Anatomy of a banking crisis – Lessons for supervision

Anatomy of a banking crisis – Lessons for supervision


Last month our family welcomed our healthy (second) baby. As my time is now even scarcer, and sleep deprivation is real, I had to look for an efficient way to meet my Finnius blogging obligation. Luckily, last month I also welcomed my column for the Dutch law review Ondernemingsrecht (here), on the recent banking crisis and the lessons for supervision. Unfortunately, in Dutch only. These lessons are too important and the book tip is too good, not to share it in English also. So, the solution to my problems is easy: here goes an English and slightly revised version of my column.

Another banking crisis

In recent months, the financial world held its breath. Giant banks went belly up in America and Switzerland. Would the ‘Great Financial Crisis’ (GFC) of 2008-09 repeat itself? For now, however, that great financial crisis has not occurred. Time for reflection. What went wrong, and why did US and Swiss banks fall but did banks in the European Union (EU) remain standing? What lessons for future EU banking supervision can we draw from this?

What happened?

First, a chronology of the crisis:


  • March 8-10, 2023: US-based Silicon Valley Bank (SVB), publishes that it is raising $2.25 billion due to losses on its bond portfolio. SVB’s stock price falls 60% and on March 9, account holders pull a staggering $ 42 billion (!) from the bank. The next day, the supervisory authority steps in and SVB is wound down by the Federal Deposit Insurance Corp (FDIC). SVB was the 16th largest commercial bank in the US with mostly tech start-ups/scale-ups as customers. Balance sheet total: $212 billion.
  • March 11-12, 2023: Just two days later, the authorities close Signature Bank, a niche bank for crypto companies and law firms. Balance sheet total: $110 billion.
  • March 13-19, 2023: Switzerland-based Credit Suisse (CS) gets a CHF 50 billion liquidity facility from its central bank. That does not reassure depositors and markets. On Sunday, March 19, UBS, that other Swiss banking behemoth, acquires CS. UBS is paying CHF 3 billion for the CS shares. The subordinated convertible bonds CS had issued (Additional Tier 1 capital (AT1)) are fully written down to CHF 0 by the Swiss authorities. More on this below. CS was one of the 30 largest banks in the world. Balance sheet total: $574 billion.
  • March 20-24, 2023: After the demise of CS, the share prices of all banks, including in the EU, plunge. On Friday, March 24, Deutsche Bank’s share price falls by 8.5%.
  • March 25-April 24, 2023: April is relatively quiet, though. Then, on April 24, in the US, First Republic Bank (FRB) publishes its quarterly report. It turns out that its customers withdrew a staggering $72 billion in Q1 2023.
  • April 25-May 1, 2023: FRB’s share price crashes, and more depositors run away. After a stressful weekend, it is announced on Monday, May 1, that FRB will be resolved by the FDIC. The FDIC sells large portions to JP Morgan Chase. FRB was America’s 14th largest bank. Balance sheet total: $229 billion.
  • May 2023-now: After these rescue weekends and soothing words from supervisory authorities, regulators, governments and various major banks, calmer times seem to have returned to the banking world (for now).


EU banks were spared. And despite the unprecedented balance sheet size of the failing banks, the crisis did not spread from Wall Street to Main Street.

Common causes

It is an oversimplification to lump together the causes of failure. Still, looking at it from a distance, some similarities can be seen.

First, the underlying concern about the failing banks was their (potential) losses. At SVB, these were losses on a sold bond portfolio for which the interest rate risk had not been hedged. That interest rate had been rising – to curb high inflation – at an unprecedented rate. CS suffered continuing losses as a result of integrity scandals and client departures. These losses were not sufficiently offset by equity buffers to reassure depositors and investors.

A second similarity: poor governance and risk management. As the Fed pointed out in its candid SVB analysis (here), “[SVB] failed because of a textbook case of mismanagement by the bank. Its senior leadership failed to manage basic interest rate and liquidity risk. Its board of directors failed to oversee senior leadership and hold them accountable.” The same flaw applied to Signature and CS. Poor governance leads to poor decisions, which can lead to significant losses and lack of trust. (Recently, the ECB drew a similar conclusion on the importance of effective governance, here)

Third, the bank run was a bank sprint. The trigger for the fall of SVB, Signature, CS and FRB was an ‘old-fashioned’ bank run; a liquidity problem. When depositors lose trust in their bank, they immediately withdraw their money; before the others do so. This is because a bank’s inherent problem is that there are never enough liquid assets to satisfy all payable, on-demand, deposits. Those short-term funds are lent long term by the bank. The recent bank run, however, was not so old-fashioned. Owing to a combination of (a) social media with a highly connected and concentrated customer base in which rumors and mistrust circulate extremely quickly, and (b) payment technology that allows money to be wiped out at the push of a button (or cellphone screen), a bank can be completely emptied in a matter of days. The crisis specifically involved banks with relatively large numbers of deposits above the guaranteed amount (above $250,000 in the US). These depositors withdraw their money even sooner; they have only an unsecured claim (and much hassle) in their bank’s insolvency.

