“[Interest Rate Swap] is a perfect example of the gains that can be realized from specialization along the lines of comparative advantage. Triple-A and single-B can together reduce their joint costs of borrowing by each borrowing in the market in which they get the best terms; then, using a swap, they can divvy up the savings they have realized, and each ends up with the type of liability they wanted in the first place.” Says Marcia Stigum (1978)
Silicon Valley Bank (SVB) was born on October 17, 1983. It was announced dead by the FDIC and other government agencies on March 1o, 2023. Nonetheless, SVB’s failure began an intriguing postmortem debate on bank runs. Many discussions have been conducted on essential issues such as the roots of the crisis, the nature of government bailouts, and recrafting bank regulation (such as making liquidity measures such as liquidity coverage ratio (LCR) time-sensitive). However, one of the standard narratives indicates that the SVB’s doomed fate is determined by the classic type of depositors who follow herd behavior and are primarily driven by human psychology and panic.
However, the March 2023 banking crisis revealed that modern depositors’ behavior, including their decision to run on a bank, is less driven by the animal spirit and more by the actual information in banks’ financial statements. In other words, modern depositors’ behavior has shifted from being uninformed and displaying herd behavior to becoming individually informed and information-derived. This distinction between traditional, uninformed depositors and modern, informed ones has essential implications for the measures to stop the bank run. Given the nature of the modern depositors, banks’ usage, and reporting of hedges, including interest rate futures and swaps, would be a better option to stop a run on a bank and a banking crisis in general than other classic solutions such as deposit insurance. In addition, accounting for the structural change in the type of depositors, from uninformed to informed, could open new avenues, such as using derivatives, to improve financial stability in the future.
Most standard narratives of the March 2023 bank run are based mainly on classic models such as the Diamond-Dybvig Model of bank runs. In this model, the assumption is that the depositors are uninformed. To such depositors, as long as they did not reach the $250,000 threshold, putting money in a bank appeared as safe as buying Treasuries. Both investments had full government backing. As a result, they do not need to evaluate the financial health of their deposit-taking institutions. As their behavior is driven by “animal spirit” rather than a particular detail in banks’ financial statements, “any” worries, imaginary or real, about the viability of such insurance, if believed by enough people, could lead to a run on a bank. In this environment, the model suggests that even though some banks tend to stop the convertibility of checkable deposits to currency, the best option to stop a bank run would be “deposit insurance.”
In contrast to the assumption of such models, the depositors who initiated the runs on SVB, Signature, Silvergate, and other regional banks are characterized as informed, finance-savvy, well-informed, and informed. They are known to use all the available information and extensively tweet about every detail and footnote in the financial statements of their bankers. For instance, in the case of SVB, they examined the bank’s balance sheet, “marked to market” its assets, and noticed its exposure to interest rate risk. Moreover, they initiated the run on the bank once the bank reported virtually no interest rate hedges on its massive bond portfolio. In other words, the run on the SVB revealed that modern depositors’ behavior is less driven by the animal spirit and psychology and is indeed informed.
This distinction between traditional, uninformed depositors and modern, rational ones has essential implications for the measures to stop the bank run. It opens the door for derivatives to be considered as a private-risk management tool that enhances financial stability. One of the most significant risks to SVB’s business model was catering to a rational group of investors who treated bank deposits as a form of investment vehicle and continuously checked the bank’s balance sheet. This change from uninformed to rational (well-informed) depositors is a continuation of the existing evolution in banking. In the first phase of this structural change, “retail” depositors were replaced by “institutional” ones such as pension funds. The main difference between these two was the amount of cash they would inject into the banking system.
