Milton Friedman said (somewhere) that money is like a refrigerator. Both are consumer durables that yield a flow of services to their possessor.
But in one important way that is a very bad analogy. Money is a medium of exchange, and refrigerators aren't. The stock of money, and the stock of refrigerators, are determined in very different ways.
What determines the stock of refrigerators held by the public? The short answer is: supply and demand. What determines the stock of money held by the public? The short answer is: supply. The suppliers of money, whether those be central banks, commercial banks, or counterfeiters, can force people to hold more money than they wish to hold. The suppliers of refrigerators can't.
There is an important sense in which all increases in the stock of money are like helicopter operations. The helicopter increases the stock of money without first persuading us we want to hold more money. We pick up the money whether we want to hold it or not. We plan to exchange that unwanted money for something we do want to hold. Friedman's refrigerator metaphor conflicts with his helicopter metaphor.
The modern debate over Quantitative Easing is a rehash of the very old debate over the Law of Reflux.
Start in equilibrium, where the existing stock of refrigerators in public hands (i.e. not counting inventories held by fridge producers) is exactly equal to the desired stock. And then the producers of refrigerators want to increase that stock.
Suppose, for example, the fridge producers have an inventory of 100 fridges in stock, that they suddenly decide they want to sell. In other words, the demand curve for fridges stays the same, but the supply curve shifts right by 100 fridges. In order to sell that extra 100 fridges, the producers will need to lower the price, in order to increase the desired stock of fridges by 100. If they don't lower the price, and if nothing else changes to increase the quantity of fridges demanded, they will fail to increase the stock of fridges at all.
Take an extreme example. Suppose the demand curve for fridges is perfectly inelastic. Every household wants to own exactly one fridge, regardless of the price of fridges and anything else. If each household were already owning one fridge, it would be totally impossible for the producer to increase the stock of fridges in public hands. You wouldn't be able even to give them away. Any extra fridges would be left lying on the sidewalk.
The actual change in the stock of fridges is determined by the short side of the market: new fridges sold equals whichever is less: additional quantity demanded; additional quantity supplied. Let Q be the stock of fridges in public hands. Delta Q = min{Qd-Q;Qs-Q}. The short side of the market determines quantity traded.
[Update: that formula doesn't work quite right for negative values. If fridge producers buy back fridges from the public, the reduction in the stock equals whichever is less in absolute value. And if Qd-Q is negative, and Qs-Q positive, or vice versa, delta Q =0. But for money, delta Q=Qs-Q works regardless of negative or positive values.]
Money is different. That's because everyone holds inventories of money. We buy money only in order to sell it again. In a monetary exchange economy we buy money when we sell any other goods, but that does not mean we plan to hold the money we buy.
Start in equilibrium, where the initial stock of money in public hands is exactly equal to the desired stock. And then the producers of money want to increase that stock, by $100. The producers of money do not need to increase the desired stock of money in order to increase the actual stock. Instead, they simply go out and buy something for $100. Suppose they buy a bicycle. The seller of the bicycle prefers the $100 to the bicycle, clearly. But that does not mean he wants to hold an extra $100 in his wallet. By assumption he doesn't, since we started at equilibrium. He wants to spend that $100 on something else. The supplier of money doesn't offer a price that makes him prefer holding the money to holding the bicycle. He offers a price that's sufficient to buy the skateboard he prefers to hold instead of the bicycle.
There is now an excess supply of money. The actual stock of money exceeds the desired stock. And that excess supply of money will hot potato around the economy, bidding up prices of goods and assets, and increasing quantities of goods bought and sold, until prices and quantities eventually rise enough so that the extra $100 is now willingly held.
Unlike refrigerators, the stock of money held by the public is supply-determined. We accept new money in exchange for goods, not because we necessarily want to hold that new money, but because we know we can always pass it on to someone else in exchange for something we do want. The seller of the bicycle didn't want the money; he wanted a skateboard. He only accepted the money because he knew it was the easiest way to get a skateboard in exchange for his bike. That's what money is supposed to do. It's the medium of exchange. We accept it only because we know someone will want to accept it from us. I might buy a fridge even if I could never sell it again. I would never buy money if I could never sell it again.
When a commercial bank offers me a loan of $100 and I accept, the bank credits my chequing account $100 as payment for my IOU. The stock of money has increased by $100, even if I have no desire to hold an extra $100. Almost certainly, I borrowed $100 not because I wanted to leave it sitting in my chequing account, but because I wanted to spend it. And when I spend it, the next person who gets it doesn't want to hold it either.
The counterfeiter doesn't have to increase the desired stock of money in order to profit from his trade. He just prints it, and spends it.
Take an extreme example. Suppose that the demand for money is perfectly price inelastic, income inelastic, and interest inelastic. Every person wants to hold exactly $100 on average, regardless of anything. And suppose every person were holding exactly $100 on average. Each individual would still accept new money, because he knows he can pass it on to someone else. They won't leave $20 bills lying on the sidewalk. They will pick them up and spend them.
For refrigerators, delta Q = min{Qd-Q;Qs-Q}. For money, delta Q = Qs-Q.
In an important sense, all money is like helicopter money. The suppliers decide how big a stock will be held, regardless of the desire to hold it. People will pick it up whether they want to hold it or not. If they don't want to hold it they will spend it, to try to get rid of it in exchange for something they do want to hold. And it hot potatos around the economy until prices rise enough or incomes rise enough that they do want to hold it.
An open market purchase of bonds is still helicopter money. It's a helicopter increase in the stock of money, plus a vacuum cleaner reduction in the stock of bonds.
