« How far out of line is the Canadian exchange rate? | Main | Bagehot on "Who owns the Fed?" »

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

"Low interest rates make people want to borrow more, but they won't be able actually to borrow more unless someone wants to lend more. There is both a demand and a supply curve of debt."

I agree. Good posts. That's it.

By the way, I've been rereading "The Methodology of Positive Economics", and I find it more to my liking than in the past.We've talked about this essay in the past. My main reason for reading "Essays in Positive Economics" is that, like Samuel Brittan, I'm a big fan of the essay "A Monetary and Fiscal Framework for Economic Stability". I'm trying to fit it into my current views.

Nick: The CB lowers the equilibrium by changing the supply schedule of loanable funds.

Thats it. Your entire discussion seems specious. You keep treating the lower-rate as deus ex machina and asking why debt increases. That's wrong. It excludes the explanation from the beginning. Which is that the CB changes the supply curve by offering to supply money at a cost below the natural equilibrium rate.

It does not matter that the CB then purchases already issued government debt. Its willingness to do so at a rate below the equilibrium alters the supply schedule.

So we get more loans at a lower cost.

Thanks Don: Would you describe Friedman's "positivism" as really prgmatism? Or instrumentalism?

Jon: That sounded plausible to me too. But the model says otherwise.

The model says that if a lender prints money, it shifts the supply curve of loans to the right. But if a borrower prints money, it shifts the demand curve for loans to the left. Fewer loans, at lower cost.

That's in the short run (fixed prices).

In the long run, with flexible prices, the demand for money shifts as well as the supply of money.

Well, I'll reserve comment until you discuss fractional (or zero, since that's what we have in Canada) reserve banking, but I'm puzzled why, if central bank money printing goes to pay off government debt, the money printing in the U.S. shows up as giant stockpiles of bank reserves while government debt explodes upwards.

Declan: In the last few months (but not during the last few years when debt was rising) it has no longer been true that nearly all central bank lending was to governments. And government debt has been rising because of fiscal expansion, plus the loss in tax revenue due to recession.

IIRC though, high-powered (base) money (the stuff created by central banks) is only about 5% of GDP in Canada (and about 10% in the US, though maybe half of that is held abroad?). And since High-powered money rises at roughly the same rate as nominal GDP, central banks printing money is just not that a big deal anyway, in terms of its effect on debt/GDP ratios, in either direction (Zimbabwes aside).

Yes, fractional reserve banking may be a much bigger part of the debt story.

What this is showing to me though is that we just have to think through these questions in terms of a model. Even if the "model" is just a story, without any maths, like my story above. It's the only way we can check on the overall consistency of the explanation, respecting adding-up constraints, and avoiding fallacies of composition. It's the discipline of general (dis?)equilibrium theorising.

Define money.

Define currency.

"Well, central banks are the ones who "find" the new money in the real world. And they are lenders, right? So if central banks increase the supply of money, and lend it out, debt will increase, right?

Wrong. Central banks are owned by governments, and governments are borrowers. It's governments who get the wealth transfer (via increased seigniorage profits) from central banks when they print more money. And central banks' assets consist (or did until very recently) almost exclusively of government bonds(/bills). That's what they buy with the freshly-printed money.

So in the real world it's the greens, not the reds, who find the new money. And they use it to buy back some of their outstanding debt.

An increase in base (outside/high-powered) money reduces debt, specifically government debt."

1) I don't think the fed is owned by the federal gov't. I thought it was private.

2) It is more likely that central banks/central bankers/investment bankers own the federal gov't.

3) I think we need some more definitions here and consider the difference between CURRCIR and bank reserves.

4) I'd like a better explanation of that paragraph that begins with "Wrong. Central banks ..."

Assume one world bank that is a monopoly with a 0% or near 0% reserve requirement and a 0% or near 0% capital requirement. Also assume full recourse loans (no default and I think that is the correct terminology) and no bankruptcy.

