The SVB Failure: Why It Happened and What It Means -- 3/15/23
Today's special edition -- from Democracy: A Journal of Ideas, "The SVB Failure: Why It Happened and What It Means", by Richard Vague.
On Friday, March 10, the Silicon Valley Bank (SVB) of Santa Clara, California, with 17 branches, $200 billion in total assets, $180 billion in total deposits, and $70 billion in loans, was closed by California regulators, who next appointed the Federal Deposit Insurance Corporation (FDIC) as receiver. This ended SVB’s remarkable 40-year run of serving the venture capital community. The failure caught the bank’s investors, customers, funders—and the country—by surprise, and sent everyone scrambling to understand its causes and consequences. This was the second-largest bank failure in U.S. history, behind the failure of Washington Mutual in 2008, and most wondered if it was a harbinger of more pervasive problems, as was the case with bank failures in the global financial crisis (GFC) of 2008.
With the FDIC’s takeover, as is typical in these cases, the bank reopened on the following Monday, March 13—reconstituted and renamed the Deposit Insurance National Bank of Santa Clara (DINB). It resumed most of SVB’s business. Yet with the failure, all SVB shareholders suffered a total loss of their investment.
SVB’s failure was primarily the result of $1.8 billion in market value losses in its large U.S. Treasury bond portfolio—after all, the Federal Reserve had raised rates by a scorching 4.5 percentage points over the previous year—rather than the more typical loan-quality issues. It had been preceded two days earlier, but with less fanfare, by the announcement that Silvergate Bank, a comparatively small bank with $12 billion in deposits that specialized in serving the cryptocurrency industry, would wind down its operations and liquidate. Two days after SVB failed, the $100 billion Signature Bank—also popular among crypto companies—was closed, such that the already beleaguered crypto industry was dealt a double blow.
This sequence caused some panic, especially among the many depositors with large operating and payroll accounts at SVB—accounts so large that they exceeded the FDIC insurance maximum of $250,000—and who now wondered if their businesses would be able to survive at all. Venture capitalist and former tech CEO David Sacks tweeted: “Where is Powell? Where is Yellen? Stop this crisis NOW. Announce that all depositors will be safe.”
Regulators vacillated, but only for a moment, burning up the phone lines and then moving quickly and decisively to stem the fallout from the SVB failure.
This made for an extraordinarily busy weekend.
Bankers then took to their phones to reassure key depositors that their funds were safe, and rumors continued to swirl around a number of small- and medium-sized banks known to be troubled.
The FDIC confirmed that an auction was being conducted to sell DINB, with a bid deadline of 2 p.m. Sunday. By Monday morning they had announced that HSBC was the successful bidder for the small UK-based arm of SVB. The winning bid was £1 for a bank with 3,300 clients, a loan book of around £5.5 billion, and deposits of around £6.7 billion. But the weekend attempt to sell the U.S. bank had not found a buyer, a disappointing result that perhaps indicated that due diligence had revealed more uncertainty, complexity, or problems in the bank than anticipated— though by Tuesday glimmers of potential buyers had emerged.
Next, federal regulators reassured depositors that they would be protected, reversing earlier indications and announcing that all depositors at Silicon Valley and Signature banks, insured and uninsured, would be paid back in full and could access their money as of Monday.
But given the adamant opposition already being expressed by certain legislators from both parties to any government bailout or bail-in, regulators were careful to underscore that “no losses associated with the resolution of Silicon Valley Bank will be borne by the taxpayer.” Instead, any losses that could not be covered by the FDIC’s $128 billion Deposit Insurance Fund would be covered by assessments of (charges to) the banking industry itself. Even President Biden was quick on Monday to insist that the U.S. banking industry was safe, that customers’ deposits will “be there when you need them,” and that “[n]o losses will be borne by the taxpayers. Let me repeat that. No losses will be borne by the taxpayers.” The goal was to avoid having this characterized as a government bailout, but since the government had taken extraordinary steps, it was not unreasonable to consider its actions as such, and to see the distinction as more a matter of semantics than substance.
