"Ironically, interest rates are probably more stable over the medium term if the central bank allows high frequency short run volatility as part of a regime that targets NGDP."
this concept of taking lots of short run micro actions to smooth out the target variable is familiar to anyone in STEM. it really is just a dynamic control systems problem.
this phenomenon is also self evident in portfolio rebalancing, delta hedging, and high frequency trading.
regulators are just not scientific thinkers despite the Fed's best efforts to be structured like a academic research institution.
In the analyst Victor Shvets’ interesting take, Arthur Burns, in having presided over the stagflationary 70s, haunts his successors for demonstrating that the better course is to be data-driven and behind the curve. While this means its policy moves tend to increase volatility, it’s an acceptable price to pay to avoid a comparison with Burns.
Without reforms to narrow banking and creating a single mandate, the Fed will continually face pressures to focus too much on interest rates. I look forward, with great alacrity, to your next post.
“Why does the quantity of reserves matter? Because the “price” at which banks trade their reserve balances, which in turn depends importantly on the total amount of reserves in the system, is the federal funds rate, which is the interest rate targeted by the Federal Open Market Committee (FOMC) in the implementation of monetary policy”
The money stock can never be properly managed by any attempt to control the cost of credit. Interest is the price of credit. The price of money is the reciprocal of the price level.
Milton Friedman had a k-percent money rule, and a nominal interest rate rule of very near zero percent. Friedman didn’t advocate targeting variable interest rates.
I have an idea for startup - It's a level targeting crypto currency.
In the beginning I plan to offer targeting the fixed exchange rate of USD. So it would be kind of semi-stable crypto currency. Later I plan to offer various targets including nominal GDP.
I plan to sell this to wide public as USDL that can be used as lender of last resort but also to startup cities to use as their own currency with target they select. I have been pitching this idea in tech circles including ycombinator but usually got response that it is a kind of a scam.
I am aware there is a misunderstanding from their side. What do you think about idea?
Paul Volcker was quoted in the WSJ in 1983 that the Fed: “as a matter of principle favors payment of interest on all reserve balances” … “on rounds of equity”. [sic]
You wonder about popular perceptions. Remunerating IBDDs reduces the real rate of interest. That alone is enough to send it into the archives. That's not interest rate targeting, that's asset price / housing price targeting.
When you look at the accomplishments of other Nobel Laureates, it's obvious that Sumner deserves the Nobel.
Here is Randal Quarles on the post-Covid mistake, illustrating many of the issues you covered in this post:
"This is maybe an eccentric belief. I don't think that it's universally shared or even widely shared on the FOMC. But my belief is that it's a separate element of the Fed religion that resulted in not moving in the fall, and it became clear that it was time to pivot to withdrawing accommodation, and it's a long-standing kind of Fed principle that you shouldn't step on the gas and the brake at the same time; meaning that you shouldn't be raising interest rates at the same time as you are still increasing the size of the balance sheet. And so we had decided that we needed to taper the balance sheet purchases. The lesson of the taper tantrum under Bernanke was that you've got to telegraph that well in advance. You've got to do that gradually, in order to avoid disrupting markets. And you have to have completed that before you can start raising interest rates so that you're not doing two conflicting things. So that was the sequencing.
With the benefit of hindsight, I think it's now clear that in circumstances such as obtained in the fall, where you had inflation printing at such a high level, unemployment that had fallen significantly below what we thought was an inflationary level, a clearly overheated demand, and the relative power, if you will, of interest rate policy versus balance sheet policy, I think that in those circumstances, for a limited period of time, to have said, "Yes, tapering the balance sheet has to happen over an extended period of time to avoid disrupting markets, but we're going to begin raising interest rates even before that taper is complete because now is the time to move more aggressively in the response to inflation," I think that would've been wise.
I'm a pilot, and I would often bore the FOMC with my analogies to flying. And one of the first thing you learn when you're a pilot is that you should always turn the rudder and the ailerons in sync so that the plane is moving in a smooth and coordinated turn when you're turning. And if you move the rudder and the ailerons in different directions, the plane will start to sink out of the sky. So usually that's a mistake, but there are occasions, like when you're trying to land the plane and you're a little too high over the runway, where you actually do that. You move the rudder and the ailerons in different directions so the plane slips into the right glide path, and then you re-coordinate and land. And I think that last fall would've been a time for the Fed to slip policy down onto the right glide path by beginning to raise interest rates even before the balance sheet taper was complete."
