Land and the liquidity trap

Bill Woolsey has a very good post, written in response to my previous post. In this post I steal Bill's thought-experiment, but change "corporate bonds" into "land".

There is an alternate universe, just like our own, with one exception. For historical reasons*, central banks do not own government bonds. They own land instead. They buy and sell land, and adjust their target price of land several times a year, to try to keep CPI inflation at the 2% target. If a central bank fears that inflation will fall below the 2% target, it buys land and raises the target price of land. If a central bank fears that inflation will rise above the 2% target, it sells land and lowers the target price of land. Sensible economists build macroeconomic models where central banks set the price of land, because that is what central banks really do.

The simplest macroeconomic models in that alternate universe are long run models. They say that if the central bank doubles the target price of land, that will eventually cause the general price level to double too, other things equal. This is known as "land price neutrality". And they say that if the central bank makes the target price of land grow at 10% per year, that will eventually cause the general inflation rate to increase to 10% per year too, other things equal. This is known as "land price superneutrality". More sophisticated macroeconomic models incorporate sticky prices and expectations, and try to say something more precise than "eventually". Those more sophisticated DSGE models also introduce other shocks, and try to say something more precise about what central banks need to do when other things are not equal.

Macroeconomists in that alternate universe argue about the monetary policy transmission mechanism. The mainstream view is all assets are substitutes, to a greater or lesser extent. If a central bank raises land prices, that raises asset prices across the economy, which makes investment in newly-produced capital goods relatively more profitable, which increases aggregate demand. An increase in land prices, and other asset prices, also increases consumption demand, via wealth effects and substitution effects. Some economists stress the role of expectations, because if the central bank raises land prices people know that all prices must rise in the new equilibrium. A few Monetarists try to insist that what is important is that the central bank is issuing more money, and the rise in the price of land is just a symptom, rather than a cause. But nobody takes them very seriously, because everybody knows that real world central banks set the price of land, and the quantity of money is endogenous.

There is a bunch of monetary cranks in that alternate universe, who call themselves "New Keynesians". The New Keynesians want central banks to hold short-term Treasury bills instead of land, and target a short-term nominal interest rate instead of land prices, adjusting that nominal interest rate target several times a year to keep CPI inflation at the 2% target. They build macroeconomic models to show how central banks could target inflation by lowering nominal interest rates when inflation looks like it will fall below target, and raising nominal interest rates when inflation looks like it will rise above target.

Orthodox central bankers don't pay much attention to the New Keynesians. They are amusing thinkers, who come up with strange but interesting thought-experiments, but that is not how monetary policy works in the real world. Because real world central banks buy and sell land, not Treasury bills. And real world central banks set land prices, not nominal interest rates.

One day, the nominal interest rate on short-term Treasury bills falls to 0%. The orthodox central bankers gently tease the New Keynesians: "So, it seems your model has hit some sort of Zero Lower Bound! You can't buy buy short-term Treasury bills to lower their nominal interest rates below 0%, because people would just hold currency instead!"

This is the end of the road for the New Keynesians. They try to save their model by introducing something they call "forward guidance" on nominal interest rates, though nobody else can figure out how it is supposed to work.

But orthodox central banking carries on just as before. The price of land is finite, so when central banks want to loosen monetary policy they simply raise the target price of land, just like before. Orthodox central bankers simply cannot understand the concept of a "liquidity trap". So what if currency and short-term Treasury bills are perfect substitutes in some circumstances? It has no implications for monetary policy. The price of land is what really matters.

[*I have tried, and failed, to cook up some vaguely plausible alternate history of how central banks came to own land rather than government bonds. John Law was clearly involved somehow.]

Comments

[Vaidas]

What happens when the real price of land gets high enough so that the -2% real return on currency becomes higher than the real expected return on holding land?

[Nick Edmonds]

Nick,

I can't follow the accounting here. For the general price level to rise, the money supply must rise, so the central bank must have actually made net purchases of land, right? So, presumably the central bank is continually acquiring more and more land? I can't see how you can have money-neutrality (or land price neutrality, for that matter), when the stock of land in private hands is changing.

[Nick Rowe]

Vaidas: That did happen once, in my alternate history. In the 1930's the Bank of France decided to announce that it wanted land prices to fall at 2% per year, to create deflation in the CPI. And everyone who held land decided they could get a better rate of return by holding currency rather than land, and everyone wanted to sell all their land to the Bank of France. So the Bank of France quickly abandoned its policy, and announced it would keep land prices constant, so that some people preferred land rents to the liquidity of holding currency. If a central bank ever found itself in danger of getting far too big, and owning all the land, that would simply mean it has chosen too low an inflation target, and needed to tell people it would be setting higher land prices in the near future.

Nick Edmonds: no. Forget accounting. The big problem with accounting is that it assumes a dollar is always worth a dollar. Yes, printing money creates profits for the central bank, which it turns over to the government. In my alternate history, the CB can give land to the government, which sells that land, and uses the proceeds to retire government bonds, or give tax cuts, or whatever.

[JKH]

Do interest rates matter at all in the CB land model?

Presumably, if the CB is making markets in land, those CB land transactions affect bank reserves.

Do bank reserves matter at all in the model?

What interest rate do bank reserves trade at?

How is that rate established?

Does the CB pay interest on bank reserves?

[The Market Fiscalist]

Lets say that the only way people can hold wealth in this model is either land or money. Lets also assume that the CB can create money in 2 ways - either by buying land for new money, or just by giving it away.

Start in equilibrium and introduce a shock that causes everyone to start saving more out of their income each month. Prices are sticky in this model so a recession begins.

If the reason that people are spending less is because they want to hold a higher % of their wealth in money rather than land then the problem can be solved by the CB buying land for money until everyone is happy with the new balance (and the price of land will follow its targeedt path).

If the reason that they are spending less is that they want to have a higher net wealth in proportion to their income then just swapping land for money may not work. Its hard to think of a mechanism that will make people's wealth increase in relation to their income just by buying up land.

On the other hand if the CB gave money away to people when they wanted to increase their wealth (and allowed them to swap the money they have been given for land in order to maintain the desired prices of land) then everything should work out well.

[Nick Rowe]

JKH: Do land prices matter at all in your New Keynesian(?) model?

Interest rates are endogenous. If the central bank changes the price of land, that will presumably affect interest rates, including interest rates on loans between banks. That interest rate is set by the market.

Economists differ on whether central banks should pay 0% interest on reserves, or pay a rate that floats with market rates of interest, plus or minus 0.25% for negative or positive balances.

[Nick Rowe]

TMF: "If the reason that they are spending less is that they want to have a higher net wealth in proportion to their income then just swapping land for money may not work. Its hard to think of a mechanism that will make people's wealth increase in relation to their income just by buying up land."

Simple. The CB just increases the price of land. (A reduction in the rate of time preference means real land prices must rise, and if we don't want to wait for the price of everything else to fall, because prices are sticky, so that would cause a recession, the CB needs to increase the nominal price of land.)

[W. Peden]

Nick Rowe,

What if there's a huge fall in the demand for land, to the point where the natural price of land falls below zero? Then conventional monetary policy is through, and any further monetary policy is purely experimental, and should be supplemented by high-divider austerity. Cuts to welfare programmes to the poor and taxes primarily levied on the poor are the most obvious forms of austerity for a New Keynesian fiscal policy in this scenario, given the propensity to consume of the poor. Although, if I understood some of your recent post, the government could create an austerity programme (in a New Keynesian model) to escape the ZLB by simply promising to spend more in the future.

