SVB and Beyond

Dear clients, friends and people who tolerate occasional emails from me,

I wanted to address the failure of Silicon Valley Bank (SVB), the 2nd largest in US history, which occurred on Friday. If you’ve read extensively on this, you may know more about this than I do, but nevertheless, I would like to share some facts and thoughts with the intent to 1) explain how this happened and 2) allay some concerns about systemic risks in the wider banking system or at Enterprise.

Source: Wikipedia

What immediately did SVB in was a good ol’ fashion Bailey Building and Loan bank run. Depositors lose confidence in the bank, en masse decide to pull out their deposits, and the bank can’t meet all the demands for withdrawals, at least in the moment. And the truth is…any bank is susceptible to this risk. No bank has all the cash on hand to meet the total amount of deposits that are held there. Banks are fractional reserve institutions (i.e. hold only a fraction of the total deposits on hand), and in addition to keeping deposits safe engage in a variety of activities like lending and investing. JP Morgan Chase, the largest US Bank with a stupid $3.7 trillion in assets, has $2.4 trillion in customer deposits (this is as of 12/31/22) and only $2.2 trillion in investment securities, which includes cash, but also longer dated investments. In theory, JPM is susceptible to a bank run. But that won’t happen as long as the public is confident in JPM’s ability to meet its obligations. JPM has plenty of assets that they can sell if need be to meet redemptions. But this is where things get dicey. First a little history.

Founded in San Jose in 1983, Silicon Valley Bank grew to be one of the largest banks in the US (#16 as of 12/31/22 per a Fed report, with $209B in assets) via partnering with VCs, taking in large amounts of deposits and filling a needed niche in banking by lending to the startups that the VCs invested in (note that by 2015, the vast majority of its portfolio was considered non-early stage borrowers). SVB had a reputation for being entrepreneurial, a rare thing in banking, and developed an expertise in serving high-growth startups, in particular via offering connections to the VCs in Silicon Valley who banked with SVB or invested in SVB. I recall bumping into SVB here in KC on occasion, when talking to some of KC’s exciting tech companies. It was hard to compete with them.

At a glance, SVB’s financial profile doesn’t look weak. At 12/31/22, it posted a $1.5B profit, slightly below that of the prior year. On the balance sheet side, it showed a 15.26% tier 1 capital ratio, which measures a bank’s equity and reserves to risk-weighted assets. So higher is better. By contrast, Bank of America posted a 13% tier 1 capital ratio, Commerce 14.3%, and UMB 10.9%. Enterprise Bank stood at 12.6%.

Digging into SVB’s assets (h/t to Genevieve Roch-Decter, @GRDecter, for much of what follows), it had a low loan to asset ratio at 35%. By contrast, Enterprise has a 74.6% loan to asset ratio, Commerce Bank 50.7%, UMB 54.1% and JPM 30.4%. One might think having less loans equates to a safer bank, but in this instance it may not. Loans weren’t the problem--not enough loans (yielding higher rates) were the problem. Instead of lending, Silicon Valley Bank invested a lot of capital into securities yielding an average of 1.87%, mainly riskless US treasuries mortgage-backed securities (this was actually mandated post the great financial crisis to improve banks’ balance sheets). Except this wasn’t riskless…while the chance of a proper default on treasuries is nil, SVB incurred interest rate risk to a larger degree than most banks. And this is where the almost 500bps Fed rate hikes over the past year come into play. Almost 48% of SVB’s securities were invested in treasuries and mortgage-backed securities with maturities of 5 years or more, far in excess of other banks. Additionally, SVB failed to hedge the interest rate risk.

Source: S&P Global Market Intelligence

As a result of the rate hikes, the majority of the securities yielded less than the current market rate and thus lost value. Now, this isn’t necessarily a problem unless a bank needs to sell the securities and realize the loss…which is what happened to SVB.

Since Q2 of 2022, total deposits have been falling at most banks and SVB was no exception. Part of that is natural run off after deposits grew in 2020/2021 behind stimulus. But the rate hikes have also provided investors, companies and individuals, more attractive places to park cash—US treasuries and money market funds.

What precipitated last week’s failure was SVB’s decision to sell $21B in securities (10% of total assets) at a loss of $1.8B in order to improve liquidity and invest in shorter-dated securities yielding higher rates. This in turn created concern among depositors. And with good reason. More than 97% of deposits at SVB exceeded FDIC insurance (likely this is a function of large, concentrated VCs parking a lot of cash there). Contrast that with 50% of Commerce’s deposits being uninsured, 65% of Enterprise’s, 71% of BofA’s, or 84% of UMB’s. The higher the percentage of uninsured FDIC deposits suggests that there are fewer and larger clients that can choose to withdraw deposits to create a bank run.

The current risk is likely not systemic in the sense that this failure will create counterparty risk and lead to the failure of other banks. In general, banks lend less to other banks than before the great financial crisis, and credit risk across the economy appears solid. The risks here appear to be somewhat idiosyncratic to SVB, though interest rate risk and proper balance sheet management remains a challenge for every bank right now, and as stated above, technically any bank could experience a run if confidence is lost. My largest concern at this point is that an estimated half of all startups worked with SVB, maybe 65,000, and if they can’t get immediate access to funds, they can’t pay employees, can’t pay vendors, and that may create some ripples in the US economy.

How will this play out? There are cries for making depositors whole at SVB. The Fed will need to decide whether or not to help conduct a bailout and backstop of the financial system if more depositors get nervous. Unfortunately, there are other institutions with large unrealized losses due to the rate hikes of the past year. (Note: Sunday evening the Fed, Treasury and FDIC announced a Bank Term Funding Program that will allow banks to borrow against available for sale and held to maturity assets at par, effectively allowing banks to meet any demands for withdrawals without incurring a hit to the balance sheet or P&L. Depositors at SVB will be made whole and have immediate access to their funds. Crisis averted(?)).

Enterprise Bank remains strong, well capitalized, and diversified amongst its client base. Having loans as a larger percentage of assets appears to be a major advantage at present. Based on my research, Enterprise assets are earning an average yield of 4.22% at 12/31/22, vs 2.47% at Commerce, 3.35% at JPM and 2.73% at SVB, putting far less pressure on any securities investments. Additionally, Enterprise is showing a relatively small unrealized loss on invested securities as a percent of equity of 18.3%, compared to 60.4% at Commerce, 19.9% at JPM, 51.4% at UMB and 108% at SVB. This means that EBT has less in the way of unrealized losses on security investments, so the financial impact of forced selling would be less. Do keep in mind that there are other ways to make money, and fee income is important for any bank (JPM, UMB and Commerce do this particularly well). ?

Disclaimers: I am just a lowly commercial banker in the middle of America. This is not the opinion of my esteemed employer, and in no way constitutes investment advice. Please reach out to me with questions, concerns, corrections, deposits, etc….

Best,

Mark?