Would this entity be able to produce UNLIMITED debt?

Who is willing to continue to lend?

According to Brad Setser's posts/blog, recently it has been central bankers.

Is debt being concentrated in fewer and fewer entities?

Nick,

Isn't this explained by the usual supply/demand curves? The way a CB lowers the interest rate is by increasing the supply of loanable funds, making more base money available.

If we take this to say that the lower price of loanable funds (lower interest rate) came about from a shift in the supply curve for loans with a fixed demand curve for loans then it means that we have moved along the demand curve. In the usual diagram if you move along a demand curve to a point with a lower price then that point will have a higher quantity.

Why is that not enough?

Adam:

Yes, it is sort of explainable by supply and demand curves for loanable funds. But since the CB is owned by the government, which is (normally, in the stock if not always the flow sense) a demander of loanable funds, I would say CB lowers the interest rate by reducing the demand for loanable funds. (I wouldn't have looked at it that way until I worked out the little model in this post).

Most people look only at the demand curve for loans, and how it is affected by a change in the rate of interest. This is clearly wrong. We need to look at both supply and demand curves. As you recognise.

But is this even enough? The aggregate quantity of borrowing and lending is a macro variable. And any change in monetary policy has macro repercussions on P, Y, as well as r. And these changes in P and Y (and other things) will generally shift the demand and supply curves for borrowing.

In short, we can only handle questions like this properly in a fully general equilibrium (or disequilibrium) model. Partial equilibrium analysis with demand and supply curves just doesn't cut it.

One of the benefits (to me) of writing the above posts (and being forced to write them by comments on my previous post) is that it has forced me to recognise and articulate this point about partial vs general equilibrium.

God, the *discipline* of economics is beautiful. It's the (implicit) *discipline* that doesn't let us get away with just BSing about "low interest rates cause debt". "Where's your model?" demands the discipline.

Too much Fed:

I must write a post on "Who owns the Fed?". Thanks for giving me the idea!

"Who is willing to continue to lend? According to Brad Setser's posts/blog, recently it has been central bankers. Is debt being concentrated in fewer and fewer entities?"

Yes, recently (last few months) it has been central banks. But has this changed the total quantity of debt, or just changed the composition of debt in public hands? CBs have been selling govt debt and buying private debt. I don't think this leads to any greater concentration of ownership. More a change in who owns which type of debt.

"Assume one world bank that is a monopoly with a 0% or near 0% reserve requirement and a 0% or near 0% capital requirement. Also assume full recourse loans (no default and I think that is the correct terminology) and no bankruptcy.Would this entity be able to produce UNLIMITED debt?"

No. Wait for my post on fractional reserve banking. (Adam: see what I mean about partial vs general equilibrium analysis?)

"Define money. Define currency."

In the context of this model, "money" means "medium of exchange". In a monetary exchange economy, with n goods (including money), there are only n-1 markets, with one good being traded in every market. That good is the "medium of exchange", and is what I mean by "money".

In the context of this model, I am using the words "currency", "base money", "high-powered money", "outside money" synomynously (damn, can't spell it). Money which is not the liability of anyone, because the "promise to pay" is a meaningless promise. It can't be redeemed on demand for anything. So it's not debt in any useful sense. That's why I prefer to think of it as shells. Paper money is like that today. In a full-bore gold standard world, where the paper currency really is redeemable on demand in gold, it's different.

Inside money, like what's in your chequing account, is different, and is debt (in some meaningful sense). That's for my third post. (God, when will I redeem all these promises to post?)

Nick,

(God, when will I redeem all these promises to post?)


it's just another small addition to net debt.

Nick, I've kind of danced around this issue but you got right to the heart of it in the very first paragragh.

BTW, I read a few comments and some people indicated that the Fed increased the supply of loanable funds when they lowered interest rates. That may be true, but it doens't explain why interest rates fall.