Finally, and perhaps most importantly, to help prevent potential deposit runs on any other banks, the Federal Reserve Board announced on Sunday that it would create a new Bank Term Funding Program (BTFP) to make additional funding available to eligible banks to meet their needs in the event of a run on their deposits. This program would provide loans of up to one year to banks that would pledge qualifying assets as collateral—all to be valued at par, a boon for institutions with underwater bond positions. Added to this, Treasury would make available up to $25 billion as a backstop for the BTFP, although the Federal Reserve does not anticipate that it will be necessary to draw on these backstop funds.
This was a powerful, even breathtaking response.
Then again, the Federal Reserve was created in the first place, back in 1913, to provide just this sort of funding in times of stress. So much attention is paid to the Fed’s current mandates and responsibilities as regards interest rates, unemployment, and inflation that most aren’t aware that it was created to provide emergency funding as the banking industry’s “lender of last resort,” in lieu of the industry’s unsettling and informal prior dependence on J.P. Morgan, who had had to serve in that role himself in major panics of the 1890s and early 1900s.
The “lender of last resort” role is in fact the most crucial and essential of all the Fed’s powers, although the least remembered. It was first articulated by Economist editor Walter Bagehot in the 1800s to describe similar actions by the Bank of England during financial panics. By stepping in with credit support, a lender of last resort gives markets the critical and indispensable gift of time to gather information, complete sales transactions, sort themselves out, and settle down. Though people generally don’t think about it this way, deposits are at bottom loans that customers make to banks, and therefore an FDIC guarantee of all deposits means that it, in addition to the Fed, is serving a huge, all-important lender of last resort role for banks—for whatever time that guarantee remains. In the GFC, the FDIC’s unlimited deposit guarantee remained in place for over a year.
Even with this strong response on Sunday, regulators braced themselves for a spate of small- and medium-sized bank closures that may yet come, and on Monday, the Federal Reserve announced that its vice chair for supervision was now leading a review of the supervision and regulation of Silicon Valley Bank.
As of Tuesday, markets were still jittery and bank stocks were suffering. Deposits continued to leave small banks for the presumably safer shores of large ones, causing many smaller banks to have to tap the Federal Home Loan Bank Board—a source of funding for banks—for an unusually large $90 billion on Monday.
Finger-pointing was quickly in full force, with Democrats blaming the Republican rollback of bank regulations, some Republicans claiming SVB failed because it was “too woke,” and some major investors lamenting that this government rescue portended the end of capitalism. And the Independent Community Bankers of America trade association was already howling against increased FDIC assessments to pay for SVB’s sins. All this was sure to lead to a renewed debate on the perennial and daunting question of fairness—whether to rescue, who to rescue, and who should pay.
But so far, the problems had remained contained.
Which brings us back to the question that was haunting the minds of Americans who lived through the GFC: Were these failures a harbinger of more pervasive and systemic national economic problems?
The short answer is no. While it will cause plenty of problems for certain constituencies (a point I will return to later), the broader economy will see nothing close to the calamity of 2008. As Moody’s highly respected chief economist Mark Zandi stated, “The system is as well-capitalized and liquid as it has ever been, [and] the banks that are now in trouble are much too small to be a meaningful threat to the broader system.”
Having dispensed with the doomsday scenario, the failure of SVB does return us to basic questions about its implications, and about risk, its stubborn persistence, and the ways that risk management fails.
Why SVB Failed: Committing a Cardinal Sin of Banking
Most commentators have attributed SVB’s failure to the losses in its large bond portfolio caused by higher interest rates, since it had purchased billions of dollars of longer-maturity Treasury and other securities in 2020 and 2021, when interest rates were low, and the market value of those bonds was crushed when interest rates rose in 2022 and early 2023. This brought to SVB unrealized losses—losses that are present, but not yet deducted from bank capital because of the specialized rules of the industry—so large that they threatened to obliterate its entire capital base.
