Black holes and Neo-Fisherites are a monetary phenomenon

Suppose you live in a Neo-Wicksellian economy, where the central bank sets a nominal rate of interest. Black holes are a theoretical possibility. Nominal demand for goods can spiral down to zero, if people expect it to. And white explosions are a theoretical possibility too. Nominal demand for goods can spiral up to infinity, if people expect it to. And Neo-Wicksellian theory says that the only thing that can prevent those things happening is a wise central bank that adjusts the nominal interest rate in the right way to prevent people expecting those things to happen.

But what is happening to the nominal stock of base money if the economy sprials down into a black hole? And what is happening to the nominal stock of base money if the economy spirals up into a white explosion? Neo-Wicksellian models are silent on that question. Because they think that the stock of money is unimportant. Which is very strange, because Neo-Wicksellian models are models of monetary exchange economies, where people buy and sell all other goods for money. If they were models of barter economies, the unemployed workers who cannot sell their labour, and the firms who cannot sell their goods, would all get together and barter their way back to full employment. Black holes would be impossible if barter were easy. I would swap my unsold apples for your unsold bananas.

So money must be in the Neo-Wicksellian model implicitly, even if it's not there explicitly.

And according to Neo-Wicksellian models, the danger of a black hole is greatest when the central bank hits the Zero Lower Bound on nominal interest rates. And that ZLB exists because it is hard in practice for central banks to set a negative nominal interest rate on currency. So any Neo-Wicksellian model with a ZLB must have currency in the model implicitly, even if it's not there explicitly.

Suppose a Neo-Wicksellian central bank hits the ZLB. If it lets the stock of base money spiral down to zero too, then of course the economy would spiral down into a black hole. But central banks don't have to do this. And central banks did not in fact do this, when they hit the ZLB. They did the opposite. Which is why their economies did not spiral down into a black hole.

If the economy did spiral into a black hole, I would buy the whole world with the $20 note in my pocket. Why wait? Except that you and another guy would see this coming, and would jump in before me, and each buy half the world with the $20 notes in your pockets. Except that I and three other guys would see this coming, and would jump in before you, and each buy a quarter of the world. And so on. The black hole theory comes apart backwards, as long as we still have some currency in our pockets.

Nominal demand can fall, but it can't fall to zero, unless the stock of base money falls to zero too. And central banks can stop it falling to zero, ZLB or not.

Neo-Wicksellian models include money implicitly. But we need to introduce money explicitly if we want to understand what happens at the ZLB. But then why not go the whole hog and include money explicitly, whether or not we are at the ZLB?

Take a related question -- the Neo-Fisherite question. If the central bank permanently raises the nominal interest rate, will this result in higher or lower inflation? If you tell me what permanently raising the nominal interest rate does to the base money supply growth rate, I can answer your question. If it causes the base money growth rate to increase permanently, like in John Cochrane's model, then inflation will increase. If it causes base money growth to decrease, then inflation will decrease. Tell me how the central bank raises interest rates, and what it is doing with base money growth when it does this.

So why not just change the question? Ask what happens to inflation and nominal interest rates if the central bank permanently increases the money growth rate? Inflation and nominal interest rates will eventually increase too. What happens to inflation and nominal interest rates immediately is a little more complicated. It depends on whether prices and inflation are sticky, and on how quickly expectations adjust and people learn that the increase is permanent.

If a question is too hard to answer clearly, that probably means you are asking the wrong question.

The Neo-Wicksellian perspective, in which central banks are seen as setting a nominal interest rate, is what leads us to ask those badly-posed questions. And since central banks' communications strategy is central to the success of monetary policy, central banks should stop doing things that cause people to ask badly-posed questions about what central banks are doing and what the results will be.

That is the real world policy implication of all this angels-on-pins talk about black holes and the Fisher relation.

[I'm just riffing off, or ripping off, Scott Sumner's very good post. Which in turn was a response, like this is, to John Cochrane's very good post. But I really must stop blogging about this stuff. Time to get out and do some last yard work, before the snow.]

Comments

[JKH]

Nick,

I have a very much half-formed comment on the Cochrane paper and how your version of things compares with that.

It seems to me that Cochrane uses the fiscal theory of the price level and you use a monetarist argument to arrive at the same sort of conclusion.

