Bernanke, Diamond & Dybvig (2022): Why Banks Fail and Why It Matters

The Most Dangerous Institution in the Economy

Banks are strange institutions. They take your money and promise you can have it back anytime. Then they lend most of it to someone else for 30 years. As long as everyone doesn’t ask for their money at the same time, this works beautifully — banks create liquidity, fund investment, and power economic growth.

But if everyone does ask at the same time, the bank is finished. It can’t recall 30-year mortgages overnight. The assets are real but illiquid; the promises are liquid but fragile. This mismatch — maturity transformation — is both banking’s greatest contribution and its fatal vulnerability.

The 2022 Nobel Prize honored three economists who explained why this vulnerability exists, why it matters far more than anyone previously understood, and what can be done about it. Their work, all published in the early 1980s, became the intellectual foundation for how the world responded to the 2008 financial crisis and the COVID-19 economic shock.

Diamond and Dybvig: Why Bank Runs Happen to Healthy Banks

Before the Diamond-Dybvig model (1983), the common understanding of bank runs was simple: banks fail because they make bad loans. Depositors run because they discover the bank is insolvent. The run is a symptom, not the disease.

Diamond and Dybvig overturned this. They showed that bank runs can happen to perfectly healthy banks — and that the run itself destroys the bank.

The setup:

• People face uncertainty about when they’ll need their money. Some will need it soon (to cover an emergency); others won’t need it for years
• Long-term investments pay higher returns than short-term ones, but you can’t get your money out early without a loss
• Without banks, people face an unpleasant choice: invest long-term and risk being stuck when an emergency hits, or keep everything in low-return liquid assets

What banks do:

• Banks pool deposits from many people and invest most of the pool in high-return long-term projects
• Because not everyone needs their money at the same time (by the law of large numbers), the bank keeps only a small reserve for withdrawals
• This lets depositors have the best of both worlds: access to their money anytime (liquidity) AND higher returns from long-term investment
• Banks create liquidity that wouldn’t otherwise exist — this is their fundamental economic function

The fatal vulnerability:

• Bank deposits are paid out on a first-come, first-served basis
• If depositors believe the bank is safe, only those who genuinely need cash withdraw. The system works perfectly
• But if depositors believe others are about to withdraw, the rational response is to run to the bank and withdraw immediately — even if you don’t need the money — because the last depositors in line get nothing
• This is self-fulfilling: the run is caused by the belief that a run will happen. A perfectly healthy bank with perfectly good loans can be destroyed by nothing more than a shift in expectations
• The bank is forced to liquidate long-term investments at a loss, destroying value that would have been there if everyone had simply stayed calm

Two equilibria:

The Diamond-Dybvig model has two stable outcomes:

1. Good equilibrium: Everyone trusts the bank, only genuine needs trigger withdrawals, and the bank functions perfectly
2. Bad equilibrium: Everyone panics, everyone runs, and the bank collapses — even though there was nothing wrong with it

What tips the economy from one equilibrium to the other? Anything. A rumor. A news headline. Another bank’s failure. The fragility is inherent in the structure of banking itself.

The solution: deposit insurance

Diamond and Dybvig showed that government-backed deposit insurance eliminates the bad equilibrium entirely. If depositors know they’ll get their money back regardless of what others do, they have no reason to run. The insurance almost never needs to be paid out — its mere existence prevents the panic.

This provided the theoretical justification for deposit insurance programs (like the FDIC in the United States, established in 1933 after thousands of bank failures during the Great Depression).

Diamond: Banks as Delegated Monitors

In a separate 1984 paper, Douglas Diamond explained another crucial function of banks: delegated monitoring.

The problem: When you lend money to a business, you need to monitor whether the borrower is using it wisely. But monitoring is expensive, and if thousands of small savers each try to monitor the same borrower, the duplication is wasteful.

The solution: Banks act as delegated monitors — they pool deposits from many savers and monitor borrowers on everyone’s behalf. This is efficient because:

• One monitor (the bank) replaces thousands of individual monitors
• The bank develops expertise in evaluating and monitoring borrowers
• By lending to many different borrowers, the bank diversifies risk — if one borrower defaults, the others’ payments cover the loss
• Diversification is so effective that depositors can be promised a fixed return — they don’t need to worry about individual loan performance

This explains why banks exist at all, rather than everyone lending directly through financial markets. Banks solve an information problem that markets handle poorly.

