Why don't we observe (macroeconomic) black holes?

Like physics, modern macroeconomic theory predicts the possibility of "black holes". The analogy with physics is close, but not exact. Because unlike physics, where it's not easy to see a black hole, there should be no difficulty in macroeconomists seeing a black hole, if one exists. At the very least, we should certainly see any economy being sucked into one. So where are they? Why don't we ever see any?

If the economy gets too close to a black hole, it can't escape, and is sucked into a deflationary death-spiral. If nominal interest rates are at or near zero, and so at their lower bound, any deficiency of aggregate demand causes increased deflation, which in turn causes increased expected deflation, which in turn causes higher real interest rates, which in turn reduce aggregate demand, which in turn causes increased deflation...and so on. The price level and real output should both fall to vanishing point. Money in a black hole should have infinite value, yet nobody will buy anything with it.

Theory does not predict that black holes will happen. But it does predict that they can happen. And commonsense (backed up by Murphy's Law, in this case) says that, sooner or later, anything that can happen will happen.

So where are they? Why can't we see them? We sure have sailed our macroeconomic spaceships close enough to the boundaries of predicted black holes plenty of times. Why didn't any economy ever get sucked into one, and collapse into an infinitely valuable pinpoint?

Either somebody up there likes us. Or there's something wrong with macroeconomic theory.

Comments

[Phillip Huggan]

After Keynes and Industrial Revolution there is always the option to junk existing financial obligations and I'd guess this behaviourally (psychology not economics) becomes more and more likely the longer you are out of work and noting the hoarding of the means of production. I guess you could use propoganda (education) or 22nd century policing powers to reduce this return-to-Keynes political demand, or you could lose the ability to mine coal or something....it doesn't seem too likely.

[Too Much Fed]

"If the economy gets too close to a black hole, it can't escape, and is sucked into a deflationary death-spiral. If nominal interest rates are at or near zero, and so at their lower bound, any deficiency of aggregate demand causes increased deflation, which in turn causes increased expected deflation, which in turn causes higher real interest rates, which in turn reduce aggregate demand, which in turn causes increased deflation...and so on. The price level and real output should both fall to vanishing point. Money in a black hole should have infinite value, yet nobody will buy anything with it."

I don't think that is right. Think about an all currency economy with no currency denominated debt defaults.

If you are thinking about the situation today in the USA, you might be able to apply some of that to lower and middle class currency denominated debt. And, someday you might be able to apply some of that to gov't currency denominated debt.

Why is it so hard to distinguish between currency and currency denominated debt???

Or, what is the "price" of currency, and what is the "price" of currency denominated debt?

[Rogue]

Extending the analogy to an 'event horizon', Wikipedia states that "any object that approaches the horizon from the observer's side appears to slow down and never quite pass through the horizon, with its image becoming more and more redshifted as time elapses. The traveling object, however, experiences no strange effects and does, in fact, pass through the horizon in a finite amount of proper time."

Maybe some economies are already in the event horizon but since we're all part of the travelling object, we do not experience the effects (or recognize it for what it is), but everything has actually been slowing down. As in an event horizon where time slows but objects stretch,we do not experience the deflation because something has been stretching to make up for it, money perhaps?

[RSJ]

" If nominal interest rates are at or near zero, and so at their lower bound, any deficiency of aggregate demand causes increased deflation, which in turn causes increased expected deflation, which in turn causes higher real interest rates, which in turn reduce aggregate demand, which in turn causes increased deflation...and so on. "

Hmm -- I look at things a little differently:

reduction in net borrowing --> reduction in demand --> reduction in output + prices --> reduction in expectations of growth --> reduction in net borrowing.

Of course, as a loop, you could start with reduction in growth expectations (which you call (risk-free) interest rates -- sorry for the dig :P).

But counteracting this trend is that wages are more sticky than prices (and people have the ability to default on debt). Therefore the worse things get, the higher real incomes are, up until people start spending again via a form of Pigou effect (but for the discounted income stream, not necessarily for stored wealth), growth improves, return prospects improve, and borrowing increases.

Merry Christmas

[Jim Sentance]

You're just asking for abuse aren't you Nick? "Or there's something wrong with macroeconomic theory" Gee.... I wonder....

More seriously, if as a non-macro question I might make a couple of suggestions. First it seems to me your black hole analogy fits best to the private economy, not really anticipating government as a sort of super actor on the side ready to push us away the other way. I think to some extent that's why we haven't slipped into the vortex in the past, as government actions (intelligently planning to do so or unintentionally blundering that way) have saved us by not following the logic of the black hole.

I think the second problem I see is the inability or the limited ability of individuals to follow the same logic. Even though there may be a compelling reason not to, we usually need to buy something - we can't reduce our consumption to zero. Nor do I think people follow the logic of the forces you describe so tightly - they aren't that rational if that's the right word. If macro assumes they do, then yes, there's something wrong with macro.

[Leigh Caldwell]

Stars only collapse into black holes if their mass is over a certain limit. Below this, electromagnetic forces or strong nuclear forces (and the Pauli exclusion principle) keep the individual particles at a non-zero distance, leading to a white dwarf or neutron star respectively.

An economy entering this kind of death spiral would presumably be prevented from total collapse by the analogous "interatomic" forces: subsistence and barter.

Now technically I'm talking about singularities, and a black hole doesn't need to be a singularity. A non-singularity black hole is simply a large enough mass in a small enough space to have an event horizon. Perhaps this equates to a subsistence economy with so many people that it can never achieve sufficient coordination to return to a monetary exchange system (see Krugman's paper on relativistic interest rates for ideas on how this could happen).

The singularity case is, of course, more interesting and more tantalising. Although we have never observed it, the hypothesis is that at some point (from memory, between 10 and 50 suns) gravitational force is greater than neutron pressure and even a neutron star cannot exist. The Pauli exclusion principle is somehow violated, or else each particle exists in some quantum gravitational state for which we currently have no accepted theory; or indeed the particles themselves may no longer have any identity. In an economic system, this may mean that all the agents in the economy begin to fetishise money and assign infinite value to it, in which case they'd spend all their time worshipping their paper currency, stop feeding themselves and die. Despite the rumours, this is not what happened at RBS.

However, the problem with the whole analogy is that unlike the other three forces, gravity has no repulsive mode and thus can never reach equilibrium once it has broken through the barrier of the strong nuclear force. Money is different. If it gains sufficient value through deflation, surely the issuer of the currency - no matter how virtuous - will eventually give in and start printing more. If not at CA$1 = US$1000, surely at CA$1 = US$1 billion?

Who says economists have physics envy?

[Mattyoung]

WW2? About as black as they get.

[Too Much Fed]

EDIT: "... Think about an all 'non-fixed' currency economy ..."

[Nick Rowe]

Leigh: Wow! You do really know some physics! (or are impressively good at faking it!)

"Perhaps this equates to a subsistence economy with so many people that it can never achieve sufficient coordination to return to a monetary exchange system..."

I think the predicted outcome would be a return to barter, which isn't quite the same as subsistence, but might be close enough in practice.

"Money is different. If it gains sufficient value through deflation, surely the issuer of the currency - no matter how virtuous - will eventually give in and start printing more. If not at CA$1 = US$1000, surely at CA$1 = US$1 billion?"

But modern macro theory says that monetary policy is powerless in a liquidity trap! (At least, the sort of modern macro theory i am implicitly criticising says it is.)

Jim: "Even though there may be a compelling reason not to, we usually need to buy something - we can't reduce our consumption to zero."

If we followed the logic of the theory, the real rate of interest would keep on rising, plus people would expect declining real incomes, so would spend less and less, both because of the incentive of high real interest rates, and because they expected to be even closer to starvation in the future. I expect when they hit the point of starvation they would say "well, I'm going to die anyway, so I might as well eat today and die tomorrow" and the economy would hit bottom.

" "Or there's something wrong with macroeconomic theory" Gee.... I wonder...."

You missed the whole point of my post ;-). Since there can't be anything wrong with macro theory, it must be that somebody up there likes us! I was making a theological argument, not a macro one. Can't you tell the difference??

RSJ: Merry Christmas!

"But counteracting this trend is that wages are more sticky than prices (and people have the ability to default on debt). Therefore the worse things get, the higher real incomes are, up until people start spending again..."

You have to be careful here. If P falls more quickly than W, real wages will rise, and wage income will rise as a percentage of total income, but real income (= real output = real aggregate demand) still falls. If the marginal propensity to consume out of wage income is higher than out of non-wage income, that might be a stabilising force though.

Rogue: Neat physics! But I think you are putting more weight on the physics analogy than it can bear. I think Statistics Canada would still be reporting accurate figures on deflation and GDP even when the economy went into a black hole. (At least until we got to total economic collapse, and by that time it would be so bad we wouldn't need StatsCan to tell us how bad it was.)

Too much Fed: "Or, what is the "price" of currency, and what is the "price" of currency denominated debt?"

It's 1 divided by the price of goods. 1/CPI, if you like.

Phillip: The escape valve would be a resort to barter. But we don't see it. Well, we do see it a bit in recessions, but not a total switch to barter.

[bill woolsey]

So, "modern" macroeconomic theory says that real expenditure can only be influenced by increases in real balances (even as they approach infinity) if they lower the real interest rate?

I would say that real balances rise, saving falls, consumption rises, real expenditure begins to rise. The natural interest will rise to meet the real interest rate.

Is there something nonmodern about this?

Is the problem that current deflation rates are always projected into the future? No one spends their infinite wealth on current consumption because they will have more than infinite wealth if they just wait for prices to fall further?

I think all it takes is for people to believe that somebody will spend some infinitely small fraction of their infinite real balances to buy enough real goods and services, that they won't project current deflation to the future and instead may even expect inflation. Has the price level undershot?

By the way, suspending currency payments would be a good idea if this was happening. Then the interest rates on money can be more negative than deflation rates.

[Nick Rowe]

Bill: "Is there something nonmodern about this?"

Yes! It's definitely non-Neo-Wicksellian! Monetary policy IS interest rates, doncha know?

OK. What I'm trying to do is force a contradiction, or at least a paradox. If there exists no mechanism whereby an increase in M/P via increasing M can rescue the economy from a liquidity trap, then there can be no mechanism whereby an increase in M/P via falling P can rescue the economy from a black hole. But we don't observe black holes. Ergo..

Matt: WW2 was certainly bad, but not a deflationary black hole. Mild repressed inflation.

[Lord]

Even more sticky than wages is debt. Those with substantial liquid assets, particularly in risk free assets, would get wealthier. While they could continue to hold out for even greater wealth in the future, they are frequently near or in retirement with shorter time horizons. Those with secure incomes such as government jobs and pensions would also experience something similar. Those with bonds would have to trade off increased real income or accelerated repayment against default. Creditors would prosper while debtors would struggle up to default. Wealth unleashes the liberal impulse. Still, while this provides a floor, it doesn't necessarily turn things around.

There have been a number of examples of reversion to local script to assist in barter under hardship conditions.

[Min]

"If the economy gets too close to a black hole, it can't escape, and is sucked into a deflationary death-spiral. If nominal interest rates are at or near zero, and so at their lower bound, any deficiency of aggregate demand causes increased deflation, which in turn causes increased expected deflation, which in turn causes higher real interest rates, which in turn reduce aggregate demand, which in turn causes increased deflation...and so on. The price level and real output should both fall to vanishing point. Money in a black hole should have infinite value, yet nobody will buy anything with it."

You have described a positive feedback cycle informally. The existence of one does not imply runaway feedback (death spiral), unless the cycle accelerates. It might approach a limit.

"Theory does not predict that black holes will happen. But it does predict that they can happen. And commonsense (backed up by Murphy's Law, in this case) says that, sooner or later, anything that can happen will happen."

That depends upon what you mean by "might". If we think that something might happen because we do not know whether it will or not, that is not enough to say that it will happen sooner or later. Argument from ignorance is invalid.

"So where are they? Why can't we see them? We sure have sailed our macroeconomic spaceships close enough to the boundaries of predicted black holes plenty of times."

We have? I have some questions about that below.

"Why didn't any economy ever get sucked into one, and collapse into an infinitely valuable pinpoint?

"Either somebody up there likes us. Or there's something wrong with macroeconomic theory."

Or it might simply be that macroeconomic theory is incomplete. There is nothing intrinsically wrong with an incomplete theory. It simply means that there are some things that we are ignorant about, and ignorance is our natural state. :)

Now, we have certainly seen runaway inflation, where the value of money has a lower limit (zero). One thing that stops it is the human impossibility of creating money fast enough. Neither printing presses nor loans can keep up the pace. I suppose that that human limitation is not part of normal macroeconomic theory, because it is irrelevant to most situations. Besides, the government can stop printing money, and people can stop lending it, when no terms of interest make sense.

What would runaway deflation look like? Stores mark down prices twice a day, but nobody comes to spend any money? Nobody bakes bread to sell, because they would pay more for the ingredients than the would get for the loaf?

I don't think so. After all, economic activity is almost always a non-zero sum game, with advantages to all or most parties: a win-win. Money facilitates it, but is not necessary.

Wouldn't runaway deflation end in barter? There would be no banks. Money might be a store of value, but it would have ceased to be a medium of exchange. (Like those huge circular stones we read about as children.)

Suppose that we were on a gold standard. In a barter economy gold would still retain some value, which would provide a floor for deflation of a gold-denominated currency. The positive feedback cycle would have a limit. This why I question the statement that we have seen close many close calls to runaway deflation. Hasn't most money been backed by something physical? Something that would retain value in a barter economy?

A fiat currency would have no such limit. However, as long as money has a value at any given point in time, it would occasionally be used in a barter economy. For instance, I might pay $1 for one year's rent. In fact, the theoretical limit for runaway deflation might be a situation where I could pay rent forever for $2: $1 the first year, $0.50 the next, $0.25 the next, etc., etc.

Now, since we would not have any banks, the only money would be physical notes and coins. In the situation I have just described, the value of that money would double each year, without human needs increasing at that rate, nor with economic activity increasing at such a rate. The extra money would be superfluous. Why in the world would everybody be twice as rich in real terms? (Since very little money is changing hands, everybody hangs on to nearly all of it, as a store of increasing value.) This picture does not add up, since the physical supply of money is not halved each year.

Years ago I saw Milton Friedman on TV, saying that during the U. S. Civil War confederate money experienced rapid inflation, except for a period of a few weeks when the printing presses were not in operation. This suggests that in the extremes, inflation and deflation of fiat currency come down to the printing press.

If that is indeed the case, then, rather than allowing the role of money as a medium of exchange to cease, wouldn't the government print more money to stabilize its value? Or even to cause some inflation? Of course, people would have to believe in the government for that to happen, but if they did not, how could you have deflation of fiat currency in the first place?

[Min]

"You missed the whole point of my post ;-). Since there can't be anything wrong with macro theory, it must be that somebody up there likes us! I was making a theological argument, not a macro one."

Thanks, Nick! That's the best proof of the existence of God that I have heard today. ;)

Wait till Thomas Aquinas hears this!

[Andy Harless]

Isn't a black hole just a special cases of a permanent bubble, where the bubble asset happens to be money? In theory, permanent bubbles can happen, but at some point they reach the limits of the theory. People can keep buying tulips for higher and higher prices in the expectation that someone else will be willing to pay an even higher price, but the process becomes more and more unstable as the difference between the market value and the fundamental value increases, as the severity of potential bubble-collapse event increases and makes participation more and more risky.

The same thing should be true for money. As money becomes more and more valuable, and as people accumulate more and more of it, the prospect that it will move toward its fundamental value becomes more and more frightening, and eventually the bubble pops. At least that's what would happen if central banks used price level targets. On the other hand, if a central bank is constantly promising to cushion the inflationary crash by bidding the value of money up to a slow downward path, then....I suppose a black hole isn't too far-fetched. But presumably we still won't observe it, because central banks will eventually learn the error of their ways, or else some cataclysmic event (e.g. WWII) will give them an excuse to suspend their policy.

[Too Much Fed]

Nick's post said: "Too much Fed: "Or, what is the "price" of currency, and what is the "price" of currency denominated debt?"

It's 1 divided by the price of goods. 1/CPI, if you like."

For currency, at first glance I don't believe that is correct.

For currency denominated debt, why isn't it the interest rate or maybe the interest payment?

[Patrick]

I'm with Leigh - Maybe black holes don't exist (or are exceedingly rare), but economic supernovae certainly exist (e.g. Easter Island, Rome). I'd add 'this sucks, I'm leaving' to subsistence and barter. The civilization dies, but the people don't. They just re-group and try again.

[Too Much Fed]

Min said: "If that is indeed the case, then, rather than allowing the role of money as a medium of exchange to cease, wouldn't the government print more money to stabilize its value? Or even to cause some inflation?"

Do you mean print more currency or attempt to create more currency denominated debt?

[Too Much Fed]

Nick and Leigh's post said: "'Money is different. If it gains sufficient value through deflation, surely the issuer of the currency - no matter how virtuous - will eventually give in and start printing more. If not at CA$1 = US$1000, surely at CA$1 = US$1 billion?'

But modern macro theory says that monetary policy is powerless in a liquidity trap! (At least, the sort of modern macro theory i am implicitly criticising says it is.)"

Notice Leigh said print more currency, while monetary policy involves interest rates and therefore currency denominated debt. See the difference???

[RSJ]

Hi Nick,

No my argument does not assume differing propensities to consume, but derives them.

For each dollar you save, you get an expected (financial) return. I claim that people desire to accumulate wealth not just to spend it, but because they obtain satisfaction from having wealth just as much as they obtain satisfaction from eating an apple. Wealth represents things like "success", freedom, security, etc, and is an end in and of itself, beyond any considerations of consumption smoothing. And many people desire to accumulate wealth with absolutely no intention of ever spending it.

But in a deflationary environment, the expected investment returns fall, whereas the amount you can consume per dollar increases. The statement that prices adjust more than wages means that even taking into consideration the inflation-adjusted wealth, still the marginal financial returns are lower from investing than from consuming, and this gap continues to widen the deeper inflation gets. At some point, the various trade-offs for enough of the population are such that it is better to spend than to save. So the net result is that deflation will eventually increase people's marginal propensity to spend once the returns on investing diverge sufficiently from the returns on consumption. At that point, prices stop falling, outlook improves, borrowing increases, and you start reflating again.

If I have time, I will write up a simple model for this, and you can then poke holes in utility functions :)

[Min]

Too Much Fed: "Min said: "If that is indeed the case, then, rather than allowing the role of money as a medium of exchange to cease, wouldn't the government print more money to stabilize its value? Or even to cause some inflation?"

"Do you mean print more currency or attempt to create more currency denominated debt?"

I was basing that comment on Friedman's remark about printing presses, so I had actual printing in mind. I would think that lending would have virtually ceased. Why lend $100 to get paid back $90?

[reason]

Wasn't the Great Depression close enough?

There are two issues here:
1. The "economy" is not partly a command economy and the black hole only applies to the market economy part.
2. People got to eat (so there is a floor at the bottom of the hole.

[reason]

oops
... 1. The "economy" is partly a command economy ....

[Jason]

No black holes, just multiple equilibria.

[Nick Rowe]

A lot of good comments here. I'm not sure I can do justice to all of them.

Lord. I ignored debt for simplicity. But if we add in nominal debt, then as you say, deflation redistributes wealth from debtors to creditors. But I would follow Irving Fisher's Debt-Deflation theory at this point: I don't see that putting a floor under the price level; I see it exacerbating the deflation.

Min: "You have described a positive feedback cycle informally. The existence of one does not imply runaway feedback (death spiral), unless the cycle accelerates. It might approach a limit."

Agreed. But it should accelerate in this case, according to theory, provided expectations catch up to reality. That's because the positive feedback exceeds one.

Unemployment is an increasing function of expected deflation.
Expected deflation eventually adjusts to equal actual deflation.
Deflation equals expected deflation plus an increasing function of unemployment.

"That depends upon what you mean by "might". If we think that something might happen because we do not know whether it will or not, that is not enough to say that it will happen sooner or later. Argument from ignorance is invalid."

Agreed. But in this case, theory says they *will* happen under certain conditions. The precise nature of those conditions depends on the particular parameters of a model. But once we hit the lower bound on nominal interest rates, and there is deflation, and expected deflation, we meet those conditions, unless some lucky shock comes along that is big enough to break the cycle. And the longer we wait, the bigger that lucky shock has to be.

I think you are correct, or at least onto something, when you say that resort to barter will eventually break the cycle. I would then re-write my AD function as:
"Unemployment is an increasing function of expected deflation and a decreasing function of barter". And add that "barter is an increasing function of unemployment".

Even though barter would put a floor under the spiral, we don't see economies hitting that floor. I think. I've not got my head 100% clear on this.

"In the situation I have just described, the value of that money would double each year, without human needs increasing at that rate, nor with economic activity increasing at such a rate. The extra money would be superfluous. Why in the world would everybody be twice as rich in real terms? (Since very little money is changing hands, everybody hangs on to nearly all of it, as a store of increasing value.) This picture does not add up, since the physical supply of money is not halved each year."

Then you are saying that M/P matters, regardless of real interest rates. In which case the standard Neo-Wicksellian macro model (in which money only matters via interest rates) is wrong. In which case, we can escape a liquidity trap with monetary policy.

Andy: "Isn't a black hole just a special cases of a permanent bubble, where the bubble asset happens to be money?" That's an interesting way of looking at it. "Yes" would be my first reaction.

" But presumably we still won't observe it, because central banks will eventually learn the error of their ways,.."

But 'theory' says there is nothing central banks can do, if nominal interest rates hit their lower bound!

RSJ: I think I am following you. But I don't think you can escape a black hole that way. Suppose that when AD hits some lower level, increasing real interest rates don't have any effect on AD thereafter. That puts a floor on output and employment. But we are still left with excess supply of output and labour. So if the standard Phillips Curve is correct, deflation should continue to accelerate.

And I think that applies to reason's comment too.

Jason: but why don't we observe some of those equilibria?

[Min]

Nick Rowe: "I think you are correct, or at least onto something, when you say that resort to barter will eventually break the cycle. I would then re-write my AD function as:
"Unemployment is an increasing function of expected deflation and a decreasing function of barter". And add that "barter is an increasing function of unemployment".

"Even though barter would put a floor under the spiral, we don't see economies hitting that floor. I think."

As Lord and others pointed out, we do see barter increase under deflationary conditions. My grandfather was a country doctor during the Great Depression, and often got paid in chickens.

The state government, perhaps unconstitutionally in the U. S., created money in the form of tokens worth less than a penny.

Moi: "In the situation I have just described, the value of that money would double each year, without human needs increasing at that rate, nor with economic activity increasing at such a rate. The extra money would be superfluous. Why in the world would everybody be twice as rich in real terms? (Since very little money is changing hands, everybody hangs on to nearly all of it, as a store of increasing value.) This picture does not add up, since the physical supply of money is not halved each year."

Nick: "Then you are saying that M/P matters, regardless of real interest rates. In which case the standard Neo-Wicksellian macro model (in which money only matters via interest rates) is wrong. In which case, we can escape a liquidity trap with monetary policy."

I did not know that I was challenging any theory. ;) Anyway, here was my idea. Does not the main value of fiat currency rest upon its being a medium of exchange? (Not entirely so, witness the huge circular stones.) If money is not really being used as a medium of exchange, but is occasionally bartered, like everything else, two questions arise. First, if you have more paper dollars than I do, why should I agree that you are richer than I? OC, the government says so, and it will take that money in payment of taxes. Also, the same could have been said of the stones, but they retained their value, anyway. People are funny. Second, why should the value of those dollars increase, if they are just sitting there? Again, most people may be a little richer, but why should they agree than people with more money than they are even richer? We are not just talking about nominal values. Why should the rich get richer for no reason at all? Sure, the value of dollars would increase if the government agreed to accept fewer of them in taxes next year in return for the same services. But why should the government do that? In fact, why should the government not print enough money so that it becomes a medium of exchange again?

The value of money is a social construct. In extremes of inflation and deflation, the social contract threatens to come apart, and the normal assumptions do not apply. Other forms of positive feedback can threaten the social fabric. Escalating conflict can lead to the dissolution of partnerships and families, to long lasting feuds and wars. Even if these conflicts eventually cease, they do not do so under the rules of the old understandings, which may contain most conflicts. There is no particular reason to expect theories that describe the normal workings of society to hold in extremis. The assumptions upon which they rest are unlikely to hold.

[winterspeak]

Fiscal

[JKH]

Why expect economists to see a black hole when they don't understand the reserve system yet?

[winterspeak]

There's definitely a black hole *somewhere* around here ; )

best line: "Either somebody up there likes us. Or there's something wrong with macroeconomic theory."

[winterspeak]

JKH: btw. not just economists. Check out this thread where we have Michael S, a bank owner, saying that his bank lends out deposits.

https://www.blogger.com/comment.g?blogID=7958140996781104565&postID=4627007655494567487

Now, the conversation is ongoing, and maybe he hasn't thought through things in this way, or maybe there is some misunderstanding over terminology. Or maybe I'm misunderstanding.

[JKH]

winterspeak,

You may be referring to:

"Let's suppose our bank has picked up $100 million in new deposits and has to decide what to do with the money."

That's generally not the way or is at least a gross simplification of the way banks or bank treasury functions work, particularly larger banks. Banks are particularly liability management sensitive in terms of the effect of deposit inflows on reserve positions. If for some reason a deposit inflow was unanticipated, it's almost certainly a short term money market flow. Bank money market desks deal with balancing inflows and outflows of this type all the time. The important point is that the focus of balancing is on the reserve account (or settlement account if you like, where reserve requirements are zero, as in Canada.) E.g. short term wholesale deposit inflow boosts reserve account by $ 100 million; purchase of $ 100 million in treasury bills produces outflow; reserve account net flat. That sort of short term management is quite separate from banking lending into longer term risk assets, which requires capital. Drawdown is accompanied by notification to the money market desk which just throws the expected reserve effect into the pot of everything else it has to deal with.

But I think your point on banks not lending deposits may be subtler than that. The fact is that the effect on micro reserves of deposit inflows and outflows has a vague similarity operationally to the effect on macro reserves of government expenditures and taxes, respectively. Deposit inflows increase reserves. Outflows reduce them. This is akin to creation and destruction from the micro world perspective of an individual bank.

The moldbug thread looks interesting. I'll have a longer gander.

[Mattyoung]

Justin Fox points to a serious deflation that occurred when the Romans de3parted Britain in 400 AD.

http://curiouscapitalist.blogs.time.com/2009/12/23/the-latest-economic-indicators-were-doing-better-than-5th-century-britain/?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+timeblogs%2Fcurious_capitalist+%28TIME%3A+The+Curious+Capitalist%29

Did the British rebels know how much of their economy depended on Roman terms of trade for Europe?

[JKH]

winterspeak,

Looking at the conversation between Michael S. and zanon more closely, zanon is more correct than Michael, but not entirely correct.

He's correct in the sense that lending requires no pre-existing reserves or deposits in either a macro or micro sense. As noted above, a large bank facing a draw down of $ 100 million on a new lending facility will simply notify its treasury desk that the draw down is coming. (This assumes immediate disbursement of the loaned funds by the borrower into the rest of the system.) Treasury can then handle the expected reserve outflow effect by offsetting it entirely through liability management if it wants to. This can be related to the "loans creates deposits" theme in the sense that new money will be created at the macro level on the day of the draw down, and the lending bank simply has to go into the market, after the fact of the draw down, to source an offsetting reserve inflow by attracting new deposits through the wholesale money market. The cause is the loan, the effect is the deposit.

As Michael S. says, that could be handled by selling liquid assets also. But the point is it doesn't have to be.

Where zanon is wrong is in his assumption that the required offsetting reserve inflow must be attracted directly from another bank. (btw, this is a common error made by PK readers.) In fact it normally isn't. The normal operation is to attract non-bank deposits where the source of the money coming in is drawn from non-bank accounts with other banks. The reserve inflow is attracted indirectly rather than directly.

It's all consistent with PK causality.

[kevin quinn]

Nick: I'm posting in a hurry and just scanned the other comments, but how about Pigou's solution? M/P and consumption?

[Lord]

Debt is a variant of the government response as it relies on government behaving differently than the rest of society. While government can counter deflation though, I see no reason it has to. I can readily imagine if Hoover had a longer term we would have had even more deflation during the depression. While barter, subsistence, and in kind services may be an explanation for the real economy, it wouldn't explain the absence of a black hole in the measured economy, why all debt wasn't repaid or repudiated and all transactions didn't cease. Can money retain its transaction value even as it ceases to be transacted? It's a mystery.

[Barry Ickes]

I am not sure what theory you refer to? I guess textbook Keynesianism. I think standard theory, at least since Patinkin, would not the real balance effect, which would cause real wealth to rise when prices fall and restores equilibrium.

I suppose you could argue that deflation also causes bankruptcies ala Fisher, but that is not really standard theory. Standard theory models do not predict what you say they do.

[winterspeak]

The moldbug post is a mix of Austrians and computer scientists, with zanon and Michael S thrown into the mix. Beware the site in general, it's ridiculous.

Michael S obviously has some operational experience. I don't think that he's PK though, although if he thinks about his operations more closely, he may get there.

Thank you for your note on how "The normal operation is to attract non-bank deposits where the source of the money coming in is drawn from non-bank accounts with other banks. The reserve inflow is attracted indirectly rather than directly."

Can you help me understand what a non-bank deposit, or non-bank account with another bank?

[JKH]

By non-bank deposit I just mean the deposit of a non-bank (e.g. Microsoft) with a bank. By non-bank account I mean a chequing or operating account of Microsoft with its bank.

In my example, suppose JP Morgan is the lender and Exxon is the borrower. Exxon draws down funds to pay for some project, disbursing funds away from JPM immediately. JPM's loan officer apprises it's treasury of the loan drawdown and JPM treasury issues a 1 month CD to Microsoft. Microsoft pays for the CD by writing a cheque on its chequing/operating account with Wells Fargo. Wells pays JPM in the reserve clearings/settlement process.

(Sorry. I do tend to be vague as well as dense in my writing. It's a powerful combination :))

[Doc merlin]

There is no such thing as a deflationary spiral. I have never ever seen it happen, never watched it, never seen it talked about in history. However, we have seen multiple times, INFLATIONARY death spirals.

Deflation is its own cure, because as prices drop, people will naturally want to consume more.

"Is the problem that current deflation rates are always projected into the future? No one spends their infinite wealth on current consumption because they will have more than infinite wealth if they just wait for prices to fall further?"

Thats baloney because of discounting. People would rather have stuff now than later.

Anyway, I can't accept Keynsian economics because it has these absurd a-physical results. Like having no inflationary death spiral, but it does have a deflationary death spiral.

[JKH]

Be clear that the operational issuance of the CD occurs following the approval of and even the draw down of the loan.

[Nick Rowe]

Kevin Quinn and Barry Ickes both come up with the "theoretically correct" answer.

Here's Kevin: "..how about Pigou's solution? M/P and consumption?"

Here's Barry: "..standard theory, at least since Patinkin, would not the real balance effect, which would cause real wealth to rise when prices fall and restores equilibrium."

Both are referring to the Pigou effect, the effect of a lower price level on raising the real quantity of base (outside) money, thereby increasing wealth, and thereby increasing consumption.

BUT:
1. We have been told that the Pigou effect is trivially small, and of no empirical significance, and we should ignore it.
2. What I am calling "modern" macro theory, the Neo-Wicksellian approach that central banks use to describe their own behaviour, and as taught in New-Keynesian grad skools, has no role for M/P in its models. Monetary aggregates play no causal role. Monetary policy can be described by interest rates alone. (And this same feature is shared by Post-Keynesian macro too.)

[winterspeak]

JKH: I see what you mean. Can't speak for zanon, but I think (s)he meant what you said. I don't think PKs are careful in their terminology when they say a bank borrows reserves from another bank. They may mean from a bank's own account, or a non-bank account at a bank. Either way, it all happens "behind the scenes".

When the loan is created, a deposit has to be create SOMEWHERE. That then has to feed into the reserve system SOMEHOW (unless it's held as cash). Once in the reserve system, it has to get to banks short reserves SOMEHOW. The number and structure of bank intermediaries in his process gets abstracted away. I don't know if anything important is lost there.

[JKH]

winterspeak,

I see what you're saying as a useful abstraction. But there is an operational distinction. The only direct MARKET for reserves is the interbank fed funds market, which is definitely not "behind the scenes". It's the only direct access to reserve "trading". That's different than the role of reserves when used as the means of settlement in the payment system, which is the clearing of reserves between banks in net settlement of customer payments. PK MMT is operationally oriented after all.

[Too Much Fed]

I think this applies here.

Titled "Conducting Monetary Policy when Interest Rates Are Near Zero"

http://www.clevelandfed.org/research/commentary/2009/1009.cfm

[Too Much Fed]

From the link above:

"It is important to stress that the extreme version of this scenario—the black hole Krugman refers to—is unlikely to occur, partly because firms anticipating a drop in demand will eventually cut production enough to stop excess supply. Nevertheless, our inability to offset a deflationary shock could conceivably prolong a period of deflation and falling output."

Productivity growth could go negative too?

Might be other reasons?

[winterspeak]

JKH: Help me understand the operational distinction.

I bank at bank A and write a check which gets deposited at bank B. When the check clears, bank A gets a liability debit and a reserve debit. Bank B gets a liability credit and a reserve credit. Bank A now may be short reserves, while bank B is long.

No new credit extension here, just a transfer.

I assumed that the reserves could balance in exactly the same ways as if there was a loan, ie. either through other non-bank account to non-bank account transfers, or via trading on the interbank market. Am I wrong?

[Nick Rowe]

Doc: "Anyway, I can't accept Keynsian economics because it has these absurd a-physical results. Like having no inflationary death spiral, but it does have a deflationary death spiral."

That's an interesting point.

If we take the standard Neo-Wicksellian model, then inflationary spirals are possible, *but it is always possible for monetary policy to stop an inflationary spiral, just by raising nominal interest rates sufficiently*. According to the same model, it is *not* possible for monetary policy to stop a deflationary spiral, once nominal interest rates hit the lower bound. So we would expect deflationary spirals to be more common that inflationary spirals.

But, as you say, we observe inflationary death-spirals (Zimbabwe etc), but we do not observe deflationary death-spirals.

Min: I like your comment, and tend to agree. But it's probably too much to ask of a merely macroeconomic theory that it incorporate a breakdown of the legal-political system endogenously.

[Nick Rowe]

Too much Fed: Yes, it was exactly reading that post http://www.clevelandfed.org/research/commentary/2009/1009.cfm
that inspired me to write this post.

I really ought to do a follow-up post to draw out explicitly the relation between the non-existence of black holes and the use of monetary policy to escape a liquidity trap.

But I'm not quite yet ready to tackle that task.

[JKH]

Winterspeak,

You’re right that the reserves could balance in exactly the same way as if you drew down a loan at bank A in the form of a cheque or bank draft or something (i.e. without a deposit account) and presented the cheque/bank draft as payment to somebody else who deposited it with their bank B. That too is a “behind the scenes” clearing of reserves to use your phrase above.

Now compare that with the inter-bank fed funds market. Again you’re right that reserves could balance in exactly the same way.

But the fed funds market is no longer “behind the scenes”. It is accessed directly by banks that clear through the fed. There is a direct transfer of reserve account funds from Bank A to Bank B. That is the operational difference.

In other words, the fed funds market is a PRINCIPAL market in reserves; the cheque and other instrument clearing market is an AGENCY market in reserves.

The Fed funds market is the ONLY market where participants have direct access to reserve account transactions. And banks that clear with the Fed are the ONLY participants as principals that can transact in that market. This is an illustration of why it is incorrect to say that banks “lend reserves” in their normal day to day operations. They only do that with each other through the fed funds market and there they are sometimes described as buying and selling reserves as terminology (rather than borrow and lend).

There is also a real time information and control advantage in the fed funds market. Banks know exactly what the net reserve effect is at the time. Conversely, it’s not always feasible to predict where Exxon is going to end up directing its money. Exxon may well be paying somebody who also banks with the lending bank, but even that’s no guarantee that the funds don't have an additional way to go through the money market. Predicting and tracking agency clearings is no trivial task.

Finally, to make things interesting, banks can also transact in the interbank Eurodollar market, which is distinct from the Fed funds market. It is again one step removed from direct transfer of reserves. Payments through the Eurodollar market will ultimately be settled in Fed funds, but only as an additional step in the process. And New York banks run “offshore” Euro books in addition to their domestic fed funds books. Very circuitous stuff.

[JKH]

too much fed / Nick / winterspeak

I trashed the Cleveland Fed report at Mosler's and Billy's.

Cleveland Fed staffers don't understand the operation of the reserve system.

[winterspeak]

JKH: Thanks for the details. From a consumer's perspective (I am not in finance, more's the pity) it's all behind the scenes!

Will check out Mosler and Billy. Makes you weep though, doesn't it? With these monkeys in the operating theatre, would you put your money in S&P?

[JKH]

Nick,

I don't recall, but maybe you've posted on the theme of how the economics profession has been "captured" by the fact that so many economists are employed by the Fed? There's certainly been a lot written about that sort of thing over the past year.

In any event, there's no reason to believe that staff economists understand operations at the Fed any better than commercial bank economists understand their own banks' operations. And I can tell you that they don't. They're too insulated. The separation is more effective than a Chinese wall.

[JKH]

winterspeak,

Yes, absolutely from the consumer's perspective, it's all behind the scenes. I see what you mean now.

[JKH]

Nick,

My point was not pejorative. There's simply no opportunity for them to get close to the actual money flows and money desk operations, and there's too much too learn relative to all their other responsibilities.

[Lord]

The real balance effect really seems too small, for those that missed Krugman, below. While housing and stocks have rents and earnings to cushion their fall, when they fall as well there is nothing to restore them.

*Somebody is going to ask, what about the real balance effect? Doesn’t a falling price level make people wealthy, by raising the real value of the money they hold. The answer is, consider the magnitudes. Before the crisis, the monetary base — the system’s “outside money” — was around $800 billion. (It’s a much more confusing situation now, so I won’t try to parse the current numbers here). This means that even a 10 percent fall in the price level, which is very hard to achieve, would raise real wealth by only $80 billion. Compare this with the effects of the decline in housing and stock prices, which reduced household wealth by $13 trillion in 2008. The real balance effect is totally trivial.

[Too Much Fed]

JKH, could you post a link(s)? Thanks!

[Too Much Fed]

JKH, could you do a short "analysis" (reserves, capital, & other) of a person getting a NEW mortgage to finance a NEW home?

Could you do it with a bank and a non-bank? Thanks in advance!

[winterspeak]

JKH: Maybe you caught this link at UR.

http://www.bis.org/publ/work292.pdf?noframes=1

It's not bad. Talks about how CBs have a signalling channel and a portfolio balance sheet channel, and goes on to say how the balance sheet channel does not work well when the CB is not the marginal monopoly supplier (they way they are with reserve balances).

[JKH]

too much fed - I'll do something in brief sometime tomorrow, here

[JKH]

winterspeak,

Yes, I saw that early this month when I was off site (from civilization).

Bill Mitchell did two commentaries on it.

Overall, it's a superb paper. There are some minor MMT quibbles about it, but nothing too serious.

Every economist should read this paper 10 times.

[Too Much Fed]

JKH, OK and thanks!

[Kaleberg]

There are a number countervailing forces that tend to stop a deflationary spiral. First of all, there is bankruptcy. When debt service gets too expensive, the debt stops getting serviced. Then, there are assets of value unrelated to the repayment of debt. Even after all debts have been written off, there are still people with money to spend. Finally, there is the spender of last resort, the government, which can issue a new currency to replace the old one and by fiat, prime the pump for economic activity.

You can probably get more insight if you consider how hyperinflation ends. There are actually fewer countervailing forces, but eventually someone puts together a new currency and the economy starts anew, but without any cash savings.

[Ramanan]

JKH,

Good points about CD issuances to corporates like Microsoft and indirectly getting the reserves. Somewhat funny - if JPM needs $10M of reserves and if Microsoft happens to have an account at JPM, the JPM issuance of CD say worth $10M to Microsoft just drains $10M deposits and it still needs to look for $9M assuming 10% RR. If M$ happens to have an account at Citi, JPM gets the reserves worth $10M.

Eurodollar complications of this market is tricky and I haven't analysed it properly. Do write if you have more to say.

Banks may also issue CDs to each other, raise term deposits - from households as well. Banks can sell each other bonds to get the reserves and can sell bonds to corporates like Microsoft. Banks also do repos with corporates to get the reserves indirectly.

Also different accounts have different RR and banks have been innovative to create accounts which have zero RR.

All this is a huge "optimization" problem since it is just a part of the more general ALM.

On a slightly different note, I completely agree with Bill/Warren on the fact that this market is a waste of lot of time - set the interest rates to zero and let the Fed lend in unlimited quantities if needed. Having said that, the market is very interesting - before landing at Billy Blog, I used to wonder why the money market is the way it is etc ....

[JKH]

Too much fed,

Highly simplified:

Banks keep surplus capital invested in treasury bills or other low risk assets, in order to be prepared before taking on new risk assets. If they make a mortgage loan of 100 and require capital of 8, and if excess capital has been invested in treasury bills, they’ll sell 8 in bills to free up capital for the loan, and raise 92 in deposits. The sale of bills and the new deposits together attract reserves to offset the reserve outflow created by the mortgage loan. For a retail bank, deposits may be wholesale initially, replaced by growth in retail deposits over time. Immediate access to the wholesale market or existing liquid assets means neither deposits nor reserves need to be in place before the loan is made. The fact that loans create deposits means that sufficient money is out in the system somewhere in order to be able to attract sufficient funds, as winterspeak describes at December 23, 2009 at 05:29 PM above. (Some of the system deposits created by the mortgage advance effectively go toward buying the treasury bills in my example, on a net basis. The rest go toward the 92 in new deposits.) The PK MMT contrast is that adequate capital unlike bank reserves and deposits must be in place before putting the asset on the books. But the capital requirement is not in the first instance a funding matter; it is a risk management requirement. It’s just that capital as a risk management requirement also becomes a partial source of funds to support the loan. If you want to get more technical, the risk protection attribute of capital rests in that it contains an embedded short put option (as well as a long call option), which can be separated conceptually from the funding effect of what is otherwise a risk free asset. The idea that capital is available to absorb losses really refers to the short put option exposure that capital has to such losses. The immediate funding impact is an additional plus. By contrast, credit default swaps only offer the put option effect up front – the funding to cover losses comes when the losses occur – or not, as we’ve seen. Risk can always be boiled down to embedded options (winterspeak – that’s also the idea behind Steve Waldman’s latest analysis of government guarantees on deposits, where the government is effectively writing put options to bank depositors in order to immunize their risk and convert their deposits to risk free status).

A non bank will do roughly the same thing as a bank, although the capital requirement and funding sources may be quite different. The regulatory requirement for non-bank capital if any will not fall under the bank regulatory scheme. Funding sources will include some sort of borrowing rather than deposits. The non-bank lender must keep a bank account with a bank, through which all lending and borrowing flows occur. It must cover its bank account outflows with inflows obviously, just as a retail customer would over time. While there is no bank reserve flow directly through this account, any activity will ultimately affect the reserve position of its bank as clearing agent. The reserve effect on its bank should net out, under the assumption that the non-bank lender is dealing with asset and liability counterparties to the transaction that themselves bank elsewhere. But these flows blend in with all sorts of other transactions occurring across the economy and across different banks. Particular transactions get lumped in with everything else from an individual bank’s reserve management perspective. Mortgage securitization vehicles work roughly like non-banks as described.

[Nick Rowe]

JKH: "Nick, I don't recall, but maybe you've posted on the theme of how the economics profession has been "captured" by the fact that so many economists are employed by the Fed?"

I remember reading a couple of blog posts on this topic a few months back, but have never posted on it myself.

My only post relevant to that topic was my post on "Churches and Central banks", which said that we don't have to accept central bankers' theory of their own behaviour.

My own take is that I am surprised by how little central banks have captured the economics profession. There's a whole school of macroeconomics (Real Business Cycle Theory) that says that central banks are basically irrelevant. And New Classical macro, which was very popular in academia in the 1970's, was never in line with CB's own view of the world.

And watching economists within the Bank of Canada, for example, for many of them it seemed their research agenda was driven more by their previous academic exposure than what was required for monetary policy to work well.

If I had to push a "capture" theory, I would stress how much macro tends to be captured by whatever is trendy in the journals and top grad skools.

I like your point about how even people working within a large organisation tend to have a very stylised view on those bits of the organisation outside of where they work.

Winterspeak: Thanks for saying I'm "charming" (I bet you say that to all the boys ;-) )...but calling this post "banal"?? Don't you endogenous money types realise that this post presents you with an empirical problem, just as much as the Neo-Wicksellians?

[JKH]

Nick,

"If I had to push a "capture" theory, I would stress how much macro tends to be captured by whatever is trendy in the journals and top grad schools."

As I recall, vaguely, the idea was that the Fed economists had sufficient reach to have a meaningful influence on what was trendy. But as you say there must be major exceptions. I doubt there are a whole lot of Austrians on the Fed staff.

[winterspeak]

JKH: I would suspect the capture goes the other way. The Academy has captured the Fed. Certainly, this is how it works in other branches of Govt.

[winterspeak]

Nick: I continue to be charmed by your graciousness. No lie! Maybe I can associate this with your being Canadian?

And you know full well why I think the post is banal.

[RebelEconomist]

Merry Christmas to all!

Nick, I note that Min (on December 22, 2009 at 11:03 PM) made the same point - but more eloquently - here that I tried to make a few posts ago about the idea that inflation accelerates without limit if the central bank tries to hold the interest rate below the natural rate. That is, that the feedback loop can converge. In your response to Min, you assert that the gain of the feedback loop is greater than one. What makes you say that?

A point that I think applies to academic economics more generally is that you make the analysis just mathematical enough to impress laymen and terrorise students, and to yield striking conclusions when stretched, but exclude details like the real balance effect which are minor in normal circumstances that would complicate the mathematics beyond the ability of most economists but which may well become vital in just the kind of extreme situations that you are using the model to explore. In this case, I agree that the fact that people have to spend some of their money to subsist, no matter how much it might be worth tomorrow, is probably the key complication that would impede the loop (and make it convergent?). I would also suggest that, if the base money in the economy was supplied by buying a real asset like gold, the value of money might be restrained by the fact that net worth of the issuing central bank would be increasingly negative as its money increased in real value.

[Scott Fullwiler]

JKH & Winterspeak . . . . interesting thread within the context of Nick's overarching theme.

Regarding PK, haven't seen many, if any, go into the details of payment settlement JKH is talking about here. That's been one of my areas of focus, though. I did a paper a few years back that got into (a bit) about the direc/indirect relations, though it was rather peripheral to the issues discussed here. I would say that JKH's description matches exactly my understanding (I'm probably more relieved than JKH is to know that) . . . and would only add that most of these "indirect" avenues use netted settlement via reserve accounts.

The Borio/Disyatat paper from BIS is quite good. Disyatat did another at BIS earlier in the year that also was pretty good. Another to look at is research by Ulrich Bindseil at the ECB; he's been at the their trading desk (or similar) and has published some pretty good stuff on cb operations and reserve management. At the Fed, William Whitesell at the Board of Governors has some good stuff on Fed operations.
Best to everyone . . . happy holidays!

[JKH]

Nick/winterspeak,

Here are some examples of what I was talking about re Fed influence on mainstream economics:

From Lawrence White:

“It is relatively straightforward to document how the Federal Reserve System’s research program pervades American monetary economics ... Although the research departments of the regional Reserve Banks seek to establish their own reputations, their incentives would seem to steer them away from research that would challenge the monetary regime status quo favored by the Board of Governors. By contrast, Fed economists are not reluctant to recommend sweeping changes in other government financial institutions, such as Fannie Mae or the Federal Deposit Insurance Corporation (for an example see Eisenbeis and Wall 2002). By extension, an academic economist who values the option to someday receive an offer from the Fed, either to become a staff economist or a visiting scholar, faces a subtle disincentive to do regime-challenging research. To repeat Fettig’s (1993) characterization of Milton Friedman’s view: “if you want to advance in the field of monetary research . . . you would be disinclined to criticize the major employer in the field.”... The Fed has an institutional interest in preserving the legal restrictions that generate its seigniorage revenues and the privileges that give it discretionary monetary policy and regulatory powers. Fed-sponsored research generally adheres to a high level of scholarship, but it does not follow that institutional bias is absent or that the appropriate level of scrutiny is zero.”

From:

http://econjwatch.org/articles/the-federal-reserve-system-s-influence-on-research-in-monetary-economics

See also:

http://www.parapundit.com/archives/006549.html

http://econlog.econlib.org/archives/2009/09/the_feds_hold_o.html

[JKH]

Scott - good to know, thanks

[JKH]

Re my 10:45 above:

Lawrence White actually comments at the econlog post

And here's a longer Huffington post article that's referenced there:

http://www.huffingtonpost.com/2009/09/07/priceless-how-the-federal_n_278805.html

[Nick Rowe]

Rebel: Merry Christmas!

" In your response to Min, you assert that the gain of the feedback loop is greater than one. What makes you say that?"

Here's how I think about it:

In {Price level, real income} space, the LRAS curve is vertical, and according to the theories I'm criticising, the AD curve is vertical too (for a given nominal interest rate, and given expected inflation/deflation.

So there are 3 possibilities:
1. AD lies exactly on top of LRAS (by sheer fluke, or perfect monetary policy). The equilibrium is like a frictionless ball on a flat table.
2. AD lies to the right of LRAS. No equilibrium. Ever rising price level, and ever accelerating inflation, given a standard Phillips Curve, where expected inflation eventually adjusts towards actual inflation. Ball on a tilted table.
3. AD lies to the left of LRAS. Same as 2, only ever accelerating deflation.

Plus, if actual inflation causes expected inflation, which lowers the real interest rate for a given nominal rate,, the AD curve shifts rightward over time in 2, and shifts leftward over time in 3. Like a ball on a curved table, that curves down from the middle.

JKH: Yes, that must have been what I read. There may be a point to what Lawrence White is saying. But thinking about the discussion papers put out by economists at the Bank of Canada, if I had to choose between two hypotheses:
1. "This is what Centre Block told me to write"
2. "This is what my thesis supervisor told me to write"
I would go for 2.
(The Bank of Canada building is in 3 parts, and Centre Block is where the power lies.)

[Nick Rowe]

Plus, from my personal experience of a sabbatical at the BoC, I found Centre Block were on average much more receptive to my (rather eccentric) ideas, even when my results could make them look bad. It's because they thought they might be relevant to what they were doing.

[RebelEconomist]

winterspeak,

Thanks for posting the link to Borio and Disyatat's paper. I did see their powerpoint slides from the Bradford Money, Macro and Finance seminar, but the paper is even better. That paper should be an example for all academics to follow: worthwhile, clear, rigorous and no superflous equations. Their discussion of the influence of reserves and how QE is supposed to work asks similar questions and reaches similar conclusions as our discussions here have over the past few months. My only slight difference with their taxonomy of QE was that I would have distinguished between term and liquidity easing. One hopes that the QE central banks read it.

[RebelEconomist]

Nick, I am afraid that your response (on December 26, 2009 at 09:29 AM) to my question (about feedback gain) simply relocates the question somewhere else; that is, why is the AD schedule vertical? It is not easy (at least not for me) to see why.

[Scott Fullwiler]

Hi Rebel,

I agree that the Disyatat/Borio paper is very good. My problem, as an academic, is that there is virtually nothing original in the paper that perhaps dozens of scholars from MMT, horizontalist, etc., camps haven't said MANY times over. If one of my students had handed in much the same paper a few weeks prior, I would have had to fail them if they hadn't given proper citation of previous works.

Best,
Scott

[Too Much Fed]

JKH at December 24, 2009 at 07:24 AM, thanks! I have some questions.

"If they make a mortgage loan of 100 and require capital of 8, and if excess capital has been invested in treasury bills, they’ll sell 8 in bills to free up capital for the loan, and raise 92 in deposits."

In money terms, what do they get for selling the 8 in bills for capital?

"Immediate access to the wholesale market or existing liquid assets means neither deposits nor reserves need to be in place before the loan is made. The fact that loans create deposits means that sufficient money is out in the system somewhere in order to be able to attract sufficient funds, as winterspeak describes at December 23, 2009 at 05:29 PM above."

What if there is not sufficient money in the system? For example, what if the fed is trying to expand currency denominated debt thru mortgages (or even borrowing against the home, as in home equity)? If a new mortgage is created, could that expand the demand for reserves, and if the fed wants to maintain the interest rate where it is, will the fed just create NEW reserves "out of thin air"?

Am I making any sense and feel free to correct anything?

[Too Much Fed]

JKH, I would like to go over the two(2) links below with you (the links are from earlier). Sound good?

http://www.clevelandfed.org/research/commentary/2009/1009.cfm

http://www.bis.org/publ/work292.pdf?noframes=1

[Too Much Fed]

RebelEconomist said: "Nick, I note that Min (on December 22, 2009 at 11:03 PM) made the same point - but more eloquently - here that I tried to make a few posts ago about the idea that inflation accelerates without limit if the central bank tries to hold the interest rate below the natural rate."

What if there are group(s) suffering "negative real earnings growth" and other groups(s) that have "positive real earnings growth"?

[JKH]

Too Much Fed

Keep in mind I said the model was highly simplified. By construction, it is an “other things equal” approach in explanation.

The reason to sell the 8 in bills is to manage the balance sheet. Banks “park” excess capital in such investments, pending deploying of capital in risk assets. When capital is required, they sell the bills in order to maintain their bill position, net of what capital has been invested in, at the same level as before.

At the level of reserve effect, the default assumption is that the 100 loan, 8 in bills, and 92 in deposits are all transactions that affect the reserve position. I.e. they are transactions where the counterparty is banking with another bank, so that funds are deposited with/withdrawn from other banks.

With that default assumption, the sale of bills results in a payment to the lending bank that is a customer transaction (e.g. cheque) drawn on another bank. The lending bank then gets credited with a corresponding reserve payment from that bank.

Any example of this type gets more complicated when you start assuming counterparties that bank with the lending bank, where reserve positions are only partially affected. It’s not necessary to get into that, given the numerous possible permutations. Banks manage their balance sheets and reserve positions ongoing. But for example, if the lending bank sells bills to its own deposit customer, both the bills and the deposit will disappear from its balance sheet without any effect on reserves. Assuming the outgoing mortgage loan had a full (negative) effect on reserves, the bank could then raise an additional deposit of 8 externally (i.e. coming in from another bank) in order to flatten its reserve position. In either case though, the bank is proactively managing the size of its bill position in response to the change in the level of surplus capital invested in bills.

BTW, this idea of surplus capital being invested in risk free assets is an important one in terms of demonstrating MMT operational realities. It’s a reflection of the fact that banks do maintain surplus capital and manage it in such a way that is very different than the way in which they manage their central bank reserve positions. Such surplus capital positions are strategic. Over a longer period of time, banks actively manage their surplus capital levels via retained earnings, dividend policy, share issues and share buybacks, and of course deployment in risk assets. They also manage their broader liquidity position on a similar strategic time frame, including by the way the deployment of surplus capital in a subset of low risk liquid assets. But the central bank reserve position itself is a very short term operational matter into which everything else feeds on a daily basis. There is no strategic build up of central bank reserves in order to “fund” lending. That’s a core message of MMT.

[JKH]

Too Much Fed

“What if there is not sufficient money in the system?”

The Fed always ensures that sufficient reserves are made available in order to prevent the overnight interest rate from permanently exceeding the target level. Banks who are short reserves can get them by selling assets to non-banks or banks, attracting deposits from non-banks, borrowing directly from other banks, or borrowing from the Fed.

BTW, you can order these alternatives to some extent on a last resort basis. The ultimate last resort is the Fed. The penultimate last resort arguably is borrowing from other banks. Banks generally prefer to attract deposits from non-banks first, demonstrating that they don’t have to rely on other banks or the Fed to prop up their reserve positions. They like to think they have a strong customer funding base apart from other banks. It’s actually one measure of liquidity strength (i.e. a liability side measure).

[JKH]

Too Much Fed

“Sound good?”

It sounds daunting. These are substantial length papers. Perhaps if you could focus on specific references to the text in either case, we could try a couple of rounds, at a measured pace. No guarantee I can hang in though.

[JKH]

Scott,

I described the Disyatat/Borio paper as “superb” earlier on, partly for the reason that in addition to being good it wasn’t coming from the MMT epicentre, i.e. it was a good thing that the right ideas were starting to come out in a comprehensive way from a source that was disconnected from the MMT core, relatively speaking.

That said, I’d be pissed if I’d been part of that center from a publishing perspective, and the attribution that was owed was in fact missing. I’ve read a lot on banking over the years, but I’ve never seen anything that tackled some of the fallacies of the monetary system as I saw for the first time with Kansas City, Mosler, Bill, and winterspeak. The timing sequence is a little too coincidental for that attribution not to be there in some way.

[Ramanan]

Scott,

Completely agree with you on referencing.

Just noticed: the BIS paper refers Basil Moore's book "Shaking The Invisible Hand". So they are indeed reading "endogenous money" literature! I haven't seen the book but I guess the style is more like an essay than a detailed research paper. In that case I can surely conclude that the BIS authors are reading your papers because without reading your papers it is difficult for someone from mainstream to get it this right!

[Nick Rowe]

Rebel: "Nick, I am afraid that your response (on December 26, 2009 at 09:29 AM) to my question (about feedback gain) simply relocates the question somewhere else; that is, why is the AD schedule vertical? It is not easy (at least not for me) to see why."

Rebel: I come at this the other way: why should it not be vertical? What reason do we have for believing a lower level of all prices would increase (or decrease) aggregate output demanded? If the price of everything you buy dropped by 10%, and the price of everything you sell dropped by 10% (as it must in macro, because everything bought is sold), then why should you want to buy, or be able to buy, any more than you did at a higher price level?

It's back to my post "What is it with Microeconomists?". If the AD curve slopes down, it doesn't do so for the same reason micro demand curves slope down.

[Too Much Fed]

JKH & From:

http://www.clevelandfed.org/research/commentary/2009/1009.cfm

"Many argue that reserve targeting (or quantitative easing when it is done in a zero-interest-rate environment), can still stimulate the economy when short-term interest rates are zero. But if quantitative easing is implemented through the purchase of short-term securities, this policy is almost certainly doomed to failure. Since banks’ cash reserves and short-term securities are perfect substitutes when nominal interest rates are at zero, banks have no incentive to lend the money out.

They are likely to simply substitute the cash they receive from the central bank for the securities they were holding in reserves. Therefore, the supply of money in circulation (that is, one common and useful definition of it, M1, which is currency held by the public plus demand and other checkable deposits) is not affected. To affect M1, banks need to lend the cash out to the private sector, which in turn will redeposit part of this cash into checking accounts, thereby increasing money in circulation. Because open market operations will not increase the money supply when short-term interest rates are zero, they can’t be used to increase either real economic activity or prices."

Before zero short term interest rates, is it possible for the fed to lower the risk free interest rate? Then mortgage rates come down and more currency denominated debt (a mortgage) is created. The home purchaser then gives the home seller the demand deposit (check), and the home seller deposits it.

Can that scenario cause the need for more central bank reserves to be created? I'm looking for a scenario to explain how more currency denominated debt leads to more spending in the present by expanding the fungible money supply.

Plus, how many central bank reserves are there?

[Too Much Fed]

JKH & From:

http://www.clevelandfed.org/research/commentary/2009/1009.cfm

"Consider, for example, a firm that decides to borrow money at a stated, or nominal, interest rate of 7 percent. If prices, including the firm’s product price, are expected to grow at 2 percent per year, then the real cost of borrowing for the firm (the real interest rate) is 5 percent per year. In principle, the real rate should be determined only by the saving and investment decisions of market participants, plus adjustments for risks, not monetary policy. In fact, a permanent change in expected inflation, say from 2 percent to 1 percent, will change only the nominal rate (in this case from 7 percent to 6 percent) and leave the real rate unchanged.

However, inflation expectations do not change instantaneously. Because they adjust over time, a policy move that decreases the nominal interest rate will also, in the short run, temporarily decrease the real rate. The decrease in the real rate will increase the willingness of banks to lend and firms to borrow. This extra lending will then temporarily stimulate output. In this scenario, a central bank could easily counteract a deflationary shock that reduces prices and expected inflation (which could potentially raise the real rate temporarily and depress the economy) by lowering the real rate, or equivalently, by lowering the nominal rate by an amount greater than the fall in prices."

I believe that scenario does not consider quantities (output) properly. Let's assume the fungible money supply grows by 2% a year and price inflation for a firm is 2% per year so that quantity growth is 0% and expected to be 0%. Why would the firm borrow at all? It seems to me that they are assuming the firm is supply constrained which may not be true.

What about firms that can expand using free cash flow and don't need to borrow?

[Too Much Fed]

JKH & From:

http://www.clevelandfed.org/research/commentary/2009/1009.cfm

"One drawback of a price-level target is that it necessitates stimulating the economy whenever prices fall—no matter what the cause. For example, an expansion driven by a positive supply shock would naturally put downward pressure on prices and upward pressure on the real rate, but few economists believe that monetary policy accommodation is helpful in such a situation. An inflation target can potentially be changed, to respond to unusual economic conditions,but a price-level target has the advantage of responding according to a very simple and easy-to-understand rule."

IMO, that positive supply shock and monetary policy accommodation "thing" would be news to greenspan, bernanke, and the rest of the fed. So, what should be done when a supply shock occurs from cheap labor, productivity, or something else?

[JKH]

Too Much Fed

There is currently about $ 1.1. trillion in Fed reserves, almost all of which is excess. Usually the Fed creates new reserves according to requirements resulting from new deposits, but there’s no need for that currently. The Fed controls the short term Fed funds rate, but won’t drop it below zero.

Re your second point, they’re basically arguing that monetary policy can work when rates are above zero, at least by temporarily lowering the real rate of interest. That’s supposed to encourage borrowing, notwithstanding previous expectations about growth. Firms with extra cash flow don’t need to borrow that amount.

I’m really not qualified to answer your question on positive supply shock response. Too much sudden policy easing in that case could overheat and lead to inflation longer term perhaps. Price level targeting could dampen the need for sudden changes in policy at times, extending existing policy for a price level target yet to be reached, compared to inflation targeting, I suppose.

[Too Much Fed]

JKH, I'm of the view that the fed used currency denominated debt (mostly mortgage) to prevent price deflation in around 2001.

I'm looking at the flow of funds. Under household mortgage debt, it says in 2001 5306.6 billion, and in 2007, it says 10485.2.

I'm trying to figure out how that shows up balance sheet wise and in the money supply. Can you help me out?

Also, how is a 100 mortgage loan broken down?
8 for capital, so much for the value of house, and other?

[Too Much Fed]

JKH & From:

http://www.clevelandfed.org/research/commentary/2009/1009.cfm

"The idea behind buying longer-term government securities is that doing so will drive up their demand and therefore the price of these securities. This will decrease their yield and therefore lower long-term interest rates. Lower long-term interest rates will end up stimulating investment and the economy. The assumption underlying this approach is that banks will not simply sit on the cash they receive from the Fed in exchange for the long-term securities, and the supply of money in circulation will actually rise in consequence. That is, banks cannot view long-term and short-term government securities as perfect substitutes. Otherwise, they will not attempt to buy other long-term securities or loan out this extra cash."

I don't see lower long-term interest rates stimulating investment in the economy when there are plenty of houses, plenty of cars, and plenty of most other products. Why do economists almost always assume more supply from the future needs to be brought to the present using currency denominated debt?

[Too Much Fed]

JKH & From:

http://www.clevelandfed.org/research/commentary/2009/1009.cfm

"We have discussed the importance of expected inflation in counteracting a deflationary spiral. If interest rates are at zero, increases in expected inflation will decrease today’s real interest rate, stimulating both the real economy and prices. Using communication to boost future inflation expectations in this environment requires policymakers to promise that they will “err” on the side of keeping interest rates low even after the economy starts to recover. In essence, this future inflation will stimulate the economy today and actually increase money today."

How did that work out in 2001 to 2007/2008? So, they want to let debt levels get out of control again?!? I don't think the people at the fed will EVER learn anything!

I'm hoping the lower and middle class experienced an internet bubble and a housing bubble and won't get SUCKERED by the fed a third time.

[JKH]

Too Much Fed,

The mortgage expansion shows up in GSE balance sheets and in securitization vehicles, and to a lesser extent, commercial banks. Only bank financing is accompanied by an increase in money supply; the former were funded by issuing debt.

A 100 mortgage represents a percentage of the value of the house at the time of the transaction, depending on the loan to value ratio. The 100 is the mortgage loan. The 8/92 split is just an example of how a bank itself funds the mortgage – 8 in bank capital; 92 in bank deposits.

The Cleveland Fed article is not very good. Again, I’m not qualified to judge how economists in general view long term rate dynamics. I think in the current case the Fed buying MBS is intended mostly to help the overall supply of mortgage financing and help mortgagors refinance at lower rates.

[Too Much Fed]

JKH, let's move on to the BIS link and from it on p.10 of the PDF.

"Fourth, central bank balance sheet policies need to be viewed as part of the consolidated government sector balance sheet. The main channel through which they affect economic activity is by altering the balance sheet of private sector agents, or influencing expectations thereof. As a result, almost any balance sheet policy that the central bank carries out can, or could be, replicated by the government; conversely, anything that the central bank does has an impact on the consolidated government sector balance sheet. In other words, the central bank has a monopoly over interest rate policy, but not over balance sheet policy. This raises tricky questions about coordination, operational independence and division of responsibilities."

I'm trying to come up with a scenario where central bank reserves are actually gov't debt in disguise similar to Fannie and Freddie. Is that possible?

Plus, I am pretty sure there is a 1-month treasury. Are there any shorter durations than 1-month?

[Too Much Fed]

JKH, were you the person who was asking why the monetary base was going up because of reserves instead of currency?

[JKH]

Too Much Fed,

The BIS paper is far superior to the Cleveland one.

Central bank excess reserves are in a sense a form of (short term) government debt, on a consolidated basis. Their primary purpose is as a liability in support of central bank assets directly. It’s not really a disguise. It’s more that people for the most part don’t understand this. It’s a bit of a stretch to compare them with Fannie and Freddie, but some of the central bank assets now are mortgage related – i.e. MBS - so you’re not far off.

I believe the 1 month Treasury bill is normally the shortest original term. There may have been special occasions where they issued shorter term bills than that.

The monetary base is going up in the form of reserves instead of currency, because the Fed has intended that the base increase (to support its own assets) and can do this most effectively through excess reserves. It can’t force currency issuance at will, because currency demand is determined by the public. The public has increased its currency holdings by only about $ 100 billion over the entire period of the crisis, compared to the Fed increasing excess reserves by $ 1.1 trillion.

[Doc Merlin]

@Kaleberg

You could also consider the flight to alternate currencies as what causes the inflation to move from just "really high" to truly black hole levels.

[Too Much Fed]

JKH, imo any currency denominated debt that the gov't won't allow to default is gov't debt in disguise, including Fannie and Freddie along with the fed itself.

Are central bank excess reserves in a sense a form of (short term) government debt, on a consolidated basis?

I think so. The gov't will probably back up the fed if needed, the reserves can be defaulted on, they have an interest rate, and they have a term (overnight). Does that apply to regular reserves?

From the BIS pdf on p 26 of the pdf, "But in order to induce banks to accept a large expansion of such balances in the context of balance sheet policy, the central bank has to make bank reserves sufficiently attractive relative to other assets (scheme 2). In effect, this renders them almost perfect substitutes with other short-term sovereign paper. This means paying an equivalent interest rate. In the process, their specialness is lost. Bank reserves become simply another claim issued by the PUBLIC SECTOR [my emphasis]. It is distinguished from others primarily by having an overnight maturity and a narrower base of potential investors."

Is there a reason why the gov't does NOT issue any debt with a maturity less than 1 month? Did the gov't ever issue shorter term debt before the fed was created?

I'm thinking that currency demand the way things are now is determined by the public thru the banking system. IMO, there are plenty of people outside the banking system demanding currency (workers and/or the unemployed), but the fed and business sector won't allow it because they like exploiting an oversupplied labor market. There are also plenty of savers demanding more currency thru interest rates, but the fed is affecting that market. IMO, the workers and the savers are getting a bad deal.

Lastly, I'm of the opinion that the fed is using excess central bank reserves to preserve bank capital so that they (the fed) can continue to try to price inflate with currency denominated debt.

What would be the difference between the fed buying with central bank excess reserves and currency?