies that would allow them to quickly restructure the balance sheets of banks in trouble, which just means to crystallise their losses and reduce the claims people have on the banks. If this is done ambitiously enough, the (now more realistically valued) assets again exceed liabilities by a safe margin: the restructured bank is now unencumbered by past losses and ready to lend again.

These bank resolution regimes are, on the whole, well-designed. But the two natural instincts to want to calm things down and to avoid imposing losses on anyone, even if the aim is to try to stop things from getting worse, militate against doing what needs to be done. My colleague Helen Thomas observes: “No matter how much work has been done on bank resolution, when push comes to shove regulators and politicians have balked at the prospect.”

With this in mind, go back over the past week’s bank rescue operations.

SVB was brought low because depositors got scared by how rising interest rates hurt the market value of the bank’s government-guaranteed assets. But rising interest rates have affected the assets of all US banks. By how much? Sebastian Mallaby and my colleague Martin Wolf cite research putting the market value losses to US banks’ total assets at $1.7tn to $2tn, almost as large as their aggregate equity cushion of $2.1tn.

That is a bigger issue than the implosion of SVB and a few other midsize banks. But it’s not the issue US authorities have addressed. Their ad hoc decision to make whole the uninsured depositors of SVB was a decision to depart from the standard resolution procedure the Federal Deposit Insurance Corporation applies to failing smaller banks. It smacks of an unwillingness to allocate losses — and that is a bad sign given the extent of unrealised losses in the banking system as a whole. It is, of course, not insurance but a bailout after the fact, which somebody has to pay for since the uninsured depositors will not. To say no taxpayer money was used is a ruse: higher FDIC fees are, for all practical purposes, a tax rise, because the banks that have to pay them have little choice but to accept. To be fair, at least the FDIC wrote down all shareholder and unsecured bondholder claims to zero, which is better than what Switzerland did, on which more below.

So US authorities are trying to gin up confidence in the US banking system rather than solve the problems that have caused the current bout of pessimism. Unless they are very lucky, that approach is unsustainable, and is already making matters worse. Everyone is now asking whether their deposits would be safe if their bank gets into trouble. Treasury secretary Janet Yellen has not given a reassuring answer: the policy seems to be that if you have a deposit above the $250,000 insurance limit, you will be bailed out if a run on your bank is likely to trigger contagion to the wider banking system — but not otherwise and not until then.

That is an invitation to large depositors in smaller banks to move their money out — either to the big ones, which will always be deemed systemically important, or to money market funds — as is already happening. Yellen’s exchange with Senator James Lankford last week (hat tip: Claudia Sahm) makes clear that the government has no answer to this. The only possible answer would be to insure all depositors, without limit, now — but as she told lawmakers yesterday, there was no immediate plan for this. And as I wrote last week, any case for guaranteeing all deposits is a case for government provision of deposits through a central bank digital currency.

The resulting negative spiral from banks to the economy is surely at work. Midsize and small US banks account for a large proportion of lending to the US economy — in particular, commercial real estate, but also mortgages and business lending. Not only will such newly risky banks become much more cautious about issuing new loans. If they also have to sell assets to cover deposit outflows, they will make one another’s solvency problems worse. Already there are deep fears in the market for mortgage-backed securities.

So what do I think US authorities could have done instead? In the specific case of SVB, they could have let the normal FDIC process run its course. In conjunction with the Federal Reserve’s new programme to lend against government securities at face value, depositor losses would have been minimal. Some have also argued the government could have invoked the special resolution powers in the Dodd-Frank rules called “orderly liquidation authority”, which would have covered SVB’s holding company and possibly “bailed in” (written down) claims on other parts of the group ahead of SVB’s uninsured depositors. Either move was available but required a greater political willingness to clarify who takes a loss.

What about the effect on other banks? As some have proposed, accounting rules could be made to immediately crystallise market value losses in the banking system, thus bringing accounting realities in line with expectations faster. That would lay bare the need for broader liquidity provision in case of bank runs and, more fundamentally, for recapitalisation. In other words, the mechanisms for writing down shareholders and bondholders put in place after the last crisis would have had to be taken out of the toolbox.

It may still come to this. The authorities are clearly hoping they can calm markets down instead — the very definition of a confidence trick. But if it doesn’t work, the pain will be greater. Again, it’s short-sightedness passing for responsible pragmatism.

