On an otherwise sleepy afternoon of Friday, March 10, 2023, we witnessed the second-largest bank collapse in US history. Around midday, the California Department of Financial Protection and Innovation (DFPI) announced that it had taken possession of Silicon Valley Bank (SVB), an institutional darling of the tech sector, and appointed the FDIC as the now-failed bank’s receiver. Ironically, the firm’s parent company, SVB Financial Group, had just been awarded Bank of the Year at a posh London gala earlier in the month. Not surprisingly, skittish investors reacted immediately to the news, as the day before, the four largest US banks lost some $52 billion of market value. More on why this happened to follow.
How did we get here, and what does it mean? Well, first, let’s take a look at how SVB operated, and what made it unique within the banking sector. While SVB was by no means the biggest dog among its banking brethren, it was still a success story, having grown to hold some $209 billion in assets since its founding nearly 40 years ago. From its inception, SVB served as a niche specialty bank, playing a critical role in providing capital for startups. As Deborah Piscione points out in her book Secrets of Silicon Valley: What Everyone Else Can Learn from the Innovation Capital of the World, at the time SVB was founded, traditional banks suffered from a fundamental misunderstanding of the needs of startup firms. Understanding that startups do not generate revenue immediately, founders Bill Biggerstaff, Robert Medearis, and Roger Smith created a risk-management model that initially connected customers with its vast network of lawyers, accountants, and venture capitalists. Once these relationships were established, they began collecting deposits from startup firms that they linked with their venture capitalist partners.
Next came a double dip, of sorts, as SVB began lending to both private equity/venture capital firms and startups. Initially, startups were required to pledge roughly half their shares to the bank as loan collateral, but as many of its loan customers repaid their loans early, the collateral requirement was lowered to seven percent. Borrowers were also required to enter an exclusive relationship with SVB – a requirement that ultimately laid the foundation for the bank’s demise. While SVB created a risk-management scheme that, for a long while, worked internally for them, it failed to allow for debtholders to diversify their own risk. This is an important factor in the recent tragedy, as the majority of the bank’s clients were susceptible to a particular kind of risk, namely interest rate risk. As a general rule, tech firms tend to fund their operations with equities sales, because the cost of obtaining this type of capital is low, relative to assuming debt. This isn’t to say that they don’t borrow; as previously noted, a large part of SVB’s business was comprised of lending to such firms. Rather, as Static Tradeoff Theory holds, firms consider the weighted cost of capital and make decisions that maximize the return on these costs.
For venture capitalists and other investors, tech stocks are often attractive, despite their risk because of firms’ growth potential. Yet they often pay low dividends and have high price-to-earnings ratios. No one really cares whether their investments are overvalued when rates are low, because expected potential future earnings are higher in a low-rate environment. When rates rise, however, investors tend to pivot toward concern for short-term profits. To wit, they start looking for safer havens to make their investments. This causes a dual difficulty for tech firms; there is less demand for their equities, and the cost of assuming debt is higher. This has a negative impact on cash flows and the ability to reinvest toward growth.
The recent diet of increases by the Federal Reserve of the Fed Funds rate caused many of SVB’s clients to begin withdrawing funds, in order to increase their liquidity and bankroll their operations. Remember, the equities of these firms were no longer in high demand, and the cost of borrowing had increased, so withdrawing their own deposits represented the lowest cost of capital. In order to honor these withdrawals, SVB had to begin selling assets – and here is where it gets interesting. While interest rates were low, SVB used a standard banking practice known as net interest margin to generate a tidy return on its clients’ deposits. In essence, with the Funds rate effectively at zero, they invested depositor funds in relatively stable securities with high yields, pocketing the difference between the interest paid on deposit accounts and the yields.
The securities purchased were a mixture of Treasury bonds and long-term mortgage-backed securities. Their exposure to the mortgage securities was high; capitalizing on yields of 1.5 percent, they invested a reported $80 billion of their $189 billion into these vehicles. There is nothing inherently risky in these mortgage bonds, but as with any bond, they’re highly sensitive to changes in interest rates. They’re also not as stable as their fellow long-term liabilities, which SVB was using to secure short-term deposits: Treasury bonds. Unable to sell these securities to meet the cash crunch, the bank was forced instead to sell its $21 billion portfolio of Treasury bonds on Wednesday, March 8. The problem was that this portfolio was now producing a yield of 1.79 percent, a significant haircut below the current 10-year Treasury yield of 3.9 percent, resulting in a shortfall of $1.8 billion. It still had obligations to meet, so the next day, SVB announced that it was selling $2.25 billion in both common and preferred convertible stock to meet the shortfall.
