Negative interest vacuum-cleaner Gesellian money for broke central banks

[I can't decide whether this thought-experiment has any policy relevance. I hope not, but who knows what the future will bring? It came to my mind after reading a (more policy-relevant) post by Simon Wren-Lewis.]

1. Helicopter money is when the central bank prints money and puts it in people's pockets. Vacuum-cleaner money is the exact opposite; it's when the central bank takes money out of people's pockets and burns it.

2. Paying 5% interest on holding money is when the central bank pays you $5 per year for every $100 you have in your pocket (or in your account at the central bank). A 5% tax on holding money is a negative 5% interest rate on money; the central bank charges you $5 per year for every $100 you have in your pocket (or in your account at the central bank).

3. Let's put 1 and 2 together. "Positive interest helicopter money" is when the central bank prints new money to pay you interest on the money in your pocket. "Negative interest vacuum-cleaner money" is when you pay the central bank interest on the money in your pocket and the central bank burns it.

That was just to explain the terminology. Now let's do the thought-experiment.

Suppose a central bank is broke (Simon says "policy insolvent"). It has no assets. And it can't borrow against its future seigniorage revenue from printing money because there won't be any. And the government won't bail it out. But the inflation rate is in danger of rising above target (which is why it won't print money and earn seigniorage revenue). The central bank cannot buy back money (do an open market sale of bonds, or "negative QE", in the silly modern terminology) because it doesn't have any assets to sell. What can it do?

It can implement negative interest vacuum-cleaner Gesellian money for broke central banks. It imposes a tax on holding central bank money, and burns the money it collects in taxes.

"Now hang on" you say. "Isn't the idea of a tax on holding money to reduce the demand for money (increase the velocity of circulation) and so raise the equilibrium price level? Isn't that why Gesell suggested his tax?"

You would be quite right. And that is what would indeed happen if the money collected in taxes on money were put straight back into circulation (as Gesell proposed). It would raise the price level, so it wouldn't be a very sensible policy if we wanted to lower inflation. But we are going to burn it instead. So a 5% tax on holding money means the stock of money will be falling at 5% per year, which will have a disinflationary effect.

The ongoing negative 5% growth rate in the stock of money causes the equilibrium inflation rate to decline by the same 5% (5 percentage points). But the 5% tax on holding money, and 5% lower expected inflation rate, have equal and opposite effects on the real stock of money demanded M/P, so P does not jump, but instead P declines at 5% per year, along with M, relative to what would otherwise have happened. It's just like an ongoing annual 100-for-105 reverse stock split. Every year you hand in 105 old shares, and get 100 new shares in return, so the shares get 5% more valuable each year (relative to what would otherwise happen).

(Just for once, the Neo-Fisherites would be right, albeit for the wrong reasons.)

Comments

[Market Fiscalist]

"But the 5% tax, and 5% lower expected inflation rate, have equal and opposite effects on the real stock of money demanded M/P"

I'm not quite getting how we could know the effects would be equal and opposite (or is that just an assumption in your model?).

I can see that there would be equilibrium where the new tax has been factored into the demand to hold money, after which we get the 5% steady deflation. But in the short term wouldn't the effects be a bit more indeterminate ? If demand to hold money is very elastic the 5% tax when introduced could cause a big one-time increase in P. Lets say P doubles. It then takes many years of 5% deflation to hit the inflation target again (during which time the CB decides that taxing money holding is a bad way of targetting inflation).

[Anon]

I think a simpler way for a 'broke' central bank to implement contractionary monetary policy is to simply raise reserve requirements of banks.

[Nick Rowe]

MF: "I'm not quite getting how we could know the effects would be equal and opposite (or is that just an assumption in your model?)."

It's a standard assumption, and seems very plausible. When you decide whether to hold an asset, you look only at the *sum* of interest plus capital gains, and don't care about the mixture (unless the two are taxed differently).

But you are right that it wouldn't hold exactly in the short run, if prices are sticky, so the inflation rate doesn't fall immediately.

Anon: Hmmm. OK. But the banks wouldn't be happy, and some of them might go bust, if you did it suddenly.

[Market Fiscalist]

"When you decide whether to hold an asset, you look only at the *sum* of interest plus capital gains, and don't care about the mixture"

This works as long as people expect the 5% tax to lead to 5% deflation. But given that the 5% tax initially leads to people to want to hold less money (unless they are all super rational and know everyone else is too) , which would lead to inflation not deflation, its not clear why they would expect that (at least at first).