Now consider Switzerland, for which the Financial Times’s tick-tock account of how the government made UBS take over the doomed Credit Suisse in a keep-it-in-the-national-family merger is indispensable. Here, too, the unwillingness to go ahead with long-laid plans for just such a situation is striking. My colleagues quote a UBS-related source as saying: “Resolution [a government-controlled wind-down] would have been a disaster for the financial system and introduced the threat of contagion around the world.”

I have been amusing myself reading the website of Finma, the Swiss financial regulator, in particular the pages on resolution and recovery schemes. These, we are told, “addressed” the risk that a bank would be too big to fail and “force a de facto government bailout funded by the taxpayer”. How is that working out for you? We are entitled to ask after Switzerland offered loss guarantees in the billions and a SFr100bn ($109bn) liquidity line to make the shotgun marriage go through.

And who benefits from these taxpayer subsidies? In complete violation of the liability “hierarchy”, where it is shareholders’ job to absorb losses first, the owners of Credit Suisse were paid $3.25bn in the UBS merger, even as some bondholders — the so-called additional tier 1 (AT1) bonds — were written down in full. We shouldn’t feel too sorry for those creditors, whose job it was to be written down when trouble hit. But being bailed in before shareholders lose everything was not the general understanding of how banks are meant to be recapitalised, even if the Swiss authorities may have the letter of the law (and of the bond contracts) on their side.

That $3.25bn is money that could have stayed with taxpayers (if UBS hadn’t had to grant it to Credit Suisse shareholders, it could have been subsidised that much less). The impression is, again, that the losses of bank shareholders will be limited, at taxpayers’ expense. If UBS faces bigger losses than thought, that will not be pretty. And again, by not going by the plans laid in quieter times, policymakers have worsened fears elsewhere in the system. EU regulators have had to insist they would never do the same as Switzerland; even so, AT1-type bonds have fallen in value globally.

And I mean “plans” quite literally. Finma’s 2022 resolution report confidently tells us that Credit Suisse had produced its required “Swiss emergency plan” for how to keep its Swiss operations going in case it failed, which Finma viewed “as ready to implement”. Finma itself had a resolution plan for Credit Suisse as a whole but said full “resolvability” would not be ready until the end of 2022. I would quite like to know what these plans were, so I wrote to Finma and asked to see them. I have not received a reply. But I think it is a fair guess they did not prescribe what was actually decided last weekend.

On both sides of the Atlantic, then, the lesson has been that policymakers in a crisis have thrown out the plans and ignored the tools that were thoughtfully prepared precisely for these eventualities. Why? Wolf suggests in the column I linked to above that “the functions of the banks in providing money and credit are too vital to allow” writing down claims to their market value. I take that to imply that regulators are right to throw out all the resolution plans and “living wills” of banks, and we were all deluded in developing them in the first place. My view is the opposite — that swift resolution is the best way to safeguard the functions of banks, and that policymakers are therefore failing us badly by not making use of them.

If they don’t quickly change tack, we are facing another lost decade like the sluggish recovery of the 2010s. So far, our regulators are proving worse than the Bourbons*. They have learnt nothing, and forgotten everything.

Other readables

• European industrialists keep pushing for the EU to “compete” with the US’s Inflation Reduction Act — in other words, to pay them more subsidies. But their lobbyists admit sotto voce that this may be more a “psychological game” in a shakedown of politicians than a real existential threat.

• Lucrezia Reichlin writes that the current crisis may lead us to adopt central bank digital currencies as a form of fully insured deposits, and Martin Wolf makes the same point in the column mentioned above. Depositors fleeing to the safety of central bank money is in fact already happening, The Economist reports, through money-market funds’ ballooning reverse repo transactions with the Federal Reserve.

• The IMF looks at the jump in digitisation caused by the pandemic: “We can see that it helped boost productivity, protect employment, and mitigate economic disruptions during the pandemic.”

Numbers news

• My colleague Anastasia Stognei reports that Russian oil-exporting companies may be under-reporting the prices they sell at, which could circumvent the western-imposed price cap but also save them tax owed to the Russian government. The finance ministry in Moscow is taking steps to limit the discount exporters can claim below the higher Brent index.

  

* This article has been corrected to make clear it was the Bourbons, not the Habsburgs, who were said to have “learnt nothing, and forgotten nothing’