Currency wars and forced dissaving

The Fed is the loan placement officer for the world's central banks. The US government is the Fed's borrower of last resort. The forced loans can be called in at any time the lender wishes.

People are different; that's why they trade. Sometimes that trade will take place between people who live in the same country, and sometimes it won't. There is intranational trade and international trade. Some of that trade is intertemporal. Some people will want to lend and others will want to borrow. Again, sometimes the borrowers and lenders will live in the same country, and sometimes they won't.

Why is international intertemporal trade ("global financial imbalances") seen as a special problem? So what if people in one country are net borrowers and people in another country are net lenders? It would be a total fluke if it didn't turn out that way. Countries are not identical. We don't expect a country's imports of apples to exactly balance its exports of apples. We don't see it as a problem if net imports of apples are balanced by net exports of bananas. Why should we expect imports of all goods today to exactly balance exports of all goods today? Why is it a problem if net imports of goods today are balanced by net exports of goods in the future?

What's the policy problem of global financial "imbalances"?

One answer is that there is a global paradox of thrift. There is a global excess of desired saving over desired investment, and so global output falls to equalise the two. Saving may be good for the individual, or individual country, but bad for the world as a whole. There is a global shortage of aggregate demand, and like a common property resource, some people and countries are hogging too much of that scarce aggregate demand for themselves, and not returning it to the common pool by spending.

That answer begs the question: why is there a global shortage of aggregate demand? Why not just print money? It's like a library, where some selfish people are hogging all the books and not returning them to the library so they can go back into circulation. Except: in this case the library can simply print more books for free. And that is what the US Fed seems to have decided to do.

What's the problem?

Pre-Keynesian monetary economists had the concept of "forced savings". When the central bank decides to print money, it forces people to save. I'm going to turn that concept on its head. When people decide to save money, it forces the issuer of money to dissave.

Nobody can force me to borrow from them. Adverts come in the mail every day offering me loans, but I don't have to accept any of them.

But it's different for a central bank. If there's an increased demand to hold Bank of Canada money, the Bank of Canada has to satisfy that demand by printing more money. If it doesn't satisfy that demand, there will be a recession and disinflation in Canada. Given that the Bank of Canada must prevent a recession and disinflation, we are forcing the Bank of Canada to print money. And if next year the demand for money falls back to normal, the Bank of Canada is forced to buy back that extra money it had previously printed, to avoid an inflationary boom.

In effect, we have forced the Bank of Canada to borrow from us. We have forced the Bank of Canada to take an interest free loan. We bought the Bank of Canada's IOUs, and next year the Bank of Canada bought back those IOUs. And it's an interest-free loan to the Bank of Canada, because Bank of Canada IOUs, at least in the form of paper currency, don't pay any interest. It's a forced loan, and the amount and duration of the loan are decided by us, not the borrower. We decide how much the Bank of Canada has to borrow from us; and we decide when the Bank of Canada must pay it back. We don't give the Bank of Canada any advance warning when we want our loan repaid.

Maybe the Bank of Canada should not be worried if we force it to borrow from us. It can use the proceeds of the forced loan to buy Canadian government bonds, and earn interest. Next year it sells the bonds again, uses the proceeds to retire the now unwanted currency, and keeps the interest as profits (which it hands over to its owner, the Government of Canada).

We may have forced to Bank of Canada to borrow from us, but we haven't yet forced it to dissave. It may turn around and buy Government of Canada bonds. Does this mean the Bank of Canada is making a loan to the Government of Canada? Only if the Government of Canada issues new bonds. Otherwise, the Bank of Canada must go out into the open market and buy already existing bonds from the people who hold them, which means that the Bank of Canada is making a loan to the people who sell it those bonds. And, ultimately, the people who sell the Bank of Canada those bonds might be the very same people who wanted to hold more Bank of Canada money. They didn't want to save more; they just wanted to hold their existing stock of savings in a different form: money rather than bonds.

Nevertheless, by demanding more money, and forcing the Bank of Canada to issue more, the immediate effect is to force the Bank of Canada to dissave. The Bank of Canada can only escape being forced to dissave if it can persuade someone else voluntarily to borrow from the Bank of Canada. Unlike me, the Bank of Canada is forced to accept the offer of a loan that comes in the mail. And if it doesn't want to dissave, it must try to find someone else who is willing to borrow from it. The Bank of Canada, in other words, is forced to act as a loan placement officer, with a quota of loans it must place. And if it fails to place the loans with willing borrowers, it must borrow and dissave itself.

Canada is a small country (in terms of the world economy), and the Canadian dollar is not (very much) a reserve currency. The US economy is large, and the US dollar is a reserve currency. The Bank of Canada is the loan placement officer for Canada, and is forced to place a quota of loans the size of which is determined by Canadians. The Fed is the loan placement officer for the world monetary system, and is forced to place a quota of loans the size of which is determined not just by US holders of currency, but by anyone in the world, including world central banks. If it fails in its attempt to place those loans, it must dissave itself, or else accept a US recession and disinflation.

Now, it's true that world central banks do not hold (much) US currency. Instead, they hold US government bonds. But that doesn't change the story. If the central banks in the rest of the world decide to save $1b more and hold S1b more US dollars (currency) in reserves, the Fed is forced to issue those dollars to prevent a recession and disinflation in the US. The Fed is forced to accept the offer of a loan. The immediate effect is to force the Fed to dissave. The Fed must then place those loans if it wishes to offset its own dissaving by lending to someone else. The Fed, as the world's loan officer, must meet its quota. And it must do whatever it takes to persuade someone to borrow from it, by selling US bonds to the Fed. And if central banks in the rest of the world subsequently switch from holding US dollars to holding US bonds, that reduces the size of the Fed's sales quota by $1b, but at the same time takes away $1b of the Fed's potential customers' demand for loans. The fact that central banks in the rest of the world decide to hold an extra $1b in US bonds rather than an extra $1b in currency doesn't make it any easier for the Fed to meet its quota for placing loans.

When central banks in the rest of the world decide to hold more reserves, and so save more, they force someone else to dissave. The Fed has the responsibility of finding a willing dissaver, or must dissave itself. The Fed is the world's loan placement officer, and must meet a monthly quota of loans determined by the world demand for US dollars. If it fails in its task of placing its quota of loans, it must borrow itself, or allow disinflation and recession. And the US government acts as the Fed's borrower of last resort, if it cannot place the loans elsewhere. And the loans don't have a fixed term. The lender can demand repayment any time, with no advance warning.

[I'm not sure this is 100% right, but I'm posting it anyway.

Is "loan placement officer" the right job title? The person at the bank who is told "your job is to get out there and find people who will borrow $10 million from us this month".]

Comments

[The Money Demand Blog (123)]

Sounds right. In late 2008, there were many borrowers who were willing to borrow from Bernanke in his credit easing operations. That demand has dried up. Unless Bernanke changes the prices at which he is willing to do credit easing again, all we got is the US government as the borrower of last resort.

[D. I. Harris]

"We don't expect imports of apples to exactly balance exports of apples" threw me for a loop for quite a while. My first thought was, "Then were are the apples coming from?"

It took me about 15 seconds to realise that you meant within a specific country.

[Nick Rowe]

123: It sort of sounds right, doesn't it. Though the Fed could persue other lenders (buy stocks?).

D.I. Harris: Good call. I have edited the post to make it clear that's what I meant. If it takes a reader 15 seconds to understand me when I'm trying to say something that's simple, it means (well, in this case it means) I didn't write it clearly enough.

[rogue]

If a central bank is buying a reserve currency by issuing its own currency, is it saving or dissaving?

[Nick Rowe]

Rogue; good question. Both. Therefore neither. What it's doing is playing "pass the parcel" to another central bank. "Here. I don't want to act as loan placement officer for this parcel of loans; you do it!"

[Determinant]

So does that parcel then get passed to the People's Bank of China then? Just thinking of a central bank that doesn't have the demand problems that the Fed does, and has a large base to operate with.

[rogue]

Sounds like pass the 'hot' potato, too. And this like the main reason why nobody wants it:

"And the loans don't have a fixed term. The lender can demand repayment any time, with no advance warning."

[Jon]

Nick: let me extend from your argument--lets see if you still like the gruel.

China has a very rapidly growing economy. Like most CB they have to gradually expand the money supply to accommodate the growth, and as a fast growing economy, they have to grow the money supply somewhat faster than usual.

Lets assume, for the sake of illustration, that world-wide 'k' is about the same when looking a broad-based 'M'--I don't assert this is true but simplifies the idea that follows.

Observe: China, unlike many countries in the world does not monetize its own sovereign debt. When looking at the balance sheet of the PBoC, we can find that the currency is ninety percent backed by US Government securities. Also observe that China's economy is now on the order-of the US economy in size.

China, however, needs to expand its money supply to accommodate its growth. So given its policy of backing its money using US debt, it must grow its holdings of US debt in order to do so.

But this demand for US debt must be paid for 'somehow'. That somehow is the unbalanced sale of goods and services back to the US. It must be so, or the aggregate current and capital account flows will not balance.

Now lets do some back-of-the-envelope calculations: in the past several years but before IOR distortions, the Fed monetized debt worth about fifty billion USD annually. Given that China's economy is about the same size, we'd then expect ceteris paribus that they'd monetize about twenty billion annually.

Now, China's required reserve ratio is about ten times larger than the effective required reserve ratio in the US. So in fact, they need a much larger monetary base to support the same amount of broad money and China is growing faster. Lets say China is growing two times as fast.

So we have 20B * 2 * 10 = 400B to adjust our ceteris paribus numbers.

And look, this is about equal to the imbalance in goods and services trade.

[Nick Rowe]

Determinant: you can't pass the parcel/hot potato to the People's Bank of China. It's doesn't play. It doesn't let foreigners hold yuan. It's the one passing the parcel to the Fed.

Rogue: yes. Only only realised that flexible term bit after I had written the post, and had to do a quick edit. (Do kids still play pass the parcel in England? But then hot potato is a better metaphor, because you don't want to get stuck holding it.