[JP Koning]

"The ongoing negative 5% growth rate in the stock of money causes the equilibrium inflation rate to decline by the same 5% "

Can you unpack this a bit? If I'm an investor and I know that the quantity of an asset with a convenience yield is going to get rarer and rarer, then I'm going to put a much higher value on those conveniences. In the long term, I'd be willing to accept a rising inflation rate given that the expected flows of convenience yield are so high.

[Scott Sumner]

Nick, Excellent post. I wouldn't say "just for once" the NeoFisherites would be right, I'd say "just for twice." In another post I talked about a crawling peg exchange rate regime where the yen fell enough each year to support higher nominal rates than in the US, and the change was accompanied by a once and for all depreciation of the yen to prevent the higher interest rates from depressing demand. Our two thought experiments have the following in common:

1. A credible change in the expected steady state inflation rate.
2. Another action to prevent a short run shock to the economy.

And perhaps one could add:

3. They are not how monetary policy works in the real world, which is exactly the problem with NeoFisherism.

[Nick Rowe]

MF: assume expected inflation stays constant in the short run. You get the one-time increase in the price level due to lower demand for money when they announce the 5% tax. But ever after you get a continuously falling price level as the quantity of money falls at 5% per year. And when expected inflation (finally) adjusts, you get a one-time decrease in the price level as the demand for money increases again. Under rational expectations, that one-time decrease happens at the same time as the one-time increase, so they exactly cancel out.

JP: Remember the big difference between a real asset like land, where the wheat yield depends on the number of acres of land, regardless of its price, and a nominal asset like money, where the convenience yield depends on the real value of the stock of money M/P, so it's proportional to the price of money in terms of goods, and not the number of dollars you hold.

Scott: thanks! It's the direction of causation the Neo-Fisherites get wrong (and hence stability of equilibrium).

In your example, the announced crawling peg causes both higher inflation and higher nominal rates, which is correct.

In my example here, the higher announced (positive/negative) nominal interest rate ***on money*** (not on bonds) is at the same time an announced higher (positive/negative) growth rate of money, and it is that announced higher (positive/negative) growth rate of money that causes higher (positive/negative) inflation.

[Peter K.]

SWL in the original post:

"If that money needs to be mopped up after a recession is over in order to control inflation, the central bank might run out of assets to do so. A good name for this is ‘policy insolvency’."

I still don't understand this. Has this happened to an actual central bank in actual history any time recently? Was Paul Volcker in danger of running out of assets when he raised rates in the early 1980s?

Plus the central bank could just call the Treasury and say "we need to fight inflation." "Cut spending and raise taxes."

[Nick Rowe]

Peter K: Ask a historian. But most historical examples would be central banks on the gold standard (or fixed exchange rates). Unless you have 100% gold reserves, there's a risk, and central banks faced with a run normally devalue *before* they run out. If interest rates rise, the prices of central bank bonds fall. If they act as lenders of last resort, they might end up holding dodgy assets.

"Plus the central bank could just call the Treasury and say "we need to fight inflation." "Cut spending and raise taxes.""

And the fiscal authority might say "No". And the central bank loses its independence.

[jonathan]

Nick: Maybe I'm confused here, but I don't see why things would work the way you expect.

Suppose there are two types of assets: bonds and money, and all money is in reserve accounts at the central bank that pay an interest rate set by the bank.

Now suppose the CB sets a -5% nominal interest rate on reserves, and shrinks the (nominal) money supply at a 5% rate also.

Under perfectly flexible prices, this should imply a -5% inflation rate, and so a -10% real rate on reserves. This should drastically reduce desired holdings of (real) reserves. I think the result is a one-time increase in the price level so that the real money supply is at the desired level. There might be some further real effects because the return on holding money differs from the return on capital.

Now suppose there's some price stickiness, so the price level can't jump, and inflation may be sluggish. The low return on reserves reduces desired reserve holdings, and but since the (real) supply of reserves doesn't fall, the interest rate (on bonds, capital, etc) falls instead. Demand increases, and there's inflation in the short-run.

Eventually this inflation, plus the contraction of the nominal money supply, reduces the *real* money supply to a level consistent with a -10% real return on money while the return on bonds is at it's natural level, and we eventually reach the equilibrium described in the flexible price case.

Okay, I might be wrong. But this is my understanding of what would come out of a simple model (say MIU, plus Calvo in the sticky price case).