Most economists disagree with me on this. They think that the stock of money is determined by supply and demand in just the same way that the stock of refrigerators is determined. And if the supply curve is perfectly elastic, as under the gold standard or when the central bank targets a rate of interest, they say that the stock of money is determined by the quantity demanded at the price set by the supplier's supply curve. They believe in the "Law of Reflux", which says there cannot be an excess supply of money, because if there were that excess would immediately return to the issuer. This ignores the fact that money, unlike refrigerators, is bought and sold in all markets. The excess supply of refrigerators will appear only in the market for refrigerators. The excess supply of money will appear in all markets, not just the single market where it is issued.
I know I'm way out of the "mainstream" in arguing against the Law of Reflux. The mainstream has not understood that money is not like other assets. We *always* accept the medium of exchange when we sell other goods. It wouldn't be the medium of exchange if we didn't. We do not ask whether we want to hold an extra $100 when we sell a bicycle. We ask whether $100 is enough to buy other goods that we want more than the bicycle. We would never accept that $100 if we had to hold it forever, even if we could get it for next to nothing.
This post is a response to David Glasner's post. The "mainstream" view is so widely held that even some Market/neo/quasi-monetarists hold it. And this is a very old debate in monetary economics. I am arguing alongside Yeager against James Tobin. This very old debate over the Law of Reflux is what is at the root of the very modern debate about whether Quantitative Easing can work. Those who argued for the Law of Reflux argued that an excess supply of money could not cause inflation [update: because any excess supply of money would immediately flow back to the issuer]. Banks could only cause inflation by lowering the price of money in terms of gold. Modern proponents of the Law of Reflux argue that banks can only cause inflation by lowering the rate of interest.
(Anyone got a good link for the Law of Reflux?)
[Update: Here is Mike Sproul (pdf) defending the Law of Reflux.]
[Update 2: My arguments in this post are very much a result of what I learned from David Laidler. I waited to get email confirmation from David that I had understood him correctly on some points before adding this. Responsibility for errors and opinions is still my own, of course. He may or may not agree with everything here.]
Worthwhile Canadian Initiative 
Oh, and just a request to commenters: please be very careful to distinguish the following distinct concepts: Demand; quantity demanded; actual quantity; quantity supplied; supply. Those ECON1000 distinctions are unusually important in this case.
For example, if you were planning to say “The supply of money is determined by demand”, then stop, and check those definitions first.
This post just seems a terrible war against using the proper units.
Everything you say makes sense when you talk about income (or revenue) which is what is really demanded, but it stops making sense when you talk about stocks of money. The only possible demand for a stock of money — in a dynamic economy — is as a buffer to smooth gaps between expenditures and revenues. But there can be an ongoing insatiable demand for revenue even though there is no excess demand for deposits, or stocks of money.
And you haven’t made the case at all that the demand for the stock (e.g. deposits) is what is in excess rather than the demand for the flow (e.g. revenue). The two are separate demands. One is always met (the demand for deposits, as these are very profitable for banks to supply to households, and banks are always ready to do this) — but the other need not always be met.
But I agree that income need not “come” from anywhere, as a small stock of money can support any income flow. It may seem that revenue drops from above — from a helicopter. Or that demand creates itself, etc. But if the government is to play a role in fueling this helicopter, or in getting the ball rolling, then it can only do this by purchasing output — fiscal policy — and addressing the right demand, rather than trying to fix what aint broken — the demand for deposits.
I’m not sure about your description of the Law of Reflux. In modern terms, the Law of Reflux states that the central bank cannot increase the money supply beyond what the public would want to hold, because the public can always bring their money to the central bank and exchange it for bonds. This is right as far as it goes, although “what the public will want to hold” depends on the monetary authorities’ market operation and policy stance.
In your example, the central bank buys a bicycle for $100, presumably in order to expand the money supply. But once the hot potato effect has played out and prices have risen, they may want to sell that bike and take their $100 back (plus any seigniorage gain/loss), which is a kind of reflux.
rsj: “The only possible demand for a stock of money — in a dynamic economy — is as a buffer to smooth gaps between expenditures and revenues.”
Yep. And that is exactly what we mean when economists talk about “the demand for money”. We mean the desired average size of that buffer stock. And we hold that stock because it’s too costly to spend all of our paycheque the minute after it hits our bank account.
anon: OK. Let’s interpret the Law of Reflux to mean: “1. banks can’t increase the stock of money beyond what people want to hold; and 2. and even if people were holding more than they wanted to hold (because the quantity demanded fell, for example), that excess supply would immediately be returned to the bank and eliminate itself that way”.
Only part 2 is the Law of Reflux in the narrow sense, but I think we can safely expand the Law of Reflux just a little to include 1 as well. It’s hard to believe 2 and not 1, though it’s logically possible? I expect with fridges, 1 is true, but 2 is false, unless fridge producers are willing to buy back unwanted fridges. Good point.
“In your example, the central bank buys a bicycle for $100, presumably in order to expand the money supply. But once the hot potato effect has played out and prices have risen, they may want to sell that bike and take their $100 back (plus any seigniorage gain/loss), which is a kind of reflux.”
Yes, I think I agree with that.
“Yep. And that is exactly what we mean when economists talk about “the demand for money”. We mean the desired average size of that buffer stock. ”
But it’s strange to think that if you really do want to hold more deposits, that you will not just sell some bonds and buy those deposits. That is easy to do, people do it all the time, and it’s impossible for those markets to not clear (in a “modern” economy with deposit insurance, etc.)
On the other hand, selling more goods in order to obtain more revenue is hard to do.
Selling goods == hard, Selling bonds == easy.
Why is this so hard to understand? No one wants more money, everyone wants more revenue.
rsj: you are ably defending the “mainstream” position. I would reply that the bond market is not the money market. Just because the bond market clears does not mean the money market clears. There is no money market. All markets are money markets, Etc. If they varied the stock of money by buying and selling bicycles, the bicycle market would always clear, provided the bank stood ready to buy and sell bicycles at (roughly) the same price.
