Let's Go for a (Bank) Run
You are reading Economic Forces, a free weekly newsletter on economics, especially price theory, without the politics. You can support our newsletter by signing up here:
One thing that banks do is serve as intermediaries between borrowers and lenders. Some people want to save. Some people want to borrow. Banks facilitate both by matching up these people. It is pretty straightforward. However, that part of what banks do isn’t really unique to banks. All financial firms do this to some extent. An insurance company collects premiums and uses those premiums to purchase a portfolio of assets, like bonds.
What makes banks different from insurance companies is in terms of their liabilities. If someone gets a life insurance policy, they pay premiums every month or every year. But if you have a $100,000 life insurance policy and you show up at the office of your insurance agent and say, “hello, I would like my $100,000”, they will look at you and they will say, “sorry, you appear to be alive. Send someone to come back when you are dead.” If you push hard enough, they will give you some money. However, the money will be substantially below $100,000 and your beneficiary would no longer get anything when you die.
In theory, it is possible to imagine that you could sell claims to your insurance policy. You could go to Walmart and buy a new television with a claim to your life insurance. You get the TV. Walmart gets 1% of the payout on your life insurance. Of course, this creates all sorts of bad incentives. If someone is entitled to money when you die, they might have an incentive make it happen sooner than it otherwise would. Not Walmart. They have too much to lose, but not everyone has too much to lose relative to the payout.
Banks on the other hand do something different. Banks take your money, but these aren’t premiums, they are deposits. Nonetheless, banks don’t just store these deposits in some deep, underground vault below the bank with a little tag with your name on it. Instead, they are in the business of lending money. Some of the bank’s assets are held as reserves in the bank vault, but only a fraction. Yet, and here is the crucial difference, banks allow you to come and get your money whenever you want. Try this out. Go to the bank with a $100 bill and deposit it. Go back to the bank the next day and ask for your money back. They will give it to you. No questions asked.
Now, you might say the insurance company will give you your money back also. However, the insurance company will ask a lot of questions and they will make you pay them a penalty. Thus, you don’t exactly get your money back. The bank, however, promises that you have the right to show up whenever you want and get your money back. In fact, they put these little machines called ATMs next to their building that let you get their money even when the bank is closed! (By the way, they are ATMS, they are not ATM machines. The “M” stands for machine.)
Why would banks do this?
Well, think of this a different way. Suppose that banks didn’t exist. Some people would still want to save. Without banks, they would have to do so by investing directly with some individual or firm or by lending directly to some individual or firm. Of course, that could be risky, but a high enough rate of return or interest rate could make it worthwhile. Nonetheless, there is another form of risk. If I want to save and I give my money to another person as an investment or a loan, they are going to spend that money. This is important. Suppose that a week goes by and I get a flat tire. I now need my money back to get a new tire. However, the person to whom I loaned my money will not have it any longer. I cannot get my money back. I saved, but I have no access to my saving.
But think about flat tires. These are idiosyncratic risks. It is therefore something that can be insured against. For example, suppose I get together with a bunch of other savers. Now, imagine that we know that one of us is likely to get a flat tire, but only one of us. What we could do is we could pool all of our money together. We could then set aside enough money to pay for a new tire and loan out the rest. When one of us gets a flat tire, that person can show up and get his or her money back. This coalition of depositors sounds a lot like a bank!
Of course, there is one possible problem here. We don’t know for sure who has a flat tire. As a result, more than one person might show up to get the money. If so, there won’t be enough money to pay out. People might be concerned about this. Fearing that too many people will show up might incentivize others to show up first to get their money. However, if everyone does this, the “bank” will be forced to claw back as much money as it can to make these payments. In the process, the bank will lose money (selling its assets at fire sale prices) and won’t be able to pay everyone back. This run on the bank will leave it insolvent. But the run is a self-fulfilling prophecy.
To prevent this from happening, maybe some third-party like the government can offer deposit insurance. If the bank doesn’t have your money, the government will give you the money. In this case, there is no reason to show up early if you don’t have a flat tire. You get your money even if there is a run and therefore there is no reason to run.
You basically now understand the Diamond-Dybvig model of banks and bank runs, the authors of which won the Nobel Prize this week.
Now, as you might imagine, this leaves a bunch of stuff out when we want to think about the real world of banking. For example, there is no equity in the model. Typically banks have shareholders whose equity value rises when assets yield higher than expected returns and whose equity value declines when the assets yield lower than expected returns or when people default. Wouldn’t a bank with a sufficient amount of equity financing prevent this sort of self-fulfilling run? Furthermore, if this type of deposit contract was susceptible to runs, wouldn’t competition between banks lead banks to offer run-proof contracts? Wouldn’t competition lead to more innovative solutions to prevent run-induced insolvency? How would other solutions, such as holding shareholders responsible for losses to depositors (as was done historically) compare with deposit insurance? Is government intervention required? Might private institutions emerge with solutions that render deposit insurance superfluous?
The model also provides fodder for empirical questions. Are bank runs really self-fulfilling prophecies? Or do bank runs typically occur when depositors have good reason to believe that the bank is insolvent? (The evidence suggests that bank runs are not self-fulfilling prophesies, by the way, but are predictable from the characteristics of a bank’s balance sheet.)
In part, these questions probably helped win the award for Diamond and Dybvig as much as the model itself. Their model provided a frame of reference and inspiration for asking important theoretical and empirical questions about the stability of banking.