Banks: quickly, then slowly

Good morning. The markets were, in relative if not absolute terms, calm yesterday. We needed it. As fun as it is to be a finance journalist when the system is under pressure, we will be glad when this weekend comes, and we suspect a lot of you will, too. In the meantime, let us know what’s on your mind: robert.armstrong@ft.com and ethan.wu@ft.com.

Marking US banks to market

Well, this doesn’t sound very good:

The US banking system’s market value of assets is $2tn lower than suggested by their book value of assets accounting for loan portfolios held to maturity. Marked-to-market bank assets have declined by an average of 10 per cent across all the banks, with the bottom fifth percentile experiencing a decline of 20 per cent

That is Erica Jiang, Gregor Matvos, Tomasz Piskorski, and Amit Seru in their recent and much-discussed paper, “Monetary Tightening and US Bank Fragility in 2023: Mark-to-Market Losses and Uninsured Depositor Runs?”

That lead sentence rounds the mark-to-market losses down: they are actually $2.2tn. That’s a big number, but the total loss figure is not the reason that the paper spooked me a bit. Instead, it’s the 10 per cent decline in mark-to-market asset values that grabs the attention. US banks, in aggregate, have a leverage ratios (equity to assets) of about 9 per cent. So a 10 per cent decline in mark-to-market asset values means the US banking system is — just barely — mark-to-market insolvent.

To put it more simply: Jiang et al argue there are $2.2tn in mark-to-market losses out there, and there is only $2.2tn in equity in the US banking system.

More specifically, the authors reckon that if all US banks’ assets were marked to market, 2,315 lenders would have negative equity, out of a total of more than 4,800.

The authors’ mark-to-market valuation technique involves using the reported fair value of banks’ securities, and then adjusting the value of their real estate loans based on the value of publicly traded real estate funds. “In some ways, our estimates are conservative, since we only marked down the value of real estate loans and other assets and securities and loans . . . rather than all assets on the bank balance sheets,” the authors write.

The paper is a strong prima facie argument for boosting the amount of equity capital in the system. But its finding are less scary than they seem. There are good reasons that banks are not required to mark their entire balance sheets to market. For starters, as the authors point out, a lot of lenders are rate hedged, and the value of those hedges are not included in the analysis.

Second, it is possible for a bank to be mark-to-market insolvent and still be “asset sensitive”, that is, more profitable as rates rise: all it needs is to have more variable-rate assets than fixed-rate assets, and a reasonably stable deposit base. Third, if the economy comes under more pressure, interest rates are likely to fall, reducing the mark-to-market losses.

Finally, as the authors point out, the mark-to-market system insolvency does not imply that many or most banks are vulnerable to runs in the way that Silicon Valley Bank was. The median bank funds its assets with 9 per cent equity, 65 per cent insured deposits, and just 27 per cent uninsured deposits and debt. Silicon Valley Bank’s assets were 80 per cent financed with uninsured deposits and debt, making it extremely — almost uniquely — vulnerable.

For bank investors, however, there is more to think about than bank runs. Those $2.2tn in mark-to-market losses represent a huge block of assets that are yielding less than market rates, which will create a big and potentially long-lasting drag on bank profitability. In a recent note, Betsy Graseck of Morgan Stanley listed a bunch of ways that the recent banking scare will add to the profit pressure:

• Industry competition for insured deposits is likely to increase, making them more expensive

• Quantitative tightening, by sopping up liquidity, will exacerbate this

• Regulators or simple prudence will dictate that banks hold more long term-debt liabilities in their capital structure, increasing the cost of funding

• The SVB scare will widen spreads on bank debt, will also increasing funding costs

• Higher cost of bank equity capital — that is, lower price/earnings or price/book ratios for bank stocks — will increase funding costs, too, at least in the short term

• Regulation and prudence will also lead to shorter-duration asset portfolios, bringing yields down

• Loan growth will slow as banks seek higher returns and tighten lending standards

“Putting it all together, it means lower net interest margins, net interest income, [and] lower return on equity across the industry,” Graseck writes.

Of course, the weaker banks, that see the most deposits leave, will have their margins decline the most as they pay up for new capital. To use a somewhat morbid metaphor, the heart attack phase of the banking crisis may be over, as the threat of runs subsides in the face of official support for banks. Now, however, we have to see if any banks are going to die slowly, killed by the cancer of declining profit.

Remember the point about asset sensitivity, though. For most banks, higher short-term interest rates means higher profit. Yes, net interest margins are falling and will continue to fall, but they are much higher now that they were even in the first half of 2022, thanks to fast-rising loans yields, as this chart from Janney’s Chris Marinac shows:

Unhedged is sceptical about the valuation and long-term performance of banking stocks as a group. But plenty of individual banks are highly profitable, and will remain so even as the collapse of SVB puts pressure on the industry.

One good read

Unhedged is rooting for Coach Yo and Mississippi tonight.