But while the bond value decline was indeed the precipitating event, there was another key factor. SVB’s business was almost entirely focused on the tech sector, and in particular on early, venture capital-dependent companies, including cryptocurrency. SVB referred to this sector as the “innovation economy.” In contrast, most banks are intentionally much more diversified (and typically focused on less risky sectors), for the very purpose of risk mitigation and diversification. Regulators and rating agencies have long been wary of “monoline banks”—those that fixate on one sector or industry—regardless of which industry that happens to be, since, by definition, they will lack risk diversification. And, in fact, there are relatively few monoline banks in the United States.
SVB might very well have protested this characterization, since it purported to have four lines of business: commercial banking, wealth management, mergers and acquisitions/investment banking, and venture capital and credit. But those four lines were focused largely on early- to mid-stage companies of the innovation economy, so in reality SVB was simply serving four facets of one business sector.
Notably, and crucially for understanding this crisis, there is less opportunity for lending to this “innovation” sector since early-stage, venture capital-backed companies are by definition less creditworthy. That’s because they are usually highly unprofitable and require continuous large infusions of new cash from their venture capital investors to reach profitability. The consequence was that SVB’s loans were only a small percentage of its deposits—just over 40 percent most recently, while for similar-sized banks, that figure was typically 75 percent or more. Here’s the problem: With loan demand that low, the bank routinely had to invest a large relative amount of its deposits in U.S. Treasury bonds and other securities. Hence, it was ultimately more—and fatally—vulnerable to the bond value decline triggered by interest rate hikes.
Further, and again because of the nature of the sector it was focused on, SVB did not have as stable and well-diversified a base of deposits as other institutions. Bankers like to have a lot of small deposits from a wide range of households, rather than fewer large deposits from corporate customers, because those small deposits are less likely to flee the bank if signs of trouble appear. SVB’s sole focus on this sector was to blame for its too low loan-to-deposit ratio and its vulnerable, business-heavy deposit base.
During the pandemic, SVB’s growth skyrocketed. SVB had been a fast-growing lender and investor for some time, but in 2020 and 2021 that momentum accelerated from an annual loan growth percentage in the mid-teens to a rate of almost 40 percent per year. This increased loan balances from $33 billion in 2019 to $74 billion in 2022. But deposits were growing even faster and pouring in—soaring from $55 billion in 2019 to $186 billion in 2022. As is often the case with booms, SVB’s earnings performance was good in the early stages of this period. This deposit growth was from the boom-related good fortune of its customers, or, as the bank put it, from “liquidity events, such as IPOs, secondary offerings, SPAC fundraising, venture capital investments, acquisitions and other fundraising activities,” along with the fact that SVB was soliciting those deposits, often as a requirement for loans or other services.
But this deposit excess brought the bond portfolio problem. Since SVB now had over $100 billion more in deposits than loans, it had to put those deposits somewhere else, and so, in keeping with customary bank practices, it bought billions of dollars’ worth of bonds in 2020 and 2021. But since interest rates had collapsed in the pandemic, its earnings on those bonds would be negligible. This would adversely impact SVB’s earnings since it would be paying rates to depositors almost as high as or perhaps even higher than the rates it was earning on bonds.
The only way out of this crunch was for SVB to buy bonds of a longer duration than most banks are willing to purchase—with maturities of five years, ten years, or even 30 years, instead of the more customary one or two years—because, at that time, the longer the bond, the higher the interest rate it paid. Because of its low loan-to-deposit ratio, SVB’s bond portfolio was larger than those of most other banks, relative to its size. And since it was chasing a higher yield, the maturity on its bond portfolio was longer than for most other banks, too. And the longer the maturity, the more the value of those bonds will get crushed should market interest rates increase. Their market value goes down as rates go up. It’s a cruel equation and can be a perilous gamble for a banker.
Nevertheless, this is exactly the gamble that SVB took. In reaching for this higher yield, it was committing one of the cardinal sins of banking: It was investing short-term customer deposits in long-term bonds, creating an egregious mismatch of maturities, and thus leaving itself highly vulnerable if interest rates went up.
If rates were to rise, the interest SVB paid customers would have to rise to match market rates, but its interest income on the bonds it had bought would not, since they had a fixed rate of interest for the years until the bonds’ maturity. As that happened, the market value of those bonds would plummet, and if the resulting losses were recognized, they would reduce SVB’s capital below regulatory minimums, jeopardizing the bank’s very existence.