Bernanke: How Bank Failures Caused the Great Depression

Ben Bernanke asked one of the most important questions in economic history: why did a moderate recession in 1929 become the worst economic catastrophe of the twentieth century?

The standard answer pointed to monetary policy — the Federal Reserve allowed the money supply to contract. But Bernanke showed that money supply alone couldn’t explain the depth and duration of the Depression. Something else was amplifying the damage.

Bernanke’s insight: The wave of bank failures between 1930 and 1933 didn’t just destroy money — it destroyed information.

• Banks spend years building knowledge about their borrowers — which businesses are creditworthy, which projects are viable, which entrepreneurs can be trusted
• When a bank fails, this information is lost. It can’t be written down or transferred. It exists in relationships, in loan officers’ judgment, in institutional memory
• New lenders can’t simply step in and replace the failed bank. They don’t know the borrowers. They can’t distinguish good risks from bad risks. So they either don’t lend at all, or they charge high interest rates to compensate for their ignorance
• The cost of credit intermediation — the cost of connecting savers with borrowers — soars after bank failures
• Businesses that depended on bank credit can’t finance their operations. Profitable projects go unfunded. Workers are laid off not because of weak demand but because their employers can’t access credit

Bernanke showed empirically that regions with more bank failures experienced deeper depressions, even controlling for other factors. The timing matched: the economy began to recover only after the bank failures stopped and the financial system stabilized.

The credit channel: Bernanke demonstrated that banks are not just passive intermediaries shuffling money from savers to borrowers. They are active participants whose health directly affects the real economy. When the banking system breaks, the entire economy breaks with it — not just because money is scarce, but because the economy’s information infrastructure is destroyed.

How Their Work Saved the World (Twice)

The research of all three laureates converged on a single message: banking crises are not just financial events — they are economic catastrophes that must be prevented or contained at almost any cost.

The 2008 financial crisis:

When Lehman Brothers collapsed in September 2008, the world faced a Diamond-Dybvig bank run in modern form. The “banks” weren’t just traditional banks — they were investment banks, money market funds, and other financial institutions that had liquid liabilities but illiquid assets. The panic was spreading exactly as the model predicted.

Ben Bernanke, then chairman of the Federal Reserve, applied the lessons of his own research:

• The Fed extended emergency lending to prevent further bank failures — acting as a lender of last resort
• Deposit insurance was expanded to cover more accounts
• The government recapitalized failing banks to restore confidence
• The Fed cut interest rates to near zero and began massive asset purchases

These actions were politically controversial but intellectually grounded in the laureates’ research. The Great Recession was severe — but it wasn’t the Great Depression. Bernanke’s own academic work on the 1930s was the playbook for preventing a repeat.

The COVID-19 crisis (2020):

When the pandemic shut down economies worldwide, central banks and governments acted even faster, drawing again on the same framework — massive lending facilities, expanded guarantees, and aggressive monetary policy to prevent a financial system collapse on top of the health crisis.

Their 2022 Nobel Prize was awarded “for research on banks and financial crises.”

Explain It to a Child

Imagine a neighborhood swimming pool that everyone shares. Everyone pays a little money each month, and the pool stays open. The pool can handle 50 people swimming at the same time — and since the whole neighborhood has 500 people but they come at different times, it works great. Now imagine someone starts a rumor: “The pool is closing tomorrow! Everyone better swim today!” If everyone shows up at once, the pool is overwhelmed, the water gets gross, and it actually does close. The rumor made itself come true. That’s a bank run. A bank takes money from lots of people and lends it out. As long as everyone doesn’t ask for their money back at the same time, it works perfectly. But if everyone panics and rushes to the bank at once, the bank runs out of cash — even though there was nothing actually wrong with it. Diamond and Dybvig figured out why this happens and how to stop it: promise everyone they’ll get their money back no matter what (that’s deposit insurance), and nobody panics. Bernanke showed that when banks DO fail, it’s not just the banks that suffer — the whole economy crashes because banks know things about businesses that nobody else knows, and when that knowledge is lost, nobody can borrow money to keep working.