Banking Crises Flashcards
What are the key points of the Diamond & Dybdig model regarding banking crises with rational behaviour?
- Banks have an important role in risk-shifting and cannot be replicated.
- They way we write contracts means that a bank run is always a possible equilibrium
- A solvent bank can still be subject to a bank run
All this leads to the conclusion that banking crises are integral to how the financial system works and if we want to get rid of bank runs, the banks become useless and lose their economic function.
Three ways of preventing bank runs in the Diamond & Dybdig paper? (will not solve the problem though, without coming at a high cost)
- Narrow banking
- Suspension of convertibility
- Deposit insurance
The bank run is a possible equilibrium in the pay-off matrix, what is it that makes this matrix a prisoner’s dilemma?
Because my optimal action depends on your action. So it might be the case that I run if I see you running, OR if I just expect that you will run. So expectations are important here.
How can narrow banking solve the problem of bank runs and what’s the problem with this solution?
Narrow banking means that you always have enough money to serve the customers. So no one needs to run to the bank. BUT this means that you as a bank does not invest in the productive asset. So your economic function of offering returns is gone, you are useless. people can just keep their money in their pockets.
Improving the contract by using ‘suspension of convertibility’ - what’s the challenge?
If you have a suspension rule that makes you able to say “you cannot get your money back” - you need to know how large the fraction of type 1 is, so that you know how much money you need to have at stake. But you will never know this. So you will have al lot of dissatisfied customers when you take your liquidity service away.
How could deposit insurance improve the contract? And what is the challenge with this solution?
It could improve by having the government insure the depositors. So when the bank uns out of money, the government steps in. The problem is two: first, you increase the liabilities for the government and second, you increase the moral hazard problem because now the bank will invest in crazy risky projects because it’s the government’s problem if they fail.
What are the two most important takeaways from D&D models?
1) It gives a strak insight into what the nature of the bank is. A bank that cannot be destroyed is not useful.
2) The only way to improve contracts either creates liabilities for someone else (depositors or the government) or reduces the function of the bank.
What is the difference between 1854 and 1857 banking panics?
54 was a local panic that started because Knickerbocker bank failed, so everyone started running. Sense-less run since all other banks were solvent.
57 was caused by a price fall in the market for western land and it led t a nationwide panic. More rational panic.
How was panic 54 and 57 different in the closing of bank accounts?
In 54, the closings increased with a sudden spike and rise quickly, the slows than and goes back to long-term average. In 57, it starts with small number of closings and then it increases.
–> This is important!! And the difference is that in 57, there were more sophisticated persons that started closing and then they great mass came after. –> more rational panic.
What is the striking thing about the origins of the Financial crises 08-09?
That it stems from such a small part of the US economy. It was basically only 6% of the US financials that were involved in creating the crisis - repackaged high risk debt (3%) and subprime mortgages. (3%).
2 arguments for why the housing market did not had a bubble in the financial crisis?
- Time preferences shift towards valuing the future more and therefore you are willing to pay more for a house, because it gives you housing services at a lower cost per each period.
- Option values - people want to buy houses because it gives you both the housing service consumption stream + insurance against future rent increase. (Cochrane argument that Palm/3Com is not the same, here renting and owning is not the same and the price increase comes form people going just crazy, irrationality).
Why did the housing prices increase according to Main & Sufi paper?
Because of a supply-effect that came from the financial market. It was a mix of loosening the requirements or taking a loan (seen by higher debt-to-income ratios for those getting a mortgage, less documentation etc) AND also increased securitisation in the financial institutions, meaning that banks did not carry the risk because they re-packaged the loans and sold them away.
Couldn’t it just be that people were better of economically and therefore got more mortgage loans accepted?
Could be yes, but it was not the case. Data shows that the largest increase in accepted mortgage applications happened in areas where there were to change in income, they were actually kind of poor, high unemployment, low education and so on. In 2006 the denial rates (acceptance rates) decreased (increased) rapidly.
How does the other paper, Keys et al. investigate the securitization?
They look at the FICO score, because there is a rule of thumb saying that people below score 620 can’t securitize, but what they see is actually drastic loosening of this rule
How did special investment vehicles fuel the securitisation?
They did so because the capital needed for banks to invest in this stuff was much lower than standard. 0.8-1.6 percent instead of 8. And these vehicles also had this credit and liquidity guarantee. So there were large incentives for banks to securitize. Thanks to these guarantees, the risk-shifting was completely non-existent.