This danger is well understood, which is why banks are required to have “asset and liability matching” policies and reporting along with liquidity requirements to prevent it. Furthermore, regulators review those regularly as a core part of their assessment of banks’ risk management practices. But somehow these guardrails did not prevent SVB’s actions.
It was a fatal misstep. In reaching for higher yield on its bonds and never envisioning or planning for anything close to the interest rate increases the Fed would undertake, SVB wittingly or unwittingly had taken an existential gamble.
Then, in late 2021, the all-but-inevitable “innovation economy” reversal began. The tech sector boom became a tech sector bust. The NASDAQ Composite Index, which had shot up from the COVID nadir of 7,300 in March 2020 to a peak of 15,900 in November 2021, had plunged back down to 11,000 by March 2023. Venture capital funding dried up, the IPO market disappeared, and thousands of not-yet-profitable companies that had been started in the good times found themselves without access to funds that would keep them alive through the bad.
With the tech sector reversal, SVB’s core business was assailed at every level, and it began to see early signs of a potential decline in its mergers and acquisitions business and the associated fees, along with a potential decline in the valuation of its own investments in tech companies, and the value of the collateral for many of its loans. To top it off, deposits began to decline—a fact that management attributed to “elevated cash burn levels from our clients as they invest in their businesses.” At this point, SVB was surely seeing the fissures that could well become cracks in the foundation of its own earnings.
A High-Tech Version of an Old-Fashioned Run
The spring of 2022 brought worse news. The nascent inflation that had resulted from COVID-impaired supply chains and supply disruption from the Ukraine war shot higher, and the Fed was compelled to begin raising interest rates. While many blamed the NASDAQ fall and “innovation economy” reversal on the Fed’s rate increase, it is important to remember that they had started to tumble a year earlier. Markets, especially tech stock markets, had simply become too overblown and too untethered from any rational valuation metrics.
With the rise in interest rates, SVB’s bond portfolio plunged in value. The math was gruesome. By 2023, SVB’s $21 billion bond portfolio was yielding a mere 1.79 percent while the ten-year Treasury yield had increased to 3.9 percent. And the unrecognized loss from the decline in the market value of SVB’s bonds was poised to wipe out its capital. An analysis showed that if those losses were fully recognized, it could take the ratio of SVB’s primary regulatory capital, the core financial buffer a bank is required to have to absorb losses and protect depositors, from a comfortable 12 percent of assets all the way down to zero, a level so far below regulatory minimums that rating agencies would downgrade SVB, and regulators might be compelled to take over the bank. A similar analysis of most other banks showed that comparable unrecognized bond losses would take their same capital ratio down from, say, around 12 percent to around 9 or 10 percent—evidence of their smaller relative bond portfolios and their shorter relative bond maturities. The fact that SVB had such a large portfolio and such long maturities compared to most banks explains the draconian impact on its capital of the interest rate increases.
Meanwhile, and in addition, deposits were continuing to decline, and SVB management was wrestling with other issues, including, in their words, “expected continued higher interest rates, [and] pressured public and private markets.” They may indeed have been wrestling with these issues but not, it seemed, with warranted speed and urgency given the magnitude of the problems they were now facing.
Then came the coup de grâce for SVB—the downgrade from the rating agencies—which was woefully belated, actually, since rates had been increasing for a year and the projected deterioration of bond values had been easy to estimate during that entire time. For banks and other companies that get funding in the form of deposits or public securities offerings, ratings are not an academic exercise, but a crucial necessity. Those ratings, primarily by Moody’s and S&P, give potential and existing depositors, funders, and investors assurance that the funds they are providing are secure. In fact, most funders have policies that require ratings at a certain level or higher (“investment grade”) in order to qualify, and when ratings are lowered by the agency, the funders are required to withdraw those deposits and that funding.
Just days before SVB’s failure, Moody’s called to advise the bank that it was considering a ratings downgrade, which, if it happened, would likely spur a mass exodus of deposits from the bank. If SVB management had not yet seemed hyper-focused on this issue, the Moody’s call jolted them into action, initiating a two-part plan.
