Public attention in the United States during the first phase of the COVID-19 crisis has been largely on the disease itself, the massive social and economic shock of the shutdown, and how we can orchestrate a safe reopening. But we also need to pay some attention to the financial side of the current situation, and the Federal Reserve’s immense reaction to it. Whatever one thinks of that reaction, it’s important to understand what the bank did, what beneficial and adverse consequences there are, and how our financial and economic system and policies might be set up better in the future.

We face a severe economic downturn of unknown duration. If it is something other than a V-shaped downturn spanning months rather than years, there will be a wave of bankruptcies, from individuals to corporations, and huge losses all over the financial system. “Well, earn returns in good times and take losses in bad times,” you may say, and I do, more often than the Fed does, but for now this is simply a fact.

Our government’s basic economic plan to confront this situation is simple: the Federal Reserve will print money to pay every bill, and guarantee every debt, for the duration. And, to a somewhat lesser approximation, the plan is also to ensure that no fixed-income investor loses money.

To be clear, my intention here is not to criticize this plan. From a combination of voluntary and imposed social distancing, the economy is collapsing. Twenty million people, more than 1 in 10 US workers, lost their jobs in the first month of the COVID-19 shutdowns. That’s more than the entire 2008–09 recession, all in the course of three weeks. A third of US apartment renters didn’t pay April rent. Run that up through the financial system: most guesses say that companies have one to three months of cash on hand, and then fail.

If you want to know why the Fed hit the panic button, it’s because every alarm went off.

Is the plan really to try to pay every bill? Yes, pretty much. This is not stimulus. It is “get-through-it-us.” People who lost jobs and businesses that have no income can’t pay their bills. When people run out of cash, they stop paying rent, mortgages, utilities, and consumer debts. In turn, the people who lent them money are in trouble. Businesses with zero income can’t pay debts, employees, rent, mortgages, or utilities either. When they stop paying, they go through bankruptcy, and their creditors get into trouble. If you want to stop a financial crisis, you have to pay all the bills, not just hand out some cash so people can buy food.

And that’s more or less the plan. There will be unemployment insurance, with 100 percent replacement of wages, for people who lose jobs, so they can pay rent, mortgages, utilities, and consumer debts. The Small Business Administration has made, and will continue to make, forgivable loans to businesses. Bailout plans are in place to make sure industrial companies such as airlines do not file for bankruptcy. (Much of this money is stuck in snafu, but that’s the plan if not the execution.) And, where the big money is, the Fed is propping up corporate bond markets, municipal bond markets, Treasury markets, money market funds, and other markets.

Is this really lending or just spending? Well, in the short run, it’s lending, but if the recession lasts more than a few months, it will turn in to spending.

Are they really printing money? Yes. Start with the Treasury. The Treasury authorized $2 trillion of spending in the first stimulus bill alone. Where is that money coming from? In normal times, that would mean selling $2 trillion in Treasury bonds and bills. But who has $2 trillion of extra income lying around that they want to use to buy Treasury debt? That’s a good question, to which we are not right now finding out the answer. Seeing ominous trouble in the Treasury market, the Fed is now buying all the debt that the Treasury is selling, and more.

When the Fed buys Treasury debt, it prints up new money and gives it to the holder of the Treasury debt. (I will say “printing money,” as that is clearer. The Fed actually creates new reserves, which are what banks hold in their accounts at the Fed, by flipping an electronic switch. Banks can convert reserves to cash and back at will.) On net, if the Treasury borrows and spends the money, and the Fed buys the Treasury debt, the government as a whole has printed up new money to spend. That’s what’s going on now.

From the March 4 and April 8 Fed H.4.1 data, we learn that the Fed increased its holdings of Treasury securities between those dates by $1,132 billion, from $2,502 billion to $3,634 billion. From Treasury data, we learn that debt held by the public (including the Fed) rose from $17,469 billion to $18,231 billion—a (huge) rise of $762 billion, or $9 trillion at an annual rate. The Fed bought all the Treasury debt, and then some, printing new money to do it. On net, the government financed the entire $762 billion by printing new money, and printed up another $370 billion to buy back existing Treasury debt.

