central bank

What happens when a bank fails

pnrj critique of neoclassical economics, current events, inequality, macroeconomics, market failure, public policy , , , , , , , , , , , , , , , ,

Mar 19 JDN 2460023

As of March 9, Silicon Valley Bank (SVB) has failed and officially been put into receivership under the FDIC. A bank that held $209 billion in assets has suddenly become insolvent.

This is the second-largest bank failure in US history, after Washington Mutual (WaMu) in 2008. In fact it will probably have more serious consequences than WaMu, for two reasons:

1. WaMu collapsed as part of the Great Recession, so there was already a lot of other things going on and a lot of policy responses already in place.

2. WaMu was mostly a conventional commercial bank that held deposits and loans for consumers, so its assets were largely protected by the FDIC, and thus its bankruptcy didn’t cause contagion the spread out to the rest of the system. (Other banks—shadow banks—did during the crash, but not so much WaMu.) SVB mostly served tech startups, so a whopping 89% of its deposits were not protected by FDIC insurance.

You’ve likely heard of many of the companies that had accounts at SVB: Roku, Roblox, Vimeo, even Vox. Stocks of the US financial industry lost $100 billion in value in two days.

The good news is that this will not be catastrophic. It probably won’t even trigger a recession (though the high interest rates we’ve been having lately potentially could drive us over that edge). Because this is commercial banking, it’s done out in the open, with transparency and reasonably good regulation. The FDIC knows what they are doing, and even though they aren’t covering all those deposits directly, they intend to find a buyer for the bank who will, and odds are good that they’ll be able to cover at least 80% of the lost funds.

In fact, while this one is exceptionally large, bank failures are not really all that uncommon. There have been nearly 100 failures of banks with assets over $1 billion in the US alone just since the 1970s. The FDIC exists to handle bank failures, and generally does the job well.

Then again, it’s worth asking whether we should really have a banking system in which failures are so routine.

The reason banks fail is kind of a dark open secret: They don’t actually have enough money to cover their deposits.

Banks loan away most of their cash, and rely upon the fact that most of their depositors will not want to withdraw their money at the same time. They are required to keep a certain ratio in reserves, but it’s usually fairly small, like 10%. This is called fractional-reserve banking.

As long as less than 10% of deposits get withdrawn at any given time, this works. But if a bunch of depositors suddenly decide to take out their money, the bank may not have enough to cover it all, and suddenly become insolvent.

In fact, the fear that a bank might become insolvent can actually cause it to become insolvent, in a self-fulfilling prophecy. Once depositors get word that the bank is about to fail, they rush to be the first to get their money out before it disappears. This is a bank run, and it’s basically what happened to SVB.

The FDIC was originally created to prevent or mitigate bank runs. Not only did they provide insurance that reduced the damage in the event of a bank failure; by assuring depositors that their money would be recovered even if the bank failed, they also reduced the chances of a bank run becoming a self-fulfilling prophecy.


Indeed, SVB is the exception that proves the rule, as they failed largely because their assets were mainly not FDIC insured.

Fractional-reserve banking effectively allows banks to create money, in the form of credit that they offer to borrowers. That credit gets deposited in other banks, which then go on to loan it out to still others; the result is that there is more money in the system than was ever actually printed by the central bank.

In most economies this commercial bank money is a far larger quantity than the central bank money actually printed by the central bank—often nearly 10 to 1. This ratio is called the money multiplier.

Indeed, it’s not a coincidence that the reserve ratio is 10% and the multiplier is 10; the theoretical maximum multiplier is always the inverse of the reserve ratio, so if you require reserves of 10%, the highest multiplier you can get is 10. Had we required 20% reserves, the multiplier would drop to 5.

Most countries have fractional-reserve banking, and have for centuries; but it’s actually a pretty weird system if you think about it.

Back when we were on the gold standard, fractional-reserve banking was a way of cheating, getting our money supply to be larger than the supply of gold would actually allow.