The first part of the plan was to sell $21 billion in securities, an action that was necessary for several reasons: SVB needed to have cash available to meet a deposit run on the bank, to better meet the Fed’s target of holding 4 to 8 percent of deposits in reserve, and to better protect against rising rates “if [the] slow fundraising environment and elevated cash burn trends persist.” It had accomplished this by Wednesday, March 8, but took a $1.8 billion loss in the process, causing its capital ratio to plunge below the regulatory requirement. The results were grim, but SVB would not have taken this action if it had had any alternative. They needed liquidity and, for most banks, the most readily saleable asset to raise cash is the bond portfolio.
The second part of the plan was to raise $2.25 billion in new capital from investors to fill this hole, and SVB already had a $500 million preliminary commitment from major investor General Atlantic to present to Moody’s and other prospective investors as evidence of progress.
Given the hope that the capital raising effort would be successful, Moody’s took its rating down only slightly. SVB hoped it had dodged a bullet, but word of its capital need spread almost instantly because of the Moody’s announcement and because SVB and the investment bank it had hired, Goldman Sachs, were reaching out to so many investors to solicit capital. Depositors were alarmed, and SVB’s stock plunged on news of the proposed share sale, ending Thursday, March 9, down 60 percent at $106.04.
Then, as the next links in the chain of calamity, the two things happened that SVB feared the most: A major run on deposits began, and the bank was unable to raise the remainder of the needed capital.
What were depositors and prospective investors hearing, and thinking? They certainly saw the Moody’s announcement and knew of the capital raise, and, as part of that, the issues with the bond portfolio and regulatory capital minimums. Perhaps they were also speculating regarding potential credit problems in the bank’s loan portfolio, or that its investment banking fee income would plunge, or that there would be a decline in the value of its warrants and equity investments in companies. Perhaps it was a combination of all of these concerns.
In SVB’s case, a substantial number of its depositors were a tight-knit community where the voices of just a few prominent venture capitalists alone could cause a run, and apparently that is largely what happened. Incidentally, this is another, subtler risk of focusing all but exclusively on one sector: Word spreads fast. Large startups, primarily in tech, rushed to withdraw their deposits.
If a large enough quantity of deposits leaves, a bank no longer has any funding, and if banks don’t have funding, they can’t operate, and if they can’t operate, they fail. As my boss explained to me when I was a novice banker, banks don’t fail because of earnings losses—they fail because deposits and other funding are pulled in anticipation of and fear of those losses. It’s a Silicon Valley high-tech twist on an old-fashioned bank run.
Since the capital raise had failed, SVB shifted gears immediately in the precious hours they had left and tried to find an outright buyer for the bank, but—given the absence of time and the many uncertainties—that effort failed just as quickly. In the middle of the business day on Friday, instead of at the end as is more typically done, California regulators closed the bank (surely a sign that the deposit outflow was accelerating), wiping out shareholders. Then they appointed the FDIC as receiver. Immediately after that, Moody’s announced it had downgraded the bank all the way to default, and S&P followed suit.
Could Management Have Prevented This Collapse?
While I can be accused of convenient Monday morning quarterbacking, it is nevertheless true that asset and liability management, liquidity management, and credit quality are the most important areas of bank management, because mistakes and mishandling in these areas are fatal—as SVB now knows indelibly. That these elements are the vital life support system for a bank was not an ineffable mystery only laid bare by this failure, but instead common knowledge among experienced bankers. So SVB could have and should have been well ahead of these issues, planning for them weeks, months, or even years in advance, and there were any number of ways that it could have mitigated the risk that doomed it.
It could have hedged its interest rate risk from the outset, it could have procured emergency sources of funding early on to tide it over in just this sort of situation, and it could even have laid off or discouraged the accumulation of deposits to avoid this very imbalance. But SVB did little if anything of the sort.
They didn’t pursue these preemptive moves because the moves are shunned; they are shunned because, like any form of insurance, they are expensive—in this case, very expensive. But in the view of conservative practitioners in this industry, that expense is a necessity.
