Banking on a Fiction in (it’s) a Wonderful Life

The essence of financial distress is loss of confidence.”

– Charles P. Kindleberger (Manias, Panics and Crashes, originally published in 1978)


Ever have the fear that when you show up to the bank to withdraw some cash, the bank won’t actually have any to give you? No? That’s good. That means that banks, with a little help from the government, are doing a good job of safeguarding, not necessarily your money, but confidence in one of banking’s great fictions.

What do I mean by that? Let’s take a peek inside George Bailey’s Building & Loan Association bank to find out…

Source: YouTube

That’s what you call a bank run, and if good ol’ George Bailey didn’t know what suffering from a liquidity crisis was like before, he certainly does now.

Where’s the Money, George?

Like George Bailey’s deposit-holders, you might be wondering why the bank doesn’t have their cash. Well, there’s a simple reason for that, but it’s different from the one that George offers. As respectable and honest a man as George appears to be, his reasoning for why his bank doesn’t have the cash to pay its deposit-holders is a bit crooked.

Now it’s possible that Frank Capra, the film’s director, is the one who is misinformed and falls prey to the common fallacy that banks lend out their deposit-holders’ cash. That is, of course, what George implies when he claims, “…your money’s in Joe’s house, that’s right next to yours [as he points to one of the other bank’s customers] … you’re lending them the money to build, and then they’re gonna pay it back to you as best they can! What are you gonna do? Foreclose on them??”

If I’ve said it once (here), I’ve said it a thousand times, banks don’t lend out other people’s cash. In fact, they don’t lend out cash at all. The misguided view portrayed in It’s a Wonderful Life might help to explain why someone created this short video, just two years later.

Source: YouTube (Thanks to Éric Tymoigne, whose textbook draft, “The Financial System and the Economy,” is where I first encountered this video.)

Did you hear that part about no money changing hands when a loan is made? That’s a different story than the one George Bailey is telling his deposit-holders. He’s trying to convince them to believe that their deposited cash is being directly lent out to others in order that these others can buy something with it, a new home for example.

People don’t generally buy homes with cash (or a lot of other things for that matter). They buy them by drawing upon the balance in their bank deposit account. While it’s true that one can exchange cash for a higher balance, banks can also increase that balance by extending the deposit-holder credit. In other words, when the bank extends a loan to a borrower, it credits the borrower’s deposit account. It does not give the borrower cash.

Let’s illustrate the two contrasting perspectives. For starters, here are two identical bank balance sheets, one for George Bailey’s Fairy-tale Bank, and the other for the Real-world Bank (not to be confused with the real World Bank).

At this initial stage, both balance sheets are identical. Each bank has $100,000 in assets matched by $80,000 in liabilities and $20,000 in equity, staying consistent with the basic accounting identity that Assets = Liabilities + Equity. Notice also that each bank has cash assets of $20,000.

Let’s assume that Joe, one of the customers of George’s bank, is looking to borrow $10,000. George Bailey’s Fairy-tale Bank would make the loan by directly handing over cash to Joe and recording a new loan asset on its balance sheet. Real-world Bank, on the other hand, would credit Joe’s deposit account while similarly recording a new loan on the asset side of the balance sheet. This process will look something like this.

The Fairy-tale Bank records a new loan asset while at the same time handing over $10,000 in cash to Joe. The Real-world Bank, however, did nothing of the sort. Instead, it recorded a new deposit liability, which is a promise to pay Joe $10,000 cash. That’s what you call credit creation, as the second video claims. It’s also what I have called money creation (here), since deposits form a part of the money supply and therefore, newly created deposits equal newly created money.

The updated balance sheets will look something like this.

There are some obvious differences. The Fairy-tale bank now has half as much cash as it did before, and the size of its balance sheet has not changed. The Real-world Bank, on the other hand, still has the same amount of cash as before, but has increased both its assets and liabilities by $10,000 each. Thus, its balance sheet has expanded by the same amount. So, what are the implications?

First of all, George’s insinuation that the bank is simply an intermediary, taking in depositors’ cash and then lending it to borrowers, while a common misconception, is exactly that, a misconception. Banks don’t merely intermediate the transfer of purchasing power from savers to borrowers. They create it.

Second of all, if the Fairy-tale Bank story were true, one can see that there is a well-defined limit to the amount of loans that a bank could make. Fairy-tale Bank’s total stash of cash represents its maximum total lending capacity. Once that stash is gone, no more loans. In comparison, all else being equal and barring any government regulation, it’s not obvious what Real-world Bank’s lending limit would be. The expansion of its balance sheet is potentially limitless.

We could speculate on a potential limit if we engaged in a thought experiment of what would happen if, as soon as the loan were made, Joe redeemed his $10,000 deposit-credit for cash.