The first case is flexible prices.

As I understand it, the fiscal theory of the price level says that the nominal value of government debt deflated by the price level equals the discounted present value of future real surpluses (no doubt somebody will correct me on that phrasing).

Cochrane goes through process of iterating that equation through time to show that the government can set the nominal interest rate simply by selling enough government debt to satisfy that equation over the full time period under consideration. Because he holds both future real surpluses and the real interest rate fixed, that forces out a quantity of debt to be sold. The way I think of debt in the case of his model is to visualize it as if it consisted of Treasury bills – which are pure discount instruments with a maturity face value that are sold at a present discounted price (visualizing his math otherwise gets a little difficult for those of us who are used to thinking in terms of bonds with coupons attached). So the quantity of face value debt to be sold is forced out by the assumption of fixed future surpluses, a fixed real interest rate which is used to discount the value of the debt, and a chosen interest rate target.

You go through your process of assuming that money created by the government grows at the same rate as the rate of interest on reserves, which is equivalent to growing money by paying interest on reserves with new money (and where that is the only new money that is created).

And I think you both have about the same view that reserves with interest can be treated the same as debt for this purpose.

But where I see the difference is that Cochrane simply folds the case of reserves with interest into the more general case of debt. And he does this without any reference that I can see to the aspect of reserves or money necessarily being used as the medium of exchange, the way you would as the assumption behind the monetarist argument. So it looks to me like his treatment of interest on reserves is just a subset of his treatment of debt. And he can do this because it all pops out of the fiscal theory of the price level.

Moreover, his emphasis and in fact the emphasis of the fiscal theory of the price level is very much an interest rate argument – whereas yours is a money argument. So it’s interesting that he is arguing from the standpoint of a view of monetary policy that is focused on interest rates rather than money – which is something you don’t particularly like. All of his arguments as he goes through the paper use a sort of interest rate arbitrage effect in forcing out various results according to the way in which the real interest rate operates in the FTPL equation.

Could we say that the fiscal theory of the price level itself is very much interest rate focused rather than money focused?

I’ll leave it there without trying to go into his treatment of sticky prices – except that the ("endogenous") adjustment in the real interest rate seems to be very much a critical factor in that case for him.

[Nick Rowe]

JKH: "It seems to me that Cochrane uses the fiscal theory of the price level and you use a monetarist argument to arrive at the same sort of conclusion."

Correct. For this particular policy experiment, where new money is paid as interest on old money, you get the exact same prediction from both theories.

"As I understand it, the fiscal theory of the price level says that the nominal value of government debt [including base money] deflated by the price level [is] equal [to and determined by] the discounted present value of future real [primary] surpluses (no doubt somebody will correct me on that phrasing)."

Yes. [Trivial clarifications done.]

"The way I think of debt in the case of his model is to visualize it as if it consisted of Treasury bills – which are pure discount instruments with a maturity face value that are sold at a present discounted price..."

Yes. That's what he intends. (I visualise it as Canada Savings Bonds, except they have a floating interest rate set daily and can be cashed in for face value at any time, because it's slightly easier for me to get my head around it that way, but it makes very little difference, if the time period is short.)

"You go through your process of assuming that money created by the government grows at the same rate as the rate of interest on reserves, which is equivalent to growing money by paying interest on reserves with new money (and where that is the only new money that is created)."

Yes. But if BoC currency exists, I'm assuming the BoC is also paying interest on currency at the same rate as interest on reserves. (I'm ignoring the practical difficulties.)

"But where I see the difference is that Cochrane simply folds the case of reserves with interest into the more general case of debt. And he does this without any reference that I can see to the aspect of reserves or money necessarily being used as the medium of exchange, the way you would as the assumption behind the monetarist argument. So it looks to me like his treatment of interest on reserves is just a subset of his treatment of debt. And he can do this because it all pops out of the fiscal theory of the price level."

Yes. And I would disagree with him on that, but I stayed silent on that question, because it makes no difference to my results **for this particular policy of paying new base money as interest on old base money)**.

"Could we say that the fiscal theory of the price level itself is very much interest rate focused rather than money focused?"

No. (And you lost me a bit on the previous paragraph)

The big difference is this: FTPL (implicitly) assumes that base money and bonds pay the same nominal rate of interest, and that the real rate of interest on money is determined exogenously by the market. So the r used in the PV calculation is exogenous.