Yes, I believe NGDP can be targeted without IOR, and it would work well most of the time—except when IOR is needed the most. Financial crises might still occur under NGDP targeting, as the financial sector would grow until risks reemerge. During these periods of stress, IOR would play a crucial role in limiting financial instability, as well as reducing the probability that the actual NGDP would deviate too much from the market forecast to one side or another.
BTW, Singapore does pay IOR, but IOR changes every day.
I disagree about the taper tantrum. The problem was not that it wasn't telegraphed far in advance (which is impossible in an efficient regime that is data dependent), the problem was that it was a bad policy--the economy needed stimulus. I don't understand why IOR is needed in a financial crisis--just inject the needed liquidity.
Without IOR, central banks would have large balance sheets only at the zero lower bound. But what if a financial crisis occurs when NGDP expectations are well anchored and interest rates are well above zero? Consider Poland in 2008–2009. In Q4 2008, NGDP growth was 5.9% year-over-year, and in Q4 2009, it was 6.5%. The IOR was reduced to 3.5% by the end of 2008 and further to 2% by mid-2009. By paying IOR, the central bank of Poland was able to support the banking sector during a period of financial stress while ensuring that NGDP did not accelerate too much.
The Polish zloty depreciated significantly against the euro during this time (EUR/PLN was 3.20 on July 1, 2008, and 4.90 on February 18, 2009), creating problems for households with Swiss franc–denominated mortgages. A counterfactual policy without IOR could have stabilized NGDP as well but at the cost of even higher exchange rate volatility, the lack of IOR would have delivered a negative aggregate supply shock to the Polish economy.
BTW, monetary policy became too tight in Poland in Q4 2009.
In "Moving Toward NGDP Targeting" Scott Sumner cites a recent Discourse column by Patrick Horan where Horan “made the case for a more flexible interest rate targeting regime:”
I don’t understand Sumner’s thesis. Supposedly the Fed does NOT target interests rates; it targets inflation, specifically, Flexible Average Inflation Targeting, FIAT. Horan seems to be talking about _instrument setting_, not _targeting? _” The idea of smaller more frequent instrument setting with no expectation that movement of an instrument in one direction in period _t_ implies further moment in the same direction in period _t+1_, is hardly new, whatever the target!
Not paying interest on reserves (IOR) is sensible if the principal policy instrument is movement in the EFFR. The two have similar effects and can easily interfere with each other if not very finely tuned jointly. Who needs that hassle? But what does use or non-use of IOR have to do with whether NGDP or inflation is targeted?
“The use of IOR gives the impression that interest rates are equivalent to monetary policy. People begin to believe that higher market interest rates represent tighter monetary policy, which is not necessarily true.”
“Give the impression?” A movement in a policy instrument is a movement in a policy instrument. People presumably have the same understanding of how movements in an instrument affects the target variable as the Fed does in moving the instrument.
“I suppose you could adjust IOR up and down each day as a tool to control NGDP”
Yes. The Fed could use either setting the EFFR or IOR to control either inflation or NGDP. (Or open market operations.) But significance of the choice of instrument on the choice of target is not explained?
“During late 2008, the Fed was distracted from stabilizing NGDP by a misguided belief that they needed to proceed methodically in cutting interest rates.”
The Fed was “distracted” (or something) from keeping inflation up to target. The Fed has never explained why it allowed inflation and inflation expectations to go negative in 2009 and then flailed even to return and keep inflation up to target 2009-2020. True it was slow to drop the EFFR to near zero in 2008, but tat was not the totality of its error.
“In late 2021 and in 2022, the Fed raised rates much too slowly, again based on the mistaken view that they needed to move methodically. You don’t need to stabilize interest rates; you need to stabilize NGDP expectations. Ironically, interest rates are probably more stable over the medium term if the central bank allows high frequency short run volatility as part of a regime that targets NGDP. That’s because stable NGDP growth tends to also stabilize the natural rate of interest.”
I agree with this criticism of the management of the EFFR but it is not more relevant to targeting NGDP than to targeting inflation. TIPS _inflation_ expectations in September 2021 flashed substantially over-target inflation.
“To summarize, abolishing IOR and moving to a highly flexible fed funds rates doesn’t magically get you to NGDP targeting. But it’s a nice first step, which creates an environment where targeting NGDP market expectations seems like the natural next step.”
