Few economic phenomena frighten people more than a run on banks. Images of long queues of desperate, angry men and women stretching for several blocks are associated indelibly with economic catastrophes such as the Great Depression. When confidence in the banking system disappears, it seems, mass panic ensues.

The technical definition of a bank run is when sufficient numbers of depositors—especially those with very large deposits — conclude that a bank won’t be unable to repay their deposits on schedule and in full, and consequently choose more-or-less concurrently to pull out their deposits with little notice. Given that banks typically keep little by way of cash-on-hand, they find themselves having to respond by quickly selling off their holdings at below-market rates. Given enough pressure from depositors, a bank can find itself unable to meet its obligations and, like any other business, will go under.

It’s not unusual for banks to fail. Between 2001 and 2023, there were 562 bank failures in the United States. But a run on a bank that is prominent enough or big enough can cause people to wonder whether other banks are able to meet their obligations. That’s when things start going south, often very quickly.

In October 1907, two New York speculators experienced enormous losses while trying to secure as much stock as they could of a copper-mining company. As a result, the banks who had backed these speculators collapsed in the wake of runs by depositors. The crisis of confidence eventually spread to trusts which, at the time, played a major role in financial markets by virtue of their deep involvement in New York’s equity markets. Within two months, 17 trusts and 25 banks had failed, as large numbers of investors and ordinary depositors started wondering whether any banks at all were solvent.

By the time the Panic had subsided, US GDP had declined by a massive 30 per cent. There was no Federal Reserve or Federal Deposit Insurance Corporation (FDIC) to step in. Instead, it was the legendary financier John Pierpont Morgan who stopped the rot by cajoling solvent banks into saving the large number of banks on the brink of collapse.

The 1907 Panic is a good example of how it’s really the spillover effects of a bank run that we should be worried about. In the vast majority of cases, a run on a particular bank doesn’t metastasize on such a scale. Even before central banks get involved, other banks will often step in to provide the necessary solvency, whether by loans or purchase of bank assets at low prices. Few runs on a bank thus turn out to precipitate a widespread financial contagion.

The odds of a wider conflagration today are heightened, however, by factors that were not as prominent in the early twentieth century. Finance and banking have always been complicated, whether in terms of the development of new financial products and ways of leveraging risk. This can make it hard to know where the trigger points associated with a run on a particular bank or series of banks might lie.

But in our time, this is even further complicated by technology. Whether it is simple deposit accounts or the functioning of banks’ investment arms, more and more banking is becoming automated. Much modern banking is driven by constantly updated algorithms. In the context of a bank run, this means that regulators can find events running away from them and constantly playing catch-up. When people can withdraw their money by pushing an app on their phone, you are in a very different world from that in which people have to line up for hours to wait for a teller to hand over bundles of hundred dollar notes.

Yet there is one constant that transcends the historical span of runs on banks, regardless of whether it is the panic of 1907, the Showa Financial Crisis in 1927 that resulted in 37 bank collapses throughout the Japanese Empire, or the banking crises that brought Argentina to its economic knees in 2001. That factor is human psychology.

The word ‘credit’ is derived from the Latin credere, meaning ‘to believe’. For all its sophistication and technological prowess, all banks—large, medium, or small—ultimately rely upon that hard-to-quantify but nonetheless real force that we call confidence. As long as banks have the faith of their investors and customers, there is little likelihood that a critical mass of depositors will suddenly want their capital back.

That collapse of belief is ultimately what brought Silicon Valley Bank (SVB) to be undone. Like many banks, SVB borrowed money in the short term and invested in the long term. But, as the Federal Reserve jacked up interest rates quickly, SVB’s assets rapidly diminished in value. Management, however, failed to address the growing gap between assets and liabilities, thereby putting SVB in an untenable position.

As soon as word leaked out that not all was as it should be in SVB, big and small investors lost confidence. They acted accordingly. Within less than two days, SVB went under. What took weeks in 1907 was over in the space of 48 hours.

Sound management and rules are important for the workings of the financial sector, but more than anything else banks rely upon trust. Without it, the delicate tapestry of banking stands on nothing. Measuring trust may be difficult, but its absence can prove calamitous. Some things never change.