In my career, whether in the role of CEO or as a young banker learning from my superiors, asset and liability matching and liquidity management were areas where we spent inordinate amounts of time in endless weekly—or daily—meetings; where we examined, forecasted, and planned for high-stress scenarios, just as a manufacturer might engage in natural disaster drills and planning.
The fact that SVB had to act with such flailing desperation meant it had not undertaken that type of risk forecasting and planning. Some blamed the panic and deposit run on management’s strategy and communication missteps in the final days, but it’s more reasonable to conclude that any actions at that point would have been unsuccessful regardless, and that the moment to act successfully had long since passed.
Is the Federal Reserve to Blame?
Perhaps, yes, although I’ve already shared here my view that all bank managements should be prepared for these contingencies. Nevertheless, it is true that higher rates are the proximate cause of this crash, and the Fed is entirely responsible for those rates. A number of economists have been imploring the Fed to forego further rate hikes, since they believe that the inflation the Fed is battling is a function of the COVID supply-chain issues and the adverse impact of the Ukraine war on the prices of oil, natural gas, wheat, iron, and other vital commodities—and that the Fed is therefore tilting at windmills with rate hikes, which will neither undo the effects of COVID nor end the war.
Even many of those who agree that the Fed had to increase rates to battle inflation feel those hikes have been overdone.
The irony, of course, is that the Fed is now rescuing banks from a problem of its own creation, and in any event the most important thing to note is that it’s fully within the Fed’s power to solve—or at least alleviate—the bond valuation problem and thus bring relief to the banking industry by lowering interest rates. With SVB’s failure, there is much current discussion about whether the Fed will need to rethink its future interest rate actions.
Why Didn’t the Rating Agencies and Regulators Act Sooner?
Interest rates had been rising, and thus decreasing bond valuations, for almost a year, and rapidly so for months. Any market participant or regulator with a basic ability to read financial statements could have seen the issues with SVB. So why didn’t the rating agency act sooner? Moody’s rating for SVB signaled “moderatecredit risk” until mere days before its failure. In a sense, the rating agencies reprised the role here that they played in the GFC—they assigned ratings that provided little of the needed insight to funders and investors. The stark truth is that, even with all the intense criticism of rating agencies and their role in the GFC, the rating process has never been truly or effectively reformed.
As for regulators, the California state regulators may not have intervened forcefully and early enough to prevent this failure either, even though assessing asset and liability matching and liquidity management is central to the regulatory examination process. But I would be inclined to give them the benefit of the doubt and, as a former state banking regulator myself, assume they were actively and vigorously discussing the matter with SVB. But I would make the generic, broader observation that regulators are not oriented toward intervening in a prophylactic or preventative manner—though I would argue that they should be, and I have proposed methods for doing so in some of my writings. Instead, regulators are best equipped to intervene after an institution has crossed below regulatory capital thresholds and the problem is fully manifest. The regulatory community overall has shown itself to be much more adept at repairing damage than preventing it.
Among many other clues of trouble, SVB had become partially reliant on the Federal Home Loan Bank for funding rather than more conventional sources, with $15 billion of outstanding loans by the end of 2022, after having had no loans the year before. Risk analysts would call this a classic red flag, which only provokes further questions about how closely regulators were monitoring developments there.
Ironically, in 2018, the banking industry had strenuously and successfully lobbied Congress to roll back oversight regulations and provide relief from stringent requirements for banks with under $250 billion in assets. These regulations and requirements had been put in place under the 2010 Dodd-Frank Act and included stress tests as well as capital and liquidity requirements. SVB management played an active part in that effort: For example, in 2015, SVB’s chief executive, Greg Becker, asked lawmakers to raise the asset threshold below which these so-called “significant regulatory burdens” from Dodd would not apply. The change passed, with 67 senators voting in favor. On the morning of March 13, Senator Elizabeth Warren, who fought against this weakening, assailed the move in a New York Times op-ed. It is reasonable to think that this very weakening contributed to the failure to prevent SVB’s demise.