The United Kingdom is abandoning pretenses. “Bank of England to directly finance UK government’s extra spending,” reads the April 9 Financial Times. Rather than have the government sell to the market, and then the bank buy it, the bank will now print money for the government to spend, and the government will print Treasury debt to give to the bank in return.

The Fed and Treasury are teaming up to provide trillions to businesses and banks, and to buy assets including money market funds, corporate bonds, municipal bonds, and mortgages. And the short answer to where those trillions are coming from is that the Fed is printing them up.

The Fed and Treasury, together but separate

In normal times, the Fed creates money (reserves) by buying Treasury bills. It has an asset, the T-bill, and a liability, the money. The money is backed by the T-bills, a good principle of noninflationary policy.

When the Fed lends money to a bank or a company, the Fed likewise prints up money, gives it to a company, and counts the company’s promise to pay back the loan as the corresponding asset. You can see the danger. The Fed is supposed to make only safe loans, to guard against inflationary finance and to keep itself politically independent. Printing money to hand gifts to well-connected companies and politically powerful interest groups is dynamite, and an independent agency will not stay independent long if it does so.

For this reason, the Fed and Treasury work together. The Treasury agrees to take the first tranche of losses, so the Fed can say this is a safe loan. Fed chair Jay Powell was, as usual, clear on this in an April 9 speech (delivered via Zoom, naturally) to the Brookings Institution:

I would stress that these are lending powers, not spending powers. The Fed is not authorized to grant money to particular beneficiaries. The Fed can only make secured loans to solvent entities with the expectation that the loans will be fully repaid.

What happens if the loans aren’t fully repaid? Well, the Treasury takes the first 5–10 percent of losses. But right now, the Treasury gets its money from the Fed. So it really comes back to printed money anyway. If losses are so severe that the Fed loses a lot of money, the Treasury will have to recapitalize the Fed with a gift of T-bills.

If the loans are not paid back, one way or another, we end up with that much more outstanding Treasury debt, either owned by the Fed, with corresponding reserves held by the public, or owned directly by people.

But this Fed versus Treasury business, while important inside baseball for Fed independence and the general function of the country’s economic plumbing, is really beside the point. The important point now is that the Fed and Treasury right now are, together, printing up trillions of dollars—$4 trillion to $6 trillion is the current guesstimate, which assumes a short, sharp recession—and handing it out. Most of it is “loans” that the Fed and Treasury hope to recoup. If they do, they can reduce the amount of money or government debt left outstanding.

Lending or spending?

As Powell emphasized, the current vision is that most of the support to date is lending, not spending. The Treasury kicks in something like $400 billion that really is spending: the anticipated loan losses (from companies that don’t survive) and forgiveness (programs that promise to forgive the loan if the company meets employment or other goals). The Fed lends $4 trillion on top of that, and gets its money back. The government as a whole has only spent $400 billion when it’s over, and the new debt (money) is soaked up again by repayment.

But is this really lending or just spending? Well, in the short run, it’s lending, but if the recession lasts more than a few months, it will turn in to spending.

Companies have no income but must pay rent, debts (interest on their corporate bonds and bank loans used to purchase now-idle plants and equipment), utilities, skeleton staff, etc. Local governments are in a similar bind. They borrow from the Fed to cover this cost. What’s wrong with that?

The sad lessons of a thousand years of government borrowing and its limits can be forgotten, but not erased.

Well, borrowing usually corresponds to a productive asset, to an increase in value. If a bakery borrows to buy an oven, the bakery will make more bread, and use the additional profits on the extra bread to pay off the loan. If it doesn’t work out, the oven is a real asset, collateral that the bank can sell to get some of its money back. A city borrowing to build a highway gets more tax revenue from greater activity, which helps to pay off the loan.

But there is no economic value to these loans. These are consumption loans, stay-afloat loans, preserve-the-business loans. They are loans against future profits, but not additional future profits. They are a transfer of the franchise value of the business to the lender—the Fed, in this case.

Clearly, at some point the business is better off shutting down than promising its entire profit stream to a lender just for the right to reopen someday. Furthermore, the government, already inclined to forgive, say, student debt, has every reason to forgive these “loans” as well. The business loans explicitly promise forgiveness if the business keeps workers on board. When we are in a sluggish recovery, and businesses are saying, “Well, I would hire more people, but we have all this extra debt because we took Fed loans to keep our employees fed while we were shut down,” let’s see just how tough the government is going to be on repayment.