1. Even if gold is used as money, a new discover of gold will reduce interest rates. But in that case there was no increase in the supply of loanable funds.

2. The increase in the supply of loanable funds from open market purchases is so tiny (often a few billion dollars or even less, even in a 14 trillion dollar economy like the US), that it could not possibly be responsible for the drop in interest rates, and the macroeconomic effects that flow from the lower rates. Instead, rates fall because nominal prices are sticky. That makes the nominal demand for money sticky at the original rate of interest. So when the supply of money rises, if prices can't immediately rise then interest rates fall. This raises aggregate demand. The higher level of aggregate demand may lead to more debt, it obviously depends on many factors. But you are right that the lower rates by themselves don't lead to more debt. When rates fall because the economy is weak (like right now) the real amount of debt does not typically increase. On the other hand debt often does rise when AD is very strong, but that is often a period of high interest rates.

"Yes, recently (last few months) it has been central banks. But has this changed the total quantity of debt, or just changed the composition of debt in public hands?"

Actually, I believe it was sooner than that.

From:

http://blogs.cfr.org/setser/2009/04/06/charting-financial-de-globalization-private-capital-flows-are-falling-faster-trade-flows/

"Two other critical points are obscured in a graph showing gross flows, but show up clearly in a plot that shows net private capital flows (inflows - outflows) against the trade balance (with the sign inverted, so a rise in deficit shows up as a large absolute number)"

graph

"First, the rise in the trade deficit — and rise in US imports relative to GDP — in the last eight years didn’t correspond with a rise in net private demand for US financial assets. That makes the expansion of the US deficit from 2002 to 2006 quite different from the expansion of the deficit in the early 1980s — or the expansion in the late 1990s. Official inflows made it possible for the US to sustain the rise in imports that was associated with the housing boom; had those flows not been around, interest rates would have had to rise to attract private flows — and the housing boom couldn’t have gotten so large. The excesses in the US financial sector and housing market simply were not, in aggregate, financed by strong private demand for US assets abroad.

Second, somehow the collapse of capital flows in the fourth quarter of 2008 produced a larger net inflow into the US than the surge in demand for US assets associated with the dot.com bubble. That, put simply, is why the dollar rallied in the crisis. Americans withdrew funds from the rest of the world faster than foreigners withdrew funds from the US."

And, "* A large share of private purchases of Treasuries from mid 07 to mid 08 will be reattributed to the official sector when the data is revised. And some “private” purchases in 05 and 08 likely came from the Gulf’s central banks, especially SAMA (which may make use of private fund managers for a portion of its Treasury portfolio)."

Banks don't care about the level of interest rates per se when they lend. They care about spreads. Spreads result from maturity and liquidity transformation - i.e. loans versus deposits.

Good post.

So perhaps the greatest quantity of lending/borrowing overall occurs not at the zero percent lower bound, but somewhere greater than zero? (Like every other market under the sun.)

Could it be that to increase lending in the current environment, interest rates should be raised?

And why the neglect of the supply side anyway? Its a bit like a stoke victim suffering from neglect of half their visual field, except that in this case its half of a Marshallian cross. Or perhaps it's just wishful thinking as to the efficacy of interest rate manipulation on economic activity that's to blame.

The central bank determines the domestic short term risk free rate.

It's got nothing to do with "loanable funds" theory.

Scott Sumner said: "BTW, I read a few comments and some people indicated that the Fed increased the supply of loanable funds when they lowered interest rates. That may be true, but it doens't explain why interest rates fall."

How about because of a productivity shock and/or cheap labor shock that lowers wage inflation and/or price inflation?

"In the context of this model, "money" means "medium of exchange". In a monetary exchange economy, with n goods (including money), there are only n-1 markets, with one good being traded in every market. That good is the "medium of exchange", and is what I mean by "money".