In any event, SVB was a monoline bank—and not just any monoline, but a monoline focused on an inherently risky business sector—and that fact alone should have automatically flagged the bank for an ongoing higher level of regulatory scrutiny.
Will Depositors Get All Their Money Back?
Yes, they will get all of it back—and instantly—as a result of the weekend’s dramatic regulatory decisions. The real question now is: How much of the funds paid out for deposits will the FDIC itself recover? It may not have to pay out much if any, and even if it does, it is likely to recover most if not all of those funds. Thus far, most are assuming that this was strictly a bond valuation issue, and that as the FDIC disposes of the loans and the other investment assets (or, in the event the bank is sold, a potential buyer meets withdrawal requests), they will all be sound and thus the FDIC or buyer will be made whole.
That’s probably at least close to plausible, yet I’m not entirely sanguine on the question. I won’t be surprised if there are more loan problems than expected or if the other investment SVB assets, such as equity in venture-capital-backed companies, are being carried at values above market.
Keep in mind that the entire business was focused on a sector that is now beleaguered, and even though SVB was experienced in weathering adverse economic cycles, with its focus on a volatile sector, it is not reasonable to believe that it has made it through the tech downturn unscathed in all areas.
For example, though the bank went to great lengths to assure its investors that its loans were of high quality, touting the low risk of such practices as “loans secured by limited partner capital commitments,” “low loan-to-value mortgage loans,” and “commercial real estate with well-margined collateral,” its loan growth was so rapid that its own loan process might well have begun to fray. Veterans of prior lending cycles can readily describe the flaws that can infect and attenuate those seemingly sound criteria, and they can testify that rapidly accelerating loan growth often portends significant credit problems.
But an inordinate number of SVB’s customers had significant amounts in uninsured deposits, a remarkable fact given that we are only 15 years out from the GFC. The FDIC only insures deposits for a given depositor up to $250,000 at a single institution, and prior to the GFC it only insured up to $100,000. So risk-minded depositors should never have had more than $250,000 at SVB in the first place. It’s one of the easiest risk management strategies anyone can adhere to.
But for some reason, most don’t. Astonishingly, of the roughly $180 billion in customer deposits at the bank prior to the run, it’s estimated that only about $5 billion were fully insured, an unusually low percentage that indirectly reveals SVB’s unusually high dependence on corporate rather than retail deposits. This shouldn’t be, since it is easy for an individual or business mindful of this FDIC limit to apportion deposits among several institutions, or to put the funds in a short-term government securities fund and thus completely eliminate any deposit risk. In a sense, the high level of uninsured deposits at SVB underscores the troubling fact that even though there have been a wealth of books and articles on this very type of risk, we have a new generation of businesspeople with little sense of risk management.
And so the FDIC and Fed have had to step in, once again.
Also once again, many are asking whether the regulatory rescue will create moral hazard, and my answer is yes and no. Bankers have always pushed the limits of risk, even well before FDIC insurance was enacted in the 1930s. Likewise, depositors disregarded the limits of FDIC insurance in 2008, and did so again here in 2023. Many of the businesspeople I have talked to since March 10 were completely unaware of FDIC limits, which is remarkable. I tend toward the view that people—both bankers and depositors—forget the lessons of risk and ignore the lessons around moral hazard, no matter what the circumstances.
Will SVB’s Troubles Spread?
Contagion typically only hurts the already weak, and the banking industry is generally healthy at the moment. Assessing the industry as a whole, the bond valuation problems resulting from higher rates are visible and transparent, and constitute a large but manageable several hundred billion dollar problem in an industry that has well over $2 trillion in capital.
From the standpoint of credit quality, which was at the heart of the 2008 banking crisis, my view is that while there are small pockets of potential credit problems, including lending for private equity transactions and information technology, and great pressure on real estate loans because of increased interest rates, there is nothing approaching the rampant overlending and credit concerns that led to the 2008 banking crisis.
But any contagion concerns are probably moot in light of the Fed’s strong and comprehensive weekend announcements. A few of the weakest banks may yet fail. And this will be challenging for certain already-troubled international banks such as Credit Suisse. But the Fed and FDIC’s actions will likely keep U.S. failures to a minimum.