I probably shouldn’t keep on saying I’m such a heterodox rebel against the mainstream. I do it just to gently tease those who are so proud of themselves for thinking they are heterodox rebels against the mainstream, when in many ways they aren’t. (I don’t mean you, rsj.)
All markets are money markets… including the bond market?
RSJ, there is no practical difference between “stock” and “flow” accounts of monetary disequilibrium, because “stocks” and “flows” are tied by accounting identities. When present and expected future money balances are in equilibrium, then money flows must be in equilibrium as well. Conversely, when either of these quantities is in disequilibrium, the others must be as well. Leland Yeager is very clear on this: see his essays in Fluttering Veil.
I think i posted a paper here, a while ago, dealing with sticky prices, the hot potato effect.
Does rational expectations hold, that the entire economy sees the money increase, and raises prices, or does the excess supply have to work itself out, because we are not sure where the bicycle seller will buy the skateboard, and the skateboard owners don’t want to scare away other buyers?
Same thing for vacuum cleaners in reverse.
O/T, but when the central bank buys bicycles isn’t it causing a redistribution of wealth towards producers and owners/re-sellers of bicycles? Does QE imply a redistribution of wealth towards the government (producer of bonds) and bond-holders? Is this redistribution significant in terms of magnitude?
“In an important sense, all money is like helicopter money.”
What about the people who say all money is debt?
if 3 skateboard sellers, each raise prices 33% of the 100$ in anticipation, or is it raise prices by the 100$ effect on the entire economy, eg, .02% or something.
Or do they raise prices by the depreciated value of the bicycle sellers excess stock, eg he doesn’t want to hold it, so the $100 dollars has negative interest rate, worth only 97 outside dollars to him, skateboard sellers raise 33% of the 3$ value. … this is o/t so i don’t care if you answer.
david: “All markets are money markets… including the bond market?”
Yes. It’s just one out of many.
anon: you’ve thought about and read about this stuff. Please stick around.
edeast: I am more or less comfortable applying rational expectations in an equilibrium, if that equilibrium is drawn from a stochastic process that hasn’t changed much for a long time, and is fairly easy for people to learn. But I don’t feel at all comfortable applying RE when I’m talking about some sort of disequilibrium process that gets us from one equilibrium to another. Sure, if people see what’s happening, and understand the process. Or even if they don’t understand the process but have seen it before, their expectations will change when the supply of money increases, and that will have direct impacts that will tend to jump the economy towards the new equilibrium, bypassing the hot potato process. But I’m wary of full-bore RE in this case. Sometimes I get the gut feel that maybe some of the old fundamentalist Keynesians were on to something, and that we can’t fully understand money in an equilibrium model.
I think this thought experiment just sort of fails, starting in the beginning.
“Start in equilibrium, where the initial stock of money in public hands is exactly equal to the desired stock.”
Under what conditions does this equilibrium exist? According to the permanent income hypothesis, what happens if some one receives an expected one time income shock- they save it, correct? Unless it’s from the Fed?
I think this is just entirely, entirely wrong in a behavioral micro sense.
Further, are reserves really money when the interest rate is above the natural rate? Or is it just like a producer’s inventory of refrigerators?
Lastly, money is not just a medium of exchange, it is also a medium of account.
Nick:
I agree with you.
Still, if we have gold redeemability, and there is an excess supply of money that is going like a hot potatoe, one of the things being demanded with that money is gold. While the price of other goods can rise, but the price of gold is fixed and the monetary authority must supply gold to meet this demand. This involves decreasing the quantity of money. The quantity of money decreases, and there is no longer an excess supply.
In equilibrium, the price of the paper money is fixed by gold redeemability. The equilibrium quantity can only rise if the demand increases. In some kind of long run (which is probably pretty short) the supply of paper money is perfectly elastic.
Click to access hastef0599.pdf
You have conditioned your argument as starting in equilibrium, but many of your arguments wouldn’t apply away from it, so it may be valid but inapplicable.
Nick, I don’t see where you and Glasner disagree. The supply of base money is exogenous and the supply of bank money (i.e. deposits) is endogenous. Don’t you both agree with that?
We all agree Tobin was wrong about base money.
So let’s go back to the continuous model in which the sum of excess demands for stocks is zero and the sum of excess demands for flows is zero.
And further assume that the markets in which stocks are bought and sold for deposits are: { bonds, equities, deposits } are perfectly flexible price markets. That seems to be the case.
But that the markets in which flows are bought and sold for revenue: {labor, output } are not perfectly flexible price markets.
In that case, everyone could hold the level of deposits that they wanted, but they might not have the revenue that they demand. It is not true that disequilibrium in the flow markets spills-sover into disequilibrium in the stock market.
It matters whether you get money by selling a bond or selling your labor. In one case you are earning revenue but in the other case you are not. Therefore there is not just a demand for “money”, there is a demand for revenue and a separate portfolio demand.
Suppose that someone with wealth has his labor hours (involuntarily) reduced by 1/2 for a year. He is not able to sell the other half of his labor. Nevertheless he is holding exactly the level of deposits that he demands. Perhaps the low employment period will make him want to obtain more deposits by selling some of his bond holdings, which he can easily do. But he cannot easily obtain more revenue by selling his labor.
In that case, if the central bank, seeing this unemployment problem, decides that what the economy really needs is more portfolio shifting services, so it offers to exchange more bonds for deposits to any takers, the rest of the economy would think the CB pretty clueless, right?
They don’t need more deposits, and they don’t want more deposits.
What they want is to be hired so that they have more labor income.
They are forward looking — if the situation continues, they will run out of assets to sell, so they need to get back to a state in which they are accumulating assets rather than dissaving as they wait to get hired.
