top of page

The Duration Chickens Come Home to Roost

When the “Si” bank (Silicon Valley, Signature, Silvergate) failures made headlines about 5 weeks ago, there was a palpable sense of anxiety about the entire U.S. financial system. Was a 2008 redux possible?

Our initial review suggested the problems were likely to be limited to growth-oriented lenders who had heavily banked the tech/VC/crypto ecosystems and startups with uninsured deposits at such banks. In other words, banks got a big influx of deposits in 2020-2021 (especially those that banked the VC ecosystem) and used that influx to buy long maturity bonds at very low yields. As the VC fundraising environment stalled, startups continued to draw down cash, and interest rates increased, everything flipped. Deposit outflows meant that banks held to sell bonds at a loss (yields up / bond prices down), which put pressure on capital ratios, triggering share price declines, which triggered depositor anxiety = spiral.

But these problems are not confined to tech/VC/crypto banks. They’re not even confined to banks. Many investors are struggling with the sharp increase in interest rates over the last 12 months. This is what’s known as a duration problem – that long dated financial assets tend to fall in price when interest rates increase.

Banks must be kicking themselves. In today’s “abundant reserves” monetary policy world(1), banks are paid interest for the excess reserves they hold with the Federal Reserve (this was not the case until after the financial crisis). In 2020-21, a bank with a deposit influx could have kept those monies at the Federal Reserve and received 0.25% interest (probably while paying zero or close to zero – positive spread but not enough to contribute much to profit). Under those conditions, I suppose 10-year Treasuries yielding 1% and 30-year Treasuries yielding 1.6% looked attractive. Today, a bank can keep deposits on reserve with the Federal Reserve and get 5%. No duration risk. Offer depositors 3.5 – 4% (reasonably competitive) and pocket the spread.

Part of the appeal for Treasuries (and agency-sponsored mortgage bonds) was that the securities were deemed “riskless” by regulators under risk-based capital requirement rules. No credit risk for Treasury bonds? Seems reasonable. But apparently everyone (including regulators) forgot about interest rate risk. In theory, any bond investor can purchase a low yielding bond and hold it until maturity (and receive par value), even as interest rates rise and the market price of the bond declines in the near-term. In practice, banks may not have that luxury when depositors show up and demand their money. And that is a BIG duration problem.

While the banking system has issues – there are reportedly ~$620 billion of mark-to-market losses across the banking sector not currently reflected on bank financial statements – it’s a much bigger problem for much smaller banks. Ultimately, banking is a scale game. To compete as a smaller bank one needs to find a niche. For a long time Silicon Valley Bank had a wonderful niche. And then one day a deposit drawdown at just the wrong time killed the bank.

Investment manager Bill Nygren recently explained the small bank / large bank dynamic:

“Forty years ago, there were 14,469 banks in the U.S. The total number of banks is now only about 4,200. That decline occurred despite public policy that has favored smaller banks…Economies of scale mean the natural path for banks is to continue consolidating. Government policy may prevent that in good times, but the economic rationale is just too strong to stop the growth of big banks…Competitively disadvantaged small banks struggle to match the breadth and depth of products and technology that large banks offer, so they can earn competitive returns for shareholders only by taking advantage of lower liquidity and capital requirements. The result is that smaller banks often have larger concentrations of risk with less capital and liquidity to protect depositors. When those risks go bad, as we saw last month, depositors move to safer big banks.”(2)

There will probably continue to be bank deposit outflows as investors find ways to generate more yield on their cash. We will probably see a contraction in aggregate bank lending, potentially contributing to recession. But it’s the larger banks – diversified, lower cost base, systemically important larger banks – that are better placed to deal with the duration mismatch wrought by the historically fast change in Fed rate policy.(3)


1. Brian Westbury, Chief Economist, First Trust (2022). Will Higher Interest Rates Tame Inflation?


bottom of page