Silicon Valley Bank Failure of 2023

On 10 March 2023, Silicon Valley Bank (SVB) collapsed after a single day of deposit withdrawals, marking the second-largest bank failure in US history. The failure exposed a duration mismatch that had festered quietly since 2021: the bank had loaded its balance sheet with long-dated, low-yielding bonds exactly as interest rates began their sharpest rise in four decades, and when depositors—mostly uninsured tech entrepreneurs and venture capitalists—rushed for the exits, there were no liquid assets to meet them.

The Prelude: 2020–2021

SVB entered the pandemic-era monetary expansion with a simple and initially profitable strategy. The Federal Reserve drove interest rates to near zero, and trillions in stimulus flooded the economy. Tech startups, awash in venture capital, deposited hundreds of billions of dollars into their main operating account. SVB’s deposits climbed from roughly $40 billion in late 2019 to nearly $200 billion by the end of 2021.

With deposits surging and yields collapsing, SVB faced a familiar problem: where to invest excess liquidity? The bank loaded heavily into mortgage-backed securities and long-dated Treasury bonds—safe assets, but with a critical vulnerability. In a world of 0.15% Fed funds rates, a 30-year Treasury yielded around 1.6%. These securities were sound, but their value was acutely sensitive to interest-rate moves. A 200-basis-point rise in rates would vaporize roughly 20% of their market value. SVB’s risk officers knew this. Senior management apparently decided the risk was acceptable.

The Trigger: Rate Rises and Deposit Attrition

In March 2022, the Federal Reserve began raising rates aggressively, eventually climbing to 5.25–5.50% by late 2022. The effect was swift: SVB’s bond portfolio, marked at fair value, hemorrhaged. By the end of 2022, the bank had accumulated $15.7 billion in unrealized losses on its securities portfolio—roughly 20% of its capital. These losses were not hidden; they appeared in the bank’s 10-K filing in February 2023 and were known to sophisticated observers.

What killed SVB was not the losses themselves, but the deposit base. Venture capital had dried up as interest-rate hikes chilled startup funding. Startups that had burned through cash no longer needed massive operating balances. More importantly, bank regulators had not stressed-tested the “uninsured deposit problem”—a phrase that would dominate post-mortem analysis. SVB’s depositors were largely above the $250,000 FDIC insurance limit; a tech entrepreneur with $20 million in cash had no insured coverage, only faith in the bank’s solvency. When that faith cracked, withdrawals began.

On 8 March 2023, SVB announced it would raise capital by selling stock and dividing its securities portfolio. The announcement was meant to shore confidence. Instead, it triggered a classic bank run. Within 48 hours, depositors withdrew $42 billion. By 10 March, SVB was closed.

The Wider Failure

Silicon Valley Bank’s collapse was not a black swan. It was a transparency problem wrapped in regulatory complacency. The unrealized losses were public. The deposit concentration (90% uninsured) was disclosed. The interest-rate sensitivity was elementary bond mathematics. Yet the bank’s supervisors—the Federal Reserve and the California Department of Financial Protection and Innovation—had approved the board’s risk tolerances and did not mandate hedging or reduce the depositor concentration when the problem became obvious.

The failure exposed a fragility that went unnoticed because rates had fallen for 40 years. In a low-rate environment, long-dated bond purchases seem safe; duration risk feels theoretical. The moment rates rise, it becomes real. SVB had $91 billion in held-to-maturity securities that it could not sell without recording massive losses. Regulators had allowed banks to hide these fair-value losses from balance-sheet visibility under accounting rules designed for less volatile eras.

Contagion and Resolution

Fear spread instantly to other regional banks. Credit Suisse, already wounded, required a forced merger into UBS. First Republic Bank failed in May. The Federal Reserve activated the Federal Reserve Credit Facility to lend against securities, effectively backstopping all banks’ bond portfolios at par. The Treasury and FDIC announced an extraordinary resolution: all SVB depositors—even the uninsured—would be made whole through emergency measures.

The resolution, though orderly, broke the principle of FDIC insurance as a limited guarantee and effectively socialized the losses. Depositors who had taken on uncompensated risk were protected. Conservative banks and ordinary savers who had kept deposits under $250,000 received no special treatment. The moral hazard was immediate and visible.

Lasting Implications

SVB’s failure killed the mythology that US banking was adequately supervised. It revealed that the Fed, despite sophisticated models, had allowed a $200 billion institution to accumulate unhedged duration risk in a rising-rate environment. It showed that uninsured deposits—once thought to impose discipline on bank risk-taking—could flee in hours, requiring government bailouts.

The failure also reasserted that tech-sector concentration, while profitable for SVB, created fragility. Startups and venture firms are cyclical; their deposits rise in booms and evaporate in downturns. A bank betting its stability on a single volatile sector learned this the hard way.

By late 2023, regional bank stocks had recovered somewhat, policy had tightened slightly (higher capital-adequacy requirements for banks over $100 billion), and rates had stabilized. But the underlying lesson remained: duration risk, interest-rate risk, and liquidity risk never disappear. They simply wait for the next regime shift.

See also