In the context of this model, I am using the words "currency", "base money", "high-powered money", "outside money" synomynously (damn, can't spell it). Money which is not the liability of anyone, because the "promise to pay" is a meaningless promise. It can't be redeemed on demand for anything. So it's not debt in any useful sense. That's why I prefer to think of it as shells. Paper money is like that today. In a full-bore gold standard world, where the paper currency really is redeemable on demand in gold, it's different."

Can we go with these terms to be clear?

Money as a medium of exchange. OK

Currency, bank reserves, and debt. OK

"base money", "high-powered money", "outside money". NO! I especially don't like base money because some people might confuse it with monetary base (currency plus bank reserves).

If money is a medium of exchange, aren't currency and debt demoninated in that currency fungible? Example: I buy a $15,000 car with $1,000 of currency and $14,000 of debt denominated in that currency.

In your apples example, I believe you need to add a corporation that earns corporate profits and workers who earn wage income.

IMO, the workers are borrowing based on current wage income/future wage income and not on apple production. Your example also gives the impression that the debt is paid off in apples (apple denominated debt). I doubt if you meant that. Isn't the debt denominated in currency and not apples?

"No. Wait for my post on fractional reserve banking. (Adam: see what I mean about partial vs general equilibrium analysis?)"

Are you a neoclassical economist?

anon said: "Banks don't care about the level of interest rates per se when they lend. They care about spreads."

Yes!

In your models, I believe you need to include the possibility that "someone" will attempt to use "cheap" debt to speculate in financial assets and produce asset bubbles if interest rates are low.

"second, if there were an increase in supply (shift in the supply curve) that caused both a fall in price and an increase in quantity traded."

If the fed is targeting the price, could they use debt (future demand) to shift the demand curve so that prices did NOT fall and there was in increase in both quantity demanded and quantity supplied (assuming I have my terms correct)?

Pictures:

http://www.raybromley.com/notes/noteimages/equilibrium/incrdemsup3.jpg

http://www.raybromley.com/notes/noteimages/equilibrium/incrdemsup2.jpg

So we don't get into the usual, I think I should add to my last post.

The rich have enough wage income to buy as many apples as they want without currency denominated debt. The rest would buy more apples but don't have enough wage income and don't want more currency denominated debt because they have enough currency denominated debt with current interest rates to ?barely? make the interest payments that the fed has kept high enough so they don't compete with the rich over apples. Once more apples become available, the fed lowers interest rates to entice the rest to go into more currency denominated debt (more currency denominated debt but the same monthly payment) to buy more apples instead of allowing more wage income for the rest.

You forgot to include "in a closed economy" in your model.

Of course in a open economy where debt is securitized and sold all over the world to nations with trade surpluses, evaluating supply and demand gets much more complicated.

Al: Agreed. It was implicit.

General equilibrium theorising can really only ever be closed economy. Open economies, small open economy models in particular, are really just a version of partial equilibrium theorising. But, AFAIK, the rise in debt has been a global phenomenon, not confined to any one country, so a global analysis seems appropriate. The world is a closed economy.

Too much Fed:
Just a few points.

Being poor, in the sense of having a lower permanent income, needn't mean you save a lower proportion of your income, let alone negative amounts. I don't think the rich countries now have a lower propensity to save than they did in previous decades, when they were poorer.

I can't think of any way the Fed could shift both the supply and demand curves of loanable funds to the right, except perhaps by making everyone permanently richer.

Every economist (including me), with the exception of a few of the more antediluvian Marxists, is a neoclassical economist, in broad perspective.

There is nothing weird about debt being denominated in apples. There probably are apples futures contracts, which are just IOUs for apples, and debt is just an IOU. Indexed bonds (called TIPS in the US?) are the exact analogy to apple bonds in my model. But an IOU for money is normally more convenient, for most people.

"base money" = "monetary base". Just two ways of saying the same thing.

Trade credit (when you get a car loan from the dealer) is a way to postpone paying with money, rather than avoid paying with money. But how you finance the payment for the car may affect your demand for money.

