Those of us who argue for monetary policy as a way for countries (and fake countries like the Eurozone) to escape a recession, even in an alleged "liquidity trap", recognise that any increase in the money supply should be permanent, and perceived as permanent, to do much good. Even some Keynesians, like Paul Krugman, recognise that monetary policy could do a lot of good if only it could be made permanent. He just doubts that any central bank could credibly commit to making it permanent.
One way for a central bank to credibly commit to increasing the money supply permanently is to increase the money supply by buying bonds, then publicly burn the bonds. Publicly trash the asset side of its own balance sheet. That means the central bank won't be able to buy back the money in future. So it's a permanent increase in the money supply, and seen as such. Helicopter money is a permanent un-backed increase in the monetary base. Burning the bonds burns the backing, and so burns the vacuum cleaner that might suck the money back up in future. It's like an invading army burning its boats.
A second way is to overpay for junk bonds. That's what the European Central Bank has just done. This second way is better than the first, because it creates a negative feedback loop between the amount of money created permanently and the amount of money that needs to be created permanently. As I shall explain. And for once, just once, it's actually an advantage that the Eurozone lacks a central fiscal authority. Because that means it's much harder for a fiscal authority to bail out the ECB if it gets into trouble, as I strongly hope it will, or be expected to. A bit of trouble, or the perceived likelihood of a bit of trouble, is just what the doctor ordered.
I had totally forgotten I had already written a post on a very similar topic over a year ago, until a comment by PierGiorgio Gawronski reminded me, by suggesting Mr Trichet publicly burn some of the ECB's assets. (My brain is failing.)
In my old post, I said that Ben Bernanke was betting on the economic recovery, using the Fed's own assets, by buying toxic waste. If there is no recovery, he loses the bet, the toxic assets become worthless, and the increase in the monetary base becomes permanent, because the Fed can't afford to retire the extra money. If there is a recovery, he wins the bet, the toxic assets are worth at least what he paid for them, and the increase in the monetary base can be temporary, because the Fed can afford to retire the extra money. That's the demand curve of recovery.
At the same time, the greater the extent to which the increase in the money supply is expected to be permanent, the greater the extent to which the economy recovers. That's the supply curve of recovery.
There's a negative feedback loop between the expected value of the assets and the expected degree of recovery. A demand curve for recovery, and a supply curve of recovery. Somewhere in the middle the two curves cross, and determine the equilibrium level of recovery, and the equilibrium degree of default on Greek bonds.
This negative feedback loop is a way around the "ketchup problem" (the central bank keeps banging the bottle trying to get some inflationary ketchup to come out, and then it all spurts out in a rush).
There was one thing wrong with my previous post. The Fed has a fiscal authority behind it, the US Federal government, that is (so far) solvent, and that might be expected to bail out the Fed if it lost the bet. Which would mean the Fed would be unable to credibly commit to increasing the money supply permanently.
Nobody understands how the ECB operates; and I think that includes the ECB itself. It's pointless looking at the rules, because the ECB, and Eurozone generally, just breaks them. It's some sort of economic/political judgment call. But there is no central fiscal authority for the Eurozone. And it's hard to get 16 governments to agree on anything, let alone stick with it. Especially if they have to come up with cold, hard, tax increases to bail out the ECB and buy back the cash. Especially given the amount of national animosity (already bad enough, from what I read and hear) that bailing out the ECB to cover a Greek default would bring.
So, it might just work for the ECB.
Worthwhile Canadian Initiative 
Central banks causing crisis since 18?? with x, trying to cure them with x, only to cause the same crisis all over again. What a sad story.
The outright purchases are the ECB’s nuclear option we talked about in previous posts. As long as they are done on the open market, they do not break the ECB’s rules. If the ECB had bought or lent directly to a government, it would be breaking the prohibition on monetary financing. But it is not doing this.
The outright purchases, while legal, are controversial because they go against the grain of the ECB’s adopted policy of extending its liabilities solely through short term refinancing operations; collateralized lending. Government debt has been accepted as collateral in these operations. Now the ECB is purchasing and bringing this debt directly onto its balance sheet. Was this sudden u-turn an independent decision made by the ECB governing council, or did they give into political pressure? If the ECB’s independence been compromised, then its no more than a crappy banana-republic central bank.
I know you think its pointless to look at rules; after all they inhibit your flights of monetary imagination. But the rules are vital.
Nick,
You assume a permanent injection creates not only inflation but growth. This is a critical assumption and you do not take into account the tail risk associated with this dynamic.
If the ECB bought and burned all PIIGS bonds the PIIGS countries would no longer have a solvency problem. However, they would still need financing for a now-smaller primary deficit. At what price would the markets finance that deficit? Given the likely devaluation caused by the ECB bail out, markets would demand a steep sovereign premium. This, in turn, would increase the PIIGS deficits, which in turn might make markets demand an even higher premium, causing an adverse positive feedback loop. Faced with that loop, the ECB would likely step in to buy and burn more bonds. What this is called is permanent monetization of structural fiscal deficits. The risk of this dynamic is what typically sends velocity through the roof as markets attempt to hedge future inflation/devaluation. The result can be hyperinflation, as we have seen in various Latin American countries.
Of course, if real growth eliminates the need for fiscal deficits, the ECB could pat itself on the back for a job well done. This might be the mid point of a forecast probability distribution. It is in no way reflective of tail risk.
Should monetary policy be formulated in the context of avoiding tail risk?
Nick:
“you really need to integrate your real Bills analysis of the price level into the AD/AS framework, so you can look at your own theory from another dimension”
That sounds like the real sticking point. I haven’t made this clear before, but I don’t see even a grain of truth in the AD/AS model. From my point of view, trying to integrate the RBD into Keynesian theory is like trying to integrate astronomy into astrology.
The RBD, while inconsistent with textbook macro theory, is consistent with classical price theory. In particular, the RBD says that money is valued according to its backing, just like any other financial security.
Nick: as I mentioned, the dividend payments are quite small. “by agreement” the Fed pays most of the money to the treasury.
JP Koning,
You are making an important point about rules. However, remember that rules are made and broken when they violate the primary objective of the institution. In this case is the financial stability of the EMU banking sector. This is what motivates them to break thie existing rules!
As far as the difference between ST refinancing operations and purchases in secondary markets is concerned, there is no difference if the banks fail to honor their repo obligations because of economic conditions and the capital erosion from the value of these securities(when you buy back your capital is eroded from mark to secondary market revals. Then the lender of last resort principle will come into play!
“…any increase in the money supply should be permanent, and perceived as permanent, to do much good.”
If AD is threatened to fall because of a drop in velocity, then money supply could be temporarily increased to maintain a constant MV. When V increases then M could be reduced.
In other words, a temporary increase in money supply would be effective.
“Luis: Sounds depressing. As I said in some previous post, nobody seems to have thought much about the lender of last resort function of central banks, when setting up the ECB. And given the lack of a central government, it’s not at all clear how, politically, it can fulfill this role.”
Nick, that is completly truth.
On the other hand, much seem to forget that “Monsieur” Trichet was the only cntral banker that rise the interest rate in 2008, just before the cataclysm triggered by Lehman failure.
Is the ECB an independent central bank, or a banana republic´s one? I think that, following what I’ve read in Nick’s, it is neither one thing nor the other, but something worse, because not having a government, but 16, against who defend their positions, there is not a power alternative face to which defend their independence. Its Council decides always by unanimity.
The ECB is defined as responsible to the European parliament; but the Euro parliament is a fictitious institution, which represents no one.
if you think about it, is an entity that has never existed in the past. A product of bureaucratic engineering, which has no other criterion that its statutes, very anti-inflation biased. A perfect example of too much independence?
or perhaps an example of supreme hectoplasma?
An historic mistake, in any case.
Sorry for wandering off my own topic, but I’ve been thinking about Mike Sproul’s comment above.
Mike: you don’t have to be a Keynesian to think of AD/AS. And any theory ought to have some sort of representation in {P,Y} space.
Suppose you halve M, while holding backing constant, or double the backing, while holding M constant. P should halve, right? What happens to Y? If it’s neutral, then all real variables, like Y, should stay the same. OK, so you are tracing out points along a vertical curve in {P,Y} space. Call it the “AS” curve. Now, what would happen if you assumed that prices of goods and services were sticky. Take an extreme case, suppose they are fixed by law, so P cannot halve. What happens to Y, according to the RBD?
Next question: suppose you hold M, and backing, constant. Suppose all resources (labour, Kapital, Land, etc.) double. So Y doubles. What happens to P? If nothing, then you are tracing out a horizontal curve. Call it an AD curve.
All asset prices are determined by demand and supply. This does not contradict RBD, which is just a theory of demand for assets. But there must be some version of something like Walras’ Law in your model. There must be some relation between an excess demand or supply for money, and an excess supply or demand for goods.
1/P is the price of money in terms of goods. P is the price of goods in terms of money. If the former is determined by the supply and demand for money, the latter must be determined by the supply and demand for goods, and those two sets of supplies and demands must be related somehow.
David: “Given the likely devaluation caused by the ECB bail out, markets would demand a steep sovereign premium. This, in turn, would increase the PIIGS deficits, which in turn might make markets demand an even higher premium, causing an adverse positive feedback loop.”
I disagree. Markets, if rational, should calculate the deficit as the primary deficit plus the real interest rate on the debt. They should ignore the inflation premium.
TMDB: “In the context of Eurozone, a run on ECB is when people demand liabilities of German commercial banks in exchange of ECB liabilities in an anticipation of an Eurozone breakup.”
Interesting. I hadn’t thought of that. And presumably people would do this expecting Euro deposits at German banks to be converted into Neu Deutsch Marks? But the ECB is not obliged to redeem its notes for liabilities of the German commercial banks.
BTW, I just enrolled your blog in the Palgrave Economics Blog roll.
http://www.econolog.net/index.php
Nick,
I meant a real premium. While governments can print and have negative short term real rates, the markets typically extract a steep, real term premium. In extreme cases (i.e., Brazil pre-Cardozo), the premium makes all but ultra-short term borrowing prohibitive.
I would put this question to you another way. If you were a Euro bond holder and saw the currency depreciate after an ECB “buy-and-burn” operation, would you,
a) sell and move your money to a more stable currency until the real, risk adjusted return on Euro bonds rises considerably
b) say, “its ok, Europe will grow, and they’ll never do it again,” and buy more
I’m not predicting the outcome, I’m saying that its useful to recognize that hedging behavior can introduce an adverse positive feedback loop, as has happened time and again in Latin America. The question is, should this tail risk be taken into account in setting policy (you can still argue the tail risk doesn’t exist, but please explain why choice a) above is highly improbable for market actors)?
Nick:
This seems like one of those cases where two views are so different as to be mutually unintelligible, but here goes anyway.
“And any theory ought to have some sort of representation in {P,Y} space.”
The usual price theory model of general equilibrium doesn’t use P,Y space, and it doesn’t try to aggregate things that can’t be aggregated. I just think of an economy as a community production-possibilities curve (CPPC). If that economy is cash-starved, maybe because of banking regulations, then people are reduced to less efficient means of trade and the CPPC shifts in. If new money is then introduced, trade becomes easier and the CPPC shifts out. If that new money is adequately backed, then money holds its value even though its quantity increases. Keynesians would see M rising with no effect on P and think “sticky prices”, but I’d answer that the value of money is not sticky. It’s just that that the new money is adequately backed.
“Suppose you halve M, while holding backing constant, or double the backing, while holding M constant. P should halve, right?”
right
“What happens to Y? If it’s neutral, then all real variables, like Y, should stay the same.”
right
“OK, so you are tracing out points along a vertical curve in {P,Y} space. Call it the “AS” curve.”
I could just as well flip a coin all day, plot the number of heads H against Y, and say I’ve traced out a vertical AS curve in H,Y space
“Now, what would happen if you assumed that prices of goods and services were sticky. Take an extreme case, suppose they are fixed by law, so P cannot halve. What happens to Y, according to the RBD?”
It would be the same if silver were held at twice its equilibrium price by law. There would be a surplus of silver and deadweight losses.
“Next question: suppose you hold M, and backing, constant. Suppose all resources (labour, Kapital, Land, etc.) double. So Y doubles. What happens to P? If nothing, then you are tracing out a horizontal curve. Call it an AD curve.”
See the coin flip example.
“All asset prices are determined by demand and supply. This does not contradict RBD, which is just a theory of demand for assets. But there must be some version of something like Walras’ Law in your model. There must be some relation between an excess demand or supply for money, and an excess supply or demand for goods.”
A stock market example might clarify. Finance prof’s don’t normally speak of excess demand or supply of IBM stock. They speak of backing. Any stock whose value doesn’t reflect backing creates arbitrage opportunities for traders, so they speak of arbitrage conditions rather than supply and demand.
“1/P is the price of money in terms of goods. P is the price of goods in terms of money. If the former is determined by the supply and demand for money, the latter must be determined by the supply and demand for goods, and those two sets of supplies and demands must be related somehow.”
The value of money is determined by backing, not supply and demand of money. Supply and demand curves are for actual commodities, produced using scarce resources. Money can be nothing but computer blips. The supply and demand of computer blips does not determine their value.
Nick’s theory of the price of IBM stock:
The price of IBM stock is determined by supply and demand. The demand curve for IBM stock (to a first approximation) is perfectly elastic, given by the forumal P=Present Value (dividends, earnings, whatever).
“I just think of an economy as a community production-possibilities curve (CPPC). If that economy is cash-starved, maybe because of banking regulations, then people are reduced to less efficient means of trade and the CPPC shifts in. If new money is then introduced, trade becomes easier and the CPPC shifts out.”
Yep. That’s what we Keynesians and Monetarists call a recession, and recovery, caused by the AD curve shifting left, due to a decline in the real money supply, then shifting right again, when the real money supply is increased.
Panayotis,
In a society that respects rule of law, and a few of these exist, central bank rules can’t be broken for the sake of “primary objectives”. Before a central bank can burn its backing, it must get the relevant section of its governing act changed. This is a long process involving senate/parliamentary committees and finally a vote. Only when all these steps are complete and the act been properly amended can the central banker burn his backing.
But you are right that rules are often broken when they violate primary objectives. I’d say this applies in the societies that don’t respect the rule of law, and therefore characterizes most African, Asian, South American, and European central banks. The Bank of Canada is better than most in respecting rule of law.
The supply curve of IBM is also perfectly elastic. If P>PV(dividends, etc) then IBM will issue huge numbers of shares. If P<PV(dividends, etc) then IBM will issue no shares and buy back existing shares. Both S & D are horizontal at the level determined by backing (or, equivalently, PV(dividends, etc)) so backing, not S&D, determines stock price.
Previous comment somehow got chopped off.
If P<PV(dividends, etc) then IBM will issue no new shares and will buy back existing shares. Thus the share price is determined by backing (equivalently, PV(Dividends, etc)) and not by supply and demand.
That one got chopped too. I’ll try later.
Educated guess: The comments got chopped when I used a “less than” sign. So once more:
If P “less than” PV(dividends, etc), then IBM will issue no new shares and will buy back old ones. Since S&D are both horizontal at the level determined by backing (equivalently, PV(dividends, etc), share price is determined by backing, not by supply and demand for shares.
Sorry you’re having troubles. Maybe you could try writing everything as a plain text file and then cut-and-paste into the message field?
Regarding runs on central banks.
I’d say that every central bank in the world has been facing at least a low-level run for decades. Because central banks consistently debase their currencies, people “run out” on their bank’s circulating liabilities by purchasing other assets.
Many central banks experience medium level runs. In certain transactions (large ones like corporate takeovers, salaries for the elite) the domestic central bank’s liabilities cease to be accepted as payment. Partial dollarization is the result.
True runs on central banks occur when even the lowliest retailer refuses to accept the bank’s circulating liabilities. This is what occurred in Zimbabwe. The run resulted in Zimbabwe dollars becoming worthless and their spontaneous substitution by rand and US dollars, though not before the decimation of Zimbabwe itself.
If you use < directly typepad thinks you’re opening and HTML tag, so it get’s confused. It may get confused by > too. Instead, use < and > (an ampersand followed by lt or gt followed by a semi-colon). More info here
Nick,
There seem to be two relevant, but unrelated, points in your post.
1) The CB can buy high beta assets.
This, as you point out, is great negative feed back. Ideally, for the sake of fairness, they should be broad indices, as well as high risk premium assets like senior CDO tranches. Buying a portfolio of senior corp/sovereign debt tranches, in particular, would provide support for credit spreads in general without benefitting any individual issuers in particular. It also has the benefit of making money creation permanent in the worst possible circumstance.
2) The CB can overpay for assets.
The ECB was already doing this by repoing Greek bonds with “actuarial” haircuts. Then in the face of downgrades, apparently, they had to suspend even that discount. PPIP is another (though more egregious) example – just a way to give money to a special interest while concealing it from the general public. It’s a fraud against the citizens and cannot seriously be advocated as responsible monetary policy.
Nick,
“And presumably people would do this expecting Euro deposits at German banks to be converted into Neu Deutsch Marks? But the ECB is not obliged to redeem its notes for liabilities of the German commercial banks.”
Yes, the expectation is that Euro deposits at German banks and German bonds will be converted into new DM. ECB is not obliged to redeem its notes for liabilities of the German banks. But ECB needs that Germany accepts ECB liabilities as a payment for German bonds. At this moment Germany is collecting premiums for this, as market rate on one week ECB deposits is equal to the yield of 1 year German bonds.
“BTW, I just enrolled your blog in the Palgrave Economics Blog roll.”
Thanks.
K: good points. I like your “high beta” way of thinking about it. I expect that when I said “overpay”, I should have said “paid more than what the fundamental value would have been if the ECB hadn’t managed, by buying them, to raise that fundamental value, by creating expectations of recovery”.
It’s a tricky point, defining fair value, when your very act of buying an asset raises its fundamental value. Which will happen, if its a central bank doing the buying.
Let’s define the fundamental value as the EMH value conditional on Eurozone inflation expectations being on CB target. I think Trichet is quite careful not to deviate too much above fundamental value so defined.
TMDB: That makes sense to me.
‘Our “interest rate fetters” bind our minds just as surely as the old “gold fetters”.’
I feel like I’m back in the sixties, except without the drugs and actually talking about something important…or reading about something important. Great post Nick.
I understand the logic of reducing real interest rates by rising expectations of future inflation. Thus, burning ECB/FED assets on the public square would make the increase in money supply permanent, and ad credibility to a policy of (future) higher inflation. But there is much more that that in my proposal. For Central banks can always reduce M if they want (increasing bank reserves is one way). Also, see Posen on Japan: the lesson he draws is that monetary policy is a very weak tool especially at the zero bound, no matter how much quantitative easing you do. So Krugman is right to say that we should rely mainly on fiscal policy. But he adds: “until it is possible”, meaning until debts are not too large. Burning Gov Bonds in the public square signals to the markets an fiscal autorities that the gov solvency equation (constraint) is being relaxed, and there is more room for fiscal expansion.
So Nick I think your idea of raising expected inflation is nice, subtle, correct. But weak. Academically nice, but we need more. Fiscal policy money financed, is much more dirct and powerful. Regards.