The central bank, by swapping some of their bond assets for deposit assets, isn’t doing anything to get households back to their desired state, because it is intervening in the market that is fully clearing, not in the market that is not clearing. The central bank is attempting to supply deposits and decrease bonds, whereas households demand more income in exchange for their labor.
The suppliers of money, whether those be central banks, commercial banks, or counterfeiters, can force people to hold more money than they wish to hold.
I’m very puzzled about this idea.
Suppose, for the sake of comparison, I am a refrigerator wholesaler who doesn’t store chilled food and has absolutely no personal use for refrigerators. The only reason I acquire refrigerators is to exchange them for other things, and the only reason I hold refrigerators rather than exchange them immediately is because I anticipate a higher reward from holding the refrigerator and exchanging it at some future time than from exchanging it right away for whatever I can presently fetch for it.
For me, refrigerators play a role similar to the role of money. They aren’t actually money, though, because although the only value I attach personally to refrigerators is as an instrument for acquiring other things; the fact that refrigerators possess this exchange value for me depends on the fact that there are a not insignificant number of others who actually use refrigerators for non-exchange purposes, and desire them on that basis.
But given that I live in a finite world, in which there is a limited desire for refrigerators and in which the marginal value of additional refrigerators rapidly decreases for those who use them; in which transaction costs for refrigerator exchange are significant; in which even well-protected stored refrigerators depreciate significantly in value on time horizons measured in months; and in which the costs of storing and protecting refrigerators is high, then there is some maximum number of refrigerators such that the marginal cost of holding, protecting and storing the refrigerator exceeds the marginal benefit of acquiring that refrigerator, so at that point there is nothing at all among my possessions I am willing to exchange in order to acquire that refrigerator. I won’t even sing a brief song for one of them. If the refrigerator elves drive up with new refrigerators, not only will I refuse to buy them at any price they ask, no matter how low, I will insist that they do not even give them to me for free. And if they steal into my warehouse at night and just leave them there, they become “hot potatoes”. I might trash them or give them away if I can do so easily. Or I might lower the price on all of my refrigerators so I can deplete my stock more rapidly than I had initially planned.
And suppose I have in my possession a valued promise from the elves for some additional number of refrigerators to be delivered to me at some point in the future, after my current stock has been depleted through exchange. I will be unwilling to exchange that promise for any number of current refrigerators, no matter how large. I prefer the refrigerator bond.
Dollars in their current, largely electronic form, however, seem somewhat different from refrigerators in many of these regards. The marginal costs of holding, storing and protecting additional dollars are so negligible that there is no practical limit to the number of dollars I am willing to acquire. Yes, it might be the case that there is nothing in my possession that I am willing to exchange for an additional dollar, but because it costs me nothing to hold the dollar, I am always willing to acquire a dollar that is simply given to me. Yes, my dollars depreciate in value when they are held; but again, since it costs me virtually nothing to hold and protect an additional dollar, the fact that they depreciate in value only gives me more reason to acquire as many as I can, without limit.
And if I have in my possession some valued promise for a certain sum of dollars to be delivered to me at some definite time in the future, there will always be some number of dollars such that I will be willing to accept that number of dollars now in exchange for the promise. I can then save these dollars securely and at negligible cost. My dollar bonds always have a price in practical terms.
So I’m not sure I can see the point at which dollars become hot potatoes.
Now, I might recognize that a general increase in the quantity of dollars in existence, and more importantly a general increase in the number and density of dollar transactions, will reduce the exchange value of the dollars in my possession. But these are general phenomena over which I have no personal control. If I am the only person acquiring additional dollars, then so as long as I don’t circulate the dollars myself, these inflationary effects will not occur. And if many other people are acquiring and exchanging additional dollars, my dollar stocks decline in exchange value, whether I acquire additional dollars or not. So it remains in my personal interest to acquire as many of these additional dollars as possible. The acquisition of addition dollars always dominates the non-acquisition of additional dollars.
So I am still having trouble understanding where is the maximum point at which the monopoly producer of dollars can force me to acquire more dollars than I wish to hold. For all practical purposes, there is no such maximum number of dollars.
Now suppose I am a dollar lender of any kind, that is, I am a possessor of dollars who exchanges dollars currently in my possession for promises of certain sums of dollars to me delivered to me at some definite point in the future – or perhaps for promises of other goods. It seems to me that we could easily imagine situations in which the market in which dollars are exchanged for promises of future dollars has already cleared. No matter how many dollars I possess, I cannot currently find a promise in that market whose value for me exceeds the value of some amount X of dollars I possess, and for which the owner of that promise is willing to exchange it for X dollars or fewer. If I acquire more dollars, I will thus prefer to hold and accumulate them, since we have already seen that dollars are not hot potatoes for me. And for any dollar bonds I hold, there will always be some number of dollars such that I am willing to accept that number of dollars in exchange for the bonds.
Now suppose the dollars-for-promises market has not cleared, but I am prevented from exchanging my dollars for promises because the authorities restrict my exchanges in accordance with a rule: I may not make these exchanges unless I have in my possession a certain number of permission slips issued by the authorities. Bank reserves function as permission slips. The fact that these permission slips are also dollars is interesting. If dollars are placed unilaterally in my reserve accounts it appears they play a dual role. They are dollars I come to possess and that could be exchanged now or later, and they are also permission slips to lend dollars that I previously possessed. But if the dollars-for-promises market has cleared, then whether conceived as permission slips or dollars, they will not prompt me to do any additional exchanges, and my preference will be simply to hold onto them at their negligible storage cost. They are hardly hot potatoes.
(I have simplified the actual situation in that the deposits and reserves held by banks are not the dollars they own, but their liabilities to depositors. They exchange these dollars on the dollars-for-promises market as agents of the depositors.)