Jon: you lost me on the back of the envelope calculations. Where did the 20 billion come from again? Otherwise what you say makes sense. Qualitatively, it's exactly what I had in mind.

[Jon]

Nick: That's an 'imagined' number based on China being like the US in almost every way except with a smaller GDP. So its Central Bank would need to monetize less money per year. Then I 'correct' that with particular recognitions of the way in which China is different.

Hokey, for sure. But the numbers seem, as I state, to have the right order of magnitude.

[Nick Rowe]

Jon: OK. I understand you now. I think you had a typo in your previous comment, when you said "Given that China's economy is about the same size,..." (as the US???) You must have meant 40% the same size.

[Determinant]

Dumb question: Why would China need to back its debt with US Dollars? I thought its currency/debt reserves were a consequence of its export policies and desire to control the yuan.

Nick: Just following your logic of the Fed passing the parcel to another central bank. If not the Fed, then who?

[jsalvati]

Off topic, but does anyone know a good discussion of formalizing the concept of "demand for money"? This concept makes a good deal of sense to me, but I realized I don't know how to formalize it. It's not the same thing as 'quantity of money held', because that remains constant unless the central bank changes it. It seems like maybe you could say the quantity people would hold if they could buy and sell as much as they wanted at current prices or something, but that seems like it runs into trouble with things like rent control. I think maybe Nick Rowe had something about this in the past, discussing constrained demand and such, but I couldn't find it. Can anyone point me to something along these lines?

[Too Much Fed]

"That answer begs the question: why is there a global shortage of aggregate demand?"

I suggest these 2 posts from Billy Blog:

"The fiscal stimulus worked but was captured by profits"

http://bilbo.economicoutlook.net/blog/?p=11911

And, "The origins of the economic crisis"

http://bilbo.economicoutlook.net/blog/?p=277

[Too Much Fed]

"Why not just print money?"

I'm going to substitute medium of exchange for money here. I believe the way the system is set up now, there is no way for new currency to be created directly. That means the only new medium of exchange is demand deposits from currency denominated loans whether gov't or private. That also means someone has to go into currency denominated debt.

[Jon]

Determinant:

If all the variables are picked by the Chinese government and the only endogenous variable is the goods and services trade. Its clear which way the process goes. The goods and services trade has to adjust to equilibrate.

The dog (PBoC) most definitely wags the tail (trade balance).

Compare this to a country that monetizes its own government debt but picks its interest-rate target to hit an exchange-rate goal. Perhaps the Canadians will speak-up because I think that's what they (used to) do. Actual direct intervention in the forex market is rare, and tends to be about bark not bite--so you need reserves but not huge reserves. Forex markets exert pressure daily but interest-rate targets change slowly.

Other countries have pegs but they operate very differently or on different scales.

China is exceptional, and so far, this conversation has been muddled by not first framing what china does differently. Its not just the peg. Its about how the peg is implemented.

[Too Much Fed]

"When people decide to save money, it forces the issuer of money to dissave."

Not exact, but it will work here for now. Bill Mitchell says something similar.

private savings = gov't deficit

[Richard W]

In some respects you are touching on the ' Triffin dilemma '. The US can't run a fiscal and current account surplus without starving the rest of the world of dollars. Therefore, the Fed is forced to constantly dissave. Moreover, when the rest of the world acquires USD assets the only way the ROW can reduce on net their USD assets is by selling less stuff to the US. If one overseas holder of USD assets wants to sell their USD assets they must find a buyer. All that is happening is one foreign holder is being swapped for another foreign holder without any net reduction. If they sell back to a US entity they are just swapping one type of USD asset for another type of USD asset.

[Nick Rowe]

Determinant: I think the parcel stops at the Fed. Does the Fed hold much in the way of forex reserves? Does it ever change? If not, the parcel stops at the Fed.

Why does China hold such large forex reserves? Two reasons come to my mind: it's scared of a financial crisis with funds fleeing abroad; and it wants to promote net exports.

jsalvati: "This concept makes a good deal of sense to me, but I realized I don't know how to formalize it. It's not the same thing as 'quantity of money held', because that remains constant unless the central bank changes it."

The "quantity of apples demanded" is very close to "quantity of apples bought". It's the quantity of apples people *desire* to buy.

Similarly, "money demanded" is very close to "money held". It's the quantity of money people *desire* to hold.

David Laidler's "The Demand for Money" is good.

Too much Fed: "private savings = gov't deficit"

That's true in a closed economy. In an open economy we need to add in or subtract borrowing/lending from/to abroad.

But that's an accounting identity. Bill Mitchell says that. So does every other economist. And an accounting identity alone says nothing about causation. I'm talking about causation. (So is Bill, but what he says about causation won't generally be the same as what I say).

Richard: I think you're right.

[Bill Woolsey]

Treating the demand for base money as lending to the central bank and the quantity of base money as borrowing by the central bank is instructive for many purposes--good.

Since a central bank is a bank--a finanical intermediary--I have trouble with the notion that an increase in the demand for money forces a central bank to dissave. I rather think it forces a central bank to intermediate. It must borrow (and since it doesn't consume, it doesn't dissave,) but rather it must also lend.

When a finanical intermediary lends by purchasing existing securities (like government bonds,) I don't think they are making a loan to those selling the securities. They are directly borrowing from them (by issuing them money.) Perhaps I am just idiosyncratic, but I see the sale of outstanding securities as changing who is lending to the issuer. I hold a newly issued one year T-bill for six months and then sell it to you, who holds it to maturity. I lent the government the money for the first six months, then you lent the money for the second six months.

You argue that the central bank shouldn't worry about being forced to dissave (or really, borrow,) because it can lend at interest (by purchasing government bonds.) This is true unless the interest rate on the bonds becomes so low that it is unprofitable. And that is the problem with the zero bound.

If you see interest bearing deposits as the "core" of money and abstract away from zero interest currency for a minute, you can see that if the interest rate the central bank can get falls so that the issue of money is unprofitable, then all it must do is lower the interest rate it pays on money enough so that it is profitable. If the interest rates the central bank can earn are low enough, then that might require negative nominal interest rates on money.

To bring this to "China," you can see that if China is accumulating too many dollars, then the solution is to charge them for doing so by making them pay to hold dollar balances.

Of course, zero-interest hand-to-hand currency is convenient for many purposes, and so issuing it on demand at par has great benefits. And that means that the interest rates that can be charged for holding deposits cannot be very negative. There is, for all intents and purposes, a zero nominal bound.

When we talk about quantitive easing, especially the purchase of long term to maturity bonds (much less equities,) we are proposing that the central bank borrow by issuing safe and short monetary instruments at perhaps zero interest and then use them to purchase longer term securities that at the very least have interest rate risk. This risk is the cost of the policy. Again, the solution is negative interest rates on the money issued, compensating for the risk, or really, allowing the central bank to purchase short and safe assets as well as issue them, but with the yields turning negative.

Of course, we can "solve" this problem with expected price inflation and so lower, and maybe negative, real interest rates. If you have a target for inflation and the target is too low to handle this, then you have problems. I favor a target for money expenditures, but at a growth rate that should ordinarily create zero inflation. This makes the issue very salient to my reform approach.

Given the money expenditures rule, as you say, the bank(s) of issue is obligated to borrow if the demand for money rises--if people want to lend to the bank(s) of issue. As banks, they turn around and lend a matching amount, perhaps by purchasing outstanding securities rather than making new loans. To keep to their rule they must always stand ready to withdraw the money from circulation--pay off the loans. Generally, they can make money on the difference between the interest rate at which they borrow and the interest rates at which they can lend. What happens if the interest rate at which they can lend gets too low? How do they issue zero interest hand to hand currency? Do they purchase riskier securities at higher yields?

As for the international capital flows, why can't other Americans accumulate Chinese securities at a faster rate than the Chinese accumulate U.S. money? By keeping to a inflation, price level, or money expenditure rule, the central bank must borrow when others wish to lend to it, including China. But suppose when the central bank purchases dollar bonds, those who sell to them purchase Chinese securities? Suppose they purchase even more on top of that? I see no reason why the U.S. cannot create the amount of money demanded and have net capital outflow to China and a trade surplus. Of course, I think it would require that Americans want to save.

How about this story? The Chinese accumulate U.S. money, the central bank issues it to them and purchases U.S. government bonds from U.S. residents. The U.S. residents invest in bank C.D.s which the banks use to fund second mortgages, which fund purchases of fancy cars, vacations, and frequent restaurant meals. Sure enough, the accumulation of U.S. money by China has funded dissaving by Americans. But all it would have required to avoid this would have been for the C.D.s to instead fund capital investment here, and so there would have been a net capital inflow to the U.S. without dissaving, or else the purchase of Chinese securities, so that there was no net capital inflow to the U.S.

[wjd123]

"We know that the renminbi is grossly undervalued, not through questionable estimates that can be endlessly debated, but on a PPE (proof of the pudding is in the eating) basis: the current value of the renminbi is consistent with massive artificial capital export, and that’s that." Paul Krugmam

In a Paradox of Thrift world this beggaring jobs by China through artificial capital exports is making us all poorer. Krugman recommends a 25% tariff on Chinese imports to us.

If China dumps it dollars our Fed can buy low and sell high, while the price of China's exports grows.

We shouldn't trust China as a reliable trade partner, besides making us all poorer it uses trade as a weapon. We just saw an example of this in its dealings with Japan over fishing rights.

It's about jobs and security. We have to stop the Chinese government's beggering.

In the words of Tim Duy, "...we have offshored so much production capacity that it becomes impossible to grow without an expansion of the trade deficit." It time to put a break on outsourcing production and it's time to change dancing partners.

[Winslow R.]

Nick wrote: (So is Bill, but what he says about causation won't generally be the same as what I say).

Yes, but the path to sameness is interesting to watch.

wdj123 wrote: "In a Paradox of Thrift world this beggaring jobs by China through artificial capital exports is making us all poorer. Krugman recommends a 25% tariff on Chinese imports to us."

