« The Globe's Konrad Yakabuski succumbs to the monetary union zombie | Main | The Conservatives adopt a scorched-earth strategy against the PBO »


Feed You can follow this conversation by subscribing to the comment feed for this post.

I would define a helicopter drop as a gift of money by the central bank (ie with nothing taken in exchange). I agree that seigniorage payments qualify (although when the central bank holds exclusively government debt, they are more like a government debt to the central bank which it forgives), but seigniorage is tiny relative to overall government revenues (ie a few percent interest on a stock of base money that is typically a few percent of GDP). And ironically, in these times of aggressive easing and massive government deficits, seigniorage may well be turning negative for those central banks that lose money on their QE assets.

Rebel: yes, a helicopter drop is a money financed tax cut/transfer increase. And yes, the wealth effect of helicopter money is very small. A serious storm probably has a bigger wealth effect. How can such small helicopters be so powerful? What gives them so much leverage? That's what I was trying to answer, and I didn't invoke the wealth effect.

I think you are mixing two different ideas Nick. Seigniorage is not what is supposed to give the central bank its power. In fact, I would argue that influencing interest rates by being the marginal market player should result in the central bank making a loss. For example, I dare say that Warren Buffett could raise deposit rates for a while by subsidising a bank that offered higher rates - until he lost his fortune!

Rebel: I am basically agreeing with you that seigniorage is not what gives a central bank its power (at least in this context).

Your Warren Buffett example is a good one. Suppose Warren Buffet is as "big" as a central bank (let's not worry about how we precisely define "big"). How come Warren Buffet can't set interest rates, but the Bank of Canada can? (Just assume WB is Canadian, for the sake of argument). My answer is that Warren Buffet promises to redeem at par his liabilities for Bank of Canada liabilities, but the Bank of Canada does not promise to redeem its liabilities for Warren Buffett liabilities. That's the asymmetry that creates the asymmetry of power. In a head-to-head war between Warren Buffet and the Bank of Canada over interest rates, where WB tried to set a lower rate than BoC, it would be WB who lost his shirt, not BoC. That's because people would borrow from WB, ask WB redeem his liabilities for BoC liabilities, then lend BoC liabilities to the BoC. So WB would need to borrow an infinite amount from the BoC in order to convert his liabilities into BoC liabilities.

I never would have thought arbitrage was such a difficult concept.


The point about Warren Buffett was meant to explain that I can believe that a large player can control interest rates for a while if they are prepared to sustain losses, rather than to argue that he can take on the BoC. You complicate the issue by introducing two institutions that are trying to control interest rates, but let's go with it anyway.

Suppose the sub-BoC interest rate that Warren Buffett was trying to establish (I am bending over backwards to avoid saying "set" to assuage JKH!) was the interest rate compatible with inflation at the BoC target, which I shall suppose is the natural rate reflecting the real investment opportunities (either capital investment or lending) available in the economy plus the target rate of inflation. In that case, people will stop investing in anything other than lending to the Bank of Canada. If Warren Buffett is daft enough to borrow from the BoC and lend people more to allow them an even bigger arbitrage on the BoC, so much the better! The BoC's liabilities will come flooding back to it and, assuming the normal relationship between money supply and inflation, inflation will fall. But if the Warren Buffett interest rate was the "right" one, this fall will take inflation below the BoC target, and the BoC will have to give up and increase the money supply again. And will the BoC be able to replace its assets as cheaply as it sold them, when the alternative opportunities all yield less? No. The BoC might not go bust, because it can call on the government and hence the entire nation for recapitalisation, so it is bigger than Warren Buffett, but neither of them are as big as the economy. As long as the BoC is genuinely pursuing its inflation target, it must follow the natural rate fairly closely.


Set is acceptable.

It's interest rates plural that's not acceptable.

Control interest rates plural is not acceptable.

Set the policy target rate is acceptable.

Control the short (risk free) rate is acceptable.

Over and out.

Nick said: "Helicopter money: the central banker prints some currency, loads it into a helicopter, and flies around the country throwing it out for people to pick up."