What will spark renewed activity in the dollar-for-promises market? Only an increase in the value of the promises capable of being produced, or a change in the willingness of the promisers to exchange currently available promises for amounts of dollars that they are not willing to accept currently. This all depends on changes in the opportunities for production and exchange of real goods in the markets outside the dollars-for-promises markets, not on the number of additional dollars placed in bank reserve accounts, which it appears do not function as hot potatoes.
If the money we used consisted exclusively in large blocks of granite, with high storage and transaction costs, maybe the situation with dollars would be more like the situation with refrigerators. But it does not. Most of it consists in electrons on computer hard drives, and it makes hardly any difference from the standpoint of cost whether you are holding a thousand or a trillion of them.
Nick:
You are not giving the logic of the law of reflux a chance to work. I want your $100 bike. You want my $100 skateboard. We both swipe our credit cards in each others’ card readers. We have each borrowed 100 credit card dollars, you from visa and me from amex. I now have your bike and you have my skateboard. The credit card companies lent us each $100 because we each gave them a $100 lien against our next paycheck. $200 of credit card dollars have been created. But the $100 we swiped to each other will not be spent again. They will sit in our accounts until we pay them off, and then they will be extinguished. It is no different than if we had bartered the bike and skateboard. You talk as if the newly-created dollars must say in circulation, but the whole idea of the law of reflux is that the dollars can be extinguished as fast as they are created.
“(Anyone got a good link for the Law of Reflux?)”
Why…yes. Yes I do. Click on my name above [below, NR] and check the last link on my real bills website.
[I haven’t read it yet, but Mike is a very clear expositor (of a perspective on money I disagree with) so I recommend you read him. NR]
Scott: I am saying that Tobin was wrong about base money, and also wrong about bank (inside) money.
The stock of base money is supply-determined. The stock of bank money is supply-determined too. Central banks, commercial banks, and counterfeiters, can all increase the stock of money by increasing the supply, and create an excess supply of money.
Whether one or the other is “endogenous” is a separate question. How much money central and commercial banks choose to create may depend on many things, which makes it endogenous if there’s a feedback loop from the amount they create to the amount they choose to create.
Here’s a comment I left on David Glasner’s post, which expands what I wanted to say further:
“While it is true of course that a bank does not literally lend its reserves, an individual bank that extends a loan by creating a deposit knows that it is very likely that that deposit will be spent, and redeposited in a second bank, and that the first bank will therefore lose reserves to the second bank.
I interpret the standard textbook story as an attempt to tell a hot potato story. Or rather, two hot potato stories at once. Starting in equilibrium, the central bank creates an excess supply of reserves. That excess supply of reserves hot potatos around the banks. But as banks pass on the hot potato of reserves, they extend loans by creating deposits. And those deposits hot potato around the public. And that second hot potato will be bigger than the first hot potato, under fractional reserve banking.
Individual banks will compete for deposits. That means they compete to get people to hold deposits. The larger the stock of deposits they can persuade people to hold with them, the larger the stock of loans they can hold on their book, ceteris paribus. It’s a competition between individual banks for market share. But the total size of that market changes as a result of the disequilibrium hot potato process the standard story attempts to describe.
What is missing from the standard story is any discussion of the public’s (stock) demand for money. And that, I think, may be at the root of what you don’t like about the standard story. And it is indeed absolutely weird that a model should talk about what determines the total stock of some asset by talking only about the supply side of the market with no mention of the demand for that asset. But, if my view is correct, that weirdness is exactly how it should be. The medium of exchange is not like other assets. The suppliers of money can increase the actual stock without having first to increase the quantity demanded.”
I expect I need to write a second post, titled “commercial banks as destroyers of money”: putting the argument here in reverse; and applying it to the present day.
I will be back later today.
OGT: “Under what conditions does this equilibrium exist? According to the permanent income hypothesis, what happens if some one receives an expected one time income shock- they save it, correct? Unless it’s from the Fed?
I think this is just entirely, entirely wrong in a behavioral micro sense.”
No. You need to distinguish spending from spending on newly-produced consumption goods. You need to distinguish hoarding money from saving. If you win a one-time $1,000, you may save most of it, but not in the form of money. You will spend most of it on assets, not on immediate consumption.
Bill: And I agree with you. But when someone buys a bicycle from the producer of money, that does not mean they prefer holding the bicycle to holding the money. It may mean they prefer holding a bicycle to holding a skateboard. (I’m just taking what I said above, and putting it in reverse, to describe how a reduction in the supply of money causes a reduction in the actual stock, regardless of the demand for money).
Nick, you said:
The stock of base money is supply-determined. The stock of bank money is supply-determined too. Central banks, commercial banks, and counterfeiters, can all increase the stock of money by increasing the supply, and create an excess supply of money.
Nick, can you please try to answer the following questions for me. How would you characterize the conditions of equilibrium (in terms of profit-maximization or some other criterion) the amount of deposits created by an individual bank? When you say that commercial banks can “increase the stock of money by increasing the supply,” what assumption are you making about whether the individual bank is or is not at a point of equilibrium? Finally, when you say that the stock of money is supply-determined, does it matter whether the supply at that moment is consistent or inconsistent with the equilibria, however defined, of the individual banks that are supplying that amount of money?
David Glasner: “Central banks, commercial banks, and counterfeiters, can all increase the stock of money by increasing the supply, and create an excess supply of money”
Consumers and businesses can all decrease the supply of money by decreasing the demand, and create an excess supply of money, is exactly equally half wrong. There’s a supply, a demand and a price which is the interest rate on bank loans. It doesn’t make any sense to focus entirely on one side of the story as if the other was totally elastic.
As I recall, in George Selgin’s “The Theory of Free Banking,” he explained how there is a separate supply and demand for outside money and a separate supply and demand for inside money. They functioned according to separate rules, but changes in either one would cause a recession.
rsj:
“The central bank is attempting to supply deposits and decrease bonds, whereas households demand more income in exchange for their labor.