[jonathan]

Oh wait, I see what I misunderstood. A -5% inflation rate implies a 0% real return on money, not -10%. So no change in the price level.

[Nick Rowe]

jonathan: yep. Weird, isn't it? The stock split is the nearest analogy I can think of. They pay you new stocks as interest, but that increases the number of stocks, so the market cap stays the same.

[jonathan]

Nick: No it makes perfect sense. I just didn't have the intuition right for the world you were describing. But I know how to model it easily enough.

Now what's interesting is that I think you would still get some inflation in the short run. Assuming some stickiness in inflation, so that the inflation rate doesn't fall immediately, the lower return on reserves will boost spending, creating *inflationary* pressure. But as the price level rises, this causes the real money supply to contract faster, leading to adjustment in this way.

[Nick Rowe]

jonathan: I think that's right. But I reckon it might depend on the particular type of stickiness of both inflation and expected inflation.

[Max]

In your hypothetical equilibrium, the central bank is still broke. So why should anyone expect this equilibrium instead of hyperinflation?

[Nick Rowe]

Max: it will presumably depend on what the central bank actually does.

[rsj]

I disapprove of this post. It gets everything wrong, as is so common with these monetarist posts, by getting everything wrong about banks and private behavior.

First, no one would hold cash, they would deposit it and then transact via debit/credit cards in order to escape the tax. The CB balance sheet would shrink to the size of required reserves as banks deposit all their excess currency -- now all currency -- with the CB and the CB is forced to drain it out of the economy in order to maintain their $50 million reserve target. Suppose the CB abandons the $50 million target, then banks will be out the spread between the target rate and 5% on the now (massive) quantity of excess reserves. Assuming that the target rate is less than 5%, then this is a tax on bank capital. If the target rate is greater than 5%, then this is a huge subsidy to bank capital.

Either way, you are creating a hole in bank balance sheets or you are adding to bank profits. In the former case, the banks go bankrupt and then the CB has to recapitalize them and put the reserves back. Or it lets them go bankrupt and we move to barter. In the latter case, this is a big subsidy to banks.

Prices don't move a lot either way, other than the sudden depression style price movement if the CB lets the banks all go under.

[Nick Rowe]

rsj: "First, no one would hold cash, they would deposit it and then transact via debit/credit cards in order to escape the tax."

Demand curves slope down, but they are not always perfectly elastic. Inflation is a tax on currency, but people still hold currency. Even when the inflation rate is quite high.

And since currency is an (imperfect) substitute for bank deposits, increasing the tax on currency would raise bank profits (though disinflation would have an equal and opposite effect).

[rsj]

"Demand curves slope down, but they are not always perfectly elastic. Inflation is a tax on currency, but people still hold currency. Even when the inflation rate is quite high."

Sure, but people hold very little currency. You have to persuade them to go to an ATM and withdraw more. It's not like people wake up with a pile of unexpected currency on their pillow. This disincentives schlepping down to the ATM. You really plan on moving the consumer price index because of the convenience yield of coins for the laundromat? It's not even worth considering (plus laundromats now take cards, too).

On the other hand, the effect on highly levered institutions such as banks of putting a hole in their balance sheet is really big. You are going to cause a banking crisis long before there is a 1 basis point change in the CPI.

[Nick Rowe]

rsj: "Sure, but people hold very little currency."

How much is "very little". Guess the average currency holding per Canadian. Or American. Before Googling. My guess is that you will be surprised.

And re-read my point above about the effect of a tax on currency on bank profits.

[Tel]

As far as I know, there's no Central Bank on Earth with the legal power to directly tax the cash out of citizen's pockets. It's simply not an option.

You may perhaps be talking about negative rates in terms of the "War on Cash" in which case various other notable fringe economics blogs got to this point before you did, but this requires active repression to maintain it. You can imagine the sort of problems that are going to crop up. Citizens might buy some alternative currency, like gold coins or something. Imagine that.

So there's still some options on the table. Government might run a surplus and burn the surplus. Good side of this is at least it's nominally legal for government to impose tax, and in effect burning the (mostly electronic) cash happens automatically if government can't think of anything to spend it on. The bad side... try to find a government right now that actually can run a surplus, not easy is it? When you bring all those elderly people who want to retire into the picture, gets even worse. Gosh, those guys think they were promised something, what will we tell them? So the answer is, yes government could do it, and no they probably won't do it.

rsj: "Sure, but people hold very little currency."

My currency holdings are nett negative, and the people I know have a book even more negative. I expect the central planners will run out of options and print money, or at least drive interest rates to rock bottom and keep them there (much the same thing). Judging from the way paintings and random curios are going up in price on a rather ballistic curve, I think someone out there must be betting even harder on the same outcome.

FWIW, here's my prediction on the Fed: the USD will gradually fall, and the Fed is happy enough at this time to let it fall. At some point, commodities are going to start rising (as from the perspective of being measured in USD) and the Fed will gently raise interest rates by a small amount, enough to control the commodity price, and prevent it rising too much. This won't fix the US economy but it will stabilise the USD and there won't be any runaway USD crash. American business will need to deal with that minor rate rise and there will be a mild shakedown, winners and losers, you know the drill.

[rsj]

Nick, it's too late! I dug this info up before writing my original post. It's interesting to note that currency is not even considered a "Key Monetary Policy Variable" by the Bank of Canada:

http://www.bankofcanada.ca/rates/indicators/key-variables/

You have to go look at the balance sheet: http://www.bankofcanada.ca/wp-content/uploads/2015/09/statement-financial-position-310815.pdf

It is about $72B in comparison to $2.6 Trillion in deposits for the top 8 domestic banks (http://www.cba.ca/contents/files/statistics/stat_bankq_en.pdf). But that doesn't include money market mutual funds and other cash-equivalents. And of course, this number will shrink very rapidly once this tax is imposed. The economic burden of this tax is borne by banks, not households.

"And re-read my point above about the effect of a tax on currency on bank profits."

No no no no no.

This is not a tax on bank _profits_, this is a tax on bank _assets_.

Banks, because they are leveraged, have a lot of assets and a small capital base to support those assets. As banks in aggregate cannot dispose of reserves, there is no way for them to escape this tax -- unlike the situation of households. This will drive the banks out of business unless the CB agrees to lift the tax.

[rsj]

And of course, go Blue Jays!

[JKH]

Nick,

Interesting post.

It could be noted that the normal mode of combined central bank and treasury operations is for treasury to “immunize” or eliminate any money supply consequence due to a central bank fiscal effect. For example, when the central bank pays positive interest on reserves, that interest becomes a marginal fiscal expense that gets passed through to the Treasury function (it is a reduction in the profit that might otherwise be remitted). Other things equal, it increases the budget deficit. And Treasury finances the deficit with bonds. In that way, the central bank’s interest expense has no net effect on money supply (i.e. reserves), because the money paid out in interest that otherwise would be left as additional bank reserves gets sucked back out of the system by Treasury’s bond issuance, which is a sort of Treasury bond vacuum cleaner.

Something comparable would be the case for negative interest as well. Negative interest is equivalent to a tax at the margin, which becomes marginal fiscal revenue that gets passed through to the Treasury function. Other things equal, it reduces the budget deficit. So it reduces required bond financing. In fact, at the margin it is equivalent to the effect of a bond redemption – a swap of money for bonds. When Treasury redeems a bond with the proceeds of the central bank interest rate tax, by swapping money for bonds – it is performing a money easing operation to offset the tightening that would otherwise result from the interest rate tax. We could extend the “silliness” of some terminology by calling this “Treasury QE” – Treasury’s swap of money for a bond.

So the norm is that Treasury “immunizes” the central bank’s money supply in this way – i.e. this is the usual institutional arrangement whereby the central bank settles its net fiscal effect (i.e. profit, including any interest expense or revenue) with the Treasury function, thereby preventing fiscal effects of the central bank from changing the money supply, other things equal.

And consistent with that normal institutional setup, the central bank is left to control the money supply (i.e. at least through bank reserves if not more) entirely through open market operations. And of course, the normal setup is that the central bank has assets that are accumulated in marginal easing and available for marginal tightening.

Your post is about what happens when those assets aren’t there.

What that means is that the central bank must now engage in direct fiscal effects in order to adjust the money supply. And in your example, the central bank wants to tighten the money supply because of inflation concerns. And it wants to do it when rates are negative. And it wants to tighten without having any assets to sell.

So it can do this by seizing the reigns of the central bank’s own fiscal effect. In the normal course, Treasury would offset the tax tightening of negative interest rates by putting money back into the system through bond redemption (or by a lower deficit compared to the counterfactual where there is no such tax revenue routed into the Treasury function).

In your case of “negative interest vacuum-cleaner Gesellian money”, the central bank takes control of the fiscal effect. By “burning” the money it takes in from the negative interest tax, the central bank prevents the routing of that money back to Treasury where it is subject to the usual Treasury easing that immunizes any potential net money supply effect. And so there is a net drain of money in your Gesellian case.

So I think the point of your post amounts to the fact that the usual money supply easing offset provided by Treasury is no longer provided. So as a result there is a net money supply tightening, which counteracts the stimulus otherwise provided by negative interest rates. Or at least - there is a relative tightening of policy compared to the counterfactual of the normal recycling of the money taken in by the tax.

Just a further point related to the glories of financial accounting around this thought experiment (No - I’m not deliberately trying to make you wince. I think this is interesting.)

The standard financial accounting for a central bank without any assets would be a central bank with negative capital in the amount of its liabilities. As many of your economist brethren (and you I think) have written, that is actually a survivable situation – to the point where you relish abandoning such accounting. But let’s stick with it for a moment.

Suppose you start out in such a position with no assets and negative capital in that same amount. Then any intake of “negative interest vacuum-cleaner Gesellian money” reduces the gross size of the balance sheet and reduces the gross size of that negative capital position – because it amounts to the central bank taking in a fiscal surplus for its own account. “Burning” your Gesellian money is the same thing as crediting the central bank equity capital account. It is a money tightening operation in the sense that reserves have been sucked out of the system by the negative interest rate tax and essentially converted to central bank equity.

You’ve spoken about an analogy with reverse stock splits. I’ve reverse-analogized your “burning” concept to that of a credit (or "deposit") to the central bank equity account.

One final point. It’s interesting that your example is one where interest rates are negative but the central bank wants to tighten the money supply on a net basis due to inflation concerns. So it resorts to the tax and burn strategy. That begs the question as to why the possibility of increasing interest rates (making them less negative) wouldn’t be considered first. (This in fact is the order of planned US Fed QE exit – interest rate adjustment before money supply adjustment.) But maybe you did consider it and decided that writing your post was more fun instead.

[Nick Rowe]

JKH: that looks like an interesting and informative comment. But I will need coffee first.

[Nick Rowe]

JKH: "So the norm is that Treasury “immunizes” the central bank’s money supply in this way – i.e. this is the usual institutional arrangement whereby the central bank settles its net fiscal effect (i.e. profit, including any interest expense or revenue) with the Treasury function, thereby preventing fiscal effects of the central bank from changing the money supply, other things equal."

I'm trying to get my head around the "usual/current institutional arrangements".

Simple case. There is $100 currency in public hands. The CB levies a 5% tax (negative 5% interest rate). So people give the central bank $5. And since that is $5 in profit for the central bank, the CB gives the government $5. And the government buys $5 worth of bonds, so that $5 goes straight back into public hands.

Got it! (I think). In which case paying negative interest on currency has no effect on the supply of currency, but does reduce the demand to hold currency, and so this causes an increase in the equilibrium price level.

So the "usual/current institutional arrangements" are the same as those that Gesell proposed.

And what I am doing in this case is imagining an alternative institutional arrangement in which a broke central bank instead uses its profits to reduce its liabilities instead of giving them to the government. It burns the $5. Which means that the same actions (negative interest on money) has different consequences.

Right?

Thanks JKH. That clarifies my thinking. My brain wasn't clear on the current institutional arrangements. But I think it is now.

[Nick Rowe]

JKH: Follow-up. I'm now starting to think my thought experiment might be a little more policy-relevant than I thought it was.

Consider a central bank that is not *totally* "broke", because it has *some* assets, but its assets are less than its liabilities. $100 in currency in public hands, but only $90 worth of bonds. This *could* happen. If it makes any profits, it probably would use those profits to reduce its negative equity position to zero, before paying anything to the government.

[JKH]

Nick,

Thanks for the response.

I just completed the following longer comment before seeing the comments you've just left.

In doing so, I've tried to make a connection of sorts between your idea and that of SWL's.

So I'll leave that here first as additional grist, and then respond to what you've just written in full context.

[JKH]

Nick,

This is a follow up comment to my last. These comments are sufficiently long and detailed to warrant their own posts elsewhere, but I’m out of that mode for now, so I hope it’s alright leaving them here.

(If coffee was helpful for first one, scotch may be a good chaser later on in having a go at this one.)

