[Dong-hyun Ahn’s Economics] Bank runs can only be prevented if viral fears are blocked

[Financial News, May 9, 23] PERI Adjunct Research Fellow (Dong-hyun Ahn / Macroeconomics·Finance)

[Dong-hyun Ahn’s Economics] Bank runs can only be prevented if viral fears are blocked

 

Bank runs, causes, and solutions

Recently, Silicon Valley Bank (SVB) went bankrupt due to a bank run, and First Republic Bank went bankrupt and was acquired by JPMorgan. With the proliferation of smartphones, mass deposit withdrawals are occurring at a faster rate than ever before, leading to the coining of the term “digital run”.

 

The nature of bank failures is contagious

As shown in the figure, bank failures have been a constant occurrence. However, bank failures tend to be concentrated at certain points in time, as seen in the Great Depression, stagflation, and the 2008 financial crisis. This suggests that a major characteristic of bank failures is their contagious nature.

There are two main causes of bank failures. The first cause is associated with issues on the debit side of the balance sheet, which directly impacts the bank’s overall health. A collapse in the value of the assets held by the bank results in a loss of capital. This happened frequently in the UK and the US in the 1800s and early 1900s. Moreover, during the 2008 financial crisis, it was predominantly obsesrved in investment banks. Credit Suisse recently faced a similar fate due to this very season. The second cause relates to bank liquidity and involves the credit side of a bank’s balance sheet. This occurs when a bank fails because it lacks sufficient cash on hand to cover large deposit withdrawals. Bank runs have occurred worldwide, such as the notable bank run in England in 1772, which led to the Boston Tea Party and Northern Rock case in the UK, and Wachovia case in the US during the financial crisis.

 

‘Contagion of fear’ tricky to counter

Bank failures have different contagion pathways. Bank failures due to asset failures on the debit side can take two main paths. The first is when the failure of one bank is transmitted to other banks through the payments network, or when the failure is contagious through interbank lending relationships. The second is through a “fire sale,” where a failing bank sells off its holdings in response, causing other banks with similar assets to fail. The 2008 financial crisis was primarily contagious through this second route.

On the other hand, the debit side of bank runs is viral panic. When a bank run occurs at one bank, the panic is contagious and spreads to other banks. To summarize, debit failures are caused by bank greed and are related to fundamentals, while credit failures are caused by depositor panic and may not be related to fundamentals.

Debit problems are related to fundamentals and are therefore relatively easy to address. In fact, much of the core financial regulation of systemic risk management by financial regulators is focused on preventing them. This includes the Bank for International Settlements'(BIS) equity capital regulation and strong prudential regulations that control the risk of holdings.

Bank runs, on the other hand, are much harder to deal with because they have to allay depositor fears. In the 17th century, the United Kingdom introduced a reserve banking system, and in 1933, the United States introduced a depository protection system, but these systems were unable to prevent bank runs.

 

Bank Runs, the Combination of the Nature of Banking and Fear

So why do bank runs happen so often? It’s fundamentally because of the nature of banking. Typically, depositors want to lend their money for the short term, while businesses and households want to borrow money for the long term. The essence of banking is to transform short-term deposits into long-term loans, a process known as liquidity transformation. The funds acquired through this transformation ultimately become the source of corporate industrial finance and household real estate and credit loans, serving as the foundation for economic growth. However, when large-scale deposit withdrawals occur, it becomes impossible to respond with loan recoveries, leading to a bank run.

The problem with bank runs is that they are based on fear, which is unrelated to fundamentals, making them unpredictable. Let’s take a theater fire as an example. The bank failures related to credit that we discussed earlier are akin to a theater catching fire. In contrast, bank runs on the debit side are different. When one or two people start running, others see them and start running too. It doesn’t really matter why they are running; witnessing others running triggers fear, leading to a collective panic. This way, it goes from a bank run to a full-blown financial crisis. Bank runs go through the same process.

What makes bank runs terrifying is the self-fulfilling prophecy. In other words, the fear of a bank run itself brings about a real bank run. Theater fires or bank runs, both are typical examples of a coordination failure in microeconomics. If coordination happens, and everyone stops running simultaneously, they can reach a ‘good equilibrium’ where no harm occurs. However, if coordination doesn’t happen, they reach a ‘bad equilibrium.’ Since fear is the mediator, it is impossible to predict or prevent when a bank run will occur. This is precisely the core of the Nobel Prize-winning research by Douglas Diamond and Philip Dybvig from last year.

In conclusion, bank runs are a product of the essence of banking, which is liquidity transformation, and the psychological phenomenon of fear. Therefore, they cannot be resolved through financial regulations. Measures such as increasing deposit insurance limits, adjusting payment settlement collateral ratios, and even measures like withdrawal restrictions have been mentioned, but these micro-level approaches provide only limited assistance. Just having fire extinguishers in a theater or firewalls won’t entirely prevent people from running. Some have suggested ‘bank holiday,’ but that could not only paralyze the economy during that period but also potentially fuel more fear. The success of President Roosevelt’s bank holiday in 1933 seems to be based on the declaration of 100% deposit insurance, which instilled confidence in the safety of deposits.

 

The role of central banks in rapid evolution

Calling the SVB crisis a “digital run” makes it seem like it’s different from a traditional bank run, but it’s economically the same except for the speed of withdrawals. Once a bank run occurs, it is similar to a wildfire, so the first response is key, and the role of central banks is paramount. Historically, central banks were introduced with the expectation that they would be the “lender of last resort” to prevent a bank run. As such, they must act quickly to provide liquidity support to banks once the fire has been lit. Whether it’s providing liquidity by pledging bank assets or depositing cash to match withdrawn deposits, swift action is needed to contain the spread of a bank run to other banks.

The restructuring of the bank can proceed after the bank run is stopped. Once again, there is no vaccine against bank runs. The key is to stop the viral fear once it starts. Ultimately, it’s a fight against fear, and building trust is the only way to overcome it.

 

SVB falls victim to outrageous report

And there’s one takeaway from the SVB bank run that hasn’t gotten much attention. At the end of last year, SVB’s BIS capital ratio was 16.05%, higher than the 14.8% average for U.S. banks. So why was the bank victimized despite its strong balance sheet?

One reason is that SVB held assets that were easily marketable, such as government bonds. A large U.S. investment bank estimated in a research report that a rise in interest rates would have resulted in a $16 billion valuation loss on SVB’s government bonds. These concerns sparked panic. But most banks hold more loans than government bonds. Loans have the same economic substance as corporate bonds, so if the 5-year Treasury declines in value by 30% under strict mark-to-market, a 5-year corporate loan should decline by more than 30%. However, loans are not easily mark-to-market, so fair value doesn’t deviate much from book value.

Here’s the problem. By the same standard, other U.S. banks are riskier in terms of impairment losses. SVB, which holds a lot of safe assets, was negatively impacted by bank runs following an outrageous report. Here is the crux of the matter. It’s not just about debit fundamentals. News or rumors suggesting problems with the assets can lead to a collapese on the debit side, and that is the essence of a bank run. In the digital world, all sorts of rumors and fake news exist. The concerning part is this: If false information triggers concerns about a bank’s health and those concerns escalate into panic, no bank can avoid a digital run. That’s why it’s urgent to take a hard line against misinformation and build trust quickly and strongly.

 

May 9, 2023.

<Dong-hyun Ahn, Adjunct Research Fellow, PERI> Professor, Department of Economics, Seoul National University