Looking back to what Real-world Bank’s balance sheet looked like after it extended the loan to Joe, and taking into account this new transaction, it’s illuminating to discover that the new balance sheet would look identical to that of the Fairy-tale Bank’s.

This thought experiment suggests that as long as people can be convinced to hold their money in the form of deposits rather than cash, banks have potentially limitless lending capacity. However, there always exists the risk that deposit-holders will come in demanding cash-redemption, a risk that increases the more banks expand their balance sheets through the creation of credit. If he was unaware of that risk before hand, George Bailey certainly understands it now.

Having lost his deposit-holders’ confidence, George tries to tap into their communal sensibilities by putting a friendly human face on the problem (mind you, not his own face): You’re lending money to each other. If you demand all your cash now, then you’re gonna foreclose on your neighbours!! And remembering the second of the two greatest commandments to “Love your…” Well, you get the picture.

Needless to say, George’s deposit-holders aren’t having any of it. They all want cash, and they all want it now. But, therein lies the problem. Although banks accept cash in exchange for deposits, they also create deposits without also procuring the cash at the same time. As both types are promises to pay cash, the bank could never make good on all of those promises at once. That these two types of promises to pay look exactly the same even though they came about by different processes is precisely the fiction on which modern banking depends.

In order to sustain that fiction, banks must maintain deposit-holders’ confidence in their ability to always make good on those promises. When confidence in the fiction dissolves, banks are at risk of suffering a similar fate.

Sustaining the Fiction

The best way banks can maintain confidence in the belief that they will always be able to make good on their promises is to actually make good on them when they are called to do so. Obviously, that’s difficult in the middle of a bank run. But bank runs are generally rare events that are preceded by a loss of confidence, rather than being a reason for a loss of confidence. It is in the day-to-day course of a bank’s business that it must prove it is capable of honouring its promises to pay.

One way to do that is to maintain a reasonably sized buffer of cash relative to deposit liabilities. A bank can draw from this buffer in the case where depositor-holders come in looking to redeem their deposit balances for cash, which is something they do, not only in bank runs, but because some of their payments are often more conveniently made in cash. This cash buffer constitutes the bank’s cash reserves. But demand for cash is just one type of payment obligation banks face.

Another obligation arises when deposit-holders make payments that rely on the transfer of deposits, rather than the exchange of cash. Although cash is still used for many small-value transactions, the greater value of transactions is done using methods of payment like cheques, debit cards, money orders and various others. In order to understand how these methods of payment impose separate obligations for banks, it’s helpful to run through a few examples.

Suppose Joe wants to use the loan from Real-world bank to buy a new car worth $10,000. To make the payment, Joe writes a cheque for the $10,000 made out to the car dealership. If we assume both Joe and the car dealership have their accounts at the Real-world Bank, the Real-world Bank will process the transaction in the following way.

That was simple. Joe got a car and the car dealership got money, and all the bank had to do was debit Joe’s deposit account and credit the car dealership’s. Instead of owing Joe $10,000, the bank now owes the car dealership $10,000. But as long as the car dealership is willing to continue holding its new wealth in the form of Real-world Bank deposit liabilities, the Real-world Bank does not have to come up with any cash to make a payment.

However, if for some odd reason the car dealership decided that cash was the only acceptable form of payment it would accept, then the Real-world Bank would be required to hand over cash to Joe, who would then use it to pay the car dealership. If the car dealership stuffs the cash under the mattress rather than depositing it in their account at the Real-world Bank, the bank’s cash reserves would be reduced. Such a reduction would increase the bank’s liquidity risk as it has less of a buffer to meet similar demands for cash-redemption.

Needless to say, the Real-world Bank prefers the non-cash form of payment rather than the cash form — less cash means less of a cash buffer, and thus greater liquidity risk. Yet, the non-cash preference of the transaction becomes a bit of a moot point as soon as we introduce another, different bank at which the car dealership does its banking. We shall call this bank the New-world Bank.

Under this new two-bank scenario, Joe’s payment to the car dealership will generate a payment obligation for the Real-world Bank, which can be settled using its deposit account at the central bank. Such deposit accounts are generally known as reserves, which along with cash, comprise a bank’s total reserves. Joe’s payment can be represented as follows.

The transaction that takes place between Joe and the car dealership now induces a transaction between the two separate banks. This transaction takes place via the balance sheet of the central bank, which debits the Real-world Bank’s reserve account and credit’s the New-world Bank’s reserve account. In this case, as in the case where Joe paid cash, the Real-world Bank’s total reserves have declined, once again increasing the bank’s liquidity risk.