Nick also said: "Rebel: yes, a helicopter drop is a money financed tax cut/transfer increase."

I'm going to assume current USA budget stats. Something is not right there. To me, currency has no interest rate attached to it and no repayment terms attached to it (like being redeemable).

If the gov't cuts taxes and/or does a transfer increase with the gov't already in deficit, that is an increase in gov't debt NOT an increase in currency. Right?

Well said JKH.

Too muc Fed: "If the gov't cuts taxes and/or does a transfer increase with the gov't already in deficit, that is an increase in gov't debt NOT an increase in currency. Right?" Yes, unless the Bank of Canada then buys that debt for currency.

Ah, I see your objection, JKH - I thought the difference between "set" and "guide to" was a bit semantic. The assumption of a single interest rate is just an assumption designed to reduce the number of moving parts, and is not unreasonable if real factors such as the underlying natural rate and term, credit and liquidity premia are regarded as constant. Of course they aren't, which has been one of the features of the financial crisis, but let's stick with the single rate assumption for the purposes of resolving the issue of central bank control. Then, when it is accepted that central banks have limited control of any economically significant interest rate, especially if they are genuine about targeting inflation (big if), considering probable shifts in the relationships between the various interest rates allows a better understanding of the financial crisis to be developed.

By the way, I still have an open mind about the issue of whether central banks control an (economically significant) interest rate, but the more that the idea that they don't is tested here, and the less discussants dispute the logic as at November 05, 2009 at 06:07 PM as opposed to objecting to the model abstractions or making glib remarks (Adam P at November 05, 2009 at 10:22 AM!), the more convinced I become that the argument that Gary Marshall and myself are making is correct.

Rebel, instead of getting offended perhaps you should read what Gary originally wrote. He was disputing Stephen's claim that the BoC has unlimited power to LOWER(!!) the exchange value of the CAD. He was wrong on this. He showed no recognition of the aymmetery (there are limits on the BoC's ability to RAISE the international value of CAD) but simply denied this point as a matter of principle. He did NOT say that "they could lower the forex value of CAD but won't because...", he simply stated that the BoC had no power at all to lower (LOWER) the CAD.

The argument then progresses into a debate about whether or not the BoC can control the level of short term INTER-BANK interest rates. At this point nobody has claimed that this rate is economically significant (yet), Gary and then you are simply denying the BoC's ability to set this rate.

Now, if you want to argue about the economic significance of the BoC's power to set this rate then do so. Certainly we can argue at length about various possilbe transmission mechanisms. Nick and I have argued about this in the past, but we haven't even got to this point because Gary refuses to think through the implications of his own agrements.

"less discussants dispute the logic as at November 05, 2009 at 06:07 PM"

Sorry, Rebel - not involved in that one. The position I'm arguing is the one I've been arguing since way back when, in our first discussion on this, and a long time before that. It's fairly simple. Never dreamed that arguing a fact could be met with such opposition. You're very free to believe what you will, of course. Warren Buffet isn't involved.

I am just making the point that the comments are not challenging the logic, which may mean that it is hard to challenge. I came into the other discussion from the side, but I know Gary Marshall's position from his comments on various blogs, notably Willem Buiter's, and I would not want to claim that I am the only one to make this case if it eventually prevails.

That's fine. I choose to draw the line at what I believe to be relevant, in order to allocate intellectual capital on a risk reward basis. That tends to pare responses to a variety of things. But who knows - you and Gary Marshall may end up becoming emininent economists. BTW, he seems to be spending quite some time at Bill Mitchell's, apparently after I posted the link here.

I will leave Gary to plough that furrow; Bill Mitchell is one of those bloggers who call to mind the quote from Blaise Pascal, "I would have written a shorter letter, but I did not have the time"!


Although I feel the same way about your Ben Friedman.

We invest in what makes sense to us.

Rebel: "As long as the BoC is genuinely pursuing its inflation target, it must follow the natural rate fairly closely." Agreed. And in the long run, if it tires to set an interest rate above/below the natural rate, the result will be accelerating deflation/inflation.