It is a nice first step to making inflation rate targeting more effective
*** ***
Sumner’s entire post could be re-written by copy and pasting “NGDP” with “inflation” and it would remain equally coherent. And the advantage cited for NGDP targeting over interest rate “targeting” (the current protocols for managing the EFFR) sems to apply equally to inflation rate targeting. I would pose the following question;
1. Why NGDP targeting rather than inflation rate targeting?
2. What is the operational difference? Which instruments are moved differently on the basis of what different information?
3. What would the difference be under the two regimes in response to:
"Why NGDP targeting rather than inflation rate targeting"
Given the Fed's dual mandate, an inflation target is not optimal, whereas an NGDP target is pretty close to optimal. If you do inflation targeting, it will inevitably get complicated, as the Fed also pursues employment goals.
What if he Fed kept interest rates "too high" but constantly built its balance sheet? That is, the Bank of Japan approach until the C19 pandemic. Seems to work.
So obviously right that I can't believe it's not happening. The question is how to get the old dog to learn the new tricks. Are there baby steps to recommend?
(Been reading you since '08 and am still befuddled why you don't have a Nobel.) Gold arn it.
Scott, you should do a post on why Bernie Fraser was the best central banker ever. He started the longest ever recorded period with no technical recessions by creating a clear policy regime that worked and his successors could keep operating.
BTW, The Russian Government has published a list of immoral countries that "impose destructive neoliberal ideological frameworks, which contradict traditional Russian spiritual and moral values". Australia tops the list. https://x.com/EerikNKross/status/1837499200436650418
"Ironically, interest rates are probably more stable over the medium term if the central bank allows high frequency short run volatility as part of a regime that targets NGDP."
this concept of taking lots of short run micro actions to smooth out the target variable is familiar to anyone in STEM. it really is just a dynamic control systems problem.
this phenomenon is also self evident in portfolio rebalancing, delta hedging, and high frequency trading.
regulators are just not scientific thinkers despite the Fed's best efforts to be structured like a academic research institution.
"dynamic control systems problem"
Milton Friedman used to use the thermostat analogy.
Brad DeLong harps on this Fed failure!
In the analyst Victor Shvets’ interesting take, Arthur Burns, in having presided over the stagflationary 70s, haunts his successors for demonstrating that the better course is to be data-driven and behind the curve. While this means its policy moves tend to increase volatility, it’s an acceptable price to pay to avoid a comparison with Burns.
Are there any members of the Federal Reserve that would be amenable to targeting NGDP?
I see getting there as a gradual process. My next post will discuss a few first steps.
Why would they wish to switch from inflation targeting to NGDP targeting?
Without reforms to narrow banking and creating a single mandate, the Fed will continually face pressures to focus too much on interest rates. I look forward, with great alacrity, to your next post.
Yeah, the FED conflates the proper volume of money with the right level of policy rates:
“when the demand for reserves is relatively strong such that borrowers are willing to pay an EQUILIBRIUM INTEREST RATE”
Bank Reserves since the Start of Quantitative Tightening | St. Louis Fed (stlouisfed.org)
They act as if the “demand for reserves” is inseparable from interest rates::
link to libertystreeteconomics.newyorkfed.org
“Why does the quantity of reserves matter? Because the “price” at which banks trade their reserve balances, which in turn depends importantly on the total amount of reserves in the system, is the federal funds rate, which is the interest rate targeted by the Federal Open Market Committee (FOMC) in the implementation of monetary policy”
The money stock can never be properly managed by any attempt to control the cost of credit. Interest is the price of credit. The price of money is the reciprocal of the price level.
Milton Friedman had a k-percent money rule, and a nominal interest rate rule of very near zero percent. Friedman didn’t advocate targeting variable interest rates.
I have an idea for startup - It's a level targeting crypto currency.
In the beginning I plan to offer targeting the fixed exchange rate of USD. So it would be kind of semi-stable crypto currency. Later I plan to offer various targets including nominal GDP.
I plan to sell this to wide public as USDL that can be used as lender of last resort but also to startup cities to use as their own currency with target they select. I have been pitching this idea in tech circles including ycombinator but usually got response that it is a kind of a scam.
I am aware there is a misunderstanding from their side. What do you think about idea?
If you, or anyone else here is interested, or want to discuss, please reach out to me at aleksandar.brajkovic.personal@gmail.com
Thanks!
Paul Volcker was quoted in the WSJ in 1983 that the Fed: “as a matter of principle favors payment of interest on all reserve balances” … “on rounds of equity”. [sic]
You wonder about popular perceptions. Remunerating IBDDs reduces the real rate of interest. That alone is enough to send it into the archives. That's not interest rate targeting, that's asset price / housing price targeting.
