Companies fail slowly (see 9fin’s Watching the Defectives) but banks fail incredibly fast. Since we published our initial 9fin explainer on SVB Financial early Friday afternoon, the planned share sale failed, trading in the bank was halted, and US regulators stepped in, all before market close on Friday. 

In the middle of Friday afternoon, SVB’s UK entity reminded clients that it “fully abides by UK regulatory requirements…and is ring-fenced from the parent and its other subsidiaries”. By the end of Friday, the Bank of England had placed it in a special form of insolvency, and before market open on Monday, HSBC had purchased it for £1. In the US, Sunday evening brought the announcement of a special funding facility for banks, allowing them to pledge government debt at par (regardless of market value), and a “systemic risk exception” to the resolution of SVB’s US entity, allowing all depositors to be made whole.

In short, it’s been a busy weekend for bank regulators and potential buyers of SVB assets. Everyone else has been chasing rumours or tweeting through it — and now it has got political.

The immediate cause of the failure of SVB was a classic bank run, with depositors scrambling to withdraw funds. 

Thanks to regulators taking over the bank, we can put some figures on this. The California state Department of Financial Protection and Innovation said that “investors and depositors reacted by initiating withdrawals of $42bn in deposits from the Bank on March 9 2023, causing a run on the bank. As of the close of business on March 9, the bank had a negative cash balance of approximately $958m”.

This is an absolutely spectacular number. 

SVB had $151bn of uninsured deposits in its US offices at the end of last year; Friday likely saw further deposits pulled out, before the Federal Deposit Insurance Corporation stepped in and took over the institution, freezing withdrawals.

Few banks could survive more than a quarter of deposits disappearing in a day — regulators require banks to put aside high quality, easily saleable assets, but that’s mostly to reassure depositors and to stop any run in its tracks. Nobody really expects a bank to be able to liquidate that much in a day.

SVB Financial actually had an unusually high proportion of valuable, easily saleable assets. However, most of these were locked away its “held to maturity” book. This is all going to be money-good; it’s mostly US Treasuries or “agency MBS” (that is, mortgage-backed securities issued by Fannie Mae and Freddie Mac, which trade as a liquid rates product in the US).

But all of this good collateral was marked at the wrong price. It was bought when interest rates were low and bond prices were high. Now interest rates have gone up, bond prices have gone down, and selling these assets would crystallise a large loss of around $15bn. In fact, thanks to bank accounting rules, selling even a proportion of this portfolio would mean marking all of the rest of it to market, effectively wiping out the bank’s equity.

As worries about SVB ricocheted around the markets, this issue came sharply into focus. Lots of US banks have large unrealised bond losses, and the more alarmist commentators started to discuss the possibility of a run on the broader banking system. It’s hard for dollar depositors to actually “run” from every US bank; there simply aren’t enough suitcases to hold bundles of dollar bills under every CEO’s bed. 

But it was terrifyingly easy to imagine a situation where uninsured depositors engaged in a “flight to quality”, placing their funds only with the largest more systemically important banks, and igniting runs on smaller regional banks with their own unrealised losses. 

On Monday, shares in San Francisco-based First Republic Bank were down 75%, prompting the announcement that no bank management team ever wants to make: “Our capital remains strong” and “Our liquidity remains strong”. If you have to say it, you’re already losing the battle. New York’s Signature Bank has already been taken over by regulators, though this links partly to their crypto exposure.

When the FDIC stepped in to SVB on Friday, it initially said that uninsured depositors would get a “receivership claim” — in effect, take a number, stand in line, and we’ll pay you when we pay you. 

There were good reasons to think that a proportion of the cash would come quickly, because the government bond portfolio is very easy to sell. $91bn is a lot to liquidate, but the US Treasury market turns over $500bn per day. A week would be plenty. So the $91bn (which was actually worth $76bn) would come in pretty fast, and depositors would get c. 40% of their money back in fairly short order. The rest of the bank would take longer to unwind, since it would essentially involve selling the assets, but the assets weren’t bad. A “motivated seller” like a bankrupt financial institution doesn’t usually get a good price, but various back-of-the-envelope numbers got to 90%+ recoveries.

I should probably note, though, that all of these figures are based on the December 2022 balance sheet — by Friday evening, a quarter of the deposit base was gone, and presumably a fair bit of the asset base too. As of December 2022, it had $13.8bn of cash. So if it had negative cash of nearly $1bn after $42bn of outflows…..that implies it did manage to sell or repo (pledge against cash) quite a lot of assets within that short period.