So, in a matter of months, these loans turn into gifts. The $4 trillion Fed lending package winds up as $4 trillion permanently added to Treasury debt.

The threat of inflation

You would think that if the Fed and Treasury are going to print up something like $1 trillion a month to pay everyone’s bills and prop up markets for the duration, we would soon be heading for inflation.

But we won’t, or at least not immediately, because reserves pay interest. Reserves are just another form of Treasury debt. (Reserves that pay interest are one of the best innovations of recent decades, and kudos to former Fed chair Ben Bernanke and everyone else involved.)

With abundant, interest-paying reserves, reserves and Treasury debt are almost exactly the same thing. In roughly functional markets, what matters is their total supply, not reserves alone. Inflation is a danger, but from the total quantity of government debt, not its split between reserves and bills. Inflation comes, basically, if the US hits a debt crisis where people don’t want to hold or roll over US debt.

(That is, so long as the Fed pays market interest on reserves, and lets the market basically have as much or as few reserves as it wants. If the Fed, and the Treasury, start worrying about interest costs of the debt, and do not pay interest on reserves and do not allow people to convert to Treasurys, inflation will come sooner. )

Why does it matter that reserves pay interest? Couldn’t the Treasury just print up T-bills, sell them for reserves, hand out the reserves, collect loans in due time, and retire the Treasurys? In the short run, it matters for a rather disturbing reason: apparently the Treasury had a hard time finding willing buyers, so printing up the reserves directly and handing them out made a difference.

Instead, the Fed ends up with a loan asset on its balance sheet against reserves, rather than the Treasury having that loan as an asset on its balance sheet against T-bills. Conveniently, also, reserves—though equivalent to Treasury debt—are not counted in the debt limit along with many other contingent liabilities.

In the long run, though, it does not matter. The Fed and Treasury print up reserves, they lend to Joe’s Laundry, Joe pays his mortgage, and the mortgage company pays its investors. If those investors are happy sitting on reserves (bank accounts backed 1:1 with reserves on the margin), the reserves sit there. If they are not happy to sit on reserves, which would be the beginning of the inflationary process, the Fed can just raise the interest rate on reserves until they are happy to hold them, thereby really, really transforming reserves into Treasury debt. Or the Fed can give people some of its stock of Treasurys and so soak up unwanted reserves. That is, so long as people want Treasurys. If people don’t want Treasurys or reserves, if the US has to promise so much interest to get people to hold them that the budget is consumed by interest payments, we get inflation.

We’re looking, for sure, at raising US debt from $22 trillion to $27 trillion, likely hitting 150 percent of GDP if this is a short and swift recession. It could be much larger if the recession goes on a year or more. Is there a demand for that much more Treasury debt in the long run? Is there a flow of that much new saving that people are willing to park with Uncle Sam? How much more can markets take?

So the chance of a global sovereign-debt crisis and accompanying inflation is not zero—but not centrally because of the fact that recession relief efforts are currently financed by printing money.

How long can this go on?

As you can see, the viability of this whole plan depends on a short recession. As I noted earlier, the Fed is printing up something like $1 trillion per month. If the recession ends up being L-shaped, those numbers will ramp up as reservoirs of private cash dry up. A few large companies need bailouts, a few more “dysfunctional” markets turn to the Fed to buy everything, and so on. The International Monetary Fund wants $1.2 trillion to bail out emerging-market economies. State and local governments, already facing pension crises, will be toast when sales- and income-tax receipts collapse.

Where is the limit? Perhaps the peasants with pitchforks, remarkably absent so far, will revolt. Perhaps the willingness to hold interest-bearing reserves or US Treasury debt will find its limit after $10 trillion. Or $20 trillion. There is no magic. The sad lessons of a thousand years of government borrowing and its limits can be forgotten, but not erased.


John H. Cochrane is a senior fellow at the Hoover Institution at Stanford University and distinguished senior fellow at Chicago Booth. This essay is adapted from a post on his blog, The Grumpy Economist.

More from Chicago Booth Review

More from Chicago Booth

Your Privacy
We want to demonstrate our commitment to your privacy. Please review Chicago Booth's privacy notice, which provides information explaining how and why we collect particular information when you visit our website.