I assume that base money is an imperfect substitute for other assets, and that there is a negative relationship (a downward sloping demand curve) between M/P and (Rb-Rm). Where Rb is the real rate of return on other assets, Rm is the real rate of return on holding money, and so (Rb-Rm) is the opportunity cost of holding money. And if you look at hyperinflations, where Rm is very very negative, while Rb is positive, but M/P, while small, is not zero, you can see the point. FTPL assumes a permanent liquidity trap.

This matters because:

1. If Rb and Rm are different, you can't do the PV(surpluses) calculation without knowing the composition of government liabilities between money and bonds.

2. Assuming there are no bonds (just to keep it simple) and all deficits are money-financed, if Rm < 0 (which it is, with 2% inflation and no interest on currency) the PV(surpluses) calculation literally does not add up. The PV of a future surplus gets bigger and bigger the further into the future it is. The sum does not converge, as time goes to infinity. Actually, the thing goes negative. You are talking about permanent primary deficits. You can finance a permanent deficit by printing negative real interest money. Even more if the economy is growing so the demand for M/P is rising over time.

And that is where FTPL really falls apart. FTPL only works in a permanent liquidity trap, where money pays the same rate of return as all other assets, forever.

[Nick Rowe]

But if you print new money to pay interest on old money, Rm is unchanged (because the increased nominal interest is exactly offset by increased inflation, just like in a stock-split), so PV(surpluses) is unchanged, and (Rb-Rm) is unchanged, so there is no difference between the predictions of FTPL and the QT, in this particular case.

[Nick Rowe]

I say that "D/P = PV(surpluses)" works (sometimes) if D is bonds, but doesn't work if D is money.

FTPL says it works for both bonds and money (because money is like bonds). MMT says it works for neither bonds nor money (because bonds are like money).

[Kevin Donoghue]

"And what is happening to the nominal stock of base money if the economy spirals up into a white explosion? Neo-Wicksellian models are silent on that question."

If the money supply is endogenous, as it must be if the central bank sets the nominal rate of interest, it explodes just like the price level. At least that's how I read Jordi Gali, whose textbook isn't silent on that question at all.

"Black holes would be impossible if barter were easy. I would swap my unsold apples for your unsold bananas."

All manner of things would be different if barter were easy. Similarly, the tourist industry would be transformed if we could build Hilbert's Hotel. It's pointless to speculate about such things.

[Nick Rowe]

Kevin: you misunderstood me about barter. Of course barter isn't easy. That's why we use money. And that's why bad monetary policy causes recessions. But Neo-Wicksellian models must be implicitly assuming that barter is very hard, and therefore implicitly assuming the existence of money. Nothing wrong with those assumptions, of course. But money needs to be made implicit.

Sure, we can bolt on a money demand function to a Neo-Wicksellian model. And make M demand-determined. And if we do that it will implode or explode along with nominal demand. But central banks don't have to let M do that. And they did not in fact let M do that when they hit the ZLB. Which is why the economy did not implode.

[Edward Lambert]

Are permanent changes in the growth of total base money the determining factor for inflation?
Wouldn't it be better to focus on the portion of base money that is or is not circulating? and why?

[Jason Smith]

I believe -- correct me if I'm wrong -- the reason the ZLB implicitly assumes money is because money is a 0% nominal interest rate asset. If you think the ZLB is important then you're assuming money.

http://informationtransfereconomics.blogspot.com/2014/11/more-goodness-from-nick-rowe.html

However, I really like the idea that you are implicitly assuming money if you say that an economy exists because that implicitly assumes you've overcome the difficulty of barter (or whatever non-monetary economic system).

[JW Mason]

Isn't this just Patinkin 1948?

[Brad DeLong]

How can a post be good if it is confused about--nay, if it works deliberately to confuse others about-- the extent to which interest rates and base money quantities are sufficient statistics for monetary policy?

Now I'm confused...

[Lord]

Didn't we try monetarism in the early 80s only to see it fail? It may have been more a measurement problem, but still.

[JKH]

Nick,

"The big difference is this: FTPL (implicitly) assumes that base money and bonds pay the same nominal rate of interest, and that the real rate of interest on money is determined exogenously by the market. So the r used in the PV calculation is exogenous."