After the Great Financial Crisis (GFC), and the strengthening of the liquidity and capital ratios, systemically important banks, such as JPMorgan Chase, Credit Suisse, and Bank of America, restructured their business models to become market makers in wholesale money markets. They minimized their deposit-taking activity. As a result, medium-sized and large regional banks changed their business model to pick up the slack. For these banks, retail deposits were a dead end. Nontheless, attracting large deposits per se was not the only goal when it came to institutional deposits. Instead, they sought “well-informed” clients such as tech, crypto, venture capitalists (VCs), and startups. Therefore, banks such as SVB, Signature, and Silvergate preferred serving highly specialized, local, and finance-savvy depositors. These regional banks structured their business models to become the Bank of Crypto, VCs, or Startups.
If the change in systemically important banks’ business model transformed the market-making business, large, regional banks’ restructuring changed the banking and deposit-taking world. On the liability side, the modern banking era started with the “death of the retail deposits.” This evolution was revealed in the balance sheet of the SVB. On the liability side, much of the recent SVB’s deposit growth was driven by VCs and startup businesses. These depositors obtained liquidity through liquidity events, such as IPOs, secondary offerings, SPAC fundraising, venture capital investments, acquisitions, and other fundraising activities.
On the asset side, SVB and other specialized banks that served rational depositors were outside the business of making loans. Indeed, startups and crypto customers do not have good collaterals like fixed assets or recurring cash flows. As a result, they are less-reliable corporate borrowers. Nonetheless, the most important reason was that these customers did not need loans. Instead, equity investors provided them with a constant supply of cash. Marcy Stigum, in the late 1970s, called this a “death of loans.” The idea is that top corporate customers have access to public and private market credit, where they can borrow more cheaply than banks. Banks have adjusted to the loss of this business by instead purchasing fixed-income securities. In the case of SVB, government bonds became a large portion of the bank portfolio.
This change from uninformed to rational depositors and the corresponding assets-liabilities strategies contributed to the ongoing banking mania in at least two hybrid ways. On the one hand, serving rational depositors who do not need loans implies the large-scale addition of long-term bonds backed by the US government to the banks’ portfolios. The result is that these specialized banks are unusually exposed to “interest-rate risk.” When interest rates go up, most banks have to pay more interest on deposits but get paid more interest on their loans and end up profiting from rising interest rates. On the other hand, banks such as SVB and Signature own a lot of long-duration bonds. These bonds’ market value goes down as rates go up. Every bank has some mix of this — every bank borrows short to lend long; that is what banking is — but many banks end up more balanced. At the same time, rational depositors, unlike traditional ones, continuously pay attention to the financial statements of the banks and the footnotes. As a result, they constantly “mark to market” the banks’ financial assets and extremely penalize the banks whenever the fair value of their assets is at a loss.
Let us use an example to understand this point and its systemic importance. Let us assume that a bank has $100 worth of deposits as its liabilities. On the asset side, the bank uses cash to purchase a government bond worth $100. In the meantime, the Fed announced a rate hike from 0% to 2%. A traditional depositor would account for the bank’s assets “at cost,” using the price the bank paid to purchase those bonds, in this case, $100. A rational depositor, in contrast, continuously marks to market the value of the bank’s assets. In this case, they account for the value of the bonds at their fair market value. If the bank has a bond with a face value of $100, this type of depositor will check the market price for that bond when the Fed announces its interest rate policy. If, for instance, it is $97, then the bank’s asset is only worth $97 on its balance sheet. The other $3 is gone, and the bank is considered insolvent. Rational people notice, tweet, and write long threads. Then, they start a bank run. This characterization explains the behavior of SVB’s depositors, mainly from the tech and VC industries. As rates went up fast, SVB’s depositors, who were all on Twitter a lot, read the footnotes on the notices on the financial statements of the SVB and started to write long threads on the topic that SVB was insolvent. As a result, they all pulled their money out at once.
Rational depositors, not unlike traders and capital market investors, use all the available information and continuously mark to market the balance sheet of financial institutions. Unfortunately, in doing so, they heightened the bank’s exposure to interest rate risks, especially for the banks that invest in fixed-income securities, such as government bonds. However, this opens up an opportunity for using interest rate hedges to stop the run on the banks. Interest rate hedges are often in the form of swaps, a financial instrument that effectively turns an investor’s fixed-rate loans or bonds into floating rates by paying a third party. These can be important for banks like SVB because many of their investments are tied to fixed-income bonds like mortgages or Treasuries.