Suppose nevertheless the BoC *tries* to set r below the natural rate, and Warren Buffet tries to set r equal to the natural rate. What happens?

My answer is that in the short run the BoC will succeed. And WB will be forced to follow the BoC, because it will be WB, not BoC, who will go bust from the infinite arbitrage if WB tries to set r above the BoC. (That's where the asymmetric redeemability comes in).

But in the long run the BoC will be forced to raise r back up to the natural rate. It's not that the BoC will go bust otherwise; it's that inflation would rise without limit otherwise.

WB must follow the BoC in the short run; he will go bust otherwise. The BoC must follow the natural rate in the long run; the monetary system will collapse otherwise.

My above comment is just conventional money/macro theory, by the way. The asymmetric redeemability part is the only bit that's not textbook.

"WB must follow the BoC in the short run; he will go bust otherwise. The BoC must follow the natural rate in the long run; the monetary system will collapse otherwise."

Believe or not, Nick, we agree on your first sentence. It's consistent with the point I've made all along about central bank operations, and probably consistent with PK/MMT/Chartalism.

I don't know enough of the theory of the natural rate to opine on your second sentence. That one may be an issue vis a vis PK/MMT/Chartalism. I suspect its a complicated debate. But bear in mind that the central bank will make an explicit decision to change its policy rate in accordance with its view of those kinds of risks (i.e. "collapse").

JKH: "Believe or not, Nick, we agree on your first sentence. It's consistent with the point I've made all along about central bank operations, and probably consistent with PK/MMT/Chartalism."

I certainly believe you believe it, JKH. I would be very surprised if you had said you didn't. (Up to, of course, your quite correct point about interest differentials on assets of varying risk and liquidity etc., which is so hard to keep saying as a caveat, so it's so easy to fall into the trap of saying the CB sets interest rates). And it is consistent with PK/Neo Chartalism too. Those theories would make no sense otherwise. I can't think of any major macro theory that would disagree to any substantive extent.

Oh, I just remembered that an extreme real business cycle theorist would claim that the CB can't affect real interest rates even in the short run, because prices are as perfectly flexible in the short run as in the long run.

Yes, it's very much the second sentence where macroeconomists disagree. Paleo Keynesians would disagree with it most. New Keynesians and monetarists (and the better Austrians) would agree with it, even if not quite 100%. It's the super-neutrality of money question.

We have some convergence, Nick, in that we seem to agree that if the central bank does not follow the natural rate, it will be forced to do so by an unacceptable divergence of inflation from its target. Since this outcome would presumably be revealed by a good inflation forecast, and would be seen to breach the mandate of a lexicographic inflation targeting central bank, in order to run an interest rate that is not the natural rate, the central bank will have to fiddle its public inflation forecast, which relates to your church post.

The problem remains that, by setting an interest rate below the natural rate (the other way round from the case that I discussed above), the BoC is opening itself up to arbitrage. This only needs to be slightly successful in terms of bringing forth extra projects that yield less than the natural rate (ie loan demand is slightly elastic) before the central bank has to increase the base money supply by such a large proportion that inflation is destined to increase unacceptably. One wonders whether it is worth the trouble.

Anyway, I sense that your patience with this discussion is exhausted, Nick, so I will desist. I will set out my argument, now with its responses to the points raised here, on my blog. Nevertheless, I hope you keep thinking about it, because the implications of the arbitrage on the central bank are potentially momentous (which might include, for example, distorted risk spreads, disintermediation of the banking system, current account imbalances, malinvestment, negative net worth central banks etc). Thanks for the discussion.

Hello Mr. Rowe,

All these scenarios are wonderful, but before running through them I must ask if the central bank has the powers which you attribute to it?

The central bank does not have the breadth or vigor to manipulate those weighty financial levers and it never has. So until the matter of central bank power is resolved, there is little point in going any further.

