Generative Data Intelligence

In the Blink of an Eye: How the Digital Age Intensifies the Risk of Bank Runs


Recently several medium-sized banks in the US (i.e. First Republic Bank, Silicon Valley Bank and Signature Bank) came into trouble, after their customers lost trust (i.e. the trust of being able to retrieve their money at any moment) in these banks. This resulted in an unstoppable and lightning-fast run on the bank.

Analysts call the collapse of Silicon Valley Bank (SVB) the first digital bankrun or also the first Twitter-fueled bank run. The incredible speed of this bankrun was in any case never witnessed before in the history of banking. In the largest bank failure in US history, i.e. the one of Washington Mutual Bank in 2008, customers took 10 days to withdraw $16.7 billion, while in the case of SVB $42 billion were withdrawn in a single day and another $100 billion were queued up for the next day.

These staggering figures have shaken up the entire worldwide financial system. In today’s interconnected world, where information travels at lightning speed and financial transactions can be executed in seconds, the risk of a “run on the bank” has intensified. Historically, it would take months for such runs to gain momentum, but the digital age has altered the dynamics significantly. Clearly no bank can withstand a bank run at such an unprecedented speed. Therefore every player in the financial industry (regulators, Fintechs, incumbent financial players…​) is now analysing what can be done to avoid this in the future.

But before looking at potential solutions, it is important to understand why those collapsed banks, which were considered (relatively) healthy before the rumours started, were more vulnerable to such a digital bank run than other banking players in the market, i.e.

  • All those banks offered their services to a specific niche customer segment, i.e. SVB to tech start-ups, Signature Bank to crypto-investors and First Republic Bank to a new generation of high-net worth individuals. Such a specific customer focus allows to grow quickly and offer very tailor-made services, but it also means that there is a lack of differentiation and that rumours spread much faster as all clients are frequenting the same (social) networks (i.e. they are strongly interconnected) and have a similar reaction to financial news.

  • The customers they were serving were forced in recent months – due to external factors – to withdraw more money than than they deposited , i.e.

    • SVB: tech start-ups are facing less and smaller VC funding rounds, meaning these companies are forced to gradually consume the deposits they raised in past funding rounds to pay for their ongoing businesses.

    • First Republic Bank: with nearly two-third of deposits not insured by the FDIC (i.e. the federal guarantee for up to $250,000 per person and per bank), these wealthy customers started fearing for their money after the collapse of SVB and Signature Bank. Therefore a lot of those customers started withdrawing all money above $250,000 and spreading it over multiple banks, which was particularly painful for First Republic Bank.
      The same phenomenon accelerated also the fall of SVB and Signature Bank, which respectively had 94% and 90% of deposits uninsured. When comparing this to the large US banks, which only have 47% uninsured deposits, you can see already a dangerous pattern.

    • Signature Bank: the customers of this bank being very active in the crypto-world faced obviously serious issues after the price collapse of crypto-currencies and especially with the bankruptcy of several large crypto-players (like FTX, Three Arrows Capital, Genesis, BlockFi, Celsius, Voyager Digital…​).

  • These banks were all well digitized and all had digital-savvy customers. This meant that withdrawals happened almost all digital, which obviously accelerated enormously the process.

  • The capital rules of those banks were less strict than larger banks in the US and banks in general in Europe. This was due to an exception made by the US government, allowing banks with balance sheets below $250 billion not having to follow all the Basel III rules. Additionally these recent events also learned that the Basel III rules are probably no longer fitting with this fast digital world and require some revision to better factor these types of exceptional, short-term liquidity needs into account.

  • All those banks had a strong asset-liability mismatch, which was due to the type of business those banks were doing, but also a direct consequence of the exceptional monetary policy of central banks in recent years, i.e. after several years of interest rates close to zero (or even negative interest rates), the central bank interest rates have been considerably increased in recent months. This situation made those (but also many other) banks very vulnerable, as they had quite some long-term assets on their balance sheets. Those long-term assets (like loans and government bonds) were acquired in times of low interest rates, which means that a forced liquidation on the secondary market (to free up liquidity for paying back deposit withdrawals) resulted in serious losses (prices of those asset dropped due to the increased interest rates). A better hedging against the interest rate risk seems to have been neglected, but obviously this is easily said in retrospect.

The above elements show some fundamental concerns with our current banking system:

  • Due to the enormous digitalization in the financial industry in recent years (strongly accelerated due to Covid), the speed of deposit withdrawals has now such a speed, that it is nearly impossible to take corrective actions (like raising additional funds, changing certain business rules, obtaining emergency funding from central bank…​) once the first signs of a bank run start to unfold. Before this digital age, banks had some time buffer to properly address liquidity challenges.

  • Rumours spreading like wildfire: social media platforms can spread rumours, often unfounded or fake, like wildfire, triggering panic and undermining the stability of even the most reputable financial institutions. Additionally once such rumours gain traction, they become nearly impossible to halt, as any communication by the bank rarely has a calming effect in these situations.

  • No bank is capable to stop such a Run (even the most profitable ones), meaning it can potentially hit everyone. This means these scenarios become critical operational risks, as these kind of events are an immediate threat to the bank’s survival. When occurring, all shareholder value can be completely destroyed in a matter of a few days (which is the case for these 3 banks).

