Direct Answer
Bank runs have accelerated dramatically in the digital era: SVB lost approximately $42 billion in deposits in a single day in 2023, a pace that is structurally impossible in the era of physical branch withdrawals. Speed depends on funding structure: banks with large concentrations of uninsured deposits or wholesale funding face faster potential outflows than retail-funded banks with insured deposits. This page presents the methodology for comparing bank runs across eras and the qualitative evidence across major historical episodes.
Bank Runs in Financial History: Speed, Funding Structure, and Evidence
Comparing bank runs across historical eras requires explicitly accounting for the technology of the withdrawal mechanism. A depositor in 1930 could only withdraw by physically visiting a branch during business hours, with cash available limited by the vault's daily holdings. A depositor in 2023 could initiate a wire transfer at 2 a.m. from a mobile application. This structural difference makes speed comparisons across eras misleading without clear methodology. This page presents the framework and qualitative evidence.
Measurement Methodology
Bank run comparisons are subject to three methodological constraints that limit cross-era speed rankings.
- Deposit flow data availability. Historical bank run statistics are often reconstructed from regulatory filings, bank statements, and congressional testimony. Exact daily withdrawal figures for 1930s bank runs are frequently unavailable or estimated from month-end balance sheet data. Only events for which reliable deposit-flow records exist can be compared quantitatively.
- Mechanism of withdrawal. Physical branch runs constrained by operating hours, cash availability, and physical queuing cannot be directly compared to digital wire transfer runs. The appropriate comparison within each era is run speed relative to the technology available at the time, not absolute speed across eras.
- Deposit insurance threshold. The presence or absence of deposit insurance fundamentally changes the incentive to run. In the U.S. before 1933, all depositors faced potential loss regardless of account size. After FDIC insurance was established, insured depositors (now covered up to $250,000 per depositor per institution) have no financial incentive to run; only uninsured depositors retain the first-mover advantage. Comparing runs across the pre-FDIC and post-FDIC eras requires acknowledging this incentive structure difference.
- Solvent vs. insolvent bank runs. A run on a fundamentally solvent bank that becomes insolvent due to the run itself is a different phenomenon from a run on an insolvent bank whose customers correctly anticipate non-payment. The SVB case had elements of both: SVB was arguably solvent on a hold-to-maturity basis but illiquid on a mark-to-market basis when the run began.
Qualitative Evidence Table
The table below compares major bank run episodes across speed, funding structure, and deposit insurance context. All figures require verification against FDIC records, Federal Reserve historical data, and primary sources for each episode.
| Episode | Approximate speed | Funding structure | Insurance context | Resolution mechanism |
|---|---|---|---|---|
| Great Depression Bank Runs (1930-1933) | Days to weeks per institution; waves across system | Retail depositors; physical branch withdrawal | No federal deposit insurance until 1933 | Bank Holiday (1933); FDIC creation; thousands of bank failures |
| S&L Crisis (1980s) | Gradual; not a classic sudden run | Retail deposits; interest rate mismatch | FSLIC insured; some fund exhaustion | RTC resolution; FSLIC recapitalization; regulatory forbearance |
| Iceland (2008) | Rapid wholesale funding withdrawal | Wholesale international funding; offshore internet deposits | Limited; EU cross-border deposit insurance dispute | Capital controls; IMF program; sovereign default on guarantee |
| 2008 Interbank Run | Overnight repo and wholesale markets (hours) | Wholesale funding (repo, money market); not retail | Not a traditional deposit run; institutional counterparty withdrawal | Fed emergency facilities; TARP; Treasury guarantees |
| Cyprus (2013) | Gradual before capital controls; immediate on reopening | Retail and offshore deposits (Russian capital) | EU insured amounts protected; above-limit depositors bailed in | Capital controls; bail-in above insurance threshold |
| SVB (March 2023) | Approx. $42bn in one day (digital transfers) | Concentrated uninsured deposits (VC-backed companies) | Majority uninsured; high first-mover incentive | FDIC bridge bank; Treasury/Fed backstop for all depositors |
Digital Banking and Run Speed
The SVB collapse in March 2023 is the clearest illustration of how digital banking infrastructure has changed the speed dynamics of bank runs. Three structural factors combined to create the fastest large-bank deposit outflow on record.
First, SVB's depositor base was unusually concentrated among venture-backed startups, many of which maintained operating accounts well above the $250,000 FDIC insurance limit. Every such depositor had a financial incentive to withdraw because their funds were at risk if SVB failed. This created a homogeneous pool of depositors with identical incentives, unlike a typical retail bank where insured depositors (the majority by count) have no reason to run.
Second, social media and private group communications (WhatsApp, Slack, Signal) allowed prominent investors to coordinate withdrawal advice across the depositor base in hours. A depositor who had not heard news of SVB's balance sheet problems by noon received a direct message from their VC firm advising withdrawal by early afternoon. This is structurally different from 1930s bank runs where information spread by word of mouth or newspaper reporting over days.
Third, digital wire transfers execute 24 hours a day, including overnight, with no physical constraint. A depositor who decided to withdraw at midnight could initiate and complete the transfer before any human at the bank was aware of the acceleration in outflows.
These three factors suggest that future bank runs involving concentrated uninsured depositor bases will be faster than anything in the historical record, and that regulatory frameworks calibrated to historical run speeds may be insufficient for the digital-era speed of outflow.
Frequently Asked Questions
What triggers a bank run?
A bank run is triggered by a loss of confidence in a bank's ability to return deposits on demand. Depositors, individually rational, withdraw funds before others can, potentially creating or accelerating the insolvency they feared. Classic triggers include a rumor of insolvency, a visible queue of withdrawers, or news of a related bank failure. Modern triggers include social media posts, news articles about a bank's concentrated exposures, or a prominent investor advising portfolio companies to withdraw. The trigger need not reflect a genuine insolvency threat; a solvent bank can be made insolvent by a sufficiently rapid run if its liquid assets cannot cover the withdrawal demand before it can liquidate illiquid assets.
How fast can a modern bank run unfold?
The SVB collapse in March 2023 is widely cited as the fastest large bank run in history by the volume of deposits withdrawn in a single day. SVB lost approximately $42 billion in deposits in a single day, driven by digital transfers and social media amplification. This is structurally faster than historical bank runs where depositors had to physically queue and withdrawal was limited by working hours and cash availability at the branch. The SVB run accelerated overnight via digital channels after a fundraising announcement triggered coordinated withdrawal advice across a concentrated depositor base of venture-backed companies. Exact figures for the rate of outflow require verification against FDIC filings.
What funding structures make banks most vulnerable to runs?
Banks with large concentrations of uninsured deposits are most vulnerable to runs because uninsured depositors have the strongest incentive to move first in a crisis. When deposits exceed the FDIC insurance limit, the depositor bears loss risk if the bank fails, creating a strong first-mover advantage in withdrawing. SVB had an extremely high share of uninsured deposits because its client base consisted largely of venture-backed companies with operating accounts well above the insurance limit. Wholesale-funded banks reliant on money market funds, repo, or interbank lending face a similar dynamic with even faster potential outflow speeds because institutional counterparties can withdraw overnight. Retail-funded banks with insured deposits are less vulnerable because insured depositors have no financial incentive to run.