”
Precisely.
That’s why all the discussions about money stocks and flows endogenous or otherwise, CB overdraft rules, cabbages and kings, while a lot of fun, are very much beside the point. The point of where that demand for “their labor” is going to come from. There is no serious discussion of that point, not even a feeble attempt unless you count MMT with their nebulous “employer of the last resort” idea.
People may have “demand for money”, but the natural way to satisfy such demand is to sell their labor in some way. And the feds contorting in a “twist” is a rather repulsive picture to watch in that context. The dance is not going to create new jobs but merely introduce new and unpredictable distortions in the already distorted market.
vjk: “People may have “demand for money”, but the natural way to satisfy such demand is to sell their labor in some way.”
When people talk about the demand for money I assume they aren’t talking about “demand for wealth” (whatever that means). I think this is totally confusing the matter. When you talk about demand for money it should be considered versus things that are almost identical (buying a t-bill or drawing from a line of credit) but which cant be used as medium of exchange. Mixing in our labour/leisure preferences really confuses things.
It is a demand for wealth, except that there is no market in which you can buy wealth, and so it is not a “demand”, in the sense of a vector of bids that can be submitted to an auctioneer. It is better described as a motivation of preference that drives all the actual bids that are submitted.
The central bank cannot supply wealth, but Treasury can. A helicopter drop of money, or a purchase order for more output, or hiring an unemployed worker — all of that will supply more wealth to the private sector. But swapping bonds for cash will not.
Lowering the short rate may do it, but it may not. Monetary policy is less reliable than fiscal policy at stimulating demand.
David: “Nick, can you please try to answer the following questions for me.”
Let me try. Though in some cases my answers might be the same as yours, or maybe, even if they are different, I might say yours are as good as mine.
“How would you characterize the conditions of equilibrium (in terms of profit-maximization or some other criterion) the amount of deposits created by an individual bank?”
The bank must be just indifferent to creating a new loan, or replacing an old loan as it is paid off. It knows that if it makes a new loan by creating a deposit that deposit will be spent, and it will lose reserves to a second bank. So it compares the expected interest revenue on the new loan to the cost of replacing the reserves it would lose if it makes the loan (or to the interest it is earning on reserves, if any). Plus transactions/intermediation costs from lending long and borrowing short. Same answer as you? Roughly?
“When you say that commercial banks can “increase the stock of money by increasing the supply,” what assumption are you making about whether the individual bank is or is not at a point of equilibrium?”
I start at full equilibrium, both for the individual bank, and all banks, and individual people are holding their desired stock of deposits too. Then something changes which makes the individual bank want to increase its supply of new loans. What is that “something”? Anything that upsets the above equilibrium. Perceived reduction in the risk of loans; increased market interest rate on loans; rightward/downward shift in the supply of reserves; plus other things I have probably forgotten.
“Finally, when you say that the stock of money is supply-determined, does it matter whether the supply at that moment is consistent or inconsistent with the equilibria, however defined, of the individual banks that are supplying that amount of money?”
Let me come at that question sideways.
Start at full equilibrium. Then an individual bank wants to increase its supply of loans, so it offers slightly better terms and makes a new loan, by creating a new deposit. This increases the total stock of deposits. But it does not (except for a day or two before the deposit is spent) increase the stock of deposits at this particular bank. The increase in the aggregate stock of deposits is an unintended by-product of the individual banks desire to increase its own supply of loans. The individual bank is back on its supply curve of loans, but some other bank(s) now have more reserves than they desire, and are off their supply curves of loans. And the public is off its demand curve for deposits. The excess desired reserves hot potato around the banks, and the excess desired deposits hot potato around the public. This is a disequilibrium process. Whether the system eventually gets back to equilibrium, and what that equilibrium looks like, is a macroeconomic question, that cannot be answered just by looking at banks. (The standard textbook story only gets the banks back into equilibrium, not the public, and only achieves that with the debatable assumption of a strictly positive desired reserve ratio.)
When individual banks compete with each other to get more deposits, that is about market shares for deposits (and loans). That is not what increases the aggregate stock of deposits.
Let me try a different tack. Would your model of banks and money apply equally well to (say) refrigerator producers?
The good part of the standard textbook story is the discussion of the double hot potato disequilibrium process. The bad part of the standard textbook story is the discussion of the equilibrium conditions themselves — M=(1/r)R.
Dan Kervick: Fridges to a fridge wholesaler (retailer?) are a buffer stock, just like money is to all of us. It’s a good comparison. The fridge retailer buy a fridge, not because he wants to hold it, but because he plans to pass it on to someone else. Just like all of us do with money.
But the fact that all of us do this with money, while only the fridge retailers do it with fridges, also explains where that metaphor must break down. The fridge retailer must persuade someone to hold an extra fridge if he wants to pass it on. We don’t have to persuade someone to hold an extra dollar to be able to pass it on.
There’s an unbroken circular chain of money retailers. The chain ends after a link or two with fridge wholesalers and retailers.
K:
“When you talk about demand for money it should be considered versus things that are almost identical (buying a t-bill or drawing from a line of credit) but which cant be used as medium of exchange
”
I pretty much subscribe, although with some reservations (transaction costs, psychological effects and so forth), to the idea that there is not much difference between bonds of different tenor and MZM since they can be more or less easily swapped one for the other.
Line of credit is obviously a different instrument from the above mentioned forms of money.
Therefore, for me, demand for money == “demand for wealth”.
Clearly, the CB activity can be, has been and will be somewhat distortionary to the extent of being partially fiscal even if unintended and thus satisfying some “demand for wealth”. But the fiscal effect of the feds monetary policy is not and cannot be substantial under the current rules of the road.
Mike: when a new loan is made by a bank (or when a bank buys something), the total stock of deposits is increased as a by-product. When, eventually, that loan is paid off (or the bank sells that thing again), the total stock of deposits declines as a by-product. I think your credit card story is consistent with my story. (But I’m not sure).
rsj: everything you say in that comment is compatible with (for example) the standard textbook ISLM model. The bond market clears, but the labour market does not clear (because prices and/or wages are sticky). But the ISLM model also says that (outside of a vertical IS or horizontal LM) an Open Market Operation can eliminate the excess supply of labour, while you say it can’t.
vjk: “Therefore, for me, demand for money == “demand for wealth”.
”
Then just say “wealth”. People (including those on this forum) have all kinds of conceptions of “demand for money”. I, for one, don’t agree that it doesn’t matter how much of your wealth is monetized. If you say people wish they were richer, that’s unequivocal. And I happen to agree the a shortage of wealth among some agents is at the core of our current troubles.
Nick,
Right, because in my thinking the effect of open market operations on the LM curve is different that the effect of those operations in your model.
So can we agree that this boils down to how we model institutions? For example, I’ve noticed in your comment to David that banks do not sell any bonds, which is why you are getting a hot potato effect of deposits among the public. In my model, banks have two “managers”, a cost of funding manager and a lending manager. The cost of funding manager notifies the lending manager how much it will cost the bank to raise funds to back any loan the lending manager makes. When the loan is made, then in the immediate short term, the bank has an extra deposit liability and an extra loan asset. But, as the customer withdraws that loan because he does not want to keep the deposit, the bank turns around and sells bonds, because it knows that it cannot attract enough new deposits. And so the public ends up holding a bond claim on the banking system rather than a deposit claim. This short circuits any hot-potato effect.
I don’t think there are any fundamental economic differences between us, and if there are, I’m certain you are right whenever there is disagreement.
The differences are in how the institutions work, and this has repercussions for which institutional interventions are effective and which are not.
One can imagine a model of financial repression, in which banks are allowed to fund themselves by rolling over overdrafts at the central bank. In that case, there would be no pressure for banks to sell bonds in response to deposit outflows. In that case, there would be a hot potato effect just as you describe. But that dynamic holds in only special circumstances, and those aren’t our circumstances. Just like one can imagine a commodity money institutional structure, or some other structure, each of which would lead to very different results of CB actions.
“And I happen to agree the a shortage of wealth among some agents is at the core of our current troubles.”
Bravo! Now if we could only convince everyone else that it is a shortage of wealth, and not money, then we might look for more effective solutions, such as redistribution, rather than quantitative easing.
Nick:
Correct, but of course if one loan is paid off as another is made, money supply is unaffected. One way I like to think of the law of reflux is to think of silver coins. If the current stock of coins is enough, and someone mints new coins anyway, then someone will find it profitable to melt those coins. That is, they will reflux to bullion just like unwanted silver spoons will be melted back to bullion. If someone instead printed new silver certificates, (100% silver reserves) then those certificates would also reflux to their issuers. Even if the certificates are backed by land, future taxes, etc, the reflux operates the same way.
By the way, in my “Law of Reflux” paper, I explain that reflux preserves the value of money, not because it limits the quantity of money, but because various channels of reflux provide the public with access to the assets backing the money. By ‘various channels’ I mean that money can reflux to the issuer in exchange for silver, for bonds, in payment of loans, payment of taxes, etc.
rsj: “I don’t think there are any fundamental economic differences between us, and if there are, I’m certain you are right whenever there is disagreement.”
I think someone has hacked into your computer! (Sorry, but this was a surprise to me, since we always argue!)
When a bank (or anyone) makes a loan, it gets an IOU in return. You could say it buys an IOU. A bond is just an IOU that is more marketable than most IOUs. So I think of buying bonds as equivalent to making a loan. So when I said “Suppose something causes a bank to want to make more loans”, I should be interpreted as saying “Suppose something causes a bank to want to buy more IOUs, including bonds”. And if the bank buys that IOU from a member of the public, it increases the total stock of deposits. And if it buys that IOU from another bank, it doesn’t. And if it buys that IOU from the central bank, the total stock of reserves falls.
When we talk about a bank making a loan and creating a deposit, what we really mean is that the bank is buying an IOU/bond from the public, not from another bank or from the central bank.
(That’s just the way my brain categorises these things.)
So in your example, where the bank decides to make a loan and fund it by selling a bond, that’s more of a shift in its preferred portfolio mix. There’s no change in the stock of deposits (provided the bond is sold to the public).
On the shortage of wealth (though this is off-topic): change in wealth = investment (roughly). If bad monetary policy causes (or fails to prevent) a recession, investment falls, and so wealth grows more slowly, or even falls. Alternatively, think of wealth as permanent income. Bad monetary policy causes recessions, which cause a fall in current and near-future income (and perhaps far-future income as well), which causes a (smaller) decline in permanent income, which is the same thing as a decline in wealth.
Mike: suppose the central bank buys and sells bicycles at a fixed price. Start in equilibrium. Then the CB doubles the price of bikes. People sell bikes to the CB, and get new money in return, even though they don’t want to hold new money, but want to buy skateborads and cars and taxi rides etc. with it. The new money hot potatos. Causing the price of everything (except bikes) to rise. As the relative price of bikes begins to fall, people start buying bikes back from the central bank. Eventually we get a new equilibrium with all prices doubled.
If prices are sticky (they are) it might takes years for the Law of Reflux to kick in.
The exogenous shock might have been a change in the productivity of bike producers. Or a fad for riding bikes.
Everyone: Now that world I have just described, where the supply of money is a by-product of the market for bikes, sounds utterly bizarre. Yet the world we live in is equally bizarre. The supply of money is a by-product of the market for financial intermediation. What nutter would ever have designed such a crazy monetary system??