Investors were not impressed, but concerned that the value of their shares would be further diluted. Share prices ended the day down 60 percent. Additionally, a number of influential venture capital funds such as Peter Thiel’s Founders Fund began advising their clients to pull their deposits out of SVB, triggering a run on the bank. By Friday morning, SVB CEO Gregory Becker began desperately looking for a way to raise capital, up to and including finding a buyer for the bank, but around noon the DFPI and FDIC pulled the plug. With the failures of crypto-banks Silvergate and Silicon Valley Bank earlier in the week, Thursday’s announced sale of SVB equities triggered a general selloff of bank shares that saw the US’s four biggest banks – Morgan Stanley, JPMorgan Chase, Bank of America and Wells Fargo – lose a combined $55 billion in market share, as bank investors worried about the ability of banks to meet their obligations if depositors began withdrawing accounts en masse. This was likely not helped by FDIC Chairman Martin Grutenberg’s noting in a speech to the Institute of International Bankers that very week that banks collectively held some $620 billion in unrealized losses – the holding of an asset that has declined in value but has not yet been sold – related to their bond holdings.
At first glance, this may all seem to paint a dire picture of the state of the banking sector. But the opposite should be the case, absent an irrational bout of mass hysteria. In order to mitigate such a possibility, officials of the FDIC, Treasury Department, and the Federal Reserve have decided that all depositors of both Silicon Valley Bank and Silvergate (Silvergate voluntarily decided to liquidate and wind down its operations) would be made whole. Typically, the FDIC’s Deposit Insurance Fund, funded by fees paid by its member institutions, guarantees only the deposits of those who have $250,000 or less in any one bank account. Uninsured depositors, or those with more than $250,000 in deposits, must wait until the bank’s assets are liquidated and sold to recover any deposits beyond the insurance limits. These limits have been dispensed with for depositors of both banks; all customers will be able to retrieve their deposits with any losses, which FDIC covered by charging a special assessment on member banks.
Additionally, the Federal Reserve is in the midst of considering a general banking facility for other financial institutions who may have sizable exposure to SVP; in essence, a type of loan scheme which makes loans available to banks which may have short-term liquidity problems, to mitigate possible runs. While some may improperly view this as a taxpayer bailout, the reality is that the Fed generates its own funds via the interest on its vast holdings of Treasury bonds and the higher rates it charges on loans to its member banks. While it is wise that the FDIC and Federal Reserve act to calm skittish investors and bank customers, the most compelling factors in favor of stability lie with banking institutions themselves.
SVB, Silvergate, and Silicon served as niche specialty banks, with their business highly concentrated within one or two industrial sectors. This concentrates risk among parities who are largely susceptible to the same types of risk. More traditional banking institutions do business across many divergent sectors, and accept deposits from a varied range of customers, spreading risk and inuring themselves to the effects of a particular singular shock. Speaking of risk, as a regional bank with fewer than $250 billion in total assets, SVB was immune from liquidity coverage ratio (LCR) regulations, which require that banks hold an amount of liquid assets commensurate with the funding of 30 days’ worth of cash outflows. This is critical, as the business model of any bank, including niche specialty banks, involves utilizing short-term deposits to extend long-term loans. Instead of assuring its own liquidity, nearly 40 percent of its holdings at the time of its collapse were comprised of mortgage bonds that were definitively illiquid, with the sale of comparably liquid Treasury bonds failing to fill the gap. More traditional banks hold a wider variety of assets that comply with LCR, making them less likely to be unable to meet required outflows.
As a result, it is highly unlikely that most traditional banks would be particularly sensitive to the unrealized losses in their portfolios from securities which must be held to maturity. Long-term bonds aren’t the foundational node of their asset holdings, which are diverse enough to hedge against such things as rising interest rates. Finally, there are enough functioning, healthy banks to purchase and absorb SVB’s assets without any great shock to the market. Again, the Treasury and mortgage bonds held by SVB were structurally sound without any great underlying risk. They simply held too many of them for too long, exposing themselves to the effects of rising interest rates. It is, in fact, fair to note that the biggest reason not to fear any great contagion effect rocking the banking sector is that SVB’s managers were just plain irresponsible. The increase in the Fed Funds Rate didn’t happen overnight, was anticipated and announced, and took absolutely no one by surprise. Their failure to take steps to hedge against imminent deflationary policy was grossly negligent, and ultimately unlikely to be repeated by bank managers whose responsibilities extend beyond a specific boutique market.