When rates go up, fixed-income bonds fall in value, as with SVB. Why do bond prices respond to interest rate fluctuations? Remember that in a competitive market, all securities offer investors fair expected rates of return. If a bond is issued with a 5% coupon when competitive yields are 5%, then it sells at par value. If the market rate raises to 8%, however, who would be willing to pay a par value for a bond that only offers a 5% coupon bond? The bond price must fall until its expected return increases to the competitive return of 8%. In this environment, the fair value of the bonds will fall. Even if the investors are intended to hold these assets until maturity, which indicates they would still earn the par value, a mark-to-market method reduces the fair value of these investors’ assets.
However, once these available-for-sale, outstanding bonds are combined with interest rate hedges, such as swaps, their value can be protected against interest rate movements. One of the swaps’ functions is to transform the nature of an asset. Consider SVB in our example. A proper swap could transform SVB’s bonds earning a fixed interest rate into an asset earning a floating interest rate. Suppose SVB owned $100 million in bonds that will provide 3.2% for two years and wishes to switch to the floating rate. It contacts a swap dealer, such as Citigroup. We assume it agrees to enter into a swap where it pays the fixed rate (3.2%) and receives floating plus 0.1%. Its position would then have three sets of cash flows. First, it earns 3.2% from the bonds. Second, it receives floating under the terms of the swap. And third, it pays 3.2% under the terms of the swap. These three sets of cash flow net out to an interest rate payment of floating plus 0.1% (or floating plus 10 basis points). Thus, for SVB, the swap could transform assets earning a fixed rate of 3.2% into assets earning the floating rate plus 10 basis points. In the real world, however, SVB reported virtually no interest rate hedges on its massive bond portfolio at the end of 2022. Instead, it terminated or let expire rate hedges on more than $14 billion of securities throughout the year, the company said in its year-end financial report. Being unhedged was SVB’s fatal sin that its depositors did not forgive.
In this piece, I shed light on a corner of the crisis that has made this run on the bank different from the classic ones. In the current banking crisis, the collective behavior of the depositors to withdraw their cash from these banks, more than being driven by pure speculation and panic, is driven by depositors’ careful examination of the banks’ financial statements. In other words, in this crisis, we are dealing with rational depositors rather than uninformed ones. Although small, this detail would change the best approach towards stopping the run on the banks, away from deposit insurance towards using derivatives. The influential Diamond-Dybvig Model posits that deposit insurance is the best way to deal with bank runs. Accordingly, in March 2023, after the SVB’s failure, the government introduced extraordinary measures to extend the deposit insurance “to all the deposits” and disregard the $250,000 threshold. Nonetheless, despite the government’s all-in approach to providing deposit insurance, such a state-backed promise did not stop the run on the other banks.
The US government’s failure to calm the market, despite offering full-blown deposit insurance, calls for a new framework beyond the classic Diamond-Dybvig Model, where the depositors choose to remain uninformed about the financial health of their bankers. The SVB’s tragedy revealed a modern financial system made by modern, well-informed depositors. These depositors treat their deposits as a form of investment vehicle. In this new financial reality, what stops the run on a bank is a proper hedging strategy that is reported in the bank’s financial statement instead of government-backed insurance. In a strange twist, the March 2023 banking crisis teaches us that the regional banking market structure has moved in the direction that a private risk-management solution, available through the derivative markets, would be a superior resolution than the government-backed insurance for financial stability. Every cloud, even if it includes a cascade of bank runs, seems to have a silver lining.
Footnote: While writing this piece, I am visiting my family in Iran. My access to the internet is nonexistent. However, a student of mine sent me all the articles through text messages. Like always, my students are my main drivers and backers in my intellectual journey.