SVB was only a little bit insolvent, luckily

Good morning. It will be an interesting day in markets today. Did the US authorities do enough to quell worries about bank runs? It looks like it to us, but the situation remains tense. Emotion, not reason, may rule the day. Send us your worries, conspiracy theories and trading strategies: [email protected] and [email protected].

Silicon Valley Bank (and the US banking system)

Until last night, there were high hopes that another bank would swoop in and buy Silicon Valley Bank, taking a problem off the FDIC’s hands and protecting unsecured depositors. But it appears a deal could not get done. At 6:15pm New York time, the Treasury, Federal Reserve and FDIC announced that they were taking matters into their own hands, not only for SVB but for Signature Bank, another lender which had suffered deposit outflows recently. Both banks would be resolved and all their depositors, insured and uninsured, would be made whole. The meat of the announcement:

[All] depositors [in SVB] will have access to all of their money starting Monday, March 13. No losses associated with the resolution of Silicon Valley Bank will be borne by the taxpayer.

We are also announcing a similar systemic risk exception for Signature Bank, New York, New York, which was closed today by its state chartering authority. All depositors of this institution will be made whole . ..

Shareholders and certain unsecured debtholders will not be protected. Senior management has also been removed. Any losses to the Deposit Insurance Fund to support uninsured depositors will be recovered by a special assessment on banks

This has been widely interpreted to mean authorities will pay uninsured depositors out of the $128bn Deposit Insurance Fund anything SVB’s assets can’t cover. The statement read a bit ambiguously to us, though, and we’re still not sure how much of a bailout this is. Details will be forthcoming in the hours and days to come. In the case of SVB, however, the fact that the bank was not very insolvent is probably a big part of the answer.

Admittedly, in banking, being a little insolvent is like being a little bit pregnant, but bear with us. Recall that SVB’s problem was not, as with most failed banks, bad credit blowing a hole in the asset side of the balance sheet. Instead, it was a pure duration mismatch between deposit liabilities and high-quality bond assets. Ken Usdin’s team at Jefferies offer an analysis under which, if SVB’s securities portfolio were sold at a 20 per cent haircut, its loans at a 2 per cent discount and its creditors and insured depositors paid, there would still be $86bn in cash left:

Jefferies estimates that there are now $91bn in uninsured deposits left at SVB after last week’s run. If that’s right, the bank is only about $5bn short. Perhaps that gap is going to be made up by leaving some or all of the banks’ other creditors, who are owed some $22bn, out in the cold?

The Federal Reserve also announced that it would make liquidity available to other banks facing withdrawals. A “bank term funding program” will offer banks loans of up to one year, taking government-backed bonds as collateral, and valuing those bonds at par, rather than marking them to market.

Will the two actions be enough to end the threat of bank runs this week? In a rational world, they would be. While the rest of the banking system suffers from some of the same illness that SVB had, the systemic illness is simply not very severe.

The illness, as a reminder, is that in the past few years banks were flooded with deposits and invested much of them in long-term government-backed bonds. Now that rates have gone up, they have significant unrealised losses on those bonds. If they were forced to sell those securities (by a bank run, say) those losses would be realised.

Banking system doomsters on Twitter have given this chart, from Michael Cembalest of JPMorgan, a workout over the weekend. It shows what would happen to the equity capital ratios of various major banks if all the unrealised losses on their portfolios were crystallised:

Yes, realising the losses would take a big bite out of some banks’ equity, as the doomsters pointed out. But notice that all the banks would still have a solid equity cushion if this happened. Even SVB (as of the end of last year) would have been solvent, and it had the highest ratio of securities to total assets of any US bank. All the other banks in the graph, under the security portfolio liquidation scenario, would have been OK. And there is no reason to think that any of them, even under extreme circumstances, would have to sell all their securities. For one thing, if they needed cash, they could repo the securities rather than sell them.

This is not to say that it’s great that banks have all these securities that they bought when rates were low. It stinks for them, and will be a drag on profits for years. But it is not an existential issue.

