One more for the Banking School. This is a thought-experiment to help us clarify our thinking about banks.
If banks bought houses, instead of lending people the money for people to buy houses, what would be different? Not much.
But we students of money and banking would avoid some common mistakes, like confusing the demand for money with the demand for loans. And we would see that financial intermediation has no necessary connection with money.
Suppose there were some prohibition (motivated by religion or politics or whatever) against banks charging interest on loans. But no prohibition on banks charging rent on houses they own. So banks stop making loans and start buying houses instead.
The asset side of banks' balance sheets would be diffferent: banks would own lumpy and illiquid houses instead of lumpy and illiquid mortgages on houses. Banks might need higher capital ratios, because houses are normally riskier than mortgages on houses. But the liability side of banks' balance sheets could be exactly the same as now. Some of the things on the liability side, like chequable demand deposits, could still be used as money, just as they are now.
People would pay rent to the bank, instead of paying interest to the bank. So banks would earn rental income from their assets instead of interest income. The revenue side of their income statements would look a little different, but banks would still make their living from the spread between the low yield on their liquid liabilities and the high yield on their illiquid assets.
So not much would change.
But it would be a lot simpler to teach money and banking.
Instead of: 1. me promising to repay a loan from the bank; 2. the bank crediting my chequing account for the amount of the loan; 3. the bank debiting my chequing account when I buy the house; 4. and the bank crediting the house seller's chequing account, — we can skip all but the last step. The bank buys the house by crediting the house seller's chequing account. Period. What happens to the money after the house seller gets it is exactly the same in both cases. And that's the important part. That money keeps on circulating around the economy. Until the bank sells a house, which destroys the money it had previously created.
It is very easy for students of money and banking to get confused between the demand and supply of money and the demand and supply of loans of money. It would be very hard for students to get confused between the demand and supply of money and the demand and supply of houses. But the two would nevertheless be related: the demand for houses by banks would be exactly the same as the supply of money by banks. (You can define them both as stocks, or both as flows, whichever you find more convenient.)
But an excess supply of money by the non-bank public wouldn't necessarily mean an excess demand for houses by the non-bank public. There are lots of other things you could buy if you wanted to hold less money.
If people wanted to hold more money, a good monetary policy would try
to ensure that this resulted in banks wanting to hold more houses. Otherwise we would get recession and deflation, as people tried to hoard money.
And
if people didn't want to hold more money, a good monetary policy would
try to ensure that banks didn't want to hold more houses. Otherwise we would get boom and inflation, as people tried to spend away the excess money.
If banks both bought and sold houses, and announced a price (or price/rental ratio, or rental yield) at which they were willing to buy or sell unlimited quantities of houses, we know that the housing market would always clear. The stock of houses demanded by the non-bank public at the price set by banks would always equal the stock of houses they actually owned. If the non-bank public had an excess demand for houses, they would immediately buy them from the bank. If the non-bank public had an excess supply of houses they would immediately sell them to the bank. But that wouldn't mean the stock of money demanded by the non-bank public would always equal the stock of money they actually owned. The market for houses is not the same as the market for money. The market for houses is just one of the many markets for money.
But if banks bought houses, instead of lending people the money to buy houses, then banks, strictly speaking, wouldn't be banks. They would act almost exactly like banks, but we wouldn't be allowed to call them "banks". Because banks, by definition, are financial intermediaries whose liabilities are used as money. And a financial intermediary both borrows and lends, and the "banks" in my thought-experiment don't lend. They buy houses instead. They create money not by making loans, but by buying houses.
But if the "banks" in my thought-experiment act like banks, but don't meet the strict definition of being banks, maybe we should change the definition? Does a bank that is left holding the houses when all its mortages default, but is still a going concern because it has adequate capital, suddenly stop being a bank, by definition? That isn't a very useful definition.
Maybe we should change the definition to: a "bank" is anything that creates money. What it spends that money on is immaterial. It could buy houses, buy financial assets backed by houses, or it could just give it away to charity, or by dropping money out of a helicopter.
What should our slogan be? "Loans create deposits!"? Maybe. "Buying houses creates deposits!"? Maybe. "Creating deposits creates deposits!". Better. "Creating money creates money!". Yep, that's it.
The one truly important and puzzling thing about banks is that people are willing to use their liabilities as a medium of exchange. But that's equally important and puzzling about any thing that creates money, whether it's a financial intermediary or not.