When you look at the accomplishments of other Nobel Laureates, it's obvious that Sumner deserves the Nobel.
Here is Randal Quarles on the post-Covid mistake, illustrating many of the issues you covered in this post:
"This is maybe an eccentric belief. I don't think that it's universally shared or even widely shared on the FOMC. But my belief is that it's a separate element of the Fed religion that resulted in not moving in the fall, and it became clear that it was time to pivot to withdrawing accommodation, and it's a long-standing kind of Fed principle that you shouldn't step on the gas and the brake at the same time; meaning that you shouldn't be raising interest rates at the same time as you are still increasing the size of the balance sheet. And so we had decided that we needed to taper the balance sheet purchases. The lesson of the taper tantrum under Bernanke was that you've got to telegraph that well in advance. You've got to do that gradually, in order to avoid disrupting markets. And you have to have completed that before you can start raising interest rates so that you're not doing two conflicting things. So that was the sequencing.
With the benefit of hindsight, I think it's now clear that in circumstances such as obtained in the fall, where you had inflation printing at such a high level, unemployment that had fallen significantly below what we thought was an inflationary level, a clearly overheated demand, and the relative power, if you will, of interest rate policy versus balance sheet policy, I think that in those circumstances, for a limited period of time, to have said, "Yes, tapering the balance sheet has to happen over an extended period of time to avoid disrupting markets, but we're going to begin raising interest rates even before that taper is complete because now is the time to move more aggressively in the response to inflation," I think that would've been wise.
I'm a pilot, and I would often bore the FOMC with my analogies to flying. And one of the first thing you learn when you're a pilot is that you should always turn the rudder and the ailerons in sync so that the plane is moving in a smooth and coordinated turn when you're turning. And if you move the rudder and the ailerons in different directions, the plane will start to sink out of the sky. So usually that's a mistake, but there are occasions, like when you're trying to land the plane and you're a little too high over the runway, where you actually do that. You move the rudder and the ailerons in different directions so the plane slips into the right glide path, and then you re-coordinate and land. And I think that last fall would've been a time for the Fed to slip policy down onto the right glide path by beginning to raise interest rates even before the balance sheet taper was complete."
Yes, I believe NGDP can be targeted without IOR, and it would work well most of the time—except when IOR is needed the most. Financial crises might still occur under NGDP targeting, as the financial sector would grow until risks reemerge. During these periods of stress, IOR would play a crucial role in limiting financial instability, as well as reducing the probability that the actual NGDP would deviate too much from the market forecast to one side or another.
BTW, Singapore does pay IOR, but IOR changes every day.
I disagree about the taper tantrum. The problem was not that it wasn't telegraphed far in advance (which is impossible in an efficient regime that is data dependent), the problem was that it was a bad policy--the economy needed stimulus. I don't understand why IOR is needed in a financial crisis--just inject the needed liquidity.
Without IOR, central banks would have large balance sheets only at the zero lower bound. But what if a financial crisis occurs when NGDP expectations are well anchored and interest rates are well above zero? Consider Poland in 2008–2009. In Q4 2008, NGDP growth was 5.9% year-over-year, and in Q4 2009, it was 6.5%. The IOR was reduced to 3.5% by the end of 2008 and further to 2% by mid-2009. By paying IOR, the central bank of Poland was able to support the banking sector during a period of financial stress while ensuring that NGDP did not accelerate too much.
The Polish zloty depreciated significantly against the euro during this time (EUR/PLN was 3.20 on July 1, 2008, and 4.90 on February 18, 2009), creating problems for households with Swiss franc–denominated mortgages. A counterfactual policy without IOR could have stabilized NGDP as well but at the cost of even higher exchange rate volatility, the lack of IOR would have delivered a negative aggregate supply shock to the Polish economy.
BTW, monetary policy became too tight in Poland in Q4 2009.
In "Moving Toward NGDP Targeting" Scott Sumner cites a recent Discourse column by Patrick Horan where Horan “made the case for a more flexible interest rate targeting regime:”
I don’t understand Sumner’s thesis. Supposedly the Fed does NOT target interests rates; it targets inflation, specifically, Flexible Average Inflation Targeting, FIAT. Horan seems to be talking about _instrument setting_, not _targeting? _” The idea of smaller more frequent instrument setting with no expectation that movement of an instrument in one direction in period _t_ implies further moment in the same direction in period _t+1_, is hardly new, whatever the target!
https://thomaslhutcheson.substack.com/p/improving-fed-decisions
Not paying interest on reserves (IOR) is sensible if the principal policy instrument is movement in the EFFR. The two have similar effects and can easily interfere with each other if not very finely tuned jointly. Who needs that hassle? But what does use or non-use of IOR have to do with whether NGDP or inflation is targeted?