Most banks still benefit, on balance, from higher rates, as their loan portfolios reprice. Bank of America is an excellent example of this. The bank has half a trillion dollars (!) of agency-backed mortgage securities down in 10 years or more, yielding 2.1 per cent. In and of itself, that’s very bad! The yield is well below market, and unless rates return to the lows of a few years ago, any sale of these things before maturity (which is a long time from now) will generate a big loss. But zoom out: the bank also has a trillion dollars in loans that pay more interest as rates rise. It also has another half trillion dollars in cash and reserves to meet withdrawals or invest at higher rates. The higher rates that sunk SVB help Bank of America.

There are regional banks that are not in nearly such a good position — their names were kicked around this weekend — but as we have argued before, no bank was as vulnerable as SVB, with its combination of rate-sensitive business deposits and an asset portfolio dominated by long bonds. It would be a shame if, despite the regulators’ actions on Sunday, there was still panic in the banking system this week. It wouldn’t make much sense.

The jobs report

Friday’s February jobs report would’ve been confusing even if a bank hadn’t failed on the same day. The data muddle included a consensus-busting 311,000 new jobs, offset by slowing wage growth, and an uptick in unemployment, but one driven by better labour force participation.

But because a bank did fail that day, it’s hard to read the market reaction to the report. Broadly, investors flew to safety: yields fell across the curve as stocks dropped. Futures markets priced in rate cuts in December (again). Yet market-watchers disagreed on whether the response was a shortlived panic or a sensible revision of rate expectations. Christian Keller at Barclays cast SVB’s failure and the jobs numbers in opposition:

If Powell’s comments midweek prepared the ground [for a 50 basis point hike this month], Friday’s nonfarm payroll numbers should have sealed the deal . . . [leaving] the 3-month average [payroll] gain at 351k — much too hot of a labour market for the Fed’s comfort. True, average hourly earnings (AHE) growth slowed to 0.2% m/m (vs. 0.3% m/m expected) [but the] measure is notoriously noisy due to its lack of controls for composition (ie, rising employment in lower-wage service industries . . . . Hence, based on the labour market data alone, we would have embraced our forecast of a 50bp hike at the upcoming March meeting with greater conviction.

However, financial sector news late in the week complicates this call . . . As this news evolved, US bonds rallied on safe-haven demand and fed funds futures priced in a likelihood of a 50bp hike below 50%, after it had risen well north of 50% following Powell’s testimony.

Taking the other side, Phoebe White of JPMorgan argued on Friday that “an improvement in labour supply and a moderation in wage inflation should take some pressure off the Fed”. The SVB-related dash-for-safety, White writes, came despite “little risk for market contagion or broader fire-sales”, suggesting that markets will soon refocus on monetary policy, especially once this Tuesday’s consumer price inflation data drops.

How to parse this? Here’s how we knit together the facts.

Contra Keller, we’d argue that the slowing wage growth trend is a meaningful disinflationary signal, and more relevant to inflation than strong jobs growth in itself. It also fits with reports from companies this quarter that hiring is getting easier. Yes, payrolls data can be noisy month to month, but the guts of recent data don’t suggest terrible distortions. As Omair Sharif of Inflation Insights notes, eight of 13 measured sectors experienced slowing wage growth over the past three months, indicating the February figure is signal, not noise. The broad trend is very clear:

White is probably right that the SVB flight-to-safety is overdone — especially now that the government has intervened. Even absent a bank run, however, the SVB mess could have a long-term effect on Fed policy and markets.

Consider Powell’s public testimony last week, in which he put a faster pace of rate hikes back on the table. That tool — the pace of increases, as opposed to the peak rate — will be most impacted by near-term economic data, including the jobs data and whatever CPI shows this week. SVB probably won’t impact this one way or another. But over time it could affect the Fed’s higher-for-longer policy stance, because it puts tightening on a timer. SVB’s double-decker bet on low rates forever was an idiosyncratic piece of stupidity. But tighter policy will expose other stupidities before long, the nature of which we can only guess. They will damage the economy, as the fallout from SVB already is. And the more things break, the less tenable higher-for-longer will be. (Ethan Wu)

One good read

How strong is the legal case against Fox News?