Probing again:

I don't know much about FTPL, but from the way his paper goes, I would have assumed that money is included only because it pays interest and the fact that it pays interest makes it equivalent to debt, which therefore fits in with bonds and the whole IOR scenario within his basic valuation equation.

And I would have thought for the same reason that the FTPL excludes consideration of money when money does not pay interest - because it is not equivalent to debt for that reason. E.g. there is no interest expense and therefore no budget cost so it never has to be repaid.

I guess that's all wrong, but why is it wrong?

Regarding negative real rates:

If r is negative, then 1 period nominal debt of B with P = 1 can be repaid with a final single real surplus of:

s = B /(1 + r)

The whole FTPL valuation is intriguing, but it seems to be a bit of a fudge with the assumption of future real surpluses. So couldn't the negative real rate problem also be fudged along those lines?

I.e. the pattern of the required real future surplus series, whatever it is, is progressively discounting due to the negative rate rather than progressively accumulating due to a positive real rate. Anyway, it can be cut off at a finite point simply by assumption.

[Nick Rowe]

Brad: I liked the post because:

It is clearly written. No poncy math you can't understand.

It acknowledges there's another side to the debate, and presents that other side properly.

It acknowledges that the short run data support that other side.

It adds an epicycle to acknowledge that short run data. (Sure, it's an ad hoc epicycle, but that's a helluva lot better than ignoring that data.)

It points to an empirical problem with the other side: that we do not observe black holes, but the Neo-Wicksellian model (without money) says that black holes are theoretically possible.

So, after reading Schmitt-Grohe Uribe, it was like a breath of fresh air.

The way I read John Cochrane, he is not in fact a Neo-Fisherite. Rather, he is exploring the problems of indeterminacy and multiplicity in standard models. Which I think are real problems. If a central bank did credibly commit to a higher inflation target, we would see both inflation and nominal interest rates rise together. But what makes it credible?

[Nick Rowe]

Edward: "Are permanent changes in the growth of total base money the determining factor for inflation?"

They are *one* of the determining factors, and they are the one the central bank can control, which makes them interesting from a policy perspective.

"Wouldn't it be better to focus on the portion of base money that is or is not circulating? and why?"

Why would anyone buy and hold money if they planned never to sell it again? Because they just liked looking at it? Or perhaps it got permanently lost behind the sofa. Leaving those few cases aside, all money circulates; some quickly and some slowly. We use money, rather than barter, precisely because money is an asset that is easy to buy and sell.

Jason: "I believe -- correct me if I'm wrong -- the reason the ZLB implicitly assumes money is because [currency] is a 0% nominal interest rate asset. If you think the ZLB is important then you're assuming money."

[I have made a very minor correction. Because chequeing accounts can and sometimes do pay interest.]

We can imagine a barter economy (and we do see some barter), but a barter economy would be a very different economy. We wouldn't see apple producers unable to sell their apples and so unable to buy bananas, plus banana producers unable to sell their bananas and so unable to buy apples. The two would just do a swap of apples for bananas, if they could barter easily.

[Nick Rowe]

JW: I only know Patinkin 56/65, not 48. But probably yes. Some old stuff is good stuff, but gets forgotten.

Lord: yes, only it was more like late 70's. But monetarism then meant "making M1 grow at a constant rate", and "fail" meant "did not stabilise inflation as well as we hoped it would". There's a difference between "increasing inflation relative to what it would otherwise have been" and "increasing inflation relative to what it was in the past". Because other things can change too.

JKH: "I guess that's all wrong, but why is it wrong?"

What you say there sounds roughly right to me. If both money and bonds pay the same positive real rate of interest, that is exogenously determined by the market, then FTPL would work fine, IMO.

Not quite sure where you are going after that. But take this simple case. Take a stationary economy where currency is 10% of NGDP, and the growth rate of currency (and the inflation rate) is 5% per year. Then the government can run a primary deficit of 0.5% of GDP every year forever. (Currency is paying -5% real interest, so currency holders are paying +5% real interest for the privilege of lending to the government.)

[Scott Sumner]

Excellent. As usual, I like the way you explain it better.