But now that we are on a pure fiat money system, it’s worth asking what fractional-reserve banking actually accomplishes. If we need more money, the central bank could just print more. Why do we delegate that task to commercial banks?

David Friedman of the Cato Institute had some especially harsh words on this, but honestly I find them hard to disagree with:

Before leaving the subject of fractional reserve systems, I should mention one particularly bizarre variant — a fractional reserve system based on fiat money. I call it bizarre because the essential function of a fractional reserve system is to reduce the resource cost of producing money, by allowing an ounce of reserves to replace, say, five ounces of currency. The resource cost of producing fiat money is zero; more precisely, it costs no more to print a five-dollar bill than a one-dollar bill, so the cost of having a larger number of dollars in circulation is zero. The cost of having more bills in circulation is not zero but small. A fractional reserve system based on fiat money thus economizes on the cost of producing something that costs nothing to produce; it adds the disadvantages of a fractional reserve system to the disadvantages of a fiat system without adding any corresponding advantages. It makes sense only as a discreet way of transferring some of the income that the government receives from producing money to the banking system, and is worth mentioning at all only because it is the system presently in use in this country.

Our banking system evolved gradually over time, and seems to have held onto many features that made more sense in an earlier era. Back when we had arbitrarily tied our central bank money supply to gold, creating a new money supply that was larger may have been a reasonable solution. But today, it just seems to be handing the reins over to private corporations, giving them more profits while forcing the rest of society to bear more risk.

The obvious alternative is full-reserve banking, where banks are simply required to hold 100% of their deposits in reserve and the multiplier drops to 1. This idea has been supported by a number of quite prominent economists, including Milton Friedman.

It’s not just a right-wing idea: The left-wing organization Positive Money is dedicated to advocating for a full-reserve banking system in the UK and EU. (The ECB VP’s criticism of the proposal is utterly baffling to me: it “would not create enough funding for investment and growth.” Um, you do know you can print more money, right? Hm, come to think of it, maybe the ECB doesn’t know that, because they think inflation is literally Hitler. There are legitimate criticisms to be had of Positive Money’s proposal, but “There won’t be enough money under this fiat money system” is a really weird take.)

There’s a relatively simple way to gradually transition from our current system to a full-reserve sytem: Simply increase the reserve ratio over time, and print more central bank money to keep the total money supply constant. If we find that it seems to be causing more problems than it solves, we could stop or reverse the trend.

Krugman has pointed out that this wouldn’t really fix the problems in the banking system, which actually seem to be much worse in the shadow banking sector than in conventional commercial banking. This is clearly right, but it isn’t really an argument against trying to improve conventional banking. I guess if stricter regulations on conventional banking push more money into the shadow banking system, that’s bad; but really that just means we should be imposing stricter regulations on the shadow banking system first (or simultaneously).

We don’t need to accept bank runs as a routine part of the financial system. There are other ways of doing things.

Is the cure for inflation worse than the disease?

pnrj current events, macroeconomics, politics, public policy , , , , , , , , , , ,

Nov 13 JDN 2459897

A lot of people seem really upset about inflation. I’ve previously discussed why this is a bit weird; inflation really just isn’t that bad. In fact, I am increasingly concerned that the usual methods for fixing inflation are considerably worse than inflation itself.

To be clear, I’m not talking about hyperinflation—if you are getting triple-digit inflation or more, you are clearly printing too much money and you need to stop. And there are places in the world where this happens.

But what about just regular, ordinary inflation, even when it’s fairly high? Prices rising at 8% or 9% or even 11% per year? What catastrophe befalls our society when this happens?

Okay, sure, if we could snap our fingers and make prices all stable without cost, that would be worth doing. But we can’t. All of our mechanisms for reducing inflation come with costs—and often very high costs.