Lessons learned

These causes of the crisis offer some good lessons for banking supervision. For example, Dutch Central Bank (DNB) President Klaas Knot said in the Dutch House of Representatives on June 7 (here), after noting that EU banks were spared due to the EU not being complacent when it comes to strict regulations and supervision, “[b]ut there is room for improvement, particularly in 1) the application of regulations in general; 2) interest rate risk; 3) liquidity risk; and 4) the orderly resolution of troubled banks.

Given Knot’s position in international supervisory policy forums (the Financial Stability Board, the Basel Committee, the ECB), these words have some prophetic value for supervisory regulation. Also, looking at his four lessons, important differences between the U.S. (and Switzerland) and the EU become immediately apparent.

  1. Application of supervisory regulations.

In principle, supervisory rules exist for exactly the common causes that I mentioned earlier: required own funds buffers, governance requirements, liquidity requirements, interest rate risk requirements, stress tests, etc. In response to the GFC, technical capital rules have been established internationally; in the Basel Accords. The EU implemented these rules – for the most part – one on one. All these requirements typically apply in full to all banks: large and small. Within the European Banking Union, compliance with these rules is closely monitored. For ‘significant’ banks (balance sheet total exceeding EUR 30 billion), the ECB conducts the direct supervision itself.

If not in the US. Through the Economic Growth, Regulatory Relief, and Consumer Protection Act of 2019, Basel standards were largely rolled back. Only the eight largest US banks are still fully supervised. Banks with balance sheet totals below USD 250 billion will receive ‘tailored’ supervision; with fewer and lower capital requirements and limited stress tests. If the supervisory authorities did identify any weaknesses at all, little was often done about them – due to that ‘tailoring approach’. In its SVB analysis, the Fed takes its own share of the blame, and calls for tighter supervision.


  1. Interest rate risk.

Banks are in the interest rate business. That makes them especially vulnerable to the current transition from low to high interest rates. If that happens too quickly, it may lead to a mismatch between interest received and interest paid. So this will have to be given sharp attention in managing interest rate risk with realistic assumptions and stress tests.


  1. Liquidity risk.

Liquidity requirements do not adequately take into account how volatile deposits are due to the digital, fintech, age. Also, the rate at which unguaranteed deposits are withdrawn has also been underestimated. Thus, liquidity weighting and assumptions will also be tightened and tested.


  1. Clear resolution.

When the banks went belly up, US and Swiss supervisors, resolution authorities and governments intervened decisively. The already available resolution plans were dusted off and implemented, allowing the banks’ critical functions to continue after a rescue weekend. However, as mentioned, things went wrong with CS’ AT1 bonds. These subordinated convertible bonds were fully written off by Swiss regulators. At the same time CS shareholders still received CHF 3 billion from UBS. Normally, AT1 holders are ranked before shareholders. The global AT1 market did not expect this ranking reversal and collapsed. That same day, EU regulators reported that they would do things differently (here): “(…) common equity instruments are the first ones to absorb losses, and only after their full use would Additional Tier One be required to be written down.” Based on EU resolution rules, however, this is not entirely evident.

History repeats

These lessons will be used in the EU for new regulations. For example, a tightening of actual, effective supervision; stricter liquidity weighting of deposits; stricter risk management requirements and stress tests for liquidity risk, interest rate risk and concentration risk; and clearer resolution rules. This is in addition to the already announced implementation of Basel 3.5 (in the EU Banking Package (here)) and revision of the EU Crisis Management and Deposit Guarantee Scheme framework (here). Of course, one can criticize overly strict and disproportionate banking supervision (so do I), but the past crisis has (again) shown the consequences of excessive deregulation and relaxation of supervision. Supervision that was strengthened and tightened precisely to prevent the previous crisis repeating.

Finally, a book tip: Trust by Hernán Díaz (Riverhead, New York 2022). This sentence struck a chord with me: “Walking around Wall Street during the weekend, one gets the impression that the world’s affairs have been settled once and for all, that the age of work is finally over and that humanity has moved on to its next stage.

But then on Mondays, everything starts all over again and history repeats.