With the recent explosive growth in the Fed’s balance sheet, there’s been a lot of misleading discussion of the Fed “monetizing the debt”.
Debt monetization occurs when a central bank prints high-powered money and uses the funds to buy interest-earning assets such as Treasury debt. In the US, high-powered money is currency (not bank reserves). The Fed does issue some new currency each year, but the revenue from this “seignorage” is small relative to the national debt (even with the recent spike in currency demand).
Many people wrongly assume that issuance of bank reserves is another way of monetizing the debt. But using interest-earning bank reserves to buy back interest-earning Treasury debt is merely exchanging one government liability for another. You could get essentially the same impact by having the Treasury issue short-term T-bills and use the funds to buy back 30-year T-bonds. This action might be profitable, ex post, but then again it might also result in a loss. Either way, it’s clearly not debt monetization in the traditional sense of the term.
If the Fed targets inflation at 2%, then the stock of high-powered money is endogenous. In that case, the Fed basically has no control over the revenue from debt monetization. You might argue that the Fed doesn’t always hit its inflation target. But in recent years they’ve mostly been undershooting 2% inflation, which means they’ve done less debt monetization than is appropriate.
Indeed if you think the Fed’s 2% inflation target correctly represents the Congressional mandate of “stable prices”, then in a sense the Fed has been breaking the law by doing too little debt monetization. That’s hardly the impression one gets reading the newspapers these days.
Perhaps the Fed will eventually let inflation creep up to 3% or 4%. But even in that case the earnings from seignorage will be trivial relative to the national debt. Debt monetization is not an important issue in the US, and is not likely to be an issue going forward. The real problems lie elsewhere; there is too little NGDP growth, which contributes to high unemployment and financial stress. (Monetary policy is not the only factor right now, but it is a factor.)
READER COMMENTS
John P Palmer
Jun 18 2020 at 9:18am
Are you arguing that the increased bank reserves will all sit as excess reserves and not be used to extend new loans and create new money?
Scott Sumner
Jun 18 2020 at 12:04pm
No, I’m saying that interest bearing reserves don’t “monetize debt”. Monetizing debt means switching from borrowing money to paying for things with an inflation tax.
cove77
Jun 18 2020 at 11:04am
Scott can you comment on obsession with Yield Curve Control in current environment? Why shouldn’t FED promote/encourage a steeper YC ? Thanks
Scott Sumner
Jun 18 2020 at 8:39pm
I’d prefer they refrain from controlling interest rates. Target NGDP growth and let markets determine the appropriate level of interest rates.
B King
Jun 18 2020 at 11:37am
Scott, can you help me think through this – let’s say congress passes a $17T debt-financed package. That doubles the national debt. But let’s say the Fed increases its balance sheet accordingly and buys $17T in treasury securities. the amount of publicly held debt stays the same, and interest payments on the newly owned $17T in assets that go to the Fed get passed back to Treasury as ‘profit’ from the Fed.
What happens to money supply and inflation? what happens to interest rates on treasury securities?
Scott Sumner
Jun 18 2020 at 12:08pm
The effect on inflation and other variables depends on the interest rate paid on reserves. If it’s lower than market rates, then there is a bit of monetization, and inflation rises somewhat.
In the past, however, the IOR has been high enough that the inflation effect is pretty trivial.
But you are right in hinting at the fact that it’s more complicated than this post suggested, I just gave “first order effects”.
B King
Jun 18 2020 at 3:16pm
If the Fed doesn’t buy an equal level of assets in my example, then the amount of publicly held debt doubles. If interest rates on treasuries stayed the same, then interest payments go from 1.75% of GDP to 3.5% of GDP (slightly higher than peak of 3.1% in 1991). Presumably interest rates would increase though since that’s a much higher debt level that is not getting recirculated through the Fed.
However, if the Fed doesn’t buy more securities, wouldn’t inflation skyrocket? That’s your whole point usually about monetary offset. Fiscal stimulus leads to inflation leads to monetary offset, the economy stays at 2% inflation. So wouldn’t the Fed HAVE to purchase a similar level of treasuries?
Presumably interest rates still go up, but then aren’t you basically at Japan-level Debt/GDP without rampant inflation? Is that unsustainable?
Brian
Jun 18 2020 at 6:46pm
Quote… “if the Fed doesn’t buy more securities, wouldn’t inflation skyrocket”.
I think as long as the private sector wants to buy the securities then there is no strong and certain inflation implication.
If neither the Fed nor the private sector wants to buy the securities then the nominal yield has to rise until they get bought by the private sector. May be the inflation rate was already high in this scenario so the Fed does not see the need to buy the securities. If there was already deflation then the private sector would be tempted to buy the securities until the yield is low enough to make them stop buying but the Fed would continue to buy the securities.
Scott Sumner
Jun 18 2020 at 8:43pm
Actually, inflation is more likely to skyrocket if the Fed does buy the extra T-bonds.
B King
Jun 20 2020 at 1:21pm
Yes, thinking through it further I see I had that basic aspect backwards. Amateur mistake. Monetary offset would be selling bounds to take money out of the system. But you are still basically stuck with the increased interest costs then.
MJ
Jun 18 2020 at 1:12pm
Sort of opposite topic of this post, but it got me thinking about if the Fed announced something drastic like an increase in their inflation target to 4%. Suppose this caused yields to immediately spike and the Fed was forced to start selling off their balance sheet at a huge loss to stabilize inflation and the deficit increases on account of the reduced remittances to the Treasury. Is this a realistic outcome and does it maybe explain some of the opposition to level targeting?
Scott Sumner
Jun 18 2020 at 8:41pm
I think the Treasury would gain much more from the higher inflation than the Fed would lose. Most T-bonds are not held by the Fed.
Mike Sproul
Jun 18 2020 at 9:47pm
“But using interest-earning bank reserves to buy back interest-earning Treasury debt is merely exchanging one government liability for another.”
And issuing federal reserve notes to buy back interest-earning bank reserves would also be merely exchanging one government liability for another.
Scott Sumner
Jun 19 2020 at 3:58pm
Yes, but zero interest currency is not a close substitute. If it were, it would earn interest. That’s what makes it high-powered money.
Erik Price
Jun 20 2020 at 2:53pm
The Fed monetizes debt when it buys securities, which is happening now, in an unprecedented fashion: by purchasing directly from the Treasury and the secondary bond markets under its newly announced facilities. Am I missing something?
Comments are closed.