I believe Krugman only recommends tariffs because our government refuses to provide enough tsy bonds to provide domestic full employment. Krugman believes we should increase government deficit spending, even more strongly than imposing tariffs. QEII will only work if there are sufficient government bonds for the Fed to purchase.

The quantity of tsy bonds must match or exceed desired savings by China etc. (private and foreign sector) to keep the world economy humming. Of course QEII could take the corrupt path and have the Fed purchase the S&P 500 or some other private sector asset.

I don't think Bill would ever say the Fed should purchase stocks from the S&P500.

[David Pearson]

The Fed is "placing the loans". The Treasury is the "willing borrower" in the form of net new issuance to finance deficits. ER's are being spent, its just that the effect on the ER/RR mix is quite small as the effective reserve ratio (given sweep accounts) is quite low -- 3% or below. For example, a bank $100b Treasury buy (which ends up being spent) shows up as a $3b reduction in ER's. This is almost imperceptible at the rate of QE the Fed is planning.

The first QE basically offset the shrinkage of the shadow banking system. That has largely run its course (barring more "accidents"). Private credit has stabilized, with most "delevering" coming in the form of charge-offs against profits rather than repayments (hence, deposits and money aggregates are unaffected). M1, M2, and M3 have all been rising since the summer. From now on, one would expect QE deficit monetization to marginally reduce ER's and increase RR's. We have seen some evidence of rising RR's in the H.3 report.

[JKH]

Bill Woolsey wrote:

“Since a central bank is a bank--a financial intermediary--I have trouble with the notion that an increase in the demand for money forces a central bank to dissave. I rather think it forces a central bank to intermediate. It must borrow (and since it doesn't consume, it doesn't dissave,) but rather it must also lend.”

Quite right – let’s get the basic economic definition of saving right - along with the accounting for it - to avoid exploring a paradigm for central bank behaviour that is anchored in the wrong terminology. It's a bit distracting.

[Jon]

Bill W. writes:

I see no reason why the U.S. cannot create the amount of money demanded and have net capital outflow to China and a trade surplus. Of course, I think it would require that Americans want to save.

No, because the PBoC wants USD assets, and the Chinese Government does not want Americans to own Chinese Assets. Indeed, they have taken several definite steps to prevent what you describe:

- Foreign possession of yuan is illegal
- Dual-listed share structures with the majority share class open to citizens only
- Foreign companies cannot operate in China except through joint ventures, where the foreign company must take a subordinate, minority position, or a majority position only if the foreign operation is contained within one of the special economic zones and produces strictly for export.

What's seriously fishy about the "American's just don't want to save enough" argument is that the trade imbalance is almost exclusively a bilateral problem with China. Somehow the rest of the world seems to be in basic equilibrium: developing countries should be net importers of capital not net exports of capital (China).

Let me stress though that the exchange-rate itself is a complete red-herring. A nominal exchange-rate peg is neutral on to itself. That does not mean there are not other issues.

[Scott Sumner]

Nick, Yes, it's possible that foreign central bank purchases of US debt have a slight deflationary effect (via reducing the US nominal interest rate, and hence increasing the real demand for cash), but the evidence from 2005-07 suggests this effect is tiny, and does not force the Fed to increase the base by any substantial amount. Of course it's a much different story when Mexican drug gangs hoard lots of cash---that does force a major increase in the monetary base.

[JKH]

Nick,

An interesting post - I have a number of comments, specific and general.

“The Fed is the loan placement officer for the world's central banks. The US government is the Fed's borrower of last resort. The forced loans can be called in at any time the lender wishes.”

The US government really is the borrower of first resort, in the sense in which you are using the term “borrower” here. The first type of asset that the Fed buys for its balance sheet is government debt. Only in “credit easing” does it turn to other types of debt, as it did in the credit crisis. Of course, you’re aware that the Fed buys government debt only from the secondary market. So you might qualify the use of the term “borrower” in the sense that it’s not a direct lender/borrower relationship at the operational level. The Fed is acting functionally in a way that is loosely analogous to the role played by ultimate buyers of mortgage securitizations, in the sense that the sequence of the origination/distribution chain for the debt originates with the government and ends with the Fed.

You then talk about global imbalances and aggregate demand and the paradox of thrift.

“Pre-Keynesian monetary economists had the concept of "forced savings". When the central bank decides to print money, it forces people to save. I'm going to turn that concept on its head. When people decide to save money, it forces the issuer of money to dissave.”

I don’t know about pre-Keynesian economics, but it is at this point that you go off definitional rails by wrapping the entire discussion in a language that abuses the economic (and accounting) definition of saving.

The central bank does not force people to save by printing money. And it does not dissave when it prints money. Since your post revolves around this theme (being titled on it), the proper identification of what’s really going on is relevant.

I’m going to assume you are stylizing your point in terms of the central bank’s issuance of currency. The central bank does not force people to save by printing money in the form of currency. And it does not dissave when it prints money.

The correct economic definition of saving is basically income not spent on consumer goods. If you don’t agree with that, then you’re making up your own world of definitions that nobody should be expected to take seriously. You may as well define black as white.

So the economic definition of saving is based on income, and the idea of saving is the idea of saving from income.

Your post is not about saving or dissaving from income. It is about financial intermediation. Financial intermediaries have assets and liabilities. They lend and acquire various other types of financial assets. They take deposits and issue various other types of liabilities.

Suppose we group the activity of financial intermediaries into broad categories of “lending” and “borrowing”, in order to simply and generalize terminology. And let’s consider asset customers as “borrowers” and liability customers as “lenders” for the same reason.

And let’s classify the central bank as a type of financial intermediary. Then, the liability activity of a central bank includes issuing (or “borrowing”) reserves and currency, and the activity of its liability customers includes “lending” reserves and currency. This just standardizes all of the terminology in the discussion. It also seems to be reasonably consistent with how you develop your discussion.

In that context:

Lending (by anybody) is not saving.

Borrowing (by anybody; and including issuing money) is not dissaving.

Lending and borrowing are balance sheet activities that stand on their own, quite separate from the saving or dissaving activity that is connected to income.

The correct economic (and accounting) link is that saving from income generates an increase in net worth or equity on the balance sheet. It is then up to the saver to decide the form in which this net worth is to be deployed. The default deployment of course typically is money – your subject. But that’s just the default asset for the deployment of saving. It’s not the saving itself, which is balance sheet equity. And the deployment can change when the saver starts to manage his asset mix away from the passive default asset of money. The saver can also repay liabilities – as per the phenomenon of “deleveraging” that is all the rage in this recession, which has also been referred to as a balance sheet recession (Koo). So this is all the language and the logic of the economics around your topic, as well as the accounting. I regret being so pedantic, particularly since I am not an economist, but it seems warranted in this case, since this is the way it works.

“In effect, we have forced the Bank of Canada to borrow from us. We have forced the Bank of Canada to take an interest free loan.”

I agree with this, insofar as it goes, and insofar as we assume it is currency you are talking about, and insofar as you did not use the language of saving or dissaving in this sentence. There is no question that the public determines the demand and the actual transactions in currency. You monetarist types tend to confuse operational causality for the different types of central bank liabilities by insisting on characterizing the behaviour of the monetary base as a whole. But here, if you are thinking clearly in terms of currency issued as the specific component, I agree.

In fact, a very interesting aspect of your post is that you have departed from the usual ambiguity of monetarists in dealing with the aspect of currency in contrast with bank reserves (again, if we can assume that it is currency you are talking about here.)

But I said, “insofar as it goes”.

That’s because currency issued by a CB is in fact paid for with bank reserves. The immediate operational effect then is that while “borrowing” has increased through currency, it has declined by an equal amount through reserves. Moreover, this still has nothing directly to do with saving or dissaving. It is a balance sheet operation, quite separate from income related activity.

So:

“Nevertheless, by demanding more money, and forcing the Bank of Canada to issue more, the immediate effect is to force the Bank of Canada to dissave.”

No. The immediate effect in terms of net borrowing is neutral, because payment is made with reserves. And even if it weren’t neutral, there is no issue of saving or dissaving in any event.

The CB may subsequently decide to acquire assets in order replace the reserves lost by issuing currency, which is along the lines of your general theme. But even that depends on the configuration of the balance sheet, noted further below.

“The Bank of Canada, in other words, is forced to act as a loan placement officer, with a quota of loans it must place. And if it fails to place the loans with willing borrowers, it must borrow and dissave itself.”

I have no problem with the loan placement officer analogy. You asked the question at the end of your post about this terminology. It’s really SOMA in the case of the Fed (the system open market account). Boiled down, it’s really a money market trader at the Fed that ends up making these kinds of residual balance sheet adjustments, including the purchase and sale of securities from SOMA.

The “quota” you refer to is a function of the Fed’s (or the Bank of Canada’s) balance sheet objective.

In normal times, it boils down to the objective for the excess reserve setting.

These days, the Fed’s reserve objective has been overtaken by its asset portfolio objectives in connection with credit easing and quantitative easing. In the case of reserves, whatever is required is required, based on its targeted asset activity, and the required amount of offsetting liabilities that natural currency demand and growth are unable to generate.

A corollary to this is that the Fed in the context of its current balance sheet strategy can absorb the secular and cyclical demand for net currency issuance through liability management alone – issuing currency and debiting reserves, without additional asset activity, due to the extreme level of excess reserves - so the currency adjustment factor is a minor detail in the context of the Fed’s balance sheet as it has come through the financial crisis. Operationally, your loan placement officer’s quota is not being driven at the margin by currency demand in these extraordinary times. It is being driven by the Fed’s asset strategy. This is actually illustrated quite nicely in the case in your example of $ 1 billion in currency demand, since excess reserves now exceed $ 1 trillion. But your point does hold operationally in normal times.

“When central banks in the rest of the world decide to hold more reserves, and so save more, they force someone else to dissave.”

No. Central bank “lending” (foreign exchange reserves in this case) is not saving. Saving is reflected in the current account surplus that typically gives rise to the related central bank currency intervention and reserve accumulation. The current account saving from income is reflected in the (balance sheet) net wealth position of the non government sector (i.e. “non” relative to that central bank and its government). That’s where the saving takes place. The CB merely swaps dollars for domestic currency so that the domestic savers can deploy their saving in local currency assets. But it’s not the central bank that’s doing the saving. The CB is a financial intermediary, not a saver. The current account surplus is the saving; the dissaving is the rest of the world’s current account deficit with that country. The central bank is not forcing someone else to dissave; it is a financial intermediary for the current account saving of its own country and the current account dissaving of the rest of the world. Lending is not saving.

“[I'm not sure this is 100% right, but I'm posting it anyway.]”

I expect what I’ve written won’t appeal to you Nick, because I believe you believe you can theorize about CB economics or any economics without paying attention to basic accounting for GDP, income, saving, balance sheets, and the flow of funds. So do a lot of other economists. We’ve been through this before.

This issue of attitude to accounting is relevant to the economics profession more generally in my view. It is an attitude that is galvanized in how few economists understand the operation and accounting for central banks and the monetary system more broadly. Coherent accounting for economic outcomes is a required constraint for the legitimacy of economists’ prognostications for those outcomes. The necessity of such a facility is something that might have figured more prominently in the string of mea culpa confessions the economics profession offered up for its role in the financial crisis - this is basic knowledge that is essential to explaining the role of financial sector. But it wasn’t offered up. My question is when are economists going to become more schooled in general on the basic accounting that is required to legitimize economic arguments about the possible future states for the world?

Once more for the road – lending is not saving; borrowing is not dissaving. It is financial intermediation and the flow of funds that you have written about here – not saving and dissaving. It is a point beyond semantics – it is a point of consistent logic, exposition, and communication – it is a point of clarity. It seems to me that you various monetarists especially should have a vested interest in the development of clarity via accounting/economics congruence wherever possible. It is you after all who are dealing most intensively with the field of monetary debits and credits, even if you don’t think about it in quite that way.

P.S. - this is not a call to “Post Keynesianism” or “Chartalism” or “MMT” at all. I don’t represent those groups. And in my view, it is completely coincidental that they happened to have anchored much of their financial analysis in the notion of operational and accounting clarity. They just happen to be right about a universal truth that connects economics with accounting and have run with it from there, along their chosen path. There’s no reason at all why you monetarists can’t pursue the congruence I’m talking about above, while at the same time disassociating yourselves from those particular groups to whatever extent you feel is necessary for your own needs.

[The Money Demand Blog (123)]

Nick, buying stocks is too risky at this time. Stocks have a very long duration, and the risk that stocks are overpriced is too high (but maybe deep out of the money puts are too expensive if markets think that there is a high risk of monetary policy mistakes that could cause 40% drop in the stock market, in this case central banks should write such puts and earn monetary catastrophe insurance premium, but on Friday VIX was below 20 and Bernanke accomplished the needed result with his speech without any actual intervention).

Central banks should purchase medium risk private sector financial assets that have low duration. A diversified portfolio of BBB corporate bonds with 3 year maturity is a perfect choice for such intervention, the price of such bonds should be higher if monetary equilibrium is restored.

Alternatively, credit easing could be restarted, central banks could lend money against eligible private sector collateral.

[Min]

"Pre-Keynesian monetary economists had the concept of "forced savings". When the central bank decides to print money, it forces people to save. I'm going to turn that concept on its head. When people decide to save money, it forces the issuer of money to dissave."

Nick, I do not mean to sound critical, but I think that your looseness of language is an asset. For most people it is not good, but I suspect that it is part of your creativity. :)

I appreciate your quoted remark, but it should be clear that you are speaking metaphorically. Strictly speaking, saving and dissaving do not make sense for the issuer of currency.

[The Money Demand Blog (123)]

China forces the Fed to engage in more risky/less profitable forms of financial intermediation. If the Fed buys long term government bonds, then China is successfully exporting socialism (just like Norway is doing it in Africa with all that foreign aid) - the cost of government spending is artificially reduced in the US.

[Phil Koop]

An aside: JKH has put his finger on exactly the issue I was not clever enough to identify during our discussion of repo and rehypothecation. When one considers the repo market as a black box, ignoring its internal mechanism, then of course it is a bank: it accepts short-term deposits on one side, and maturity-transforms these into long-term loans via securitized assets on the other. These loans expand money in the ordinary way: as they are spent, they are deposited in banks. Some of these "banks" are the repo market, where the deposits may fund more securitizations. Lending is not saving, and borrowing is not dissaving; yes, just so. And by the same token, creating credit does not create credit-money; it is the spending of the credit that does so. This is why rehypothecation is not like credit money creation; it is merely a balance sheet activity, in JKH's words. It may make the black box more fragile, because it multiplies the number of moving parts inside it, but it does not finance any real economic activity. It merely transfers money from one pocket to another.

[wjd123]

Winslow R. writes:

"I believe Krugman only recommends tariffs because our government refuses to provide enough tsy bonds to provide domestic full employment. Krugman believes we should increase government deficit spending, even more strongly than imposing tariffs. QEII will only work if there are sufficient government bonds for the Fed to purchase."

Winslow R.

I don't see this. Krugman has written many times that the political will of Americans for more stimulus isn't there. Politically is a dead issue.

What isn't a dead issue is the American will to make China fly right with the rest of the world. Congress after seven years of trying to work with China has just declared it a currency manipulator. This didn't happen because Congress finally saw the light but because the American people are sick of their stalling while American jobs and factories are moved to China.

Here there is a political opening for tariffs on China because the American public opinion isn't against it. Of course multinational corporations are against it, but they aren't to popular with the American people today.

This is a perfect time to stand up to China. Krugman has cleared the way with his explanations of how a trade war would shake out: not well for China.

Of course Krugman doesn't wish that events will esculate to a trade war. He hopes that China will appricate it's currecy by 25% and the tariffs taken of and trade resumed.

It's also true that most economists are running away for the idea because they are scared of unintended consequences.

However in don't believe that American public opinion uses an economic criteria in solving economic questions. I believe that they filter questions about jobs losses through nominal criteria. There is also a common sense filter in thier thinking. They subjects economic questions to the duck test: If it looks, walks and quacks like a duck, it's a duck. This isn't much different than Krugman's the proof of the pudding is in the tasting test.

I believe Krugman's move toward tariffs is a move in the right direction when it comes to getting results. It's is in tune with American society as opposed to American government in so far as our government has been out of tune with our society for so long. He has worked it out economically which puts more pressure on our government to act.

In my opinion Krugman's stance is a call for other economist to join him. Most of them have run away. That doesn't help our unemployed.

[jsalvati]

Nick: maybe I am just being dense, but it seems to me that unlike for other goods, people almost always have some way of increasing their money holdings by selling assets or reducing expenditure, so their demand for money will equal their quantity of money held. When we see someone increase their money holdings for some exogenous reason by cutting down on their expenditures, we say there's been an increase in the demand for money. When the producer who had their sales cut because of that action cuts down on their production enough so that they hold less money why don't we call that a decrease in the demand for money? It seems like there is something missing.

[Nick Rowe]

There's a lot of meat in these comments. I am still trying to get my head around it all.

I'm going to start with jsalvati's comment. Both because I can answer it, and because I think it's at the root of my answer to the other comments.

"maybe I am just being dense, but it seems to me that unlike for other goods, people almost always have some way of increasing their money holdings by selling assets or reducing expenditure, so their demand for money will equal their quantity of money held."

That's true for the individual, but it's not true for all individuals together. Each individual can always get rid of money by selling more or
buying less, but the population as a whole cannot. They just play "pass the hot potato". And because an individual can always get rid of excess money, there is never any difficulty in a central bank increasing the stock of money held, even if there is no increase in the demand to hold money. An individual will accept the new money from the central bank, even if he does not wish to hold it. He accepts it because he knows he can just pass it on to someone else. But in aggregate they can't get rid of it. The central bank can *force* people in aggregate to hold more money than they wish to, even if it can't force any individual to accept more money. Money is not like other goods; it is weird.

The old 1920's 1930's British monetary economists (Robertson, Hawtrey, etc., and I include Hayek in there too) spoke of this as "forced savings". I'm not sure they were ever coherent in this doctrine, from a National Income Accounting definition of "Saving", but they were onto something. (Anyone who has ever tried to puzzle through Dennis Robertson's horrible taxonomy of forced/automatic/induced saving/stinting/lacking whatever will know how I feel!).

I'm exploring the flip-side of that forced lending to the central bank. When the demand for money increases, and the central bank needs to accommodate that increased demand for money to prevent nasty macro consequences if it doesn't, it is forced to borrow. And it is forced to borrow in a way that I would be if I were forced to accept every offer to lend me money that comes through the mail (provided only that the rate of interest offered is the market rate).

Bill/JKH and others: yes, money is issued by central banks, and they are financial intermediaries. But money doesn't have to be issued by financial intermediaries. There is *absolutely nothing* in accounting conventions that can force us to assert that it is logically necessary for an increase in the supply of money be matched by an increase in the assets of the money issuer. For example, the money issuer could be a counterfeiter, who spends the proceeds on booze. He doesn't have to *lend* his newly-printed money. Similarly, it is not logically impossible for the Bank of Canada to give money to a charity, or spend newly-printed money on a conference. Or we could simply consolidate the government and central bank.

There is nothing *logically* wrong in saying that the issuer of new money must either spend it or lend it. And if the issuer is *forced* to issue new money, then the issuer is forced to either spend or lend. (Hence my loan placement officer metaphor, who has a quota he must meet, and must borrow himself, if he fails to meet his quota.) And if the issuer is successful in lending, then all that happens, is that he passes on the obligation to the next person. who likewise is either forced to spend or lend. Ultimately, someone is forced to spend.

I'll come back to this later.

[RebelEconomist]

I think you are overcomplicating this Nick. The Chinese are merely ensuring that current supply and demand for renminbi/dollar matches at their set exchange rate. Since that exchange rate leaves a private sector excess demand to sell dollars for renminbi, the Chinese authorities fill that demand, and, since dollar currency does not have much utility to them, they buy something with the dollars. The involvement of currency is transient. The Chinese might prefer to buy US armaments (a consumption item?) or oil companies, but given what they are allowed by the US to buy, they buy treasuries. None of this need force the Fed to do anything, although the resulting lower long-term interest rates could be expected to crowd in some private sector borrowing. Moreover, if the Chinese had only bought real assets and index-linked treasuries, they could just laugh in the face of QE, just shifting the peg a little to adjust for any inflation that the Fed generates in the US (unless of course the US decided to restrict China's ability to buy US assets altogether, as I suggested on my blog a couple of years ago: http://reservedplace.blogspot.com/2008/10/just-say-no.html ).

There is a lot of confused misinformation being written on this subject by people with fully-baked opinions and half-baked ideas based on textbook macroeconomics instead of simple accounting. As usual, JKH gets it right I think.

[Bill Woolsey]

Nick:

OK. Someone in the rest of the world must spend in response to the Chinese government's saving because otherwise, money expenditures in the rest of the world would drop to an undesirable level. I think your "dissave" lingo throws me a bit because you are ignoring investment (and probably said that in your post.) Generally, we don't have to dissave, but can save by purchasing and producing capital goods, even if the Chinese accumulate U.S. government bonds. Naturally, we might want to save less than otherwise in response to their policy.

Jon:

My point was that a foreign central bank can shift between domestic assets (like loans or bonds) and foreign assets (say, U.S. Treasuries) without forcing the people in the rest of the world to dissave. I wasn't really focusing on real investment (as above,) but rather a pathway by which the rest of the world purchases other finanical assets from the country whose central bank is shifting away from those financial assets to Treasuries.

On second thought, however, such a pathway would also offset any tendency for the shift in the balance sheet of the central bank to raise its exchange rate. And I can see how restrictions on foreign investment would help manipulate the exchange rate by manipulating capital outflows--block offsetting inflows to better control the net outflow.

Mercantilist policies are bad for the world, but especially bad for the country following them. I don't think an argument--but they are forcing us to borrow, really makes much sense. If there is too much lending going on, then nominal interest rates should be negative. Wouldn't this provide a clear signal of the costs of their policy?

In reality, I think the structural deficit in the U.S. shows that the "problem" isn't too much lending going on, but rather too much borrowing and prospective borrowing by the U.S. Or more exactly, there isn't enough saving by the U.S.

[The Money Demand Blog (123)]

Nick,
without any loss of generality, counterfeiter, who spends the proceeds on booze is a financial intermediary with 100% profit margin who happens to spend his profits on booze.
The real problem is government/market failure in financial intermediation.

[Too Much Fed]

Nick's post said: "Too much Fed: "private savings = gov't deficit"

That's true in a closed economy. In an open economy we need to add in or subtract borrowing/lending from/to abroad.

But that's an accounting identity. Bill Mitchell says that. So does every other economist. And an accounting identity alone says nothing about causation. I'm talking about causation. (So is Bill, but what he says about causation won't generally be the same as what I say)."

I can't remember if he says explicitly, but I am pretty sure he means that the private savings drive(s) the gov't deficit.

[Too Much Fed]

Bill Woolsey said: "In reality, I think the structural deficit in the U.S. shows that the "problem" isn't too much lending going on, but rather too much borrowing and prospective borrowing by the U.S. Or more exactly, there isn't enough saving by the U.S."

Can this show up as a lower overall effective capital requirement at the "banks"?

[Jon]

Bill writes:

Mercantilist policies are bad for the world, but especially bad for the country following them.

Yes. I think China's people are harmed by the policy. Scott has posted two ideas of late on his blog:
1) Revaluing the currency will do nothing (agree)
2) The Chinese are poor and shouldn't be asked to hurt themselves with a tight monetary policy (agree in part, dissent in part).

Its not obvious to me the effect of a tighter monetary policy, on the balance, with an immediate wealth revaluation of every Chinese person--that's a fuller perspective on the impact of the exchange-rate policy choice.

I don't think an argument--but they are forcing us to borrow, really makes much sense. If there is too much lending going on, then nominal interest rates should be negative. Wouldn't this provide a clear signal of the costs of their policy?

I don't know that they are forcing us to borrow. What at issue is who owns the debt that exists by whatever means. The Chinese policy is making the debt external whereas without their policy, I'd expect production to higher domestically and there to be more domestically held debt.

In so much as their demand for debt lowers rates, I agree they create a bit more supply. I don't see when this would make interest-rates negative. It just makes them less than they would be ceteris paribus.

In reality, I think the structural deficit in the U.S. shows that the "problem" isn't too much lending going on, but rather too much borrowing and prospective borrowing by the U.S. Or more exactly, there isn't enough saving by the U.S.

No amount of increased saving by the U.S. will eliminate the deficit. It will only make it more costly for the Chinese to maintain.

[Nick Rowe]

TMDB: "Nick,
without any loss of generality, counterfeiter, who spends the proceeds on booze is a financial intermediary with 100% profit margin who happens to spend his profits on booze.
The real problem is government/market failure in financial intermediation."

If the increase in the demand for money is permanent, then the counterfeiter has accepted a permanent interest-free loan. Sure, he can spend it all on booze immediately. But if the loan is temporary, he must lend or invest the proceeds, and only spend his interest earnings on booze. If the loan can be recalled any time, he can only lend or invest in very liquid assets (ugh! Sorry, that was not intended).

Bill: yes, I should have been more explicit about investment. Either hold it constant, or else define my "saving" as "saving minus investment".

[The Money Demand Blog (123)]

If demand for money is temporary, the central bank cannot take on duration risk. This is why it is a bad idea to buy stocks. Short term lending against illiquid long term assets is OK if appropriate haircut is applied.

[Phil Koop]

"Ultimately, someone is forced to spend."

That conclusion does not follow from your premises. For instance: the CB issues money and forces it onto the banking system. The banking system has no use for it and deposits it as reserves at the CB. The CB is thus both ultimate borrower and ultimate lender, and nothing is spent. Is this scenario not the case made by Sumner and others for negative interest rates? At present, when the CB is "forced" to borrow, it, too, seems to do this in a way that does not necessarily result in increased spending: it uses its new money to buy close money substitutes.

[Nick Rowe]

TMDB: "If demand for money is temporary, the central bank cannot take on duration risk."

But I think that's part of the problem. Unless there are real (as opposed to financial) investments that are very liquid, so the physical investment can be reversed quickly, someone, somewhere, has to take on duration risk. And probably won't want to.

Phil. Total (global) spending on newly-produced goods and services must equal total income from the sale of newly-produced goods and services. (National Income Accounting Identity). If someone *wishes* to spend less than his income, he needs to find a borrower who is *willing* to spend more than his income. And if he can't find someone willing to borrow, he can't spend less than his income. Except in the case of a monetary exchange economy, where I can spend less than my income just by hoarding money. I don't need to find a willing borrower. I then put the onus on the central bank to find a willing borrower. I force someone to spend more than his income, and it's up to the central bank, and ultimately the Fed, to find that someone.

[vjk]

Phil:

the CB issues money and forces it onto the banking system. The banking system has no use for it and deposits it as reserves at the CB.

How does the CB force money on the banking system, exactly ? What is the coercion mechanism of forcing cash on the resisting banks ?

[The Money Demand Blog (123)]

Nick, I see no problem at all. The Fed can increase the liquidity of real investments without taking on duration risk. Just lend 50% of the market value of investment for one year and you have increased the liquidity of investment without any significant transfer of duration risk. I have no confidence that the Fed has a better model of duration risk than markets have. Bernanke should not say "we have to take on duration risk", he should supply liquidity until monetary disequilibrium disappears. Household balance sheets have too little liquidity at this time, driving down the price of 10 year treasury duration risk with QE2 does very little to solve the problem.

[vimothy]

Wait a sec--are we confusing what the CB does here. When the CB prints money, it doesn't directly increase or decrease saving, it merely changes the composition and term structure of the existing portfolio of risk free savings. Right? Nick are you thinking of government spending funded by CB money printing?

[RSJ]

Cash is a liability of the central bank, and an asset to the holder. When the CB issues more, it does so by purchasing an interest bearing bond. The interest from that bond is seignorage income to the CB, so the currency users are the ones lending at zero rates to the currency issuer, by parting with interest bearing assets in exchange for a non-interest bearing asset.

Supposedly this is offset by the transactional utility of the cash. That's fine, both households and banks have a certain need for cash, and are willing to hold small amounts of it, and even to part with interest in order to have it. When the CB creates more cash than is necessary to meet this transactional or precautionary demand, it is receiving seignorage income on a larger stock of cash than is economically justified. As households are never forced to hold excess cash but can always deposit it with the banking system, any excess cash sits idle within the banking system as excess reserves.

This would reduce bank income and increase CB seignorage income, if the QE was very large. The payment of interest on reserves undoes this shift.

"And if the issuer is successful in lending, then all that happens, is that he passes on the obligation to the next person. who likewise is either forced to spend or lend. Ultimately, someone is forced to spend."

Someone is forced to spend, but not necessarily on goods, and certainly not on final output. Suppose a bank has, as assets, $1 in loans, $1 in government bonds and $1 in cash. The liabilities of this bank are $1 of stock, and $2 of deposits. Now the central bank buys the government bond. The bank is stuck with excess reserves -- it now has $2 of cash when it only needs $1. It would have preferred to park the funds in an interest bearing government bond, but as the cash was not drained, the banking system is forced to hold onto it as excess reserves.

What is the flow of spending that this operation is supposed to trigger?

All that happened was that you pushed the banking system into a reserve excess position. Perhaps deposit interest will fall a bit and/or borrowers will be charged a bit more as the bank now has only $1 of interest earning assets against $3 of interest paying liabilities. But no one is going to "spend" one dollar more because the CB created an additional dollar and destroyed $1 of bonds.

I think the actual institutions matter: the CB is not there to supply "the economy" with cash, but to supply its member banks with cash. The financial system is an intermediary to the household and non-financial business sector. The latter sectors hold exactly the composition of assets that they want. They can't choose the total net-worth of their portfolio, but they can choose how much of that portfolio is cash, how much is bonds, deposits, stock, etc. And the financial system has to adjust to supply those changing quantities of assets on its liability side against the assets that are available. If households want to, in aggregate, increase their holdings of deposits and decrease their holdings of bonds, then they can do this, with the financial sector buying more bonds from the household sector and selling more deposits to them. But if households want to hold fewer short term liabilities and more long term liabilities than are available, then the financial sector is put in a difficult position, being forced to lend short and borrow long. In order to prevent this, the consolidated government/CB, when it issues a certain bond/cash mix of liabilities, does this so that banks are not stuck on the asset side of their balance sheets with non-interest bearing cash excess to their own transactional needs.

Basically the financial system needs safe, liquid, long term assets that it can hold against short term liabilities. It does not want to hold cash against interest bearing liabilities.

And the CB, by removing those assets and replacing them with zero duration cash, causes problems for the financial sector. Even if banks pay zero interest on checkable deposits and are willing to provide these at a loss in the hopes of selling other products, still the bulk of deposits is in the form of money market or savings accounts, and banks must compete with commercial paper, other mutual funds, or non-bank money market accounts. Given that constraint, there is little wiggle room -- the composition of cash and debt issued by the consolidated government is determined in such a way as to prevent banks from having excess non-interest bearing assets.

But to see this, you need a financial sector as well as a government/CB sector in your model. You'll get a completely different model if you don't include a financial sector. Simple things like deposit insurance, bank regulation, the existence of a discount window, paying interest on reserves, the ability to buy an asset on credit, or the ability to re-sell an asset make a big difference in the model.

[Phil Koop]

Nick, I would have replied in more detail, but RSJ has beaten me to it.

Nobody is in disagreement when we are talking about real spending in the real economy, lending by forgoing spending, or borrowing to fund spending. But the latter activity accounts for only a trivial fraction of lending and borrowing.

Your description of how hoarding by me forces the CB to force ultimate borrowing by someone else is fine as a prescriptive model of how you would like the CB to operate, but it does not describe what the CB actually does. It would be necessary to unify fiscal and monetary authority to achieve your counter-factual.

[vjk]

RSJ:

You and Phil's apparently consider the commercial bank an utterly passive and unwilling agent in the ongoing QE game. E.g.

And the CB, by removing those assets and replacing them with zero duration cash, causes problems for the financial sector

or

the CB issues money and forces it onto the banking system

Leaving aside household cash preferences for simplicity, what is the coercive mechanism to force banks hold the cash they do not want to hold ? Do Bernanke's special forces descend from their helicopters on the poor banks, snatch banks' T-bonds and leave behind equivalent bags of cash ?

Who is the enforcer ? My answer is : no one. What is your answer ?

[Phil Koop]

vjk, this notion of coercion is the premise of Nick's piece. It is uncontroversial, as you can gauge from the run of the comments.

[vjk]

Phil:

If the notion of coercing the bank into selling its 10 year T-bond holding is uncontroversial, it should be easy for you to describe the procedure of extracting such a bond from an unwilling bank.

What is the procedure of making the unwilling bank comply with the Fed wish to extract the bond ?

[RSJ]

VJK,

The bank doesn't need to sell the bond to the CB, it sells to "the market" just as the CB buys from the market. But the effect of the sale is that the financial sector balance sheet changes. It loses an interest bearing asset and gains cash.

This could be because they incur a short position in an interest bearing asset that is matched by a long position in cash, or because they reduce a long position in an interest bearing asset and replace that with cash -- it doesn't matter; the financial sector balance sheet is hit, rather than households or firms.

Assuming that the CB sale does not change household or firm preferences for the types of assets that they wish to hold -- e.g., that the act of selling a treasury does not cause households to want to hold more deposits, then the financial sector liability side will remain unchanged. This means that effectively the financial sector has relinquished an interest bearing asset and replaced it with cash, regardless who the CB bought the bond from.

The specific asset(s) relinquished by the financial sector need not be the ones that the CB purchased -- the specific assets trade at indifference prices and diffuse throughout the entire system.

So, for example, the CB buys a treasury form PIMCO, who rolls the proceeds over to the purchase of an agency from a bank. Or the CB buys a treasury from a pension fund, that pays out the proceeds to a household that repays a bank loan. The possibilities are endless, and each transaction triggers a chain of many other transactions, but the net effect is that the financial sector loses an interest bearing asset and gains non-interest bearing cash.

The net effect is *not* that the CB purchases a bond and the proceeds are spent on goods. The proceeds are spent on some other asset, and at the end of the day, that is all that happens -- an asset swap between the CB and the financial sector, with no goods purchases involved.

[RebelEconomist]

vjk,

You are right; when dealing in a market economy, the central bank cannot force the private sector to do anything. It has to find terms which INDUCE the private sector to do the transaction desired by the central bank. Arbitrage ensures that those terms propagate from the institution that the central bank actually deals with, so there is no sharp divide between the financial sector and the rest. Those terms will depend on the expected return on base money as well as its (expected) utility to meet reserve requirements and for facilitating transactions. This expected return will depend on interest paid on reserves as well as the likelihood that an inflexible buyer of base money - ie the central bank - will come to market before long.

Again, academic economists tend to tie themselves in knots trying to analyse events in terms of their "standard toolkit" instead of investigating how the economy actually works in practical detail.

[Nick Rowe]

Suppose Fred decides he wants to save more.

Barter economy: the onus is on Fred to induce someone to borrow from Fred, by offering favourable enough terms that someone wants to borrow from Fred. If Fred can't find someone willing to borrow from him, he can't save (unless of course he directly invests himself in real assets).

Monetary exchange economy. Fred just stops spending, and starts to hoard cash. He can save without needing to induce someone to borrow from him. The central bank has to increase the supply of cash to prevent a recession. The onus is on the central bank to induce someone to borrow from it. The central bank has to find someone willing to borrow from it (unless of course the central bank itself, or it's owner the government, directly invests in real assets).

Metaphor: in a monetary exchange economy, the central bank is Fred's loan placement officer, that is forced to find a willing borrower for Fred to lend to.

[Nick Rowe]

Take an extreme case: suppose nobody wanted to borrow from Fred? In a barter economy, Fred would be unable to lend. In a monetary exchange economy, Fred would always be able to lend. The central bank is forced to borrow from him, and then pass that loan on to someone else.

[vjk]

RebelEconmist:

when dealing in a market economy, the central bank cannot force the private sector to do anything.

That was my simple point, thanks.

I was rather interested in the non-bank sector's, mainly households' I'd imagine, motivation to swap 10 year T-bonds for cash. The non-bank financial sector(mutual/pension funds, brokerages, etc) can be considered in the first approximation as a households' proxy for such dealings.

[Nick Rowe]

Rebel: "vjk, You are right; when dealing in a market economy, the central bank cannot force the private sector to do anything. It has to find terms which INDUCE the private sector to do the transaction desired by the central bank."

Not quite right. The central bank can force people to hold more money, even if nobody wants to. Money is different. Sure, the central bank has to *induce* someone to accept cash in exchange for some asset. But just because someone willingly accepts money doesn't mean they want to *hold* more. Each individual accepts more money even if he doesn't want to hold more money because he can always pass it on to another individual. Each individual can get rid of it, but in aggregate they can't.

[vjk]

RSJ:

at the end of the day, that is all that happens -- an asset swap between the CB and the financial sector, with no goods purchases involved

Right, the word "forcing" was what I found amusing, as well as the implied bank sector pain and suffering of being unable to extract even more rent in addition to what the Feds are already paying for "forcing" the bank hold excess cash.

[RebelEconomist]

Spot on, Nick. It is the Fed which is forced, by virtue of its function (although of course the authorities always have powers to force the market if they really want, such as by changing reserve requirements).

Maybe you can build from this to the effect of Chinese intervention, although I cannot see how, because that intervention is money-neutral (I should have said "money" not "currency" in my comment above). The Chinese are basically swapping goods for (dollar denominated) IOUs.

[Nick Rowe]

I just willingly sold my labour to Carleton University in exchange for money. But I had absolutely no desire to hold a bigger stock of money.

[vjk]

Nick:

The central bank has to increase the supply of cash to prevent a recession

I am not sure if supply of freshly printed cash has anything to do with preventing a recession -- witness $1 trillion of cash in "excess" that is already there.

Perhaps you mean direct give-away's to households and firms rather than mindless swaps of government paper of different tenors ? The Feds unfortunately (or fortunately) are prevented from engaging into that sort of activity by law, the helicopter Ben metaphor notwithstanding.

[Nick Rowe]

vjk: I don't know. This whole post started out with a vision, and I'm trying to articulate that vision, explore its implications, and get my head straight on it, by reading and arguing with you guys.

[RebelEconomist]

Nick (at 08.18),

Now you are stretching it. I don't think it matters whether the central bank counterparty does want to hold whatever the central bank is offering. They may value it because they expect they can trade it for something else that they do want to hold. Besides its transactional utility as the medium of exchange, I don't see that money is so different (okay, okay, you can force people to take money using legal tender laws, but I doubt whether that is ever a binding constraint).

[The Money Demand Blog (123)]

vjk,
there is no need for helicopter drops. Just lend more money to the private sector...

[RebelEconomist]

vjk at 08.29,

Now here I am with Nick. As you know, demand for base money derives both from its role as the medium of exchange and as an unbeatably safe, liquid store of value. Briefly, in a panic, the store of value demand increases greatly, and if the central bank does not accommodate it, the shortfall of supply to serve the medium of exchange purpose gums up the economy. So I think it was reasonable for the Fed to allow a dramatic expansion of base money in 2008. More recently, however, the purpose of QE has become to drive down longer term interest rates by an asset swap of interest-bearing reserves for long term bonds. I would question whether that is appropriate, and, as you say, it does not seem to have been effective either.

[vjk]

RebelEconomist:

Briefly, in a panic, the store of value demand increases greatly,

That's a good point, I think, cash as sort of gold for those who cannot afford real stuff. Irrational, perhaps. Would you speculate that the panicky mood still persists thus assuring "success" of the apparently forthcoming QE2 ?

It would be interesting to know who(commercial banks, actual households, their financial proxies, etc) made specific choices with respect to government paper maturity distribution in their portfolios. That would allow one to try and quantify, as it were, behavioral aspects of cash biased portfolio composition in our turbulent times. I assume such information is not readily available if at all.

and if the central bank does not accommodate it, the shortfall of supply to serve the medium of exchange purpose gums up the economy.

That, I am not so sure of unless one assumes a totally broken interbank market which it was for a while. I believe it recovered at least partially *before* QE1. Has the ib market fully recovered by now (judging by the bank I am associated with it has, but one sample is statistically meaningless) ?

the purpose of QE has become to drive down longer term interest rates
Well, I am sceptical about the Fed having enough paper power to accomplish that regardless of desirability of lowering long-term rates in principle.

[vimothy]

Cash is a subset of reserves. Maybe the demand for notes and coin has risen during this recession but it can hardly be a driving factor. Perhaps its better to think of the CB as supplying liquidity rather than "printing money" as its too easy to mistake printing money with something different--supplying saving, for instance. Wouldn't a more realistic story be something like the demand for liquid assets (demand for a particular type of savings) has risen alongside a demand for increased net worth (demand for increased saving)? The former the CB can help with directly but the latter requires somebody else as well at the very least.

[vjk]

vimothy:

CB as supplying liquidity
I do not know what the above means because the CB supplies "cash" or FRNs that can be exchanged for government obligations, not some nameless "liquidity". Different people mean different things by "liquidity" thereby making the word rather meaningless without providing context.

I do know what "cash", "coins", "currency, i.e. Federal Reserve Notes" are.

Clearly, "base money" == "cash" + FRNs -- that's a trivial and unimportant point.

"Coins", by the way, are not issued by the Feds and therefore are not their liability, but that's unimportant too for trying to understand why one would prefer to hoard cash or FRNs rather than differently colored government paper.

[Jon]

I just willingly sold my labour to Carleton University in exchange for money. But I had absolutely no desire to hold a bigger stock of money.

Hahaha.

Its not fair to poke holes like that. Lets correct vjk's argument: They traded goods for government bonds. The mystery here is not about a stock of money (which they do not have). Its about their choice to acquire government bonds rather than investment goods.

That's why the trade is unbalanced. In so much as the Chinese have a propensity to save, it's in yuan not dollars. Direct dollar holdings are negligible.

[vjk]

Nick:

I just willingly sold my labour to Carleton University in exchange for money. But I had absolutely no desire to hold a bigger stock of money.

That's your personal preference that only you can rationalize ;)

Perhaps, one can consider your case statistically insignificant.

[vimothy]

vjk--sorry for the confusion, my comment was directed at Nick in response to "The central bank has to increase the supply of cash to prevent a recession".

You responded with: "I am not sure if supply of freshly printed cash has anything to do with preventing a recession -- witness $1 trillion of cash in "excess" that is already there."

I agree with that. Just saying that demand for cash is probably irrelevant, and the CB's "cash" to the bank sector is just an asset swap, to make existing stock of savings more liquid, not a flow of new saving. Probably just ignore it as it's old news and not phrased particularly well in any case.

[vimothy]

Actually, I'm not sure that I do agree with that. Anyway...

[RebelEconomist]

vjk,

It was the seizure of the inter-bank market in 2008 that I had in mind. After that, I doubt whether there has been a shortage of base money. I know Nick takes the view that the present recession is due to a paradox of thrift style shortage of the medium of exchange, but, while I can see that that can happen in theory, I have not seen any empirical evidence that that is what is happening in practice.

I have no doubt that the Fed has the power to lower long term interest rates if that is what they want to do and they are not deterred by the prospect of hyperinflation or taking big losses when the time comes to unwind the expansion.

By the way, "cash" means different things in different fields (eg including t-bills in asset allocation literature). As you define it above, cash is always "reserves" (as in a bank current account balance at the central bank, to avoid confusion with foreign exchange reserves!).

[vjk]

including t-bills in asset allocation

That's odd as you can repo t-bills, they are indistinguishable from zero-coupon bonds, and you cannot buy shoes or gas for them.

In fact, t-bills is the Fed's favorite toy in OMOs. Or used to be.

Repo'ing cash for cash sounds of odd, but, if that's the accepted terminology in asset allocation, who am I to argue ;) ?

[Andy Harless]

Nick:

"I just willingly sold my labour to Carleton University in exchange for money. But I had absolutely no desire to hold a bigger stock of money."

How is that different from any other kind of inventory? You may have some ideally preferred stock of money, but you're willing to let your stock fluctuate with inflows and outflows, just as a fruit stand is willing to let its stock of apples fluctuate. "The orchard guy came by this morning, and I willingly purchased a bushel of apples. I had absolutely no desire to hold a bigger stock of apples, but I figured I could probably unload them within a day or two."

If there's a bumper crop of apples, we don't say that it forces people to hold (or eat) more apples. Rather, it temporarily swells inventories above the ideal level and then (or simultaneously) changes the price of apples (and the prices and quantities of other goods like pears and cupcakes) until people are satisfied with their new level of apple consumption. Same way with money. A bumper crop of OMO's temporarily swells money balances above the ideal level and then (or simultaneously) changes asset prices (and goods prices and the quantity of aggregate output) until people are satisfied with their new money holdings.

[RSJ]

Nick, you believe that

1) "money" is simultaneously the medium of exchange and the unit of account
2) the central bank controls the size of this money stock (although it may be "forced" to adjust it based on changing preferences)
3) Unless the CB intervenes, the stock of money is fixed, and therefore households are not able, in aggregate, to increase or decrease their stock of money. That is where the paradox comes in.
4) Households have a demand for money, or for M/P.

The problem is that there is no "money" that satisfies all of the above. So you need a different definition that makes 1) hold, but then makes 3) false. Or if 4) is true, then 2) is false. Etc. So you are thrashing about. And each time someone points out that whatever definition you use to support argument #X makes argument #Y false, you then change the definition to make Y true, but then X becomes false.

All of this work, just to keep the flawed model going.

1) and 2) are only true if money is cash. All goods and services are bought and sold for cash. The cash settlement period may be a few days after, but it is a cash settlement unless the parties specifically agree to a swap. And even in that case, cash is always valid for the settlement of debts. Deposit accounts are not a medium of exchange, only cash is. And the CB only controls the amount of cash in the private sector as a whole. It does not control the stock of deposit accounts, bonds, or anything else. And you are right, in that "control" is misleading, as the CB reacts to the needs of the financial sector.

But 3) is false if money is cash.

First, households are able, both on an individual level, and in aggregate, to increase or decrease their cash holdings. They can withdraw cash from banks or deposit cash into banks. The private sector is not households, but households, firms, and the financial sector. The private sector as a whole has no demand or utility for anything. There is no "representative private sector". There are representative households and firms, that interact with banks and the government. Households can change their asset allocations both individually and in aggregate, by engaging in transactions with the financial sector. That is why when the CB reduces bonds and increases cash, it is not households that have their bond and cash holdings changed, but the financial sector has its balance sheet changed. 3) is only true if you define "money" as all financial assets. Then, a portfolio shift wont help. But as long as you define money as one type of asset, then households can, in aggregate, do a portfolio shift to hold less of that asset and more of some other asset, and they will succeed in the shift. These shifts happen all the time.

So if you want to define money that way, you have to accept the limitations of the quantity based view of monetary policy. The only quantities changing are of one form of money with another, and the only economic actors who see this change are the financial actors. That leaves only the short term interest rate mechanism, which isn't a very good mechanism in general, but is impotent in a zero rate world.

And 4) is true for financial assets but irrelevant for cash. Walking around money is not significant in any sense, and has no place in a simple economic model. Pollution is more likely to be a cause of our recession than a sudden desire to hold more precautionary cash. Just make the model cash-less, and abandon the cash in advance constraint that is causing so many problems. If you could abandon that constraint, then you would also be abandoning the budget constraints in your walrassian model, and so there would be no paradox of excess demand. And I would argue that a true "monetary" model would assume that the economy is cash-less and view *both* the unit of account *and* the medium of exchange as debt instruments. That allows investment to be self-funding and there can be "bubbles". If I create a debt instrument, I just created money. No one needed to save. By accounting identity, someone did save, but that savings could have been forced. All that happened was that I convinced the other market participants to sell their own debt instruments for my newly created one. That is a balance sheet operation that leaves the net worth of the participants unchanged, and so does not require any savings. There is no cash in advance constraint here.

If debt instruments are both the unit of account and the medium of exchange, then firms and households are creating debt instruments, and then paying for goods with them, so that you cannot argue that the sum of excess demands must be zero, but the sum of excess demands will be equal to zero plus the sum of net new debt instruments created. Therefore during a period of credit contraction, you must see an overall deficiency of demand. Now we can argue whether this simple model is true or not, or whether it is more true than the cash-in-advance hoarding explanation -- but at least if you are going to defend the old model, then make a commitment, define what you mean by "money", and then simultaneously defend all of the suppositions using the same definition.

[Currency wars 2010]

So far, the Bank of Canada has said nothing about the possibility of quantitative easing on this side of the border. Barring some sudden and ridiculous spike in the loonie, the Bank is unlikely to intervene. The loonie may fly higher yet.

[Greg Ransom]

"Forced savings" was language introduced by Hayek.

It meant changes in the price and availability of money and credit that changed the level and structure of investment in production processes -- i.e. more was invested for longer periods, i.e. "savings" in the sense of building stuff creating output for consumption on down the road.

The expression wasn't about stuffing money under the matress for future use.