[JKH]

Nick,

There is no such thing as a market rate for excess bank reserves.

There can only be a CB administered rate.

That’s because the traded rate for excess bank reserves will go to zero if the central bank does not manage that rate (i.e. manage it by controlling the supply of reserves according to a targeted trading rate, or by paying interest on excess reserves).

Warren Mosler calls this effect the “natural rate of interest is zero”.

That's wrong, but for more than the reason you think.

It’s wrong because the fact that the interest rate goes to zero is the result of the central bank paying an administered rate of interest of zero in that case. It’s a policy choice. There’s nothing natural about it. Plus what you’ve pointed out in the past about it being an abuse of terminology that’s used conventionally in a different way.

This is the problem with an endogenous rate view.

The central bank must determine the interest rate on excess banks reserves either by:

a) Controlling the quantity of excess reserves – which contradicts your model because the CB can’t simultaneously control the price of land and the price of bank reserves – the zone of inelastic reserve demand is far too narrow for this to be the case

b) Paying interest on reserves

c) Going the Mosler route – which amounts to a conscious policy of paying an interest rate of zero on reserves

[Nick Edmonds]

"In my alternate history, the CB can give land to the government, which sells that land, and uses the proceeds to retire government bonds, or give tax cuts, or whatever."

Have you not just collapsed the whole thing to conventional monetary policy (if retiring bonds) or fiscal policy (if giving tax cuts)?

[W. Peden]

Nick Edmonds,

I'm pretty sure that central banks already give the profits of their transactions to the government.

[W. Peden]

(So there's an equivalence between (i) the CB trading bonds and giving the proceeds to the government, and (ii) the CB trading land and giving it to the government to sell and take the proceeds. Therefore, if the latter can be fiscal policy, so is the former.)

[Nick Rowe]

W Peden: "What if there's a huge fall in the demand for land, to the point where the natural price of land falls below zero?"

Hmmm. If a change in technology (hydroponics) causes land to become a free good? About the same as happens now, I expect, if the government defaults on all its bonds, so government bonds become a free good. The assets of the CB are worthless, and it would have to target something else. Which one is more likely?

JKH: central banks once set the price of gold, or the exchange rate, and let interest rates adjust to preserve the price of gold, or the exchange rate. I'm having them set the price of land instead. Commercial banks existed before central banks, as we now know them, existed. The Lender of Last Resort function always has some conflict with the central bank's other objectives. Even in Canada today, the actual overnight rate does not exactly equal the Bank of Canada's target for the overnight rate.

Things weren't always, and don't have to be, and won't be in future, the way they seem to be now. Economics is not about monetary realism. It's about monetary unrealism.

Nick Edmonds: what W Peden said. Yep, it's called "seigniorage/the inflation tax". Government-owned central banks do in fact currently give their profits to the government. In that sense, government-owned central banks mean that monetary and fiscal policy must be linked.

[W. Peden]

Nick Rowe,

I was trying to get at some sort of analogue for the ZLB in the kind of world you present. There's a lower limit on the price of land, just as with the price of bonds, but it would be a scenario of the demand for land falling so low that it's below zero.

In this parallel world, New Keynesians are austerians who worry about a situation where conventional monetary policy can't do anything to depress demand.

[Odie]

"In my alternate history, the CB can give land to the government, which sells that land, and uses the proceeds to retire government bonds, or give tax cuts, or whatever."

That is the same as saying the Fed just credits the Treasury account with $1 trillion or so no questions asked. Bankrupt governments tend to fund themselves that way but it is not how the Fed operates. (Or the Treasury issues a debt note which will increase the government deficit; now it becomes a fiscal operation.)

The difference between bonds and land is that bonds have a fixed dollar value (their price fluctuations in the secondary market non-withstanding). A T-bill for $10,000 at 5% interest and 1-year to maturity has a value of $10,500. It helps me to think of a fixed-rate bond as a check over a certain, pre-determined amount with a cash-in date at the day of maturity. If the Fed wants to lower rates it pays e. g. $10,250 for it and interest rates will fall to ~2.5%. If it pays more than $10,500 the Fed is going into negative interest territory essentially "donating" money to the private sector and assuming a capital loss.

So what the Fed does is buying a monetary asset with a fixed max $-value by issuing $ as deposit (which can be converted into cash). It does not buy non-monetary assets of undetermined $-value like land, goods or services etc. Essentially, it buys my CD and deposits that money in my checking account. Whether I spend that money or not is entirely my decision. Since I probably saved that money for a reason I will just look for another savings opportunity. And if I would be a financial adviser and that money would not be mine I would definitively not spend it but invest it again. That describes banks and their reserve accounts.

In your "CB buys land"-model it would swap land for "land-to-be-used-later" with future land usually being bigger (positive interest). It hits the ZLB when it swaps 1 acre for 1 acre and pays negative interest when it swaps 1.1 acre of land for 1 acre of "future land".

[Nick Rowe]

W Peden: I think Vaidas was also getting at the analogue to the liquidity trap. It's when the central bank needs to buy all the land to prevent the price from falling. That would happen if the central bank created expectations of a price of land that was falling too quickly (or rising too slowly). It's like too low an inflation target for land prices.

[W. Peden]

Nick Rowe,

What if the parallel central bank is targeting NGDP at, say, 3% growth per year?

[Nick Rowe]

W Peden: what you are getting at is this: there is in fact *no* difference between my alternate universe and the current one. Except the central bank's communications strategy. My central banks communicate via land prices. Real world central banks communicate via nominal interest rates. (What they hold in their balance sheets is pretty much irrelevant.)

[The Market Fiscalist]

"Simple. The CB just increases the price of land. (A reduction in the rate of time preference means real land prices must rise, and if we don't want to wait for the price of everything else to fall, because prices are sticky, so that would cause a recession, the CB needs to increase the nominal price of land.)"

In your view the CB can always hit the inflation target by increasing the price of land sufficiently. I suppose that if time preference has changed then this may work. The CB can raise the price of land and avoid a recession that would otherwise happen while the general price level fell. However if people's TP hasn't changed but they just want to build up a bigger stock of savings (wealth) as a precaution against a feared recession then I can see problems. There will be a price of land where the return on it is so small that everyone would rather hold cash than land and so any further sales at that price will not increase nominal wealth (ZLB on land?). If that price is lower than that which would generate the spending required(via the wealth effect) to hit the inflation target then the CB is at an impasse

Of course the CB could , if it was allowed to, start buying up other assets with higher returns (and higher risks) than land. But at what point does this all become too complex and too risky compared with a simple injection of new money in response to the initial desire for a greater stock of savings ?

[Odie]

Or to put money in that context: Each note would e. g. say 1/100 of an acre and the CB would pay more than 100 of those notes per acre = negative interest.

[Mike Sproul]

"I have tried, and failed, to cook up some vaguely plausible alternate history of how central banks came to own land rather than government bonds. John Law was clearly involved somehow."

Not implausible at all. First the colonists in Massachusetts (1690) issued paper shillings based on land. They mostly issued shillings on loan, backed by land, but if anyone defaulted, the colony became the owner of the land. Next came John Law, (inspired by the colonists) who backed his paper money (livres, I think) with the land that we nowadays call the Louisiana purchase. Next, in the French revolution, the National Assembly seized the land of the Catholic church, and backed its assignats with that land.

One problem though: If the central bank sets $1=1 square foot of land, then issuing another dollar $100 and buyiong another 100 square feet will still leave $1=1 square foot.