Worthwhile Canadian Initiative 
Tom B: say a used car dealer likes to keep an inventory of around 1 month’s sales of cars on the lot, so the average car is sold one month after he buys it. His desired velocity of circulation of cars is 12 times per year. The desired velocity of circulation of money is the same thing, except it’s money not cars. The car dealer only buys cars because he wants to sell them again. We only buy money because we want to sell it again. We are all dealers in money.
Learn from Mike Sproul. You can learn a lot from him, even though he’s totally wrong about money!
Nick, thanks. How about my questions 2 and 3 above? Am I on the right track there?
Nick,
You say:”If banks set the price of houses, and buys whatever quantity of houses the public wants to sell, the increase in the quantity of money is determined by the quantity of houses the public wants to sell. But the quantity of houses supplied by the non-bank public is NOT the same as the quantity of money demanded by the non-bank public.”
Is it not so that the newly created money by the banks is just added to the already preexisting stock of money? If I translate that to cars I interpret your answer like this: The increase in the number of (new) cars is determined by the quantity of money the public wants to spend. But the amount of money spend by the non-bank public is NOT the same as the number of cars demanded (driven) by the non-bank public.
Still, it would look to me like it is the car buyers who decide how many cars they want to buy at the price that was set and less the car manufacturer how many he wants to sell (assuming he makes a profit at that price). The thing is no one has to give up their old car just because someone is buying a new one (even more so for money than cars).
Alex,
I would argue that Qs* (given a price, the theoretical quantity that suppliers are willing to sell) is for bank deposits (money) always larger than Qs/Qt/Qd. Banks will always be willing to create a low-yield liability to receive a high-yield asset as from that spread (and fees) they generate their income. If tomorrow 1 million homeowners go to their bank and demand a mortgage or home-equity loan the banks will be happy to comply as long as the loan risk is perceived to be small.
Tom B: 2 and 3 yes.
Odie: now imagine a world in which only car dealers buy cars, because they plan to sell them to other car dealers. Nobody buys a car because he plans to drive the car. We use cars as money. Nobody buys money because he wants to drive it, or look at it.
Nick, other question: Would you say in your picture here that a house is the equivalent of an Odie or Nick Rowe IOU in real life?
Nick, now I am getting confused (and really hope also for an answer for my preceding post to clarify that). In the sentence I made from yours, cars were sold by the car manufacturers (=banks) to the public. However, you say:”..only car dealers buy cars, because they plan to sell them to other car dealers.” Do you want to say that everyone is a car dealer, now? That would be fine, so nobody drives the cars but instead they exchange the cars for some other good and service (equivalent to money as you said next:”We use cars as money.”) Once the car manufacturer buys the car back (exchanging the car for “money” again) the car gets destroyed. The total stock of cars goes down. If that is correct why do you think only the car manufacturer and not the public decides how many cars are around? Especially as the car manufacturer receives rental fees for his cars meaning he has a vital interest to sell as many cars as there is demand.
Someone above said it seemed like double counting to include both the debt and deficit in my graph of things available to contribute to hyperinflation. I was not clear on why I did that. I was sort of thinking about potential monetization over the coming year when I made the graph. I rewrote things some to try to make that more clear. Thanks.
http://howfiatdies.blogspot.com/2013/09/usa-hyperinflation-risk.html
I’ve given this some more thought and come to following conslusion:
The dispute is between those who consider money emission a result of a tripartite agreement (bank + 2 customers) and those who see it as a purely bilateral affair (bank +1 customer).
In the tripartite view, we have a buyer and a seller who have previously agreed on a price for the purchase of some product but the buyer has no, or wishes not to spend, previously earned money (note: the bank does not set the price!). The bank credits the seller’s account and takes on the buyer as a new creditor. It functions as underwriter in the name of the currency issuer and demands interest payment to cover its costs – which include potential losses – and to hopefully generate a profit. Money, in this case, is a relative measure of that which the seller has but the buyer has not yet produced. All money in circulation is thus a mirror image of promises made but not yet honoured. It is at once positive and negative. Money = debt. Money in this sense is also generic, in that the seller may use it to buy any good or pay his own debts and the buyer may earn it back through any (legal) method he/she wishes. Only with earned money (income), can debts be repