What Have We Learn from the Crisis and Subsequent Policies Response?
The Unnecessary Banking Crisis
You have probably heard that the U.S. banking is in crisis, sparked by the failure of Silicon Valley Bank (SVB) on March 10th along with the closing of Signature Bank (SB) on March 12th which created concerns about the regional banking industry generally. You have probably not heard a legitimate summary of what is going on, what mistakes were made, what lessons need to be learned, and what is necessary to end the pattern of periodic financial/banking crises.
SVB had a business model of collecting low-cost, high-touch deposits from start-up companies and the venture capital community and investing these deposits in low-risk securities. Most of these deposits were large accounts that are mostly not guaranteed by the FDIC (essentially the Federal government). With short-term interest rates close to zero, SVB shifted towards investing these uninsured deposits into longer-duration assets. When interest rates rose rapidly last year, the value of these longer-duration assets fell, and the bank’s solvency came into question. Depositors responded by rapidly withdrawing their deposits and the bank failed.
Was SVB Bankrupt?
On March 10th, absolutely. On March 8th, arguably not at all. And this is the key to understanding what is currently happening in the regional bank sector.
There is an underlying tendency to slot SVB’s problems into the framework that “rising rates killed the U.S. Savings & Loan industry in the 1980s” but there is a key difference. S&Ls basically went from paying 3% on deposits and making 6% on mortgage loans to paying 9% on deposits while continuing to collect 6% on their loans. Time was not going to solve the S&L problem - the industry was both insolvent and unprofitable (which means the industry is getting more insolvent every day even after interest rates peak).
In contrast, time would have solved SVB’s problems: this is the key difference between a solvency crisis and a liquidity crisis. Collecting low-cost, high-touch deposits to invest in short-term Treasuries at 0.3% is unprofitable (hence why the SVB had shifted to longer-duration assets) but doing the same to invest in short-term Treasuries at 4.5% is fantastically profitable. Once interest rates peaked, SVB would rapidly be getting less insolvent every day. The best way to think about this is that yes, SVB had longer-term assets that were hurt by rising interest rates, but the bank also had longer-term deposits that benefited (the bank) from rising rates. At least that is what everyone thought.
On March 8th SVB still had low-cost, high-touch deposits and on March 10th deposits were not leaving in search of a higher interest rate. If SVB’s deposits are sticky, the longer-duration assets don't lead to insolvency because they are offset by what are effectively longer-duration deposits. Although technically insolvent, with stable deposits SVB is economically solvent despite rising rates eroding the value of the bank's assets. However, if SVB’s deposits turn skittish, the bank's longer-duration assets would suddenly be matched with short-duration deposits (that don’t benefit the bank in a rising rate environment) and SVB would instantly become economically insolvent.
SVB was technically insolvent on March 8th based on the market value of the bank’s assets, ignoring the market value of the bank’s deposits (which on March 8th everyone assumed was quite high; sticky low-cost deposits are worth a lot in a rising-rate environment). Whether the bank was economically insolvent depended on whether deposits were in fact skittish or long-duration. By March 10th it became clear deposits were skittish and SVB failed.
This is not a new lesson in banking – every bank today has a “confidence switch” that if it gets flipped turns the bank from healthy to bankrupt. No bank survives a deposit run. The category “core deposits” exists to give a sense of how sticky a bank’s deposits are versus, say, brokered CDs that jump from bank to bank looking for the highest rate. This is important because core deposits are not supposed to run.
What is new is that SVB’s failure undermined our collective faith in the underlying value of at least one kind of core deposit and with that, the regional bank business model. Deposits that a month ago were considered some of the most valuable in a regional bank now seem potentially toxic. Possibly low-cost uninsured corporate deposits only work for national “too-big-to-fail” banks. Which creates its own set of problems.
As an aside, a lot has been written about how SVB’s management made bad decisions and took ill-advised risks. And while we agree with both sentiments, the bank has plenty of company. SVB essentially made a big bet on Team Transitory and, if inflation had come down organically in 2021, none of us would be talking about SVB today. Blaming SVB management in isolation misses potentially important lessons from this crisis. Although blaming SVB management initially did have the possible side benefit of pretending that SVB’s problems were a unique isolated case. Unfortunately, this opportunity was lost through an unforced regulatory error.
What Fueled the Banking Crisis?
On March 12th the potential banking crisis turned into an actual crisis. Signature Bank’s (SB) closure two days after SVB’s failure poured gasoline on uninsured deposit fears. SB was neither technically nor economically insolvent – the bank was not invested heavily in longer-term assets – but most of the bank’s deposits were large uninsured accounts (only slightly less than SVB). SB’s closure meant that SVB was not an isolated case and confirmed the risk of uninsured deposits, which created a self-fulfilling prophecy – if uninsured deposits are viewed as risky for regional banks, then uninsured deposits start to leave regional banks, making them risky for regional banks.