[Ralph Musgrave]

The big problem with a central bank buying and selling land would be the distortionary effect: the price of land gyrating relative to other assets, even if in the long run an X% rise in the price of land was matched by an equal rise in the price of other assets.

Exactly the same distortionary flaw applies to conventional CBs that buy and sell government debt (and perhaps high quality private bonds). That is, it’s only the behaviour of the asset rich that is affected. E.g. in the case of QE a flood of money goes to other countries in search of yield, much to the annoyance of those countries.

I see no reason, given a need for stimulus, for one set of assets or one sector of the economy to be boosted in preference to any other.

[Donald A. Coffin]

Let me begin by saying two things: (1) I have not yet read all the comments. (2) I found this extraordinarily interesting.

But how much of the "work" here is being done by the fact that "land" in any specific country is in essentially perfectly inelastic supply? Suppose the central bank targeted the price of some other asset, for example, automobiles. If the CB raised its target price (or raised its target for the time path of prices) of cars, we would expect a supply response, wouldn't we? And how would this change the outcomes we would expect?

[W. Peden]

Nick Rowe,

Exactly. If the austerian New Keynesians of this parallel sound silly, then so should contemporary New Keynesians calling for fiscal stimulus. Of course, this is your point, but I just find the image of an inverted "liquidity trap" amusing.

[Frank Restly]

"One day, the nominal interest rate on short-term Treasury bills falls to 0%. The orthodox central bankers gently tease the New Keynesians: "So, it seems your model has hit some sort of Zero Lower Bound! You can't buy buy short-term Treasury bills to lower their nominal interest rates below 0%, because people would just hold currency instead!"

"The price of land is finite, so when central banks want to loosen monetary policy they simply raise the target price of land, just like before."

And presumably when central banks want to tighten monetary policy, they simply lower the target price of land - except on one day they discover they have no land left to sell.

How is a price floor (central bank is limited to zero lower bound on interest rates) any different than a quantity ceiling (central bank is limited to its own land holdings available for sale)?

[W. Peden]

Frank Restly,

They could do what a central bank could do if it ran out of assets to exchange for base money- raise land requirements for commercial banks, increasing the demand for land.

[Nick Rowe]

Mike: "Not implausible at all."

I figured you could do it. I'm too bad at history. I was having John Law only wound that guy in the duel, so he didn't need to escape to France, and then persuade the Bank of England to buy land near London Ontario. Then I realised my dates were off by a century!

"One problem though: If the central bank sets $1=1 square foot of land, then issuing another dollar $100 and buyiong another 100 square feet will still leave $1=1 square foot."

Only if land and dollars are perfect substitutes in portfolios. If land pays rent, but dollars are more liquid, they won't be.

(BTW off-topic. Did you know that UK government bonds were once less liquid than land? Because government bonds were basically lottery winnings ("$100 dollars a year for life/X years!"), and had your name attached, so it was hard to sell them. So UK government bondholders were very pleased to swap bonds for the much more liquid shares in the South Sea Company.)

[Nick Rowe]

Ralph: the government of Canada currently owns land. The Bank of Canada currently owns government bonds. Would it make much difference to anything if the government of Canada, and the Bank of Canada, just did a swap? Keeping the 2% CPI inflation target the same?

[Gene Callahan]

@Ralph Musgrave: "The big problem with a central bank buying and selling land would be the distortionary effect: the price of land gyrating relative to other assets, even if in the long run an X% rise in the price of land was matched by an equal rise in the price of other assets."

Well, this is the Wicksell business cycle theory with land instead of loanable funds as the manipulated price.

Nick, Mike's scenarios are good, but his problem arises because he assumes that the currency is backed by land. The bank could have gotten into land price manipulation through, saying, having been granted tons of federal land (in the US, say) in the 1800s, and deciding to use that to influence the macroeconomy.

[Nick Rowe]

Donald: I don't think it would change things much at all, if land were not perfectly inelastic in supply. Compare gold, for example. And, in practice, new land can be drained and cleared.

Frank: a central bank can always lose all its assets, if the government takes them, or it makes some very unlucky investments. What would it do if it wants to tighten, and has no assets to sell? Same as now. It could always sell land short, basically borrowing against its future seigniorage profits.

Odie: "That is the same as saying the Fed just credits the Treasury account with $1 trillion or so no questions asked. Bankrupt governments tend to fund themselves that way but it is not how the Fed operates."

That is exactly how the Bank of Canada operates. Every year it calculates its profits, and gives them to the Government. And the Canadian government is very far from bankrupt. And I am almost sure the Fed operates the same way.

Now, if the government *chose* how much the Bank of Canada must give it, then the BoC is no longer independent, and if the government chooses too much, and the BoC cannot say "no", then we have Zimbabwe.

[Nick Rowe]

Gene: "Well, this is the Wicksell business cycle theory with land instead of loanable funds as the manipulated price."

Yes but: suppose you did not hear the central bank's communication strategy. Suppose you only had annual data on: nominal interest rates, land prices, monetary aggregates, and the CPI. You would be able to figure out that the central bank was targeting 2% inflation (if your data series was long enough). Would you be able to figure out whether it was using nominal interest rates, land prices, or base money, to hit that inflation target?

[Min]

Ralph Musgrave: "I see no reason, given a need for stimulus, for one set of assets or one sector of the economy to be boosted in preference to any other."

What about stimulating the economy by putting money into the hands of the poor, who will spend it, so that it spreads to the rest of the economy? Compare that with putting it into the hands of the rich, who cannot be counted on to spend it, and if they do spend it, cannot be counted on to spend it in the domestic economy.

[Nick Rowe]

Min:
1. Given that half the time the economy is going too slow, and the other half too fast, would you be happy to take money out of the hands of the poor whenever we need to slow it down?

2. Don't answer. That was a rhetorical question, as well as off-topic.

[Lord]

Really sounds like post independence America when the price was set at $2 an acre. It was purchasable on credit but could be returned for a refund if it could not be completed.

[Odie]

"That is exactly how the Bank of Canada operates. Every year it calculates its profits, and gives them to the Government. And the Canadian government is very far from bankrupt. And I am almost sure the Fed operates the same way."

Profits it makes by buying and selling securities or through interest spreads between assets and liabilities. Thus, profits coming at the expense of the other sectors not just by "printing money". It can also happen that the CB makes losses (e. g. ECB: 1999, 2003, 2004) which results in no transfer.

[ATR]

JKH, Nick:

Clearly, independent of its land purchases, the CB still has the choice to set the floor rate (IOR, could be 0% if desired) and the ceiling rate (central bank lending facility). If the operation of targeting land prices would cause the quantity of reserves to be such that banking system would have such an excess (in the case of large purchases) or shortage of reserves (in the case of land sales) relative to reserve requirements and any other source of reserve demand, then the central bank would effectively be running a floor or ceiling system, respectively. In other words, interbank rates would go to the floor or ceiling, and so the central bank would control rates by choosing where to set the floor and ceiling.

There’s a bit of talking past each other on whether the market or central bank controls rates. Of course, banks ultimately decide what rate they want to trade reserves at given market conditions. But the central bank can steer the market where it wants it by controlling market conditions, at least in the capacity of controlling a combination of rates (IOR, discount window, etc.), reserve quantities (via OMO), and other tools like reserve requirements (if chosen). That’s enough to steer rates where the CB wants.