So the balance sheet over the weekend, in short, looked pretty different to the balance sheet at the end of the year.

Anyway, the point is, having a “receivership claim” and a probable cash inflow of 40% and maybe a cash inflow of the rest at some undefined point is a lot less good than having a bank account you can use to make payroll today. If there’s even a small chance this happens to your company, you get out — hence the bank runs.

The goal of the twin interventions from the US government over the weekend was to stop precisely this move, in two ways. 

Invoking the “Systemic Risk Exception” to the rules of bank resolution, and agreeing to make SVB US depositors whole, even when they are uninsured, is a straightforward bailout, though predictably there are running battles under way over the semantics. Treasury Secretary Janet Yellen is describing it as “not a bailout”, and it’s true that the move won’t help equity holders. But the crucial point is, if SVB depositors were able to access funds on Monday morning, courtesy of Uncle Sam, there should be no need to keep running.

The other intervention was a special scheme to provide funding against government bonds, and mortgage-backed securities from the federal agencies (Fannie Mae, Freddie Mac, Ginnie Mae), which are effectively guaranteed.

It’s already extremely easy to raise funding against these securities. Daily aggregate repo outstanding is $4.3trn, of which nearly all is USTs, TIPS (inflation-linked treasuries) and agency MBS.

But the crucial difference with the new facility is that it will value securities at par (100%), rather than market value. 

That’s aimed directly at the “overvalued bonds” issue that weakened SVB’s balance sheet; even if the bonds are underwater because interest rates have risen, you can get 100% of their face value. This won’t completely fix the problem, as some bonds will have been bought above par, and this only offers par. But it still opens a huge liquidity tap for any bank which gets into trouble.

Across the pond

We should also turn to SVB UK, a far smaller institution, which took a different path from its US parent through the chaos. 

As we noted above, mid-afternoon on Friday, SVB UK’s management were still emphasising the bank’s separation from its parent, and the Bank of England placed it into bank insolvency on Friday evening. This is a separate procedure from “resolution”, a process created after the financial crisis to give an orderly way for regulators to parcel out the pain from a bank failure, and is much more akin to an ordinary corporate insolvency — the plan was to pay back the depositors from the sale of the assets.

The Bank did this because it’s a small institution, even given the smaller size of the UK market. It was also effectively a brand new bank, having being incorporated as a subsidiary in August 2022.

But depositors in SVB UK had far less information than depositors in SVB US. The entity had filed no accounts since becoming a subsidiary, and very little information could be gleaned from the consolidated balance sheet. The only tidbit in the (out of date) accounts was a signal that the parent had injected £10m of equity in May.

The rumours, however, flew out on Friday and over the weekend. The Financial Times reported that SVB UK had asked for £1.8bn from the Bank of England’s Discount Window Facility (the emergency liquidity tap). 

Anyway, the weekend brought a hurried sale process, which concluded in the happy ending announcement before market open on Monday that HSBC had purchased the bank. HSBC’s announcement disclosed that SVB UK was fairly sizeable — loans of £5.5bn and deposits of £6.7bn as of March 10, with tangible equity of around £1.4bn. This puts it in “medium sized challenger bank” territory, but it’s a minnow in the broader HSBC group, which has a balance sheet of nearly $3trn.

This has stopped the run, though there are clearly still jitters in the European banking system. Shares of perennial problem children Commerzbank and Credit Suisse took a dive on Monday, for example, though not to terrifying levels.

The rules on marking bond portfolios work a little differently in Europe and the UK, and both jurisdictions require bank runs regulations (the “Liquidity Coverage Ratio” and the “Net Stable Funding Ratio”) to apply to all institutions, not just the largest banks. These are crude measures, but basically force banks to have more liquid assets on hand to meet outflows, and limit the extent to which a bank can borrow short to lend long.

Even if the interventions over the weekend successfully stop bank run fears on both sides of the Atlantic, the political fallout will take a long time to settle. Some truly deranged takes have already done the rounds; the Wall Street Journal’s claim that it was “distracted by diversity demands” probably tops the table for most overtly unhinged in a mainstream publication.

Many have relished the sight of venture capital hubris being punctured, and of the modern masters of the universe begging for government bailouts. The bailout did come, and will be re-litigated for years in opinion columns; some question whether a bank that served only farmers would have received such generous treatment. The regulatory pendulum will swing once again, and banks with concentrated deposit bases and mismanaged interest risks will be under greater scrutiny. Banks which are already too big to fail will get even bigger and even more systemically important; better hope the belt-and-braces regulatory regime which covers the JP Morgans and HSBCs of this world is up to scratch.