“The use of IOR gives the impression that interest rates are equivalent to monetary policy. People begin to believe that higher market interest rates represent tighter monetary policy, which is not necessarily true.”
“Give the impression?” A movement in a policy instrument is a movement in a policy instrument. People presumably have the same understanding of how movements in an instrument affects the target variable as the Fed does in moving the instrument.
“I suppose you could adjust IOR up and down each day as a tool to control NGDP”
Yes. The Fed could use either setting the EFFR or IOR to control either inflation or NGDP. (Or open market operations.) But significance of the choice of instrument on the choice of target is not explained?
“During late 2008, the Fed was distracted from stabilizing NGDP by a misguided belief that they needed to proceed methodically in cutting interest rates.”
The Fed was “distracted” (or something) from keeping inflation up to target. The Fed has never explained why it allowed inflation and inflation expectations to go negative in 2009 and then flailed even to return and keep inflation up to target 2009-2020. True it was slow to drop the EFFR to near zero in 2008, but tat was not the totality of its error.
“In late 2021 and in 2022, the Fed raised rates much too slowly, again based on the mistaken view that they needed to move methodically. You don’t need to stabilize interest rates; you need to stabilize NGDP expectations. Ironically, interest rates are probably more stable over the medium term if the central bank allows high frequency short run volatility as part of a regime that targets NGDP. That’s because stable NGDP growth tends to also stabilize the natural rate of interest.”
I agree with this criticism of the management of the EFFR but it is not more relevant to targeting NGDP than to targeting inflation. TIPS _inflation_ expectations in September 2021 flashed substantially over-target inflation.
“To summarize, abolishing IOR and moving to a highly flexible fed funds rates doesn’t magically get you to NGDP targeting. But it’s a nice first step, which creates an environment where targeting NGDP market expectations seems like the natural next step.”
It is a nice first step to making inflation rate targeting more effective
*** ***
Sumner’s entire post could be re-written by copy and pasting “NGDP” with “inflation” and it would remain equally coherent. And the advantage cited for NGDP targeting over interest rate “targeting” (the current protocols for managing the EFFR) sems to apply equally to inflation rate targeting. I would pose the following question;
1. Why NGDP targeting rather than inflation rate targeting?
2. What is the operational difference? Which instruments are moved differently on the basis of what different information?
3. What would the difference be under the two regimes in response to:
a. A positive sectoral supply shock?
b. A positive sectoral demand shock?
c. A negative sectoral supply shock?
d. A negative sectoral demand shock?
"Why NGDP targeting rather than inflation rate targeting"
Given the Fed's dual mandate, an inflation target is not optimal, whereas an NGDP target is pretty close to optimal. If you do inflation targeting, it will inevitably get complicated, as the Fed also pursues employment goals.
NGDP targeting, good. End interest on reserves.
Should the Fed sell off its balance sheet?
What if he Fed kept interest rates "too high" but constantly built its balance sheet? That is, the Bank of Japan approach until the C19 pandemic. Seems to work.
So obviously right that I can't believe it's not happening. The question is how to get the old dog to learn the new tricks. Are there baby steps to recommend?
(Been reading you since '08 and am still befuddled why you don't have a Nobel.) Gold arn it.
"Are there baby steps to recommend?"
My next post.
"am still befuddled why you don't have a Nobel"
I consumed way too much lead as a child---took 5 or 10 points off my IQ. :)
Scott, you should do a post on why Bernie Fraser was the best central banker ever. He started the longest ever recorded period with no technical recessions by creating a clear policy regime that worked and his successors could keep operating.
Stephen Kirchner would be the one to do that post. Maybe he already has.
BTW, The Russian Government has published a list of immoral countries that "impose destructive neoliberal ideological frameworks, which contradict traditional Russian spiritual and moral values". Australia tops the list. https://x.com/EerikNKross/status/1837499200436650418
Australia should be honored to top that list.
Oh yes. I certainly feel so. Indeed, I have said so on FB, X and Substack Notes.
https://substack.com/@lorenzofromoz/note/c-69857796