The chief mechanism by which inflation is currently controlled is open-market operations by central banks such as the Federal Reserve, the Bank of England, and the European Central Bank. These central banks try to reduce inflation by selling bonds, which lowers the price of bonds and reduces capital available to banks, and thereby increases interest rates. This also effectively removes money from the economy, as banks are using that money to buy bonds instead of lending it out. (It is chiefly in this odd indirect sense that the central bank manages the “money supply”.)

But how does this actually reduce inflation? It’s remarkably indirect. It’s actually the higher interest rates which prevent people from buying houses and prevent companies from hiring workers which result in reduced economic growth—or even economic recession—which then is supposed to bring down prices. There’s actually a lot we still don’t know about how this works or how long it should be expected to take. What we do know is that the pain hits quickly and the benefits arise only months or even years later.

As Krugman has rightfully pointed out, the worst pain of the 1970s was not the double-digit inflation; it was the recessions that Paul Volcker’s economic policy triggered in response to that inflation. The inflation wasn’t exactly a good thing; but for most people, the cure was much worse than the disease.

Most laypeople seem to think that prices somehow go up without wages going up, but that simply isn’t how it works. Prices and wages rise at close to the same rate in most countries most of the time. In fact, inflation is often driven chiefly by rising wages rather than the other way around. There are often lags between when the inflation hits and when people see their wages rise; but these lags can actually be in either direction—inflation first or wages first—and for moderate amounts of inflation they are clearly less harmful than the high rates of unemployment that we would get if we fought inflation more aggressively with monetary policy.

Economists are also notoriously vague about exactly how they expect the central bank to reduce inflation. They use complex jargon or broad euphemisms. But when they do actually come out and say they want to reduce wages, it tends to outrage people. Well, that’s one of three main ways that interest rates actually reduce inflation: They reduce wages, they cause unemployment, or they stop people from buying houses. That’s pretty much all that central banks can do.

There may be other ways to reduce inflation, like windfall profits taxes, antitrust action, or even price controls. The first two are basically no-brainers; we should always be taxing windfall profits (if they really are due to a windfall outside a corporation’s control, there’s no incentive to distort), and we should absolutely be increasing antitrust action (why did we reduce it in the first place?). Price controls are riskier—they really do create shortages—but then again, is that really worse than lower wages or unemployment? Because the usual strategy involves lower wages and unemployment.

It’s a little ironic: The people who are usually all about laissez-faire are the ones who panic about inflation and want the government to take drastic action; meanwhile, I’m usually in favor of government intervention, but when it comes to moderate inflation, I think maybe we should just let it be.

What does a central bank actually do?

pnrj macroeconomics, public policy , , , , , , , , , , ,

Aug 26 JDN 2458357

Though central banks are a cornerstone of the modern financial system, I don’t think most people have a clear understanding of how they actually function. (I think this may be by design; there are many ways we could make central banking more transparent, but policymakers seem reluctant to show their hand.)

I’ve even seen famous economists make really severe errors in their understanding of monetary policy, as John Taylor did when he characterized low-interest-rate policy as a “price ceiling”.

Central banks “print money” and “set interest rates”. But how exactly do they do these things, and what on Earth do they have to do with each other?

The first thing to understand is that most central banks don’t actually print money. In the US, cash is actually printed by the Department of the Treasury. But cash is only a small part of the money in circulation. The monetary base consists of cash in vaults and in circulation; the US monetary base is about $3.6 trillion. The money supply can be measured a few different ways, but the standard way is to include checking accounts, traveler’s checks, savings accounts, money market accounts, short-term certified deposits, and basically anything that can be easily withdrawn and spent as money. This is called the M2 money supply, and in the US it is currently over $14.1 trillion. That means that only 25% of our money supply is in actual, physical cash—the rest is all digital. This is actually a relatively high proportion for actual cash, as the monetary base was greatly increased in response to the Great Recession. When we say that the Fed “prints money”, what we really mean is that they are increasing the money supply—but typically they do so in a way that involves little if any actual printing of cash.