[anon]

Isn't it the standard view that the stock of money/"M" is irrelevant in a liquidity trap because money and short-term bonds are perfect substitutes, and expected to be perfect substitutes for the foreseeable future? So if you want a consistent view of "M" for this case, I think it needs to include _both_ "money" and these government bonds. You could retire all denominations of currency by introducing new ones at the same time, but we don't consider that to be "driving M to zero" because, for all relevant purposes, it's not.

[Nick Rowe]

Thanks Scott. But I read your way of explaining it first. Which brought me back to money. Neo-Fisherianism is a Neo-Wicksellian heresy.

anon: yes, that's the "standard" Old Keynesian ISLM view. But I don't think it's right. But that's another argument. Call me when the central bank runs out of things to buy, then we can talk about needing fiscal policy. And the credible *threat* to buy until it runs out of things to buy, or NGDP rises to a sensible pre-announced level, (whichever comes first), can change expectations, so that threat need never be carried out.

[Fred Fnord]

"Call me when the central bank runs out of things to buy."

I'm confused. Do you think that quantitative easing was just as effective as a bigger stimulus would have been? Or perhaps you simply think we needed a few orders of magnitude more quantitative easing? (Never mind the argument about whether any central bank currently in existence could credibly threaten to even approach such a thing, because clearly we are not currently operating at a practical but instead at a wholly theoretical level.)

[Nick Rowe]

Fred: "Do you think that quantitative easing was just as effective as a bigger [fiscal] stimulus would have been?"

Monetary and fiscal policy don't have the same units. We measure one in $, and the other in $ per year.

The effect of a change in fiscal policy depends on whether it leads to the central bank doing a different monetary policy.

The effect of QE depends on whether it is expected to be temporary or permanent. And that depends on what the central bank is (expected to be) targeting. To change expectations about the target, it matters less what the central bank actually does, but what people expect the central bank *would do* if the economy did not hit the target.

My local garage parked a (US) cop car on their forecourt (on the main street in a Canadian village). (I don't know why). That cop car actually did nothing. But everybody slowed down to the speed limit (until they realised it was a US cop car) because they expected the cop car *would* do something *if* they didn't slow down.

Concrete steppes.

Chuck Norris

[Nick Rowe]

And if you use the word "stimulus" as a synonym for "fiscal loosening" you are biasing the question right from the beginning, by implicitly assuming that fiscal policy works, and by implicitly assuming monetary loosening is not "stimulus".

Did actual QE (in the US) do more for aggregate demand than actual fiscal loosening? Yes, probably (because we saw little effect when fiscal got tightened again). Which does not mean that fiscal loosening was a bad thing.

What was needed was not *more* QE, but an explicit NGDP level path target ("We want NGDP to be 5% above where it is now, and then to grow at 5% thereafter") plus the *threat* to do QE **and not reverse it** (make it permanent) until that target is hit.

[Jeff]

"Why would anyone buy and hold money if they planned never to sell it again?"

To "corner the market"? (For a time, in implicit collusion with other hoarders, to manipulate government decision-making/policy in their favor.) I'm assuming you meant "never" hyperbolically, but I certainly could be misunderstanding the model.

[JKH]

Nick,

I’m still confused, I guess.

Returning to your very first comment:

“Assuming there are no bonds (just to keep it simple) and all deficits are money-financed, if Rm < 0 (which it is, with 2% inflation and no interest on currency) the PV (surpluses) calculation literally does not add up. The PV of a future surplus gets bigger and bigger the further into the future it is. The sum does not converge, as time goes to infinity. Actually, the thing goes negative. You are talking about permanent primary deficits. You can finance a permanent deficit by printing negative real interest money. Even more if the economy is growing so the demand for M/P is rising over time. And that is where FTPL really falls apart. FTPL only works in a permanent liquidity trap, where money pays the same rate of return as all other assets, forever.”

But the FTPL valuation equation that Cochrane uses, which I assume is correct, equates the real value of current outstanding debt to the present value of future real primary surpluses. It doesn’t value the permanent primary deficits that money financing is capable of accommodating in the future. It values outstanding debt, which in the money case would be assumed to be the outstanding stock of money that pays zero interest – if money was even allowed into the TFPL valuation model. (I thought you agreed that it wasn’t intended to be part of TFPL in your second last comment, but I’m assuming here that it does because of your last comment.)