  • Trust is lost in a day, but takes very long to (re)gain. This means that even a small rumour can have enormous consequences, as even in the positive case that a run on the bank can be avoided, it will require enormous investments to regain trust from customers.

  • The asset-liability mismatch (i.e. banks primarily invest long-term in illiquid assets but face the need to satisfy immediate demands for cash withdrawals), which is at the foundation of banking, can be very problematic in this digital age and in this economic situation, with fast changing central bank interest rates.

As a result, changes to the banking system are required to protect banks against this new digital evolution.
Some examples could be:

  • Banks can consider implementing measures to limit the amount of funds that can be held in certain types of accounts (like current and saving accounts). For example, they can impose maximum limits on deposit amounts or apply lower or even negative interest rates for large deposits. This encourages customers to spread their funds across multiple banks or invest in longer-term deposits or in off-balance securities, reducing the risk of sudden withdrawals.

  • Several Fintechs are already exploring (or even offering) solutions to easily spread money over different banks (when the deposited amounts surpass the government bank guarantees). Via some kind of aggregating BFM/PFM app, which has access to all your bank accounts, accounts with too much deposits can be easily identified and the excess money can be automatically moved to banks where the guarantee limit is not yet reached. Additionally the app could allow easy opening of additional accounts at other banks, when the limit is reached on all accounts at current banks. Especially for businesses, which typically have large cash pools, such tooling can become a necessity.

  • Governments or collective actions among banks can explore options to increase the level of deposit insurance or guarantees. Such actions can help increasing confidence in the banking system and reduce the likelihood of runs on the bank.

  • Effective communication and transparency: fast, transparent and honest communication with depositors, shareholders, and regulators becomes essential. Banks should proactively disseminate accurate and timely information to counter false rumors and allay concerns. Additionally they should put in place systems to continuously monitor social media (several platforms exist on the market for this, like Hootsuite, Mentionlytics, Reddit, Brandwatch, HubSpot, Keyhole…​), allowing to pick-up on rumours or other potential harmful communications instantly, so that immediate action can be taken.

  • Enhancing liquidity management: banks should develop robust liquidity risk management frameworks to ensure sufficient reserves are available to meet depositor demands during periods of stress. Maintaining a diverse funding base and regularly stress-testing liquidity positions can help identify vulnerabilities and facilitate proactive measures.

  • Changes in Basel III calculations: currently the Basel risk-weighted capital requirements are mainly based on the credit quality of a bank’s asset portfolio. In the case of SVB, this was positive, as it had a lot of long-term government-backed securities with very low credit risk in their portfolio. Specialists are therefore suggesting to take interest rate risk more into account in the risk-weighting formulas.
    Additionally experts suggest to review the rules for calculating LCR (Liquidity Coverage Ratio, i.e. the quality and liquidity of assets) and NSRF (Net Stable Funding Ratio, i.e. the quality and stability of liabilities, meaning the funding sources). Clearly the recent events have showed that the definitions of a stress-situation might have to be reconsidered and also the definition of quality needs to look more at the interest rate risk.

  • Asset-liability mismatch: as an asset-liability mismatch is at the heart of banking, the delicate mechanics of this mismatch should be carefully monitored. This means banks have a strong responsibility to better monitor this and hedge themselves against certain risks (like interest rate risk), but at the same time there is also an important responsibility for Central Banks and politicians in this story.

    • Central Banks might want to be more careful in using the central bank interest rate as a tool to intervene in the economy. Although the central bank interest rate is considered by economists worldwide as a powerful tool to reduce economic fluctuations, we see now that a too dynamic policy of the central banks (via changing interest rates and via quantitative easing) also has quite some adverse effects on the long-term.

    • Politicians need to be careful with certain regulation in the financial sector aimed to help consumers, but in the long-term potentially increasing the risk in the financial situation. E.g. in Belgium there is a strong push from the government upon banks to rapidly increase the interest rates on saving accounts. Currently the 4 major banks in Belgium provide an interest rate on saving accounts of around 0.5%, while if they park the money of those saving accounts at the ECB they get themselves an interest of 3.25%. This difference has never been so big, hence the public push on all banks to rapidly increase this interest rate.
      Despite the fact that banks in Belgium, due to a lack of competition and a lack of mobility of savers, the delicate balance of the the asset-liability mismatch should not be overlooked.
      A lot of assets of the bank (i.e. credits and investments in instruments like government bonds) were originated/purchased a few years ago, when interest rates for credits were still at 1% and less and a lot of those are quite long-term. This means the Belgian banks are still carrying these long-term, low-interest assets for quite some time. When increasing the interest rates on saving accounts too quickly, this can create a bigger mismatch.
      This shows that the short-term demands of customers rarely align with the long-term stability concerns of a bank.

Clearly these are interesting times, with dynamics changing faster than ever before. This means that banking management and regulation need to rapidly evolve as well. Unfortunately the complexity of the financial system makes it very difficult to understand all dynamics in detail, meaning that we probably need events like the bankruptcy of SVB, First Republic Bank and Signature Bank, to better understand and adjust certain rules and guidelines to our modern digital times.

Check out all my blogs on


Latest Intelligence