I think Nick’s point is that he doesn’t care where rates wind up, as long as he achieves his land price target. He thinks the land price target is enough for the CB to achieve its objectives. So whatever the combination of floor/ceiling rates and whatever rules the CB sets up, he’s determined to hit his land target and live with wherever rates land.

But what if it would possible to control the price of land AND interest rates? And if so, what if it would be desirable to do so?

For example, suppose the CB would want a rate of X and price of land of Y. Suppose for whatever reason the price land is not affected by the rate. Then the central bank can simply buy/sell land until its price moves to Y. Then given the quantity of reserves available in the system as a result of this purchase, it can set reserve requirements, the floor rate, and the ceiling rates in order to hit its rate target. A given interest rate is always possible to achieve through various combinations of rates, quantities, and reserve requirements. (See my blog for models from Woodford or Bindseil to see how this works. It’s all about the central bank manipulating the probability banks end short or long reserves via OMOs and in conjunction setting the rates corresponding to banks ending short or long rates – the discount and IOR. The probabilistic weighted average rate equals the rate it wants. What determines if banks end short or long depends on reserve requirements, which the CB also controls.)

If instead the price of land is affected by interest rates, it might still be possible to control rates AND the price land, depending on how they interact. For example, suppose we start with banks awash with excess reserves such that the rate is at the floor the CB sets. Then say the CB determines it would take a land purchase worth X reserves, given the current interest rate, to boost land prices to their target. So the central bank does that, and the interbank rate doesn’t change since it will stay at the floor. Then maybe the CB decides it wants to keep the land price the same but raise interest rates by 1%. It could raise the floor rate 1%, but raising interest rate raises the price of land. To counteract that, the CB would have to purchase land. But that would be okay, because the injection of reserves accompanying that purchase would not move rates below the floor. So it achieves its new target land price and target rate. So there could be a variety of reserve quantities and rates that can achieve the same land price and interbank rate.

[Nick Rowe]

Lord: interesting. If the central bank had been the one buying and selling land, it would be like my alternate world.

Odie: sure. But all assets in the central bank's balance sheet ultimately come from printing money to buy those assets. The higher the rate of money printing, the higher is inflation, and the higher are nominal interest rates, and the bigger the spread between the interest it earns in its assets and the 0% interest it pays on its currency liabilities. Yes, it can also be thought of as a tax. "The inflation tax".

[JKH]

Nick,

"central banks once set the price of gold, or the exchange rate, and let interest rates adjust to preserve the price of gold, or the exchange rate. I'm having them set the price of land instead."

Didn't you do a post once comparing the way the CB sets the policy rate according to an inflation target with the way it happens with a gold price target? I quite liked that one.

The CB has to price excess reserves in either case. You can say that market interest rates adjust, but you'll still have the CB setting a policy rate based on how it sees market forces and administering it. It's a choice. Even at the ZB, it decides on 25 basis points or 0 basis points for example. They set administered policy rates during the gold standard did they not? And I thought the post I'm thinking of had those things happening not fundamentally differently in either type of regime.

[Phil Koop]

"UK government bondholders were very pleased to swap bonds for the much more liquid shares in the South Sea Company."

You are doing more than a little violence to history here, not least with the phrase "government bonds", which elides the enormous differences displayed by multifarious government obligations circulating at the dawn of the 18th century. Still, the basic problem is with your representation of the South Sea Company, which appears to be conflating different periods.

The point of the initial 1711 flotation was to convert short-term government debt - which it had not adequate revenue to service - into long-term debt. The way this worked was that the government granted 6% perpetual annuities to the South Sea Company matching the short-term debt tendered for exchange into shares. The land holdings of the company were nil, and the shares traded on the basis of the annuity income of the Company. In no sense did the take up of this offering demonstrate the liquidity of land. It is true that the Company was also granted monopoly right of trade in the South Atlantic, but since this trade was negligible it had little impact on the share price.

The lottery loans were 32-year loans; they were not the loans tendered for South Sea Company shares. And it is unlikely that holders of those loans would have been keen to exchange them for 6% perpetuals; they paid a minimum 6.84% return but the lottery payouts raised the total payout to 8%.

[Vaidas]

Nick: "That did happen once, in my alternate history. In the 1930's the Bank of France decided to announce that it wanted land prices to fall at 2% per year, to create deflation in the CPI. And everyone who held land decided they could get a better rate of return by holding currency rather than land, and everyone wanted to sell all their land to the Bank of France. So the Bank of France quickly abandoned its policy, and announced it would keep land prices constant..."

Wow, that was quick. It took 20 years for Japan to take the same route.

In your example you are describing a permanent liquidity trap. However, since the neutral rate is fluctuating, the more realistic scenario is a temporary liquidity trap. How do most central banks react to this? They don't change their inflation target. They announce a lower price of land than needed to hit the inflation target, and they undershoot their existing inflation target instead.

[JKH]

ATR,

Good comment, but you're really just describing (quite well) how a CB controls the policy rate and the volatility of rates around the policy rate (e.g. fed effective versus fed target; IOR; etc.). That's standard stuff. No need for Woodford or Bindseil et al.

That's my point. The CB has to make a choice.

And it doesn't matter if its targeting gold or land on balance sheet or CPI off balance sheet.

A degree of potential balance sheet disruption in the former cases - but the CB still has to choose where to set the policy rate. Markets don't set policy rates for gold or land cases any more than they do for CPI targeters. CB's do.

[Frank Restly]

W. Peden,

"They could do what a central bank could do if it ran out of assets to exchange for base money- raise land requirements for commercial banks, increasing the demand for land."

When a central bank wants to put downward pressure on land prices, it would sell land. If raising land requirements for commercial banks increases the demand for land, this would put upward pressure on prices.

Nick,

"Frank: a central bank can always lose all its assets, if the government takes them, or it makes some very unlucky investments. What would it do if it wants to tighten, and has no assets to sell? Same as now. It could always sell land short, basically borrowing against its future seigniorage profits."

A central bank could certainly buy short land positions from someone else selling short land positions. But if neither the buyer (central bank) or seller has land to settle the position, then this would have no effect on land prices.

If only the seller of the short land positions has land, then he is in an enviable position - he can sell as many short positions to the central bank as they are willing to buy, all of the while using the proceeds to bid up the price of his own land.

How does that result in central bank tightening?

[Frank Restly]

Nick,

Contrast that with a central bank that can set an interest rate by decree - for instance - "The central bank of the United States has raised its interbank lending rate from 0% to 3%". In that instance, no assets are bought or sold. Tightening has occurred with no transfer of assets.

A central bank is a strange creature - it has two means of affecting interest rates - open market operations (bond buying and selling) and by decree. In the U. S. these two means were given to the central bank 20 years apart from each other (Federal Reserve Act of 1913 and Banking Act of 1933). Can you guess why two methods exist?

[ATR]

JKH, in your first comment, you asked Nick how the CB would control interbank rates in his world, so I threw out a few options and examples.

"What interest rate do bank reserves trade at?

How is that rate established?"

You also seemed to be focusing on floor rates and the floor system as a method for central bank rate administration and rate control, but I wanted to make clear there are variety of other options. There's also a ceiling system. There’s also altering reserve requirements in conjunction with floor/ceiling rates to achieve a particular interest rate even with varying land price targets (and their accompanying reserve quantity changes). OMOs to independently move rates between the floor and ceiling also could be used with an appropriate reserve requirement, but the CB would have to accept some price variation around its land price target (probably small though, if the reserve demand curve is inelastic enough, as you note)

[JKH]

ATR,

Thx - you were comprehensive in looking at different techniques.

I think we're saying something similar in general.

Point being there's an issue there around reserve management and policy interest rate setting that can't be ignored as part of the CB's management of land prices as well. I don't see that question as being unrealistic in context.

[ATR]

Also, in my opinion, it's not standard stuff. At the very least, I haven't seen it adequately explained in the Post-Keynesian/MR/etc. corner of the blogosphere nor in the Fed’s more popular papers/primers (nor anywhere else in the literature, blogosphere, media, etc.). Based on how you've described what the CB does here and elsewhere, I’m not sure I 100% agree with your views, or rather, am not sure you see it as Bindseil and Woodford do. But no need to get any more technical here for the purposes of this post and the questions you posed Nick.

Off-topic: On related note, I sent you a response to a conversation we were having on MR through the MR contact form, since the post’s comments section closed. Did you ever receive it? No need to respond to it if you don’t want to or don’t have the time, just wanted to make sure you received it.

[ATR]

"Point being there's an issue there around reserve management and policy interest rate setting that can't be ignored as part of the CB's management of land prices as well. I don't see that question as being unrealistic in context."

Definitely not unrealistic. And the question must be confronted: will the central bank care about rates, and their volatility, in that world?

Nick's default setting was probably something like 0% IOR and some ceiling rate. In that world, three things would happen to rates:

a) in a period of enough excess reserves, rates would stay 0%
b) reserves might move to a level where they become volatile in between 0% and whatever the ceiling is
c) in a period of enough deficient reserves, rates would stay at the ceiling

If he literally didn't care about rates, he'd live with the rates and volatility implied a), b), and c) and any movement between them across time.

On CBs and the gold standard: Nick's saying that's essentially what CBs did when they were on gold standards. I think the history is a bit more complicated though. I need to revisit my Mehrling, but with minimal threat of gold drains to external sectors, for example, I think CBs were willing to manipulate rates a bit by changing the floor/ceiling and/or using OMOs - that is, without worrying about defending the gold standard.

[JKH]

ATR - didn't see that contact form, I'll look into it

[Mike Sproul]

Nick:

"Only if land and dollars are perfect substitutes in portfolios. If land pays rent, but dollars are more liquid, they won't be."

But real banks wouldn't push things to the limit like that. If the bank's land purchases start affecting the price of land, they'll start buying other stuff. If the bank floods the market with too much money, the money would start refluxing to the bank, and the bank would stop issuing.

[JP Koning]

In the gold standard days, a central bank like the BOE converted notes into gold and vice versa as well as setting its discount rate -- the rate at which it would provide a loan upon certain sorts of collateral. If the CB set the discount rate too low relative to other banks, prices wouldn't change -- the price level was determined by gold. However, the central bank would lose gold reserves to the others. That's what I get from reading David Glasner, at least.

[JP Koning]

"There is an alternate universe, just like our own, with one exception. For historical reasons*, central banks do not own government bonds. They own land instead. They buy and sell land, and adjust their target price of land several times a year, to try to keep CPI inflation at the 2% target. If a central bank fears that inflation will fall below the 2% target, it buys land and raises the target price of land."

I don't think these are precisely symmetrical. A modern bank doesn't adjust the target price for bonds in order to keep inflation on target, it adjusts the target for the rental rate on reserves. If it did set government bond prices as its target, it would be in a different sort of universe than the one we're in now. So you're alternate universe is like our own, but with several exceptions. A land-owning central bank could also buy and sell land in a way to hit its target for the rental rate on reserves, in which case it would be exactly like a modern central bank, save for the one exception that it held land instead of government bonds.

[ATR]

JP, good points, but I didn't completely follow your train of thought at the end. First you said Nick's world was quite different (and I followed your logic), but then I read you as saying it wasn't that much different (which I might agree with - see below).

If Nick had said the central bank's operational target remains the interbank rate and all the CB changes is that they'll conduct OMOs with land instead of govt bonds (and there's no land price or bond price operational targeting to speak of), then his world would be quite similar. Indeed, the asset of choice for OMOs was not always government bonds. For example, government bonds didn't become the OMO asset of choice in the U.S. until after WW2, when the financial system became flush with them. Before that, there were "real bills," bills of exchange, etc.

But if we now say the CB is going to try to target the price of land AND the interbank rate through OMOs of land, that'd be like saying the CB will target the price of bonds and the interbank rate via OMOs of bonds. Well, I suppose both are possible. We're sort of doing that today in the U.S., and the Fed did it before the U.S. Treasury Accord. By setting the right combo of a) floor and ceiling rates, b) reserve requirements, and c) the size of the land or bond OMO, the CB may be able to do both. So I suppose I agree with you, even though I'm not 100% following what you said :).

[Too Much Fed]

For Mike Sproul and JKH:

The way central banks are run now, they don't refuse to take back currency for gov't bonds? Right?

[Min]

Nick Rowe: "Given that half the time the economy is going too slow, and the other half too fast, would you be happy to take money out of the hands of the poor whenever we need to slow it down?"

Weren't you the one who let me know that skimming off the top is the least distortionary way? (One Big Union)

[Diego Espinosa]

"It could always sell land short, basically borrowing against its future seigniorage profits."

Future seigniorage profits are an asset of the Treasury, not the central bank. In the U.S., the CBO makes baseline budget projections that include seigniorage profits. Falling short of this baseline increases the projected deficit, which increases the taxpayers' tax liability.

[ATR]

Nick, how would the land short sell work exactly? Also, I was going to say that if the Fed ran out of assets today, it could always sell its own bonds or offer its own time deposits. But that concept doesn't translate directly to land, since the Fed can't create land.

[Ludovic Coval]

First I'm not economist so I apologize if my question seem both simple and dumb.

Is not land a limited resource ? What happen if

#1 There is no more land to buy for CB ?
#2 There is no more land to sell for CB ?

Will not those two case be identical in your alternate universe to a money ZLB in our ?

Second, I'm under impression that you set an erroneous link between ZLB and Liquidity Trap. For Keynes, a LT can occur at any CB interest rate but later is a mean to avert eventual LP effect. When ZLB is hit LP effect can not be more offset. It is LP => ZLB not the other way around.

[JKH]

ATR,

By standard, I meant something like this:

http://www.ny.frb.org/research/epr/08v14n2/0809keis.pdf

[JKH]

JP,

"In the gold standard days, a central bank like the BOE converted notes into gold and vice versa as well as setting its discount rate ...

I don't think these are precisely symmetrical. A modern bank doesn't adjust the target price for bonds in order to keep inflation on target, it adjusts the target for the rental rate on reserves"

Right. They do both.

Even if today one makes the case that under QE the CB targets the quantity of bonds on its balance sheet without necessarily targeting the price of bonds directly or effectively, it has to do both asset and liability targeting.

The CB can transition from exclusive liability price targeting (conventional reserve management, with or without IOR, with various rate boundary management arrangements optional) to combined asset price targeting (gold/land/QE pricing) and liability price targeting (the price of reserves, one way or another).

But with asset price targeting, the CB must do both - i.e. it must continue to do separate but co-ordinated liability price targeting. The CB sets the policy rate. The market follows what it thinks the CB will do with the policy rate.

[Ralph Musgrave]

Nick,

I don’t think that simply saying “Would it make much difference…?” answers my point. So I’ll repeat my point.

Having the central bank purchase just one asset (land, government debt, or whatever) is distortionary. Strikes me that’s obvious from the effect of QE: QE has boosted the price of assets that are similar to government debt, i.e. shares. As to the effect on spending on Main Street, that’s been more muted. Plus spending by the US government on infrastructure during the crisis actually DECLINED: completely barmy. It should at the very least have remained constant, or preferably risen a bit.

And I’m not arguing for pure fiscal policy with no monetary element. The problem with fiscal alone is crowding out. Plus having the government / central bank machine borrow money when it can print the stuff strikes me as lunacy. I therefor favor the policy advocated by most MMTers and by Positive Money: simply create new money and spend it (and/or cut taxes) in a recession.

Min,

As you’ll see from my above answer to Nick, I pretty much agree with you. Though I don’t agree with SPECIFICALLY targeting the poor, any more than agree with specifically targeting government debt, or land (as per Nick’s suggestion).

[Adam P]

"if the central bank doubles the target price of land, that will eventually cause the general price level to double too, other things equal. This is known as "land price neutrality". And they say that if the central bank makes the target price of land grow at 10% per year, that will eventually cause the general inflation rate to increase to 10% per year too, other things equal. This is known as "land price superneutrality". More sophisticated macroeconomic models incorporate sticky prices and expectations, and try to say something more precise than "eventually". "

This paragraph is patently false in all but the most knife edge case. It is only true if the rate of productivity growth is identical for all goods traded all the time.

This would never be true, for example the rate of producitivity growth in services tends to be lower than in manufactures.

The conclusion of this post is simply wrong. It is not true that the central bank can control inflation, on any timescale, by controlling the price of a single traded good or asset unless that implies control of the real interest rate.

[Nick Rowe]

Adam P: It is simply the standard neutrality/superneutrality/quantity theory of money restated.

Sure, if the equilibrium (natural) real value of land is growing at 1% per year, then a 10% growth rate in the nominal price of land causes 9% inflation. That's why i said "other things equal". Or, we could restate it as: "a 1 percentage point increase in the growth rate of the price of land causes a 1 percentage point increase in the inflation rate." Same difference.

"The conclusion of this post is simply wrong. It is not true that the central bank can control inflation, on any timescale, by controlling the price of a single traded good or asset unless that implies control of the real interest rate."

The conclusion is totally right. Actually, it's a lot easier to see how a central bank can control the price level by controlling the nominal price of land than it is to see how it could do it by controlling the nominal interest rate. Because you don't get the Steve Williamson/Kocherlakota problem. It's totally standard classical dichotomy, for the long run result. Real factors determine relative prices; so if the CB controls one nominal price that determines all nominal prices. Sure, nominal interest rates will change too, if the CB changes the growth rate of land prices. And real interest rates will change too, in the short run, if the CB changes the price of land and prices are sticky. The Price/rent ratio for land *is* (the reciprocal of) a (very long term) real interest rate.

[Nick Rowe]

ATR: "Nick, how would the land short sell work exactly?"

I confess I'm not exactly sure. Because land isn't fungible (each acre is different). But I vaguely remember hearing there was some way to short houses, which also aren't fungible)? But if you promise to pay a stream of payments, indexed to land rents, you have effectively created synthetic land.

[Nick Rowe]

ATR: "I think Nick’s point is that he doesn’t care where rates wind up, as long as he achieves his land price target. He thinks the land price target is enough for the CB to achieve its objectives."

Yep.

[JKH]

TMF,

"The way central banks are run now, they don't refuse to take back currency for gov't bonds? Right?"

There's no obligation for a CB to be on the offer for bonds. If they're not in the market for other purposes, they're just as likely to steer some lost customer to a bond dealer.

[JKH]

ATR,

Policy rates "wind up" where the CB sets them.

[Odie]

What I am really wondering about now: If the CB can just pass land to the government why can it not also pass on money and let the government buy the land? Where do you draw the line between monetary and fiscal policy?

[Barkley Rosser]

I do not think their banks directly bought and sold land, but historically certain nations at times had land be the nominal base of their currencies. Probably the most famous case was the German rentenmark, although hyperinflation in the 1920s led to its demise, with this supposed land price anchor providing no stability in the end.

[Max]

When you talk about buying land as a solution to the ZLB, you're talking about the purchase as a trade (like Warren Buffet's financial crisis investments), not as a means of permanently shifting the price level, correct?

Which would make sense if a "flight from land" was the reason for hitting the ZLB. But suppose that land is considered a safe haven asset, or for some reason happens to be overvalued when the ZLB is hit. What good would buying it do?

(I'm not arguing about whether the CB can control the price level via land - just whether it's a solution to the ZLB).

[JP Koning]

ATR/JKH

I think we're on the same page. Nick's land example seems to me to be less like our modern world and more like the WWII scenario you mention, when the Fed directly set bond and bill prices (it had ceased to enforce the gold peg).

If Nick's example were to be like our world, then the Fed would be buying and selling land to enforce the rental rate on reserves (the overnight rate), not to enforce the price of land, which would stay floating. I don't see the modern day Fed as targeting bond prices, just setting an amount it'll buy and letting prices bob about.

Barkley, the rentenmark was implemented to help stop the inflation -- it had a stable value thanks to its claim on property.

http://jpkoning.blogspot.ca/2012/12/how-to-stop-hyperinflation.html
http://jpkoning.blogspot.ca/2013/01/rudolph-havenstein-independent-central.html

[Min]

Nick Rowe: "Real factors determine relative prices; so if the CB controls one nominal price that determines all nominal prices."

That assumes that price ratios remain the same when the price of one thing changes. Something that we do not see in real life. Something that ignores psychology. Something that assumes no constraint on money.

Given a monetary constraint, for instance, a rise in the nominal price of land might lead to a rise in nominal housing rents, which might lead to a lower ratio of the nominal price of movie tickets to the nominal housing rents, as a monetary constraint constrains household budgets. Movie tickets might go up, but by not as much as the price of land.

[Min]

Ralph Musgrave: "Min,

"As you’ll see from my above answer to Nick, I pretty much agree with you. Though I don’t agree with SPECIFICALLY targeting the poor,"

Thanks, Ralph. Me either. :)

[Min]

Nick Rowe: "Real factors determine relative prices"

I guess one thing that I am trying to say is that there is no such thing as a real price ratio between movie tickets and land, not even in the long run -- especially in the long run. I don't mean that there are no real price ratios, but, as J. P. Morgan said, "Prices fluctuate." They are nominal, they are psychological, they are ephemeral.

[ATR]

JKH - okay, I stand corrected, that Fed primer is good, but that's from the same family of thought to which Woodford/Bindseil belong. In other words, I could have said 'see this paper' instead of see 'Woodford / Bindseil.' For example, the Whitesell papers it cites to support its explanation of how it works themselves cite Woodford 2001. Others in that family of research cite Bindseil. Moreover, that was published only recently in 2008, and though Poole was on the right track in 1968, that line of thought practically vanished from the literature until the late 1990s/early 2000s. From what I can tell from general discourse, it's still not widely known or understood outside of that circle of authors or people who research / work in that area. Lastly, I think getting inside of the math helps add a level of insight not provided in that paper, but that's just my opinion.

[JKH]

ATR

Not trying to correct you, or be dismissive in using the word "standard".

Whether Keister, Woodford, or Bindseil or a few others - its' all good and its how central banks work - but not known nearly well enough, as you say.

I use the word "standard" as meaning the guys who get the facts right.

:)

[Too Much Fed]

JKH, I don't think I said that right. Assume nobody wants currency anymore. People turn it into the commercial banks for demand deposits. The commercial banks turn it into the fed. The fed does not say no. The fed will exchange currency for gov't bonds.

With QE and if commercial banks try to exchange central bank reserves for gov't bonds, the fed says no to the exchange of central bank reserves for gov't bonds.

Correct?

[Peter N]

J P Koning

"the rentenmark was implemented to help stop the inflation -- it had a stable value thanks to its claim on property"

The "it had a stable value thanks to its claim on property" apparently isn't quite correct. The story is rather complicated.

The stability came from it being a (dollar) indexed "wertbeständiges" currency in an economy that was by now dominated by them and by the extremely tight monetary policy of early 1924.

However,

"While Germany enjoyed monthly small trade surpluses of 68 million and 93 million gold marks in the last two quarters of 1923, in the first two quarters of 1924 average monthly trade deficits of 212 million and 323 million gold marks were observed."

http://www.econ.puc-rio.br/gfranco/Ch10.PDF

The rentenmark bought time until the Dawes loan, after which capital inflows financed the trade deficit.

This story really makes a great deal more sense than the "magic of the rentenmark".

"It was only in 1925 that there was a massive return of funds, as measured by the very large unaccounted capital inflow of approximately US$ 481 million shown in Table 10-5. This seems to suggest that the decisive factor in securing confidence on the stabilization was not the spell of dear money enforced in the spring of 1924 but some event taking place later in 1924, most likely the approval of the Dawes plan in August and the floatation of the Dawes loan in October"

BTW the paper argues that the cause of the transition to hyperinflation was the marginalization of the mark by the huge amount of better "wertbeständiges" currency - a fall in demand.

[JP Koning]

PeterN, leave a comment on my blog if you want to discuss the rentenmark, Nick is probably grinding his teeth at this off-topicness.

[Nick Rowe]

JP: Nick is fine. I get annoyed with off-topicness early in the comments, but once the main topic has been discussed thoroughly, we can wander a bit. Some very interesting sub-themes here. Carry on about the rentenmark.

[JKH]

TMF:

The Fed exchanges currency for bank reserves on demand.

As a separate step, the Fed has the option of selling bonds or doing something else to drain bank reserves if necessary.

The commercial banks can’t buy bonds from the Fed unless the Fed decides first it wants to sell bonds.

But the banks on demand can buy currency in exchange for reserves.

[Mike Sproul]

ATR:
"The way central banks are run now, they don't refuse to take back currency for gov't bonds? Right?"

Correct. I like to think of a case where a bunch of holders of paper dollars are on their way to the bank, aiming to redeem their dollars for gold. But the bank knows they are coming and uses its bonds to buy back a bunch of its dollars. This soaks up the unwanted dollars, and the dollar holders decide they don't want to redeem their dollars for gold after all.

[Mike Sproul]

Barkley Rosser:
"Probably the most famous case was the German rentenmark, although hyperinflation in the 1920s led to its demise, with this supposed land price anchor providing no stability in the end."

Land makes perfectly good backing for money, and if you issue $100 against land that is worth 100 oz. of silver, then $1=1 oz. The problem is that the Germans issued 100 million rentenmarks against land that was worth only 100 oz, so of course there was inflation.

[primedprimate]

Nick would any tradable asset work in place of land? What about something that occupies an almost inconsequential part of the economy?

Could the CB, for instance, buy and sell autographed baseballs by Babe Ruth to control the economy?

[Nick Rowe]

primed; good question. For inflation targets of around 2%, currency demanded is around 5% to 10% of GDP, in most advanced economies. So you would need some asset whose total value was at least 10% of GDP. Plus, it would be nice if the central bank earned some income from owning that asset, or renting it out. So land would work OK, but autographed baseballs probably wouldn't.

[Too Much Fed]

Nick's post said: "ATR: "I think Nick’s point is that he doesn’t care where rates wind up, as long as he achieves his land price target. He thinks the land price target is enough for the CB to achieve its objectives."

Yep."

I think it will be more complicated than that.

Monetary base could go up. Demand deposits could go down.

Monetary base could go down. Demand deposits could go up.

In both cases and with 1 to 1 convertibility (relative pricing) between currency and demand deposits, both currency and demand deposits are MOA and MOE. MOA does not have to be monetary base.

Also, let's assume Apple and Warren Buffett have accumulated a lot of land and gov't bonds because of excessively positive real earnings growth and saving. Start at year 0 with both Apple and Buffett not wanting to spend on goods/services for at least 20 years. They are both using land and gov't bonds as savings vehicles. In year 2, debt defaults and debt repayments cause demand deposits to fall so that MOA falls. The fed buys land and bonds in the exact amount that demand deposits fell, probably overpaying for the assets. Apple and Buffett both yawn about the demand deposits and hold them as a savings vehicle. MOA in circulation remains at where it has fallen too. There is still a shortage.

[Too Much Fed]

Mike Sproul at 5:22, I think that was my question.

[Too Much Fed]

JKH, let's say the fed funds target is 1.5%. People no longer want to hold any currency. They swap currency for demand deposits. [The commercial banks swap the currency for central bank reserves.] The [] part always happens. The fed does not refuse to take back the currency?

Next, the fed funds rate starts falling towards zero. Most likely, the fed will swap a treasury for the central bank reserves. The fed could pay IOR? The fed could raise the reserve requirement?

JKH, do you think demand deposits are MOA?

[JKH]

TMF,

- no the Fed does not refuse
- yes, any of those
- I'm unschooled in MOA; Scott Sumner probably has the best definition

[Squeeky Wheel]

JKH, ATR, Nick,

Regarding the setting of interest rates for reserve liquidity when the CB does not target rates:

The CB of Singapore (Monetary Authority of Singapore) targets SGD exchange rates (technically a nominal effective exchange rate on a basket of currencies within a defined band with a defined drift). Hence domestic interest rates should be 'market determined'. The precise specification is
"6 The MAS Standing Facility is unique [1] in that interest rates under the Facility will be based on a market-determined rate. This is determined on a daily basis via an auction of overnight clean borrowing from Primary Dealers in the morning of each working day through MAS' daily money market operations. The weighted average of successful bids will form the reference rate for that day. The rate at which Primary Dealers may borrow from the facility will be 50 basis points above the day's reference rate, while the deposit rate will be 50 basis points below the reference rate.

7 The introduction of the Standing Facility does not represent a departure from MAS' exchange-rate based monetary policy, and does not constitute a mechanism for interest rate targeting. " (http://www.mas.gov.sg/news-and-publications/press-releases/2006/mas-launches-electronic-trading-platform-for-sgs.aspx)

Clearly there is a daily floor and ceiling, but those rates are not set by the CB (at least not directly - the expectations of SGD NEER does tend to set rates relative to Fed rates). Today the deposit rate is 0.00 and the borrow rate is 0.62.

[reason]

Nick,
it seems to me a corollary of this is that real factors must determine the real interest rate. I may at one time have believed this, but I don't anymore - the evidence is simply against it. I think this is a dilemma for monetary theorists, but I guess that the key lies in the anomolous behaviour of exchange rates. Quite simply, international financial flows do not equalise real rates of return in the real world, so that an increase in leverage pushes down real rates of return (partly because of the bias in lending policies towards lending for purchase of assets - regardless almost of the rate of return).