Assume current outstanding “debt” is just money. The real value of money is M/P. The value of M is its face value and P is whatever it is. So that is straightforward.

Because this is the pure money case, assume r is defined as the negative of the inflation rate – as you have done in your example - since “debt” pays zero interest in this case.

Then there are still many surplus series that satisfy the FTPL equation, no matter how negative r becomes. I gave a one in my comment above – a single bullet real surplus of B/(1 + r), assuming for simplicity of illustration that P = 1. That happens to be a single bullet nominal surplus of B, which also makes sense under such assumptions.

There’s no requirement that the FTPL assume an infinite series of surpluses whose present value “explodes” because of the potential future deficits that M accommodates and the negative interest rate.

It appears to me like the FTPL valuation equation does not value future deficits. It values the real value of outstanding debt, which is finite and well defined by B (or M) and P.

The fact that B is only money and that money accommodates deficits doesn’t prevent the government from raising taxes to make TFPL work.

I.e. the fact that M is money financed and accommodates a deficit does not mean there can’t be a future surplus net of that marginal deficit capacity. Just add taxes. And a surplus can be assumed that satisfies the FTPL equation.

This would seem to suggest that M be “repaid” with taxes. Now maybe that’s a practical problem to consider, but in fact Cochrane covers that issue more generally in the paper by contending that in fact TFPL doesn’t have to assume that at all for even the general case of bonds – that it’s an equilibrium relationship and not a budget constraint relationship.

But I guess I must be wrong on this somewhere.

[Majromax]

@Nick:

> The effect of QE depends on whether it is expected to be temporary or permanent.

"The money printing will continue until the price-level improves."

[Min]

Nick Rowe: "And if you use the word "stimulus" as a synonym for "fiscal loosening" you are biasing the question right from the beginning, by implicitly assuming that fiscal policy works, and by implicitly assuming monetary loosening is not "stimulus"."

Well, there is stimulus and there is response. Stimuli do not always "work" by eliciting the right responses. :)

And metaphorically, anyway, loosening is a very weak stimulus. All it does is to allow responses to happen that have been constrained. Loosening does not encourage those responses or reward them, or provide energy or motivation, or do anything else of an active, positive nature.

[Min]

Nick Rowe: "Call me when the central bank runs out of things to buy, then we can talk about needing fiscal policy."

Pardon me, but that is like the doctor saying call me when the pills run out. No, doctors say things like call me if the condition gets worse, or if it has not gotten better in a couple of days.

In the US, as of the end of September, the core inflation rate has remained below target for 22 of 24 months. Isn't it about time to call the doctor? (Except that the doctor believes in blood-letting. ;))

[Majromax]

@Min:

> Pardon me, but that is like the doctor saying call me when the pills run out. No, doctors say things like call me if the condition gets worse, or if it has not gotten better in a couple of days.

I'm not so sure that's the best metaphor. QE's problem is that the treatment is temporary -- as Nick has pointed out in other articles the US Fed has implicitly threatened to yank the money back when inflation picks up.

If we have to stick with this imagery, the current therapy is "Look at these pills, and all they could do for you. Now, if you don't start feeling better I'm going to give you more pills to look at."

If you'd like a Chuck Norris metaphor, our Fed Norris isn't beating people up, nor is he threatening to; he's threatening to threaten to beat people up maybe.

[Jacques René Giguère]

"Monetary and fiscal policy don't have the same units. We measure one in $, and the other in $ per year."
QE may be in X$ this year and 0$ next year if it's your policy. But it is in $/year. Dimensionnal analysis( whixh I teach to my students, help wonders with structuring their homeworks) must lead to consistency in your units ( my physicist side).

"Call me when the central bank runs out of things to buy."
Why wait? I asked my students this morning what they would do with helicipter money. Unanimous answer"Save it" What if they were hired to dig ditches of unknown utility? "Spend it, it's regular income."

[TallDave]

Great post Nick, I had roughly the same response. I don't think you can usefully model anything without considering that people have expectations tied to the Fed's inflation target, not the Fed's interest rates.

"If the central bank permanently raises the nominal interest rate, will this result in higher or lower inflation? If you tell me what permanently raising the nominal interest rate does to the base money supply growth rate, I can answer your question."