Introducing corporations into the model would make little difference. A corporation is like an apple growers' cooperative, except they pool their capital rather than their land or labour.

"Being poor, in the sense of having a lower permanent income, needn't mean you save a lower proportion of your income, let alone negative amounts."

I am going to put wage income in place of income in the two places.

Someone makes $20,000 per year and needs to spend it all in the first year. In the second year, this person's price inflation is 4% and gets a raise of 2%. To maintain the same standard of living, this person needs to borrow and/or sell financial assets.

And, "I don't think the rich countries now have a lower propensity to save than they did in previous decades, when they were poorer."

I believe that the savings rate in the USA has fallen from about 10% to 12% in the early 1980's to a negative savings rate around 2006 or 2007. I don't believe that savings rate takes into account wealth/income inequality that has increased since the early 1980's.

"I can't think of any way the Fed could shift both the supply and demand curves of loanable funds to the right, except perhaps by making everyone permanently richer."

Hmm... I am thinking I did not explain that correctly then.

Regular supply demand curve for apples; excess supply of loanable funds which is limited on the demand side by the interest rate and proper underwriting.

Supply curve shifts for apples (both a fall in price and an increase in quantity traded). Fed lowers interest rates which allows for more debt (there is plenty of supply of loanable funds). The rest buy more apples with currency denominated debt (future demand) assuming current wage income or increasing wage income in the future. Does that shift the demand curve?

If so, both the demand and supply curves for APPLES shift (I think to the right).

Sound better?

"There is nothing weird about debt being denominated in apples."

I thought gov'ts made debts payable in currency?

"In the context of this model, I am using the words "currency", "base money", "high-powered money", "outside money" synomynously (damn, can't spell it)."

And, "base money" = "monetary base". Just two ways of saying the same thing."

By my definitions and I think the fed's too, no way! Currency plus bank reserves equals monetary base. IMO, base money should not be used because of too much confusion.

"Trade credit (when you get a car loan from the dealer) is a way to postpone paying with money, rather than avoid paying with money. But how you finance the payment for the car may affect your demand for money."

How about changing money to currency those 3 times?

Trade credit??? Trade credit sounds like debt to me?

"Introducing corporations into the model would make little difference. A corporation is like an apple growers' cooperative, except they pool their capital rather than their land or labour."

I think it could make a big difference. The gov't allows policies to oversupply the labor market. Workers have borrowed in currency denominated debt against current wage income/future wage income. The corporation can now cut wage income to increase corporate profits. Now, the workers can't make the payments on the currency denominated debt.

Nick:

Perhaps I mentioned this before, but monetary policy impacts the price level, and the price level impacts nominal debt. The vast levels of nominal debt has been caused by monetary policy. If the quantity of money hadn't increased, nominal debt would have been much lower.

Of course, this doesn't explain high ratios of debt to to GDP.

Now, seriously....

I think your political economy is steering you wrong. Base money doesn't create new lending because it is issued by the government which is a debtor.

First of all, suppose the monetary authority only buys private debt? Say, it follows the real bills doctrine? Or maybe there is no national debt or budget deficit because the goverment is run by libertarian fiscal conservatives? How does it work then?

Anyway, I think that even under the status quo, where the monetary authority buys government bonds, this creates an increase in the supply of credit. The way to see this is to think about what those who sold the goverment bonds (or those who would have bought bonds from the government) do with the money. If they purchase private bonds, then that is an increase in the supply of credit to the private market. (Crowding out has been avoided.) This is matched on private balance sheets by the newly issued base money. This could be currency held by the public or reserve balances of banks at the monetary authority. This is lending of a sort (which you are failing to see, again because of political economy reasons.) Anyway, that additional lending to the monetary authority is not necessarily voluntary. And there you go, the increase in the supply of debt matches the excess supply of base money. (And the same thing can be true of banks issuing debt instruments that can be used of money.)

Bill Woolsey said: "This could be currency held by the public or reserve balances of banks at the monetary authority."