Your post uses the scenario of an asset-less central bank tightening the money supply (i.e. bank reserves) when rates are negative. Negative interest paid on reserves (in effect interest received/“taxed”) can be “burned” (credited to equity in financial accounting terms as I describe in my first comment), which shrinks bank reserves. It is assumed there is no transfer of this money to Treasury as would be the case under normal institutional procedure, and therefore no subsequent transfer of money from Treasury back out to bank reserves. There is a net contraction of money supply. As you describe it, the stimulative effect of negative interest rates in reducing the demand for money is offset by the tightening effect of reducing the money supply by the amount of interest/tax collected by the central bank.

This revised institutional procedure would be problematic for an asset-less central bank with the same tightening objective when interest rates are positive. Positive interest paid on bank reserves increases the level of bank reserves, other things equal. This is an expense and a direct financial accounting transfer from central bank equity to bank reserves. Under normal institutional circumstances, with an asset rich balance sheet, the central bank receives interest revenue from assets at a level that typically dominates the interest expense on reserves, with a net profit result. That consolidated profit is in effect instead a net drain on bank reserves, other things equal. Under normal institutional procedure, that profit is remitted to Treasury. That is depicted as a financial accounting transfer from central bank equity to Treasury’s cash account with the central bank. Treasury then spends that money from its account as part of its overall budget. That is depicted as a net financial accounting transfer from the Treasury account back to bank reserves.

However, an asset-less central bank that pays positive interest on bank reserves will incur a loss because of that net interest payment. Bank reserves increase on a net basis, which poses a dilemma for a task of money supply tightening. Not only are there no assets to sell in order to effect a potential tightening policy, but the problem faced in a tightening scenario gets worse as interest paid on reserves increases the level of reserves, and that is an easing effect – the opposite of what is desired.

I think there could be two ways for an asset-less central bank in such a positive interest rate environment to tighten the money supply (reserves) at this point, depending on whether one is looking to conventional institutional procedure or unconventional procedure (notwithstanding the fixed assumption of an unconventional asset-less balance sheet).

The conventional procedure (albeit with an unconventional balance sheet) would imply that if an asset-less central bank makes a loss due to paying interest on reserves, the Treasury should immediately recapitalize it to cover that loss. In the recapitalization process, Treasury provides an asset (likely a Treasury bond) to the central bank. Then the central bank can sell the bond to drain reserves. If additional tightening is desired beyond that point of immediate loss, then Treasury could increase the capital injection with more bonds so that the bank can sell those.

This capitalization solution might be considered to contradict the core assumption of the original thought experiment – i.e. maybe we are considering a central bank that is asset-less in both good and bad times. If that is the case, we would not assume a capital injection solves the problem of how to tighten the money supply.

Instead, the central bank could seize the fiscal reins in a way that is comparable to the negative interest rate case. In order to tighten the money supply (bank reserves), the central bank could impose some kind of tax of its own (not an interest rate tax - because interest rates are now positive) such that bank reserves are drained by the tax, even net of the money expansion provided by positive interest paid on reserves if desired.

Both the capitalization and tax routes under positive interest rates might be considered infusions of equity into the central bank balance sheet. They both reduce bank reserves and they both increase the equity position of the central bank (make it less negative) by reducing the absolute size of what is in effect a negative equity balance sheet equity position.

I think the Simon Wren-Lewis post also applies to this case of positive interest rates, although not with your more general assumption of an asset-less central bank by construction. He is proposing helicopter money issuance by the central bank, combined with a contingent capitalization commitment. The contingency for capitalization is that the central bank may run out of assets to sell in the event of a tightening program. Capitalization would provide bonds as assets to sell for tightening purposes.

This SWL case is in effect a variation on the capitalization option noted above for an asset-less central bank. The central bank in the SWL case is not asset-less to begin with, but is in danger of become asset-less part way through a tightening program due to sales of what assets there are. Hence the contingency for capitalization.

I do question two of SWL’s more general points.

I think central bankers don’t want to deploy helicopter money – not because they would fear government reneging on the kind of contingent capitalization commitment he describes – but because helicopter money fits neither the job description nor the DNA of a modern central banker. If the likes of Bernanke and Yellen are reluctant to advise directly on fiscal policy in front of Congress or at a press conference (and they are reluctant), why would we think that they would want to take charge of a helicopter money operation? If they dare not even speak to fiscal policy in public, why would they want to take operational charge of something that is a form of government expenditure (a tax expenditure in effect)?