Regulators have come through quickly and thoroughly. As Treasury Secretary Janet Yellen has stated, “We want to make sure that the troubles that exist at one bank don’t create contagion to others that are sound.”
Will There Be Broader Economic Damage?
The answer regarding broader damage is most certainly yes for the sectors of venture capital, early-stage tech, and the already pummeled crypto (which was shoddily financed to begin with), since the SVB failure has further chilled the high-tech and innovation economy sectors. So early-stage companies that were already having trouble raising funds or obtaining loans over the past year will certainly have more trouble now. Ro Khanna, a congressman whose district includes some of Silicon Valley, has stated that SVB was “the lifeblood of the tech ecosystem,” which, if even partially true, will mean greater difficulty over the near term for hundreds of companies. But those sectors are only a moderate component of the overall U.S. economy, particularly in contrast to the massive size of the real estate industry in 2008, when it was the culprit in the GFC. As a proxy for size, in 2008, real estate loans were roughly 50 percent of total private sector debt, while today the information technology sector is a mere 3 percent of total U.S. private sector loans, and only 6 percent of all business loans. According to SVB itself, what it called the innovation economy—its sector of choice—is only 10.2 percent of the total U.S. economy.
In terms of the stock market, the answer is probably yes as well, although the unpredictability of that market is well known. In 2010, tech was 23 percent of the market value of the S&P 500, while more recently it has been almost 40 percent, which, since it is so high, may suggest that tech stocks are still very highly priced, and thus potentially more vulnerable to further market correction.
And the answer is probably yes for a small number of small- to medium-sized banks that may have entered the period of rising interest rates already vulnerable, and thus have limited capacity to absorb any market value decline in their bond portfolio.
But the damage will not approach the level of the systemic nationwide U.S. banking risk seen in 2008 or a crisis in the broader economy. SVB is small, albeit a top-20 bank. The banking industry has $17 trillion in deposits, its biggest banks have $2 trillion in deposits, and SVB had a comparatively very small $200 billion in deposits and $70 billion in loans.
Further, my researchers and I study aggregate credit trends across the economy, and the concerns we see now don’t come close to the concerns of 2008.
So it may take a bit of time to corral all of this, but there’s no need to go hide in a bunker.
Why Don’t We Ever Learn?
Why aren’t we more cognizant of risk after the lessons of the GFC and the pandemic? The truth is, we will never fully learn to give risk its proper due in our dealings, and in those brief moments when we seem to—when risk is illuminated for us like a blinking neon sign against the dark night of a crisis—afterwards we forget all too quickly. On risk, our knowledge is all but intrinsically ephemeral.
In large part, this is because, as the songwriter Chet Baker wrote, we fall in love too easily, we fall in love too fast. We should follow caution and balance, but instead we fall in love—with crypto and WeWork, Elon Musk and FTX. We fall in love because love can make us rich.
Then again, perhaps it’s more deification than love. And once we’ve deified these companies and their leaders, mere rules—the wearisome rules of risk management and discounted cash flow valuation—no longer apply. We want excitement and heroes. We’re taught that markets are efficient and rational. Instead, they’re more like teenagers at a Justin Bieber concert, or collegians at a keg party—wild and ecstatic; bleary-eyed and hungover the next day—who swear they’ll never overindulge again…but are soon ready for new highs.
But more than this—and perhaps because of this—the institutional structures we build to protect us from risk are poorly designed, neglected, underfunded, and too often simply ignored.
Bond rating agencies are paid by the industry. Who believes that they can ever truly work? And the actions and authority of regulators are more designed to act after a calamity is underway, with little appetite or support for acting preemptively.
Again and again, when risk exacts its vengeance, we mask our culpability behind phrases like “perfect storm” and “black swan”—even when risks like these are excruciatingly foreseeable and preventable.
We never give ourselves over to the tethering lessons of risk, and so neglectingrisk management, rather than heeding sound practices, is likely to remain our habit.
|The SVB Failure: Why It Happened and What It Means|
|Democracy: A Journal of Ideas|