When information makes banks less sustainable: What Silicon Valley Bank reveals about bank fragility
Non-technical summary
By Eva Schliephake, Research Associate at CSF
July 2026
Bank runs are often described as sudden episodes of collective panic. Nevertheless, many bank runs develop gradually as depositors learn new information - from financial news, social media, and observed behavior of fellow depositors. Some depositors are better informed, better connected or more exposed. They move first. Others observe their actions and infer that the early movers may know something. Our central finding is deliberately counterintuitive. More informed depositors can make a bank less stable because they change their behavior changes incentives of everyone else.
Take the run on Silicon Valley Bank is an example. The run was led largely by venture capital firms and technology companies with large uninsured deposits. These were not passive retail depositors waiting for a newspaper headline. They had stronger incentives to monitor the bank, better access to information networks, and a greater reason to react quickly to early signs of distress. SVB held long term fixed rate securities that might have paid out if held to maturity, but whose market value had fallen after interest rates rose. Once withdrawals forced the sale of the assets, losses became real pushing SVB into default despite being overall safe investments.
This is the mechanism at the heart of our research insight. Informed depositors can identify weak bank assets and withdraw early. Less informed depositors face a difficult choice. They can withdraw now based on their incomplete information, or they can wait and learn from the behavior of other better-informed investors. The ability to wait and learn sounds stabilizing, but in fact it can have the opposite effect. As bad news become only visible only after the better-informed depositors have already withdrawn, waiting for information may mean being too late. The prospect of learning tomorrow bad news can therefore trigger withdrawals today.
The logic fits SVB closely. The problem was not merely that some depositors had information. It was that they publicly discussed their withdrawal decisions. SVB depositors observing other investors moving their funds had to had to ask themselves this only a false alarm, or am I about to be last in line? The answer need not be certain. In a first come first served setting, uncertainty itself is enough. We call this mechanism the fear of missing out. The larger the share of informed depositors, the more exposed the uninformed group feels. If informed creditors discover that the bank is weak, they will leave quickly and reduce what remains for those who wait. Anticipating this, less informed depositors may run preemptively, even when the bank could otherwise survive. This means that a depositor base made up of sophisticated, uninsured investors can raise liquidity risk, not merely reflect it.
This contradicts the common view is that better informed depositors improve discipline. They monitor banks, punish weak balance sheets and stay invested when fundamentals are sound. That is partly true. But there is also an important disadvantage. When only some depositors are informed, the uninformed know they may be disadvantaged if bad news emerges. And the more informed depositors there are, the greater the fear that informed creditors will extract value first. This fear of missing out makes preemptive withdrawal more attractive.
The policy lesson is not that supervisors should hide information. The lesson is that information design matters. Disclosures, stress tests and deposit insurance rules shape not only what depositors know, but when they know it and how they interpret others. Bank fragility is not only about capital, liquidity or asset quality. It is also about the structure of the depositor base which dictates how dispersed concern converts into a coordinated bank run.
Cover image: Sean Pollock/ Unsplash