The real economy made up of real stuff matters -- and is at the heart of "forced savings".

[Adam P]

"It meant changes in the price and availability of money and credit that changed the level and structure of investment in production processes "

that's not what it meant, not even close.

[Greg Ransom]

Adam, read Hayek on Bentham and others on forced savings here:

http://mises.org/books/profits_interest_hayek.pdf

Read Hayek's original discussion of forced savings here:

http://mises.org/books/pricesproduction.pdf

Educate yourself, Adam.

[RSJ]

I think Hayek attributes the "forced savings" concept to Bentham, at the end of the 18th century, and it refers to someone creating money and then bidding away goods from others with it, therefore forcing the others to spend less (and consequently save more) than they wanted. The main targets for scorn by the Austrians was bank money creation via lending, but you can also blame monarchs who debase the currency, etc. I.e., this is a crowding out of consumption goods story.

In all cases in order to have forced savings you need someone to violate their cash in advance constraint. The government can violate it because they make their own cash, banks violate it because they lend first and then get any required reserves later, and it can be violated in the credit markets by the following transaction:

A sells a bond to B in order to buy goods from C. B sells his own bond, say a government bond to C, and uses the proceeds to buy A's bond. C uses the proceeds of A's purchase to buy the bond from B.

All three transactions occur simultaneously, or at least settle simultaneously.

Here, B prices the bonds but neither saves nor dissaves -- he is shifting his portfolio. A dissaves and C saves, but C obtains the income necessary to save from the borrower. A funds C, not the other way around -- investment is self-funding.

If C doesn't want to save, then he will, effectively, sell his bond to X, use the proceeds to buy goods from X, and X will use the proceeds to buy the bond that C sold.

The bonds are just tokens that can be passed around in exchange for goods. The IOUs are money.

If everyone rushes to dispose of the bond, then goods prices rise up until savings is forced. If demand curves slope down, then this can be stimulative to some degree and you have a boom, with prices rising and real interest rates falling, up until the market begins to reprice the IOUs and/or there is a financial crisis in which the IOUs stop being liquid assets. Then, it is as if the money supply suddenly shrank. The central bank, by purchasing the safe assets that everyone is rushing to buy is not going to cause the money supply to increase. They would need to restore the value of the junk assets to the excess exuberance levels in order to restore the effective money supply to the boom period. Even then, there is no guarantee that the level of dissaving will be the same, as expectations will be different.

The idea that IOUs can be used for payment, and that therefore investors do not require a pre-existing pool of financial savings from which to borrow is not new. Also at the end of the 18th century, Alexander Hamilton, argued that the new constitutional government should consolidate on its books the states' revolutionary war debt, creating a large federal debt, and one of his arguments was that

"..there is a consequence of this, less obvious, though not less true, in which every other citizen is interested. It is a well known fact, that in countries in which the national debt is properly funded, and an object of established confidence, it answers most of the purposes of money. Transfers of stock or public debt are therefore equivalent to payments in specie; or in other words, stock, in the principal transactions of business, passes current as specie. The same thing would, in all probability happen here, under the like circumstances."

And it need not be only government debt. Confidence and/or the institutional arrangements determine when and what types of IOUs become liquid assets. Do banks turn away borrowers and/or raise the interest rate charged to borrowers while they wait to acquire enough reserves in advance, or do they lend first and then worry about reserves later? Do households, after receiving income, store it in the form of currency in advance of their expenditures or are the proceeds of sales automatically used to purchase deposit accounts or money market fund shares, which are really claims on longer maturity assets?

To the degree that the latter case is true, then transfers of stocks of IOUs do pass as payment in specie, and you need a different way of thinking about what equilibrates planned savings and planned investment if investments, or a large subset of investments, are self-funding.

Strangely this was known 200 years ago, but when economics bothered to introduce money into the model at all, they did it in the obvious way, in which money is only specie that is needed in advance for payment, and not in the "less obvious" way, in which promises to pay specie could also pass for payment in specie. To my mind, any model that enforces a cash-in-advance constraint is not really a monetary model, or at least, it is a non-financial model.

[Nick Rowe]

Interesting comment RSJ. Out of curiosity, did you read up on this in response to this post and comments?

But I think you are too hard on the CIA constraint. It was seen as a crude way to prevent barter exchanges in economic models; not as a final word on all monetary exchange.

[Adam P]

"it refers to someone creating money and then bidding away goods from others with it, therefore forcing the others to spend less (and consequently save more) than they wanted."

yes, that's basically correct. I would have said it operates through inflation, if their is a leverage financed spending boom this causes inflation that expropriates some of the real purchasing power of money holders. This forces them to consume less in real terms.

For the term "savings" to be appropriate it should refer credit financed investment that causes the inflation. Since investment must be matched by savings somewhere else. In this case the investment spending causes inflation and the expropriation of the real wealth of money holders is the savings that funds the investment. Of course these people aren't willing savers in this case, they were holding the money to finance their consumption, if they wanted to fund investment they'd have bought bonds.

Nothing at all to do with the structure of investment or investment for longer periods, everything to do with inflation. The only part Greg got right was that it is related to an increase in credit availability so I guess saying "not even close" was too strong.

[RSJ]

I agree, Adam, that inflation is the most common driver, but then you have Michael Pettis' critique of the asian development model, in which you don't necessarily see consumer price inflation, but a transfer of income from households to state owned enterprises, via forced household subsidies to the banking sector. Maybe you can split that up into a forced savings component and an inflation fighting component.

[RSJ]

Nick,

Mostly yes, although the treatise on credit has been a favorite read of mine for some time. Hamilton also goes on to argue that the creation of government debt and subsequent increase in broader money will drive up asset prices, specifically real estate, which had fallen significantly during the revolutionary war.

I think the endogenous money view is a very attractive alternative. All of the "fringe" groups from the Austrians to the PKers agree on this, and it has the added advantage of being the mainstream view of central bankers and others on the front lines when the industrial revolution was taking off and these financial crises as well as depressions were more common.

There is no a priori reason why you can't incorporate this view in a model that includes intertemporal utility maximization, and I've been puzzling about that, but there are other people here who could really do it right more quickly. I was actually hoping that you would do it :)

But the main thing is that a medium of exchange by itself should not make a real difference to the model. If you go from transactions of the form A is sold for B to A is sold for M and then M is sold for B, then the situation is isomorphic. M is unnecessary, except for changing the transactional overhead, and the model should not rely on this overhead to get fundamentally new effects from the A <-->B model.

In order to get a model with different outcomes, you need to allow for fundamentally different transactions, not isomorphic transactions. The tri-party transactions are an example. If you combine that with a financial sector that does maturity transformation, then effectively you can issue long term bonds to an intermediary, use the proceeds to buy goods, and the seller of the goods, who does not want to save, ends up purchasing a transactional account from the same intermediary that is backed by the bond that you issued. He is not purchasing a long term investment because the financial sector does the maturity transformation, so he is just receiving "money" for his good. But now you've introduced a non-isomorphic transaction, so the model can have different outcomes.

And the same argument goes for "why don't firms just pay their workers in the goods that they produce"? Again, it shouldn't because of M, since that model, apart from the frictions, is isomorphic. But if you allow for investment that does not immediately produce consumption output, then workers producing capital goods can't be paid in what they produce, since they demand consumption goods in the present period. At best the firm can pay them in IOUs for the future consumption goods that their present labor is producing, and these IOUs will be sold for present consumption. But now if the long term cost of capital does not fall, then everyone agrees that the IOUs are not worth much, and there is less demand in the economy as a whole -- less demand for labor, less demand for capital goods, and less demand for consumption. It was as if a large part of the economy had decided that it's current efforts were uneconomic and not worth doing, and as a result there are idle resources. And it seems reasonable to assume that long term rates will not adjust as quickly as short term rates, or that they can remain "too high" for a very long time before it is realized that they are too high. Here, too, I would prefer to look to a non-isomorphic model to explain the new effect, in this case, the introduction of investment that does not produce consumption during the present period.

So all of this is part of a program to try to find better models that are still very simple. I hope the program does not come across as a warpath, but I think that everything -- from sticky prices to demand failures, should be explainable in real terms, and in relatively simple terms, and you should not turn to frictions such as menu costs or ad hoc external constraints such as Calvo mechanisms to hammer the observed results out of your model. Anyways, I find these discussions very enjoyable.

[Nick Rowe]

Damn! It ate my comment!

Trying again:

RSJ: "But if you allow for investment that does not immediately produce consumption output, then workers producing capital goods can't be paid in what they produce, since they demand consumption goods in the present period."

Workers producing capital goods don't want to be paid in capital goods, because they would then have the hassle of swapping those goods for the consumption goods they do want. Agreed. But workers producing apples also don't want to be paid in apples, because they then have the hassle of swapping 99.9% of those apples for other consumption goods.

In both cases, it's more efficient for the firm to specialise in selling its goods for money, and giving the workers money.

[Greg Ransom]

adam, it's clear you didn't read hayek or Bentham.

In other words, your comments are comming out of the back of your pants.

[Adam P]

Greg, it's equally clear you didn't understand Hayek or Bentham and as such it's unknown where your comments are coming from.

[RSJ]

Nick, no that was not my argument. You argued that Keynesian economics makes no sense in a non-monetary world because "workers want goods but don't have the money, firms would hire the workers if they could sell the goods -- why not pay the workers in goods".

Whereas Keynes always, or primarily, argued in real terms -- "wage units" -- and he argued that investment was independent of savings and a determinant of it. Therefore the interest rate did not necessarily equilibrate planned savings and planned investment. It has nothing to do with the convenience or inconvenience of paying workers in goods. The workers paid in capital goods IOUs find those heavily discounted, and so overall demand declines because the investment rate of interest is too high for the markets to clear.

[Greg Ransom]

I invite you to actually read some Hayek -- and the Bentham quoted by Hayek.

There is no evidence that you have any knowledge of Hayek's work -- or Bentham's for that matter.

But your behavior does fit the definition of a "troll" to a T.