Some Misconceptions About Banks

People do not understand how different banking used to be

Many people hold the view that the banks are too free. They are not regulated enough, and it is this lack of regulation which leads to periodic financial crises and recessions. I do not think this is true. In fact, I think that basically everyone does not grasp how much, and how badly, banks were regulated. The history of U.S. banking has been one where, out of fear that banks would be too big to fail, we left them too small to succeed.

The first sin of U.S. banking was leaving the states largely in charge of it. National charters were extremely rare, so if you wanted to be in business, you would have to do it all in one state. Nor was that all. Many of those states forbade even opening branches. You could have one office, in one town, and that was all. The banks were too darn small.

I cannot emphasize how insane this is enough. The whole point of banking is to spread out risk over a wide pool so that people can access capital for improvements when they need it, just like how insurance works. If a bank is forced to be in one place, then all of the negative shocks are correlated with each other. If it is a bad year for one corn farmer, then it is probably a bad year for all corn farmers, and down goes the local bank. On the eve of the Great Depression, there were over 24,000 different banks. In the state of Nebraska alone, there were 1,146 banks in 1919. Illinois had almost 2,000. They were like tinder before the blaze. If the banks had been larger, they could have held off the bank runs and the Great Depression would have been much less bad.

I will give some background on banks, and bank runs, before we go on. The job of a bank is to allocate capital. It finds loan opportunities, and raises funds to make those loans. To make a profit, it demands interest to be paid; in order to raise funds, it pays interest to its depositors. What makes a bank special is that its depositors can demand repayment at any point. If everyone does this, though, the bank would fold. The bank had lent out the money to do tasks which would take a longer time to complete than its depositors could accept. Importantly, once the number of depositors seeking their money reaches a high enough threshold, it is now privately optimal for everyone to rush out and get their money before the bank fails — hence, a bank run.

The small size and the local character of the banks left them doubly vulnerable to negative shocks. The fewer the depositors in a bank, the less cushion there is to avert a bank run, and the correlation of negative shocks makes it more likely that a bank will be made insolvent. If we contrast it to a place with national banking, like Canada, we will find that bank collapses were far less common there than elsewhere. There have been occasional collapses, but those were due to the mismanagement of the particular enterprise, and not a general collapse of the banking system. There, the federal government had the power from the beginning to regulate and charter banks. Even if you believe that regulation is necessary to keep banks from systemic risks, leaving it up to the states pushes capital out of banks and into securities markets, which are much harder to regulate.

The bans on interstate banking would not be fully repealed until 1994, although banks at the state level were increasingly deregulated during the 70s and 80s. Twelve years earlier, another incredibly damaging regulation would be effectively deregulated — Regulation Q.

Regulation Q banned paying interest on any deposit account. It was simply illegal for a bank to pay people to lend them money. The reasons why it was passed are so bafflingly confused that I do not wish to believe anyone actually believed them. They thought, at the time, that being able to pay interest led to excessive competition between banks, and that they took excessive risks. As an aside, I think it is underrated how primitive an understanding of economics the Roosevelt administration had. Their basic problem was that the demand for gold elsewhere, in particular France, had increased. Since the price of gold was $21 dollars an ounce, the only way for gold to become more expensive in real terms is for the price of everything else to fall. Real wages sharply increased, and companies simply ran out of money and failed. Both Hoover and FDR confused the symptom of loose money (high prices) with the cause of loose money, and as a result kept markets from clearing. Hoover’s industrial policy kept wages from falling to match the new supply of money, and FDR’s NIRA explicitly existed to prevent “ruinous competition” and falling prices, when that is exactly what needed to occur.

In any event, Regulation Q is passed, and interest on any account which you can get on demand is banned. This includes savings accounts, by the way. While the cap was adjusted somewhat up – around 5% on savings accounts, with some small variations – the natural result is that banks have less in the way of deposits, and now the real effect of the cap on interest is dependent upon monetary policy.

There is a very plausible story where this worsens inflation, and is responsible for the stagflation of the 1970s. The primary source for this is a recent paper by Itamar Drechsler, Alexi Savov, and Phillipp Schnabl, called “Credit Crunches and the Great Stagflation”. This is a revisionist story, to be clear, and I do not think it explains everything which went on in the 1970s, but I do think it is true, and cleanly explains many of the facts of the period. This is not even a particularly new story. Timothy Cook, in 1978, showed that when the market rate of interest bumped over the legal threshold, depositors pulled their money out of the banking system and lent it on their own – in other words, it largely went into the stock market.

This does two things. First off, the main mechanism by which the Federal Reserve is trying to control the money supply no longer actually touches most consumers. They are raising rates, and it doesn’t increase the return to saving. Second, by reducing the amount available to banks, many otherwise profitable loans are unable to be made. Firms find themselves unable to borrow, and they become unable to fill the orders which they have received. Nor could they replace bank loans with corporate bonds – only some firms had access to this, and while bond issuance did increase during credit crunches, it does not balance out the reduction in loans.

Alternative explanations are not fully adequate. It is striking that the oil shocks, which are often taken as a cause of lower than expected supply, generally came after the change in prices. Nor is it accurate to argue that stagflation was a global experience, and Regulation Q unique to the US. The specific regulation was unique, but other countries also engaged in extensive capital controls, with similar results. I do think the ending of stagflation was due to Volcker, but Regulation Q certainly didn’t help.

Nominal price caps are going to have very different effects depending on the macroeconomic environment. This ties into prior work on this blog. We are now considering interest rate caps on credit cards. Otherwise sensible pundits, like Matt Yglesias, have endorsed them, at least on the grounds that it would be extremely popular. Indeed it would be, but it would also be economically disastrous. I would like, very much, for us to learn from our mistakes. Popularism must reject insanity, just as much as it reject unpopular policies.