The Bank of Canada is a bank like any other, with several differences. It is comparatively small when measured against Canada's big commercial banks. It does not pay interest on deposits and its lending is negligible. So it really is not much of a lending institution or rather market player.

The source of the BofC's assets and liabilities lies chiefly in one prerogative, the issuance of currency. It alone supplies and meets the demand for Canadian currency. Those notes, assuming a small cost of production, provide all of its power. I think we are agreed on this point.

There is about $50 billion in currency in circulation amounting to about say 4%-5% of the stock of money of say $1 trillion. So one might say that just 1 out of every 20 financial transactions requires currency. The markets determine the usage of currency, not the central bank. If demand rises or falls, the commercial banks will adjust their currency holdings by requesting additional notes or returning the excess.

When the commercial banks receive those new currency notes, they are paid for with that money I call ethereal, intangible and existing only in bank records. Currency is outside money created by the Bank of Canada, and ethereal money is inside money created by commercial banks in the lending process. The Bank of Canada may also create ethereal money, but rarely ever does, often in small quantities, and only for brief periods. So to obtain currency, a BofC created liability, the commercial banks surrender ethereal money, a commercial bank created liability.

The BofC invests this windfall generally in Government of Canada securities from which it earns interest. The interest, after deducting a large agency fee for its operations, is then reinvested or returned to the Government of Canada.

It is safe to say that when currency is issued or returned, there is mutual redemption in that a currency dollar note goes one way and an equivalent ethereal dollar goes the other.

Imagine that the commercial banks discover an excess of currency of about $200 million in their system. They convey it back to the BofC. Upon acceptance, the BofC will have to return an equivalent amount of commercial bank liabilities. If the ethereal money is fully invested, the BofC will have to sell a government bond. If not, then it can draw down its ethereal money reserves.

I have said in previous postings that currency is an inferior form of money, comprising only 4% of the total. It has a physical state and earns no interest. For these reasons currency is used for immediate purposes, passed as quickly to another.

If the Federal Government were to have the BofC print up $50 billion in notes to pay its bills, which consist of about 40% of the GDP, it would not get very far, perhaps about 2 month's worth of expenditures. The BofC would then print up and distribute currency to match the Government's returning cheques, deposited by recipients in member banks and submitted for clearance. The BofC would now have $100 billion in liabilities and only $50 billion in government securities covering them.

The commercial banks would return most of the currency immediately having use for only a small fraction of it. To redeem the notes, the BofC would have to sell all its ethereal money government bonds. Or it could compel the Government to come up with a $50 billion dollar bond to balance its liabilities. The BofC would then have to sell the bond on the financial markets in exchange for ethereal money to redeem the notes.

The BofC could just refuse to accept the notes. However, this would generate inflation of 100%. The market established balance between currency and money is 4% or about 1:20. Doubling the currency supply to $100 billion without reason will double the entire money supply to $2 trillion as banks rid themselves of the excess of currency by making loans to restore the former equilibrium.

But why would the Government create such havoc, destroying the value of existing financial assets and transforming expectations for the fleeting pleasure of avoiding 2 months of disbursements? All it really had to do was float a bond and obtain the same results without the financial chaos.

In attempts to influence the exchange rate or the interest rate by printing up currency notes the result would be the same. Though the Bank of Canada has this privilege, it is severely restricted in its ability to use it.

The only method of profit in currency issuance comes from the receipt of commercial bank liabilities, i.e. ethereal money, which is used to buy interest bearing government securities. With these bonds, about $50 billion in total, the BofC can influence the exchange rate or interest rates in the medium, overnight or long term markets, but only to that amount and only in actually doing so. Empty threats will not bring about enduring changes in either. One must move out into the markets to create the desired effects.

But even $50 billion dollars will not go far against much larger competitors and in financial markets in the $trillions. It will not control interest or exchange rates when the other market players can draw upon sums far more immediate and ponderous.

Now all of this may be required reading for any economics novice, but I never saw a word of it. And had I, there would not now be a field of study in economics known as monetary policy.

Gary Marshall

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad