A lot of commentary, including my own, is critical of the Federal Reserve. In the spirit of the season, I thought it would be good to say something nice for a change.
This post advertises is a Wall Street Journal OpEd, in praise of interest on reserves and lots of reserves. I can’t post the whole thing for 30 days, but here are excerpts and extra comments.
The issue: In 2008 the Fed started paying banks interest on reserve accounts that banks hold at the Fed. And the volume of those reserve accounts has shot up from less than $50 billion to $3,000 billion. The Fed controls interest rates by changing the rate it pays on reserves, and the rate it charges to borrow reserves.
In my view, this is a great and good thing. (Yes, I’ve written about these issues before, but putting things together in a different way is sometimes useful and part of the job.)
Why is it a good idea?
Milton Friedman described the “optimal quantity of money” as that situation in which money and short-term investments pay the same interest rate. Then banks, people and businesses hold a lot of cash, and waste less effort economizing on its use. Since the Fed can costlessly buy bonds and issue interest-paying money, providing such “liquidity” is free. So, provide it in abundance. Money is the oil in the economy’s engine, and it’s free. Fill ’er up.
Ample reserves are also great for financial stability. Bank runs happen when banks hold illiquid interest-bearing assets to back their deposits. When banks funnel deposits into reserves, a run can’t break out. All financial institutions holding lots of interest-paying cash more easily stay out of trouble.
****
A lot of the essay counters standard objections. One common objection is that if banks funnel deposits into reserves, they will have less money to lend. That’s a great question for an economics 101 midterm. No.
Banks holding lots of reserves don’t lend less. If the Fed buys Treasury bonds to create reserves, banks hold more in reserves and less in Treasurys. Money available for lending is the same.
The deepest issue among economists is whether inflation becomes uncontrollable under the new system. For that reason, a lot of economists think the Fed should go back to the old system: no interest on reserves, banks hold as little as they can get away with, and the Fed controls the supply of reserves rather than directly fix the interest rate. To raise the interest rate, the Fed reduces the quantity of reserves, and then banks pay each other higher interest to borrow scarce reserves. The higher interest rate is really just an indicator of a tighter supply of money, in this view, and if you don’t control the money supply you don’t control inflation. Contrariwise, when money pays the same interest as other investments, money supply is meaningless. The Fed has been setting interest rates for decades, but this Cheshire Cat of monetarism lingers. It lingers even in new-Keynesian models, which study a cashless limit but not the limit point. (I just discovered Mark Gertler’s excellent notes, where this underlying monetarism is clear.)
Even Milton Friedman, though he famously described this “optimum quantity of money,” did not advocate its implementation, either through an interest rate peg at zero, or by paying interest on reserves. Instead, he argued for a money growth rule. Why? I believe he thought the former would lead to unstable inflation. (This view is a little inconsistent with the monetarist view that zero interest is not a “liquidity trap,” and still has “stable velocity,” but I digress.) Of course, I like fiscal theory, and one reason is that it can determine the price level in an economy with no distinction between cash and bonds.
We did not really know what would happen. Now we do. 15 years of interest on reserves and an ample balance sheet worked:
Economists worried that paying interest on reserves would unhinge the price level by eliminating the separation between money and bonds. Many predicted inflation or deflation spirals. We learned that simply isn’t the case. Inflation trundled along from 2008-21 at 2% or so. Inflation spiked recently, but nobody thinks the Fed was primarily responsible. The Fed bought a lot of assets during the pandemic, but the same purchases had no effect in the 2010s. The cause of inflation was massive fiscal stimulus, not interest on ample reserves.
*****
The next section is more speculative but more novel, and I think there is some good research to be done here. (Thesis topic alert!)
Why did the Fed’s 5 percentage points of interest rate rise not produce a recession, or any softening of the economy? Why did inflation come down without that recession? We are witnessing yet another challenging data point!
The change in operating procedures is an intriguing possibility:
…back in the day, to raise interest rates the Fed reduced the quantity of reserves, and with that, via reserve requirements, the amount banks could lend and deposits people could hold. Scarce money and credit, arguably, caused the economy to tank. Now, raising interest rates has no such credit and quantitative effect. The Fed used to slow the car by draining oil. Now it just eases off the gas.
I’ve been exploring a new-Keynesian model with money, and then maybe with financial frictions. My hunch is that a higher interest rate lowers output more with these frictions than when money pays interest, is abundant, and there is no money multiplier. That would explain why 1980 was so much more painful than now. You’re welcome to scoop me on this if it’s right.
But in the standard story, the Fed raises rates, maybe lowers the money supply, the economy softens, and then via the Phillips curve inflation declines. If reducing money supply and lending are no longer direct consequences of Fed policy, maybe it also loses some power to control inflation? Maybe all the Fed had was the oil supply.
Well, that all depends on how you see the current easing of inflation. If you think the Fed’s rate rises were the central cause of inflation coming down, then you should really cheer:
…the new regime will have doubly proved itself, eliminating painful credit crunches and recessions as needless casualties of Fed action.
In my view,
the Fed has always had less control over inflation than most thought.
Inflation came from the pandemic fiscal blowout, and would have eased on its own. The Fed helps on the margin; raising rates lowers inflation now at the cost of a bit more future inflation. (See Fiscal Histories and Inflation Past Present and Future.)
Or, perhaps, you think that some remarkable speechifying and precommitment ability has made the Fed much more credible than it was in 1980, so we are seeing a classic Lucas-Sargent painless disinflation from rationally expected declining inflation expectations.
Whatever your explanation for why inflation comes down forever,
that doesn’t mean we need to return to the old way of deliberately inducing a credit crunch to control inflation
Which reason for inflation’s decline is true is another great question.
*****
Of course nothing is perfect.
The Fed still does not embrace its own innovation, and both allow and encourage banks to take deposits, park them in reserves, and thereby be completely immune from runs. Stop the war on narrow banks.
There is no economic reason for the Fed to limit the quantity of reserves. Many other central banks basically announce deposit and borrowing rates, and let banks have whatever they want at those prices.
A flat supply curve is a flat supply curve. The Fed’s efforts to limit the balance sheet has led to Treasury market disruption and other hiccups. Why?
Dallas Fed President Lorie Logan also gave an excellent speech about interest on / ample reserves, addressing a lot more financial plumbing issues. She also adds an interesting point that the Fed and banks need to get the plumbing in better order to borrow reserves.
The Fed offers money market funds access to reserves via a reverse repo program, but pays them a lower rate. There is no reason for a continuing subsidy to banks.
I grant many critic’s view that the asset side of the balance sheet poses some economic and larger political trouble. (Charles Plosser has been eloquent on this point.)
The Fed’s choice of which assets to buy, including long-term Treasurys and mortgage-backed securities, has had negative consequences. Shortening the maturity structure has made interest costs on the debt spike quickly, multiplying the Treasury’s awful decision to bet on low rates by borrowing short. Buying other assets distorts credit allocation.
All of these problems would be made easier
if the Treasury issued overnight, fixed-value, floating-rate debt directly. Then the Fed could worry only about immediate liquidity to the banking system, and not be in the business of providing the safe asset for the whole economy or managing interest rate risk for the federal government.
Here I am plugging again A New Structure for US Federal Debt. It won’t be the last time!
But these are little fixes, which just make a good system better.
Well done, Fed.
And, the new year will soon come and we can go back to complaining!
Long time, first time.
In this post and prior posts I’ve seen a repeated conflation of money and reserves. I’m wondering why you often write about them as if they are the same?
Whereas “reserves” are an asset of the banking system (narrowly defined as banks with a federal reserve account), “money” is representative of the short term liabilities of the financial sector (broadly defined). In this view they are distinct concepts.
If banks could park excess deposits with the FED and earn interest, why did so many banks pile into bonds when interest rates were low and now have unrealized losses on those bonds?