Exactly! It seems really obvious that the Volcker/Greenspan regime of tighter money led to lower interest rates, as people gradually accepted that the Fed was no longer targeting unemployment as strongly and hence there would be less inflation (just as the opposite happened during the Great Inflation). But it took a sharp recession to get there, and I'm sure a lot of people in the early 1980s viewed Volcker's regime as radical and destructive.

[TallDave]

"What was needed was not *more* QE, but an explicit NGDP level path target ("We want NGDP to be 5% above where it is now, and then to grow at 5% thereafter") plus the *threat* to do QE **and not reverse it** (make it permanent) until that target is hit."

A thousand times yes. This seems so obvious once you understand the role of expectations! What we need today is a sort of "reverse Volcker" move in policy, one that seems radical at first but actually just finds a middle ground between the Great Inflation and the current Great Disinflation, one that will maximize employment and RGDP while also providing reasonable price level stability. NGDPLT is that policy.

Really, the only question today is what the NGDP target should be.

[Too Much Fed]

Nick said: "Suppose a Neo-Wicksellian central bank hits the ZLB. If it lets the stock of base money spiral down to zero too, then of course the economy would spiral down into a black hole. But central banks don't have to do this. And central banks did not in fact do this, when they hit the ZLB. They did the opposite. Which is why their economies did not spiral down into a black hole."

JKH, I think I have a scenario where that is not true. Would you be willing to look at it if I put it up? Thanks!

[JKH]

TMF: why not

(if base money goes to zero, the interest rate on interbank lending of reserves goes to infinity, which is not pretty)

[JKH]

Nick,

From your Chuck Norris post:

"Every 6 weeks, the Bank of Canada announces its overnight rate target for the next 6 weeks. And the actual overnight rate instantly moves to where the Bank of Canada wants it to be. It's pure Chuck Norris. The Bank doesn't actually do anything. It just lets the market know where it wants the market to move, and the market moves there. The market moves because of the Bank of Canada's threat. If the overnight rate is above the Bank's target, the Bank will add however much settlement balances as are needed for as long as is needed until the overnight rate falls to the target. And if the overnight rate is below the Bank's target, the Bank will subtract however much settlement balances as are needed for as long as is needed until the overnight rate rises to the target. And because this threat is credible, the Bank doesn't need to carry it out."

Right.

There actually aren't that many people that understand this. The Fed worked in the same way pre-crisis.

I've made this point before, but I think you're making a bit of a category error in describing this concrete steppes stuff. What you've described above is the role of expectations (or counterfactual expectations or counterfactual conditional expectations or something along those lines) where the expectation factor can only be formed on the basis of what might be a concrete steppes sequence in a counterfactual where everybody is stupid and hasn't learned from previous experience.

The point is I think that there is a difference between a state of obtuseness with respect to the role of expectations and the formation of expectations that have a concrete steppes counterfactual at the heart of that formation. So I think you're casting too broad a net on this idea of concrete steppes, at least the way I read.

For example, I think there's a huge difference between the way you've described the role of expectations above as it refers to the nature of the reaction to CB target rate changes and the much bigger story of how market monetarism explains the alleged effectiveness of what amounts to massive QE - the "nothing left to buy" sort of stuff you guys like to refer to. I think you've assumed your conclusion in that argument but have barely scratched the surface with a comparable precision of the kind that can easily explain the case I quoted above.

[Hólmsteinn Jónasson]

How is that such lucid analy­sis does not get con­sid­er­ably more atten­tion ?
"Submission to the Financial System Inquiry (pdf) (html) - Includes a new formula for the relationship between bank credit and inflation." http://fsi.gov.au/files/2014/04/Harkness_Leigh.pdf
See: "THE OPTIMUM EXCHANGE RATE SYSTEM" by Leigh Harkness http://www.buoyanteconomies.com/PAPER3.pdf
"Formulas for the Current Account Balance" http://www.buoyanteconomies.com/Formulas%20for%20the%20Current%20Account%20Balance.pdf

Buoyant Economies. Explaining what is happening to our economies and what can we do about it? http://www.buoyanteconomies.com/

[Too Much Fed]

JKH, let's assume the desired demand for currency is zero, the desired demand for central bank reserves is zero, and the required demand (reserve requirement) for central bank reserves is zero.