Nick: “What nutter would ever have designed such a crazy monetary system??”
Wait! Let me think… Who earns all the the seignorage profits??
Nick, no when a bank buys a bond, it is not increasing its deposits.
You are mixing concepts here.
The first concept is balance sheet expansion — increasing both your assets and your liabilities. That is very different from purchasing an asset. Banks expand balance sheets all the time in response to non-bank loan demand, and what constrains them is not reserve requirements but regulation (e.g. capital requirements).
No bank ever increased its balance sheet because it had too many or too few reserves.
The second concept is portfolio shifts of bank assets and liabilities. Whether a bank’s liabilities take the form of deposits or bonds is not determined by the bank, but by the bank’s creditors.
Deposits are free money for banks (banks earn seignorage income from deposits). And the willingness of the non-bank sector to supply free money to banks is limited by their demand to hold deposit claims on banks rather than bond claims. Banks, of course, always want free money. Therefore its not a negotiation — the non-bank sector says “Yes, I have a loan for $100; You can earn seignorage income on $20, but for the remaining $80, I demand interest.” The banking sector says “Thank you for the $20” and then sells a bond for $80.
Deposits are a cost for the non-bank sector in the same way that reserves are a cost for the bank sector. But unlike reserves, the non-bank sector can increase and decrease its deposit claims on banks merely by purchasing other bank liabilities with their deposits.
The non-bank sector, by withdrawing deposits from individual banks and using the proceeds to purchase bonds, can force the banking sector as a whole to offer more net bond liabilities as they struggle to obtain non-goverment funding (which regulations require them to obtain except for very short periods of time (e.g. over-drafts).
In this way, there is no hot potato effect due to the public sector having too many deposits.
“So in your example, where the bank decides to make a loan and fund it by selling a bond, that’s more of a shift in its preferred portfolio mix. There’s no change in the stock of deposits (provided the bond is sold to the public).”
Absolutely not. When a bank sells a bond to the non-bank public, that decreases deposits by the same amount. Think through what is happening here.
Seriously, all of our disagreements versus monetary policy is due to the wierd model you have of how banks actually work, and how the underlying accounting works. This has nothing to do with economics, but with what sorts of kooky institutional structures you think are at work. That is why you keep getting push-back from knowledgeable people on here, and why in response to this push-back, you retreat to gold coins and models without any banks in order to defend your theories.
Nick:
The bank can sell a loan only if it can find a buyer who is able to buy (creditworthiness). When the bank sells a loan, it gets future labor the buyer sells for the bank’s money, intermediated, as it were, by the buyer employer (or some other source of “wealth”).
So, it’s not lack of the bank’s money that is the problem. The bank is more than willing to sell the loan, it just can’t find a buyer with an asset the buyer can swap for the loan. The real issue is the lack of sources of “wealth” that would make a potential loan buyer a real one.
The CB by itself cannot create new sources of wealth that the private sector is unable to manufacture, sua sponte, in the current environment for various reasons. Thus, the feds playing with differently painted pieces of paper can have only distortionary and largely psychological effect on market lemmings and speculators, as has been the case until now and most likely will be during and after the “operation twist” (swapping short bonds for long ones in a pointless attempt to lower the price of bank loans of certain tenors).
Nick,
I don’t get this:
“The medium of exchange is not like other assets. The suppliers of money can increase the actual stock without having first to increase the quantity demanded.”
So what, that is not in fact what they do. They lend to people who want to borrow at the rates offered – that is what they do do. And this will happen as long as the bank thinks the return is worth the risk. (And both the return and the perception of risk vary considerably with time.)
Nick, I agree that the banks can create an increased supply of deposit money, and that this might well be inflationary (depending on the monetary regime–certainly inflationary in a fixed monetary base regime.) I assume Glasner would agree too–as that effectively reduces the demand for base money.
Unlike the Tobin view, I don’t think Glasner’s view differs in terms of any testable implications.
Nick:
(I think this comment must have disappeared into your spam folder 2 hours ago.)
Here is the backing theory view of how the bike scenario would play out: The bank has assets worth 100 bikes as backing for $100, so each dollar is worth 1 bike. Then the bank prints and spends $200 on 100 more bikes. The bank now has 200 bikes backing $300, so 1 bike=$1.50. If this continued, then pretty soon the bank would have 1 million bikes backing about $2 million, for 1 bike=$2, but presumably the bank would come to its senses and devalue the dollar long before. So assume that the bank devalues to $1.50/bike. This is an equilibrium price, since $300/200 bikes=$1.50/bike.
But the community was used to having a total money supply with a real value of 100 bikes. Now they have a money supply worth 200 bikes, so 100 bikes’ worth of money ($150) will reflux to the bank. Now we are back to an equilibrium quantity of money.
But suppose the bank starts paying $1.51/bike. Bikes will flow in to the bank and dollars will flow out. Conversely, if the bank starts trading bikes at $1.49, then dollars will flow in to the bank while bikes flow out. By the time $149 has returned to the bank, and 100 bikes have come out of the bank, the bank will have no assets, and the one remaining dollar in the town will be unbacked and worthless. The money supply has disapeared, so trade will slow. Keynesians would call it a failure of aggregate demand, but I’d call it a money shortage, caused by the bank setting the price of money below equilibrium. Quantity theorists would be surprised that the reduction of the quantity of money did not cause deflation, so they would claim that prices and wages are downwardly sticky. But the backing theory says prices weren’t sticky at all. It’s just that the bank’s assets fell by more than the money supply.
Scott: “I agree that the banks can create an increased supply of deposit money”
Do you mean with or without a change in their lending rates? If they independently lower rates more people will borrow and there will be more money. But, presumably that’s an uncontroversial statement.