Broadening the picture a bit, it’s worth remembering that payments like this (of values both larger and smaller) are happening all day long. In some economist’s ideal fantasy world, all of the payment flows between banks might net to zero over some relatively short time period, a day for example. If that were the case, bankers would know that any liquidity risk arising from an imbalance of payment flows throughout the day would be temporary. But of course, we live in the real world, where daily imbalances are commonplace.

Yet, while the imbalance of payment flows that leads to an imbalance in reserve flows appears to pose a problem, it also suggests a possible solution. Except for the rare case where deposit-holders are redeeming their deposits for cash and stuffing it under their mattresses, one bank’s outflows are another bank’s inflows. Now if only there were some market where banks experiencing outflows could borrow from those experiencing inflows. It turns out there is. It’s called the money market.

In reality there are different types of money markets, but the idea is simple: it is a space where those looking for short-term funding can find those with excess liquid funds looking to earn a competitive interest rate. In this space, we find banks competing for individual and corporate deposits. We find institutional investors, like pension and endowment funds, looking for a short-term place to park their liquid funds in between their longer-term investments. Yes, we even find interbank overnight lending markets, where banks deal directly among themselves knowing that if they are a lender today, they could just as well be a borrower tomorrow.

If banks are able to reshuffle cash and reserves back and forth among themselves so that any loss of liquid assets due to payment flow imbalances are temporary, then liquidity risk is kept to a minimum, or at least it appears that way. Indeed, if this reshuffling remains smooth and frictionless even as the proportion of banks’ reserves relative to their liabilities diminishes through the expansion of their balance sheets (as illustrated below),

then liquidity risk starts to fade into the background, a fleeting memory of crises past…

…and, back to the real world, where frictions can and sometimes do arise. That’s why it always helps to have a backup source of liquidity. You know, like this…

Source: YouTube

Hmmm…dipping into the honeymoon fund — that’s one dedicated banker, and banker’s wife for that matter! But for better or worse, right?

And for Richer or…

While today’s bankers may be able to afford some fairly extravagant honeymoons, they have means other than their honeymoon fund for dealing with temporary liquidity crises. They have the government, which has an interest in ensuring that the banking system continues to function smoothly as it plays an important role in financing new entrepreneurial enterprise and economic growth.

Through the central bank, an agent of the government, banks may obtain short-term backup funding in the case where they are unable to obtain it due to various frictions that sometimes arise in money markets. These standing liquidity facilities, as they are often called, are intended as a way for solvent banks to obtain liquidity when they are suffering from a temporary shortage. Such liquidity backups keep a small hiccup in the financial system from cascading into an all out crisis. It’s all about confidence management, and maintaining a smooth functioning payment system helps to sustain confidence.

Deposit insurance is another way to help sustain it. Such insurance is an explicit government guarantee to deposit-holders where, in the case of a bank failure, the government reimburses them for their loss. Simply having such insurance in place helps to reassure deposit-holders that their money is indeed safe. That reassurance alone may serve to prevent a run on the bank, and the subsequent increase in risk of bank failure that is brought about by such an event.

Of course, one bank failure is not so bad. In fact, it’s to be expected in a competitive market economy where the strong and competent are rewarded, while weak and blundering amateurs are punished. Yet, the problem with one bank failure is that it can trigger other bank failures due to the financial system’s interconnectedness. One bank failure, not so bad. Complete financial meltdown, not so good.

Meltdown is precisely what was threatening the U.S. financial system around ten years ago after the failure of Lehman Brothers, something I’ve discussed here. To prevent such a catastrophe from happening, the U.S. government stepped-in to bailout the banks, laying bare the implicit guarantee to offer much more than the usual short-term liquidity support in times of crisis. This wasn’t just some temporary liquidity crisis (although one could postulate that it could have been if the government stepped in sooner and not let Lehman fail; but hey, start giving freebies to one blundering amateur then you have to give freebies to all of them — wait, what?!).

The problem with such guarantees, whether an explicit guarantee in the form of deposit insurance or an implicit guarantee in the form of bank bailouts, is that they can actually create a little too much confidence. That is, both deposit-holders, when they exchange their cash for bank deposit liabilities, and bankers, in the normal course of their business, can become complacent about the risks they are assuming, because the government has got their backs. That’s the moral hazard problem.

For that reason, the government requires banks to adhere to its rules in exchange for those guarantees, or as Goodhart and Schoenmaker claim, “He who pays the piper, calls the tune.” One such rule is a liquidity requirement, which stipulates that banks must maintain a minimum level of highly liquid assets, like cash and reserves, on hand in case of an abnormal redemption or transfer of deposits. I mean hey, if such a profitable alliance is going to last, both partners have to be willing to make compromises, right?