Even if SWL is correct that central bankers would fear the risk that governments might renege on “a delayed commitment to provide central bank with assets to sell” (contingent capitalization), I think there is also an open question in suggesting that the financial and political/institutional risk associated with such a commitment collapse is no greater than the risk due to potential losses under normal QE as it already exists.

Normal QE loss risk is related to the possibility of negative interest margins – as might happen if the central bank raised the short term rate it has to pay on bank reserves before the long term fixed interest rates it receives on its assets have time to adjust upward. (This is the classic bank “short funding” risk that used to be the concern of banks everywhere 20 to 30 years ago.) I think this interest margin risk is probably of a lesser scale than might be the case with the immediate interest margin cost and future balance sheet risk of moving to helicopter operations. But it does all depend on the numbers one assumes for each case and the analysis is definitely too complicated to get into in this comment.

[JKH]

Nick,

“Simple case”

Yes. Succinct as always.

And easy to visualize – but easiest in the case where you imagine that this is the only thing in the government budget. Then the negative interest becomes the consolidated surplus, and the government can use it to buy back or redeem bonds on a net basis.

I think it’s probably more confusing to visualize when you consider that the profit/surplus impact from central bank operations becomes a relatively small piece of the overall budget that’s submerged in larger budget operations that in total typically produce a deficit, with no government reduction of debt on balance.

That’s when I think it becomes very useful to consider what is happening on the central bank balance sheet, as follows:

a) With respect to the CB’s interest/tax effect, you can visualize that as a transfer from bank reserves to CB equity (retained earnings for a temporary period).

b) With respect to the CB’s remittance to Treasury, you can visualize that as a transfer from CB equity to Treasury’s account with the central bank.

c) With respect to Treasury using that money, you can visualize that as an expenditure of cash from its cash account, producing a shift of money from the Treasury account back to bank reserves.

All of this happens as depicted on the CB balance sheet, and all of it happens regardless of whatever else goes on in the budget, including the typical case where there is in fact an overall deficit net of the CB effect, and no net retirement of debt, but instead a net expansion of debt.

“And what I am doing in this case is imagining an alternative institutional arrangement …”

Yes.

[Nick Rowe]

JKH: Yep. Gonna need Scotch. But it's too early for Scotch now.

[JKH]

Nick,

“Consider a central bank that is not *totally* "broke" …”

Yes to your point.

The example could be the profile of a CB that starts out with a "semi-conventional" looking $ 90 balance sheet (bonds matched with currency - although no bank reserve balance, no treasury balance, and no capital) but which has just done a $ 10 helicopter currency drop.

So now it has negative equity of $ (10) corresponding to the asset-less helicopter drop operation - currency issued now exceeds assets.

With your simplified balance sheet construction, you could assume that Treasury takes in currency by taxing and borrowing, and pays the CB interest on the bonds in the form of currency. So the balance sheet would actually shrink, with a reduction in currency issued by the CB gradually reducing the negative equity position – until currency issued matched the original asset level of $ 90 – all other things equal of course.

The “real world” case where there are bank reserves and electronic payments would be a little different to explain, but would amount to the same net effect.

And so yes – the CB could use internally generated earnings to reduce its negative equity position over time.

This would be a further institutional arrangement/procedure/understanding between the CB and Treasury, contrary to the usual regular remittance of profit.

Your point is somewhat parallel to SWL’s idea of contingent capitalization.

His idea concerns contingent “external” capitalization by Treasury - injecting capital and bonds as needed if the CB runs out of bonds. But his idea is not aimed at re-capitalizing the CB to point of eliminating negative equity - it’s only aimed at ensuring a supply of bonds at the margin for policy tightening purposes.

Your point on the other hand concerns steady “internal” re-capitalization through retained earnings, with the eventual implied objective of eliminating negative equity.

[Nathan W]

I thought helicopter money was literally handing out money.

I'm not sure the concept can apply so well to (above zero) interest rate policy. For every dollar in savings, there is a loan (ignoring reserve requirements/practices). So a higher interest rate just means that savers earn higher interest and borrowers pay more interest. There's no helicopter effect.

[Nick Rowe]

Nathan: Suppose the central bank prints $5 million and gives it away. That's a helicopter operation. Suppose it gives it away to the people who already own currency, in proportion to how much currency they own. That's still a helicopter operation.