This is a huge regulatory error given that simultaneously the Fed/FDIC guaranteed uninsured deposits at SVB and SB (if you were not going to cover uninsured deposits at SVB, preemptively closing SB to prevent a likely run would arguably have been prudent). In 2023 uninsured deposits are a major fact of life – they represent more than half of the deposits at many large banks. Uninsured deposits represent one of those confidence kill switches. With confidence the bank is healthy and profitable; without confidence the bank is dead (without massive support).
First Republic Bank (FRB) is the case study for this. FRB was a lot like SVB, technically insolvent (because FRB has more longer-duration loans while SVB had longer-duration securities, but the idea is exactly the same) but profitable and economically solvent. At least until the confidence switch is flipped. Which the SB closure did.
At that moment FRB was dead absent massive support and the banking crisis began. FRB failing likely flips more confidence switches, more banks fail, more switches flip, until something breaks the cycle, or the regional bank sector is gone. Regulators are trying to save FRB (various forms of support together exceed $100b so far) but it is very hard to stop a banking crisis that has basis in fact (SVB and FRB were technically insolvent), even if a more nuanced read of the banks’ condition supports continued confidence. This is the core challenge in banking – both confidence/healthy banks and no confidence/dead banks are legitimate rational equilibria at virtually every moment in time.
It is hard to overstate how costly an error it was to close SB on March 12th. We have no special insight into why a solvent bank was closed but it clearly made saving FRB harder (maybe impossible). How do we know SB was solvent? First, estimating the market value of a fixed income security in response to rising interest rates is straightforward math (it is harder to adjust market values for changing credit conditions). Second, on March 19th SB was “bought” by a peer, New York Community Bank, with FDIC reporting a cost of $2.5 billion. That is, a week after SB was closed by regulators, precipitating a banking crisis, the bank was barely insolvent. Over that same week FRB’s parent company lost over $20 billion of equity value. And FRB is only the next domino in the regional bank chain.
Where does this end?
We know that some regional banks are technically insolvent due to interest rate risk in their assets – but very few are economically insolvent if their deposits prove to be sticky. However, uninsured deposits continue to leave regional banks, making them less profitable and driving the most marginally solvent banks (both technically and economically) to insolvency. If this cycle continues, much of the regional banking industry is likely to disappear.
We also know that shifting interest rate expectations have reduced the immediate pressure on banks with longer-duration assets. Long-term interest rates jumped in early March as investors moved towards expecting a 50-bps hike later in the month and the Fed ending the year between 550-575 bps. Investors now consider a 25-bps as most likely in March (with nearly a 38% chance of no hike) and project ending the year between 375-400 bps. Long-term rates have come down to reflect updated expectations, supporting longer-term asset values and capital levels just as regional banks may be needing to sell securities and loans to raise cash in the face of uninsured deposit instability. Banks being able to easily sell assets to cover deposit outflows alone could quell the sector's instability. Returning confidence can also be a self-fulfilling prophecy, this time in regional banks’ favor.
Making Progress from Making Mistakes
From an Ingenuism perspective, the important question is what have we learned from the crisis and the subsequent policy response? Progress comes from making mistakes and learning from them, so you do something different (and more effective) in the future.
First, it appears that a 21st Century banking crisis moves even faster than past crises. Panic can spread extremely rapidly when we are all connected and moving your deposits takes less than a minute on your iPhone. In particular, the illusion bank regulators have tried to create, to have depositors believe that deposit insurance is unlimited while investors believe that it is limited, is broken.
What comes next is unclear but there is no reason to believe it will prove any more effective if it is a single solution designed by experts to address the current crisis. We happen to be experts on this here at Ingenuism – and we don’t imagine to know the appropriate response to the financial instabilities that the ongoing crisis has revealed – and neither do other experts. Only our collective ingenuity is up to answering that question.
Which requires trial and error plus learning. Of course, errors can be very costly in the financial system. The Ingenuism approach is to allow experimentation where it is low cost and build from there. For example, smaller banks could be allowed to opt-out of deposit insurance while regulators simultaneously raise the cost (both financial and regulatory) to insured banks so that it reflects the actual subsidy banks receive from being guaranteed by the taxpayer. This sets the table for innovation that produces more valuable/stable relationships. As these evolve over time the opt-out ceiling rises and more banks (and their customers) can experiment and take advantage of the fruits of past experimentation.
And this is only one obvious idea! In its current form banking seems to periodically deliver unimaginable crises. We trust that in a supportive environment innovation would produce a more effective financial system that is unimaginable today. We also believe that the current environment will produce more of the same. You choose.
The broader monetary policy question is harder. SVB’s failure has its roots in past policy of zero-interest rates. Given the dislocations from Covid, were zero-interest rates worth it? We have no way of knowing without trying different alternatives. Ingenuism applies here as well but the path to experimentation and learning seems even less likely than with bank regulation. However, we can continue to point out that top down solutions will at best only improve slowly while in other equally-complicated areas our collective ingenuity has changed the world.