Now can the fed set the fed funds rate with that scenario and the monetary base (currency plus central bank reserves) equal to zero?

[Too Much Fed]

JKH said: "Nick,

From your Chuck Norris post:

"Every 6 weeks, the Bank of Canada announces its overnight rate target for the next 6 weeks. And the actual overnight rate instantly moves to where the Bank of Canada wants it to be. It's pure Chuck Norris. The Bank doesn't actually do anything. It just lets the market know where it wants the market to move, and the market moves there. The market moves because of the Bank of Canada's threat. If the overnight rate is above the Bank's target, the Bank will add however much settlement balances as are needed for as long as is needed until the overnight rate falls to the target. And if the overnight rate is below the Bank's target, the Bank will subtract however much settlement balances as are needed for as long as is needed until the overnight rate rises to the target. And because this threat is credible, the Bank doesn't need to carry it out."

Right.

There actually aren't that many people that understand this. The Fed worked in the same way pre-crisis."

So the fed funds rate can change without the monetary base changing and can other interest rates change without changing the monetary base?

[Too Much Fed]

JKH, sorry it took so long to reply.

[JKH]

Too MF

Assuming reserves don’t pay interest, the CB generally needs to supply at least a small cushion of excess reserves to help smooth out frictions in interbank clearings. Otherwise the effective trading rate for the targeted rate will likely drift up toward to the penalty rate for loans from the CB, because some banks will inevitably end up short reserves (negative balances).

Assuming reserves don’t pay interest, the Fed’s announcement of a rate change is enough (mostly) to move the effective rate to the new target rate. Otherwise, if banks collectively resist the intention of the announcement, the Fed can add or drain reserves as required to move the rate by force. The banks know this so they don’t resist. That’s the Chuck Norris effect (I think as Nick means it).

It’s a bit different when reserves pay interest, depending on whether it’s a channel system or a floor system. If it’s a channel system, it’s not that different. If it’s a floor system, the CB sets the target rate at the floor, and that prevents rates from dropping further, while the chronic reserve excess prevents them from increasing.

[Too Much Fed]

"Assuming reserves don’t pay interest, the CB generally needs to supply at least a small cushion of excess reserves to help smooth out frictions in interbank clearings. Otherwise the effective trading rate for the targeted rate will likely drift up toward to the penalty rate for loans from the CB, because some banks will inevitably end up short reserves (negative balances)."

If some banks end up short central bank reserves (negative balances), wouldn't there be other banks that have excess central bank reserves (positive balances) that would exactly offset the negative balances?

"Assuming reserves don’t pay interest, the Fed’s announcement of a rate change is enough (mostly) to move the effective rate to the new target rate. Otherwise, if banks collectively resist the intention of the announcement, the Fed can add or drain reserves as required to move the rate by force. The banks know this so they don’t resist."

I agree with that. If the fed wanted to lower the fed funds rate, add reserves. Once the fed funds rate is lower, take them back so the monetary base is unchanged.

"That’s the Chuck Norris effect (I think as Nick means it)."

Nick usually wants to veer off to some kind of hot potato effect. I'm trying to set up a system of zero currency that yields 0% and zero central bank reserves that yield 0% so that all the fed does is change the fed funds rate (an interest rate). The fed then hopes that entities in the private sector react to interest rate changes to set how much money (demand deposits) there is and how much of it circulates in the real economy.

I'm assuming the fed is in a passive mode, not an active mode.

QE would be considered an active mode.

I believe Nick has said QE is just another name for open market operations the fed has done in the past. I don't agree with that. QE means the fed won't take back excess central bank reserves. In a non-QE mode, the fed will take back excess central bank reserves for treasuries (usually treasuries).

[JKH]

Nick,

My thoughts at Cochrane’s blog on his spread targeting idea.

I referred to your original post on paying interest with new money, and the question you posed at the end of it (second comment).

Cochrane’s way of thinking about institutional separation of Treasury and the central bank seemed like a good fit for my way of answering your question - so far.

http://johnhcochrane.blogspot.ca/2014/11/target-spread.html?showComment=1417354643756#c2984633605804124015

http://johnhcochrane.blogspot.ca/2014/11/target-spread.html?showComment=1417354750622#c3261118998850857172