The second thing to understand is that central banks don’t exactly set interest rates either. They target interest rates. What’s the difference, you ask?

Well, setting interest rates would mean that they made a law or something saying you have to charge exactly 2.7%, and you get fined or something if you don’t do that.

Targeting interest rates is a subtler art. The Federal Reserve decides what interest rates they want banks to charge, and then they engage in what are called open-market operations to try to make that happen. Banks hold reserves—money that they are required to keep as collateral for their loans. Since we are in a fractional-reserve system, they are allowed to keep only a certain proportion (usually about 10%). In open-market operations, the Fed buys and sells assets (usually US Treasury bonds) in order to either increase or decrease the amount of reserves available to banks, to try to get them to lend to each other at the targeted interest rates.

Why not simply set the interest rate by law? Because then it wouldn’t be the market-clearing interest rate. There would be shortages or gluts of assets.

It might be easier to grasp this if we step away from money for a moment and just think about the market for some other good, like televisions.

Suppose that the government wants to set the price of a television in the market to a particular value, say $500. (Why? Who knows. Let’s just run with it for a minute.)

If they simply declared by law that the price of a television must be $500, here’s what would happen: Either that would be too low, in which case there would be a shortage of televisions as demand exceeded supply; or that would be too high, in which case there would be a glut of televisions as supply exceeded demand. Only if they got spectacularly lucky and the market price already was $500 per television would they not have to worry about such things (and then, why bother?).

But suppose the government had the power to create and destroy televisions virtually at will with minimal cost.
Now, they have a better way; they can target the price of a television, and buy and sell televisions as needed to bring the market price to that target. If the price is too low, the government can buy and destroy a lot of televisions, to bring the price up. If the price is too high, the government can make and sell a lot of televisions, to bring the price down.

Now, let’s go back to money. This power to create and destroy at will is hard to believe for televisions, but absolutely true for money. The government can create and destroy almost any amount of money at will—they are limited only by the very inflation and deflation the central bank is trying to affect.

This allows central banks to intervene in the market without creating shortages or gluts; even though they are effectively controlling the interest rate, they are doing so in a way that avoids having a lot of banks wanting to take loans they can’t get or wanting to give loans they can’t find anyone to take.

The goal of all this manipulation is ultimately to reduce inflation and unemployment. Unfortunately it’s basically impossible to eliminate both simultaneously; the Phillips curve describes the relationship generally found that decreased inflation usually comes with increased unemployment and vice-versa. But the basic idea is that we set reasonable targets for each (usually about 2% inflation and 5% unemployment; frankly I’d prefer we swap the two, which was more or less what we did in the 1950s), and then if inflation is too high we raise interest rate targets, while if unemployment is too high we lower interest rate targets.

What if they’re both too high? Then we’re in trouble. This has happened; it is called stagflation. The money supply isn’t the other thing affecting inflation and unemployment, and sometimes we get hit with a bad shock that makes both of them high at once. In that situation, there isn’t much that monetary policy can do; we need to find other solutions.

But how does targeting interest rates lead to inflation? To be quite honest, we don’t actually know.

The basic idea is that lower interest rates should lead to more borrowing, which leads to more spending, which leads to more inflation. But beyond that, we don’t actually understand how interest rates translate into prices—this is the so-called transmission mechanism, which remains an unsolved problem in macroeconomics. Based on the empirical data, I lean toward the view that the mechanism is primarily via housing prices; lower interest rates lead to more mortgages, which raises the price of real estate, which raises the price of everything else. This also makes sense theoretically, as real estate consists of large, illiquid assets for which the long-term interest rate is very important. Your decision to buy an apple or even a television is probably not greatly affected by interest rates—but your decision to buy a house surely is.

If that is indeed the case, it’s worth thinking about whether this is really the right way to intervene on inflation and unemployment. High housing prices are an international crisis; maybe we need to be looking at ways to decrease unemployment without affecting housing prices. But that is a tale for another time.