What happens if the fed CHOOSES to spike bank reserves but NOT currency? What are they trying to accomplish?

The increase in the supply of loanable funds from open market purchases is so tiny (often a few billion dollars or even less, even in a 14 trillion dollar economy like the US), that it could not possibly be responsible for the drop in interest rates, and the macroeconomic effects that flow from the lower rates.
I disagree. Open Market Purchases are large compared to MZM. Therefore they are significant determinates of present supply. People are not in the market to borrow assets generally. They are in the market to borrow something generally fungible, i.e., money.

Jon: what is 'MZM'?

Bill: I have been thinking about your 7.46.

First, agreed, its is not surprising if nominal debt rises in proportion to nominal GDP, and monetary policy can influence nominal GDP. But we are trying to explain why debt seems to have increased relative to GDP.

Suppose we had a net creditor government, running a budget surplus. (It both owns private debt, and buys more each year). If the government then prints more money, and buys even more private debt, it will be increasing the demand for that debt still further (increasing the supply of loans to the private sector still further), and (assuming fixed price level) increase the total amount of debt, as interest rates fall and the private sector borrows more in response. So debt increases when money growth increases in this case.

Government is then like the red apple growers finding more shells.

I guess that didn't really make sense. What I was really thinking was about is more in tune with excess reserves. Excess reserves and vault cash determine the supply of bank loanable funds at a given moment. OMO are large in comparison.

Long-rates must be risk-adjust averages of short-rates. Therefore, the OMO purchases do control the interest-rate across the yield curve.

MZM: money with zero maturity. M2 less time deposits, plus all money market funds.

Jon, any thoughts about these two articles.

http://seekingalpha.com/article/155367-fed-exit-strategy-another-bank-handout?source=article_lb_articles

http://www.marketskeptics.com/2009/03/us-banks-operate-without-reserve.html

The section on "deposit reclassification" is dead-on. Banks are not subject to reserve requirements in a meaningful way. ... There is a pattern in the global housing bubble that relates to which countries suspended reserve requirements when. The US effectively suspended reserve requirements in the mid 90s--many other hold-outs followed suit once the US validated the idea.

The primary limitations on banks are 1) cleaning balances and 2) capital requirements. Clearing balances now so low--along with the reserve ratio--that the difference between zero and the present value is not relevant. Remember: the money multiplier is 1/reserve ratio. We've already done from a 10x to a 100x constraint.

That business about vault cash struck me as irrelevant.

The business about raising reserve requirements to defeat the excess of liquidity is a valid point. That it hasn't been mentioned as a potential fall-back to dampen fears of disorder reveals a real intellectual rot at the Fed.

The trouble is that among the intelligentsia like Paul Krugman, there is a general ignorance as to the current reserve-requirements and history of what happened when. Krugman doesn't even understand how the Fed conducts OMOs. He's repeated the incendiary rumor that 'conventional policy' only involves TBill transactions. This is false.

The rot in economics is not just about theory; its about lacking basic knowledge of the facts.

A tangent: I like M.Carney's intention to deviate from inflation targetting and prioritize bubble-targetting.
Previously I'd only considered if the Gini becomes gross enough to be inefficient, a wise banker could give rich people the middle finger here. But Mark's idea might work for real estate bubbles, if we blew $100B on Iraq, maybe if 1993 bankruptcy became a real possibility (no will to raise taxes), any obvious inefficiencies (to little or too large reserve requirements)...
Inflation targetting seems to work at 2-3%/yr because that is *around* the underlying rate of growth. But taking it to the next level might uncover what is genuinely responsible for this growth. At the very least it corrects the fact GDP doesn't measure federal debt. It seems the USA BofC equivalent did the exact opposite last decade: bubble-building. But I trust our business schools here especially from a 2009 perspective.

The comments to this entry are closed.

Search this site

  • Google

    WWW
    worthwhile.typepad.com
Blog powered by Typepad