Who Was Holding the Bag in 2023 — and Why It Matters Now
The participant-composition story of the Silicon Valley Bank collapse is, at its core, a story about who buys into a business model during a liquidity boom without asking what happens when the boom ends. SVB's deposit base was not a cross-section of the American economy. It was a highly concentrated cohort: venture capital firms, their portfolio companies, and the founders those firms backed. When the Federal Reserve began raising rates aggressively in 2022, that cohort faced a simultaneous funding crunch — startup burn rates were rising, new VC rounds were drying up, and companies were drawing down deposits rather than replenishing them.
This is the participant-composition paradox that distinguishes the 2023 crisis from prior banking failures. In the savings-and-loan collapse of the 1980s, the buyers at the peak were largely retail depositors and unsophisticated thrift managers chasing yield. In 2023, the buyers — and the runners — were the smartest money in the room. The sophistication of the depositor base did not reduce systemic risk; it amplified it, because sophisticated actors can coordinate and exit faster than any regulatory early-warning system can respond.
Duration Mismatch: The Structural Fault That Rate Cycles Always Find
SVB's management made a decision that, in retrospect, looks reckless but was, in the context of 2021, almost conventional. Faced with a significant tripling of deposits in the years prior to 2022, the bank deployed a large portion into long-dated held-to-maturity securities, primarily agency mortgage-backed securities and US Treasuries. When rates rose, the market value of that portfolio collapsed. The unrealised loss on the HTM book reached a substantial level — a figure that, when disclosed in March 2023 alongside an equity raise, triggered the run.
The mechanism is not novel. Duration mismatch — borrowing short and lending or investing long — is the original sin of banking crises. What changes across cycles is the specific asset class that concentrates the mismatch and the catalyst that makes it visible. In the savings-and-loan crisis, it was junk bonds and commercial real estate loans. In 2008, it was mortgage-backed securities. In 2023, it was the safest assets imaginable — US government paper — made dangerous not by credit risk but by interest rate risk at scale and speed.
The current concern, as the 10-year yield revisits 5%, is that the next iteration of this mechanism is already loaded. Regional banks across the United States hold significant exposure to commercial real estate loans originated during the low-rate era. Estimates of CRE debt maturing through 2027 run to trillions of dollars, much of it concentrated in office, retail, and multifamily segments that have experienced material valuation declines. Unlike the SVB HTM portfolio, CRE exposure is not marked-to-market on most regional bank balance sheets, which means the losses are present but not yet visible — precisely the condition that existed in SVB's books in 2022. For broader macro context on how rate cycles create banking stress, the pattern is consistent across decades: extended accommodation followed by rapid tightening produces hidden duration and credit risk that surfaces only under pressure.
The Regulatory Moral Hazard Left Behind
The Federal Reserve's emergency response in March 2023 — the Bank Term Funding Program, a large emergency facility established within days — effectively backstopped the regional banking sector by allowing banks to borrow against HTM securities at par value rather than market value. This was a rational crisis response, but it created a precedent with lasting consequences. Uninsured depositors at SVB and Signature Bank were made whole through a systemic risk exception, a decision that implicitly extended deposit insurance beyond the statutory $250,000 limit for institutions deemed systemically important by regulatory discretion.
The moral hazard this creates is not abstract. If sophisticated depositors believe that regulators will protect uninsured balances at institutions deemed too interconnected to fail, they have reduced incentive to perform due diligence on the funding stability and asset quality of the banks where they park operating capital. The very behaviour that made the SVB run so sudden — concentrated, sophisticated depositors moving in coordination — is the behaviour that the BTFP precedent may inadvertently encourage in the next cycle, because those depositors now have reason to believe they will be protected even if they do not monitor their bank's balance sheet.
Social Media and the Velocity Problem
Perhaps the most consequential structural change the 2023 crisis revealed is the permanent alteration of bank run dynamics by digital communication. The 24-hour withdrawal of $42 billion from SVB was not the product of a slow-moving information cascade through newspaper headlines and branch queues. It was the product of real-time coordination among a networked community of founders and investors who received the same signal — SVB's equity raise and HTM loss disclosure — simultaneously and acted on it collectively within hours.
Traditional bank examination frameworks assume that deterioration in a bank's position becomes visible gradually, allowing supervisors time to intervene. The SVB episode demonstrated that for institutions with concentrated, digitally connected depositor bases, the interval between a triggering event and a fatal run can be measured in hours rather than weeks. No examination schedule or early-warning indicator can close that gap. The implication for current conditions is direct: any regional bank with concentrated deposit exposure — whether to a specific industry, geography, or depositor type — carries a structural vulnerability that standard regulatory metrics do not adequately capture.
Investors assessing current regional bank exposure should consult the Bubblepedia Bubble Watch indicators for ongoing monitoring of banking sector stress signals, and review the full bubble encyclopedia for historical context on how prior banking crises resolved. The comparison between the Japanese property bubble and the US housing crisis of 2008 is also instructive for understanding how CRE exposure can remain latent on bank balance sheets for years before forcing recognition.
Why the 2023 Crisis Remains Unfinished Business
The narrative that emerged from the spring of 2023 — that the BTFP and regulatory intervention had contained the crisis — was accurate in the narrow sense that no additional major bank failed after First Republic in May. But containment is not resolution. The substantial volume of maturing CRE debt, the moral hazard embedded in the BTFP precedent, the unresolved question of deposit insurance reform, and the structural vulnerability of concentrated depositor bases to social-media-accelerated runs all remain present in the system. The return of 5% Treasury yields does not create these vulnerabilities; it illuminates them.
Read the full Silicon Valley Bank and the Regional Banking Crisis of 2023 entry in the Bubblepedia Encyclopedia for a comprehensive account of the crisis mechanics, regulatory failures, and historical precedents that shaped the most significant banking episode since 2008.