[reason]

Oops
this was commenting on something that nick wrote in a comment

" Actually, it's a lot easier to see how a central bank can control the price level by controlling the nominal price of land than it is to see how it could do it by controlling the nominal interest rate. Because you don't get the Steve Williamson/Kocherlakota problem. It's totally standard classical dichotomy, for the long run result. Real factors determine relative prices; so if the CB controls one nominal price that determines all nominal prices. Sure, nominal interest rates will change too, if the CB changes the growth rate of land prices. And real interest rates will change too, in the short run, if the CB changes the price of land and prices are sticky. The Price/rent ratio for land *is* (the reciprocal of) a (very long term) real interest rate."

[reason]

Min
"
I guess one thing that I am trying to say is that there is no such thing as a real price ratio between movie tickets and land, not even in the long run -- especially in the long run. I don't mean that there are no real price ratios, but, as J. P. Morgan said, "Prices fluctuate." They are nominal, they are psychological, they are ephemeral."

My guess is that land is a REALLY lousy example. Land is a positional good. It gobbles up excess (so that part of its value is its price - and increasing taxes will definitely reduce the relative price of land). And there is no such thing as "the price of land" (- every piece of land is unique, whereas every dollar is the same).

[reason]

By the way, I think this whole paragraph is completely nuts:
"If a central bank raises land prices, that raises asset prices across the economy, which makes investment in newly-produced capital goods relatively more profitable, which increases aggregate demand. An increase in land prices, and other asset prices, also increases consumption demand, via wealth effects and substitution effects. Some economists stress the role of expectations, because if the central bank raises land prices people know that all prices must rise in the new equilibrium. A few Monetarists try to insist that what is important is that the central bank is issuing more money, and the rise in the price of land is just a symptom, rather than a cause. But nobody takes them very seriously, because everybody knows that real world central banks set the price of land, and the quantity of money is endogenous."

If a central bank raises land prices - then everybody then it will mean that people buying houses will have to go deeper into debt and so will spend less on other things - making investment in newly produced capital goods relatively LESS profitable (and also drying up investment funds for other purposes). An increase in land prices (which will eventually feed through to rents) is exactly like a tax increase. And there is no such thing as equlibrium (because the change in relatively asset prices will change the distribution of wealth - which will then agains change the "equilibrium" ad infinitum).

[reason]

Oops
That last sentence came out completely mangled.
"And there is no such thing as equilibrium (because the change in relative asset prices will change the distribution of wealth - which will then again change the "equilibrium" ad infinitum). Non-independence of the the mechanism and the target happens with monetary policy as well as with fiscal policy. I simply don't like the concept of "equilibrium" - we are never at equilibrium and will never be at equilibrium. We can push things one way or the other, but it is better to keep on eye on where we actually are than to assume we know where we are going.

[JKH]

Squeeky W.,

I believe Singapore is a currency board arrangement?

If so, that's very different, and the exception when it comes to interest rate policy targeting.

[Nick Rowe]

Squeeky: good comment. That's almost exactly the sort of thing I had in mind. Just replace Singapore's basket of currencies with land. (Though land is a bit less liquid. But hey, I only meant it as a metaphor anyway.)

[Too Much Fed]

JKH said: "I'm unschooled in MOA; Scott Sumner probably has the best definition"

I cringe just hearing that one. Let's work on it because Scott Sumner says MOA is monetary base (currency plus central bank reserves). If he is right, then banking should not matter. I believe banking matters. I think you believe banking matters too.

Let's start here:

http://jpkoning.blogspot.com/2012/11/discussions-of-medium-of-account-could.html

"JP Koning December 9, 2012 at 1:13 PM

The definitions I'm using for medium of account, unit of account, and medium of exchange come from NME (see comment above). Here is an example:

"A medium of exchange is an asset that is widely accepted in trade and to settle financial obligations. Currency notes or transferable bank deposits are typical examples. A medium of account is the commodity defining the unit of account. A unit of account is a specific amount of the medium of account. For example, for the gold standard the medium of account is gold, while the unit of account might be one ounce or one pound of gold of specific purity. A unit of account is the unit in which the medium of exchange and other assets are denominated and in which other values and prices are expressed."

Warren Coats, 1994."

I'm not completely sure about the last sentence.

Also, http://worthwhile.typepad.com/worthwhile_canadian_initi/2012/10/medium-of-account-vs-medium-of-exchange.html

"[Update: just to clarify terminology: in my model, gold is the medium of account; and (say) an ounce of gold is the unit of account.]"

I agree.

And, "All prices are quoted in ounces of gold."

Let's start there. I agree. Nick might be shocked I agree with him.

Using a gold standard with no banks, gold is MOA and one ounce of gold is UOA. There are 1,000 ounces of gold. Gold is not used as MOE at first. If V of gold = 0, then NGDP = 0. That is a problem. Solution: Add currency as MOE. Fix convertibility at 1 ounce to $1 (relative pricing). Everyone turns in their gold for $1,000 in currency. The currency entity gets the 1,000 ounces of gold. Everyone now quotes prices in terms of gold and currency at the same price because of the 1 to 1 convertibility. The $1,000 in currency starts circulating with a velocity of 2. NGDP = $2,000. Sound good so far?

[JKH]

TMF,

For convertible currencies, MOA is the object of conversion (e.g. gold)?

For non-convertible currencies, MOA is the monetary base, because there is no other objection of conversion? It seems like a default in that case.

Also, banks are guaranteeing customers convertibility of their deposits into currency, and the CB is guaranteeing the banks convertibility of their reserves into currency. So what the banks offer to customers as deposits depends on the monetary base as MOA.

That's my guess.

[Max]

JKH,

Base money is the MOA for bank money, that's clear. But in my opinion it doesn't make sense to say that base money is the MOA for base money (or that there's no MOA). The MOA for base money is whatever the central bank is targeting (e.g. a price index).

In theory, the Fed could offer gold convertibility (not at a fixed price!) without changing its target in any way. The MOA would remain a price index, and the gold market would have no effect on prices in general (unlike the gold standard).

[Too Much Fed]

JKH said: “For convertible currencies, MOA is the object of conversion (e.g. gold)?”

In my example (fixed/pegged convertibility), gold and currency are both MOA. There is a dual MOA. Everyone quotes prices in terms of gold and currency at the same price because of the 1 to 1 convertibility.

From my example above (fixed/pegged convertibility), 1,000 ounces of gold have a velocity of zero at the currency entity, and $1,000 in currency have a velocity of 2.

Now mine another 1,000 ounces of gold and have gold be MOE as well as MOA. Everyone keeps their original $1,000 in currency. Another $1,000 in currency needs to be available to maintain the fixed 1 to 1 convertibility. 1,000 ounces of gold have a velocity of zero at the currency entity and 1,000 ounces of gold have a velocity of 2. $1,000 in currency have a velocity of 2, and $1,000 in currency have a velocity of zero at the currency entity. NGDP = 4,000.

Now have everyone panic and not want to hold any currency/use it as MOE. They swap for gold. 2,000 ounces of gold have a velocity of 2. $2,000 in currency have a velocity of zero at the currency entity. NGDP = 4,000.

Now have everyone panic and not want to hold any gold/use it as MOE. They swap for currency. 2,000 ounces of gold have a velocity of zero at the currency entity, and $2,000 in currency have a velocity of 2. NGDP = 4,000.

Notice the fixed/pegged convertibility leads to a dual MOA. Sound good?

Now make it more interesting. Get rid of the gold standard and replace it with a demand deposit standard where “mining” more demand deposits relates to the capital requirement/ratio.