What 2023 Added: Speed, Concentration, and Insurance

The 2023 failures of Silicon Valley Bank, Signature Bank, and First Republic Bank provide a modern stress test of the pass-through model. The FDIC’s account-level study asks who held the deposits, how quickly they moved, how insurance affected behavior, and what continued after regulators intervened. The cases make the historical lesson more specific: technology changes the time available for response, but depositor concentration, bank condition, and confidence still determine whether a run becomes a failure.

Three banks, three deposit bases

The three institutions did not enter the spring of 2023 with identical funding structures. Business deposits were the largest category at all three, but their proportions and surrounding deposit types differed in ways that reflected the banks’ business models.

At Silicon Valley Bank, business deposits represented 94 to 95 percent of total deposits in the months before the run. That is an unusually concentrated base in one broad category, and it meant that the bank’s liquidity depended heavily on business customers’ decisions. At Signature Bank, business deposits were roughly half of deposits. Passive escrow deposits—such as mortgage-servicing escrows, apartment security deposits, and lawyers’ trust accounts—were also significant, as were active escrow deposits holding customer funds associated with investment companies and financial-technology businesses. At First Republic, business deposits were again about half of deposits, while consumer and trust deposits were each sizable and active escrow deposits were meaningful.

These categories matter because the label deposit does not tell us who can control the funds, how directly they can withdraw them, or how insurance applies. A business operating account, a pooled escrow account, and a fully insured retail account may all appear as bank liabilities, but the people making withdrawal decisions have different incentives and different access to information. The FDIC study created categories for analysis from account ownership, beneficial ownership, account titles, and internal records. They are not identical to every regulatory category, so the findings should be read as a detailed case description rather than a universal classification system.

The contrast with the historical evidence is useful. The New York Fed study could compare thousands of bank episodes and link them to balance sheets, but it generally could not observe the identity and account structure of every depositor. The FDIC study covers only three failed banks, but it can examine the accounts and transactions that produced the outflow. One source supplies breadth; the other supplies depth.

Why a small group can control the run

The FDIC study defines top depositors as the top 0.5 percent of depositors ranked by total deposits. That group was small in number but large in funding share. On March 6, 2023, fewer than 500 top depositors at SVB held about 39 percent of domestic deposits. At Signature, fewer than 600 held 62 percent. At First Republic, roughly 3,000 held 50 percent.

Concentration matters because run risk is about dollars and timing, not only the number of people who withdraw. A bank may have hundreds of thousands of accounts, but if a few hundred or a few thousand depositors control a large share of balances, their decisions can exhaust liquidity quickly.