The collapse of Silicon Valley, Signature and First Republic banks means that 2023 — already, and it’s only June — has set a record. Although far more banks failed in 2008 during the financial crisis, the crisis of 2023 is worse as measured by total assets: Those three institutions held nearly $550 billion, compared with $530 billion in the banks that failed then.
That’s $550 billion and counting, of course, because billions more in bank losses still loom, especially given the head winds battering the $20 trillion commercial real estate market.
How bad will it get? Let’s start with the grim news.
First, the line from commercial real estate losses to banks is a direct and short one. Financial institutions hold 60 percent of commercial real estate loans, and losses on these loans without question raise the likelihood of bank failures, as significant trouble in commercial real estate threatens to push the value of bank assets (e.g., loans, securities and other investments) below total liabilities (e.g., what they owe depositors). That’s simple, unavoidable arithmetic.
In fact, the developing crisis in commercial real estate is one you can actually see: Walk around most major cities across the country, and you’ll spot a lot of large, vacant office buildings. The purchases of those buildings were largely financed by bank loans, and empty office buildings are less valuable than full ones. Vacancies were understandably on the rise throughout the pandemic, but the problem is, because so many people are still working remotely, they haven’t come back down.
Meanwhile, interest rates are rising, raising the risk that borrowers will struggle to refinance their loans in the years ahead. In many cases, commercial landlords are already choosing default.
And here’s the really bad part: Two-thirds of bank commercial real estate loans are held by regional and smaller banks — the very institutions that have experienced the most turbulence in the wake of Silicon Valley Bank’s collapse this spring.
That all seems … scary! But even so, to paraphrase from the great Taylor Swift: We (probably) need to calm down. Because a closer look reveals a much more nuanced, and less definitive, picture.
First and foremost, not all commercial real estate is created equal. This is a broad category, and it includes loans on office, apartment, retail and industrial buildings such as warehouses. Indeed, office loans — the ones hardest hit by the remote work revolution — make up just 18 percent of regional banks’ commercial real estate portfolios and just 3 percent of total loans. And vacancies in non-office commercial real estate have fallen, not risen, in recent years.
Furthermore, so far, commercial real estate loans aren’t all that risky to the banks that hold them. One key indicator of risk is the “loan-to-value ratio,” which can be understood this way:
Suppose a building is worth $100 million, and a bank has made a loan of $10 million to help finance its purchase. That would make for a loan-to-value ratio of 10 percent. The bank stands to lose only if there is a 90 percent decline in the building’s value. If the loan is larger, it will see losses sooner.
Loan-to-value ratios for commercial real estate today sit in the 50 to 60 percent range, at least on average, which is well in line with historical norms (not to mention the rest of bank portfolios). That creates a healthy buffer for weathering tough times.
That’s not to say things won’t get worse: One way to grapple with the scale of potential losses is to follow the Federal Reserve’s stress tests, which judge how well-equipped big banks are for downturns. The 2023 stress tests released this week, just like prior year’s tests, consider what would happen if commercial real estate prices fell by about 40 percent.
Note: That’s a lot! Indeed, part of the reason commercial real estate is disproportionately held by smaller banks (which don’t have to undergo Fed stress tests) is because it’s hard to pass a test that severe. If commercial real estate lost that much value, banks with more than $50 billion in assets would see their capital levels drop by 13.5 percent on average, with the most exposed banks losing double that, or more.
Such a severely adverse scenario is unlikely: Analysts estimate that even the riskiest office commercial real estate loans won’t lose as much as the stress tests assume — and it will take a decade for the losses to accumulate because borrowers can buy time by taking advantage of opportunities to extend their loans.
Still, given the grim news we started with, it certainly isn’t unreasonable to ask financial institutions to prepare for the risks on the horizon.
Any new commercial real estate-related capital losses would add only to the roughly $600 billion in capital that has been wiped out by the interest rate losses that led to bank failures this spring. And yet Goldman Sachs reports that payouts to shareholders in 2023 will outpace disbursements in 2022 by around 40 percent — meaning around $100 billion will leave the system, rather than backfilling for these losses. There’s a dangerous disconnect here.
I hope we’ll see more prudence than that. JPMorgan Chase chair and chief executive Jamie Dimon is likely right: Banks are going to experience more losses in the months ahead, just as they do in every credit cycle, and “the off-sides in this case will probably be real estate. It’ll be certain locations, certain office properties, certain construction loans.”
But because no one can predict which locations, which properties and which loans, the financial sector writ large should plan for the worst, even as we all hope for the best. That means bolstering capital levels to safeguard against potential losses now, before it’s too late.
Is it possible some banks could end up with more capital than they need, much to the chagrin of shareholders who would prefer bigger dividends? Sure. But as banks weigh the frustration of shareholders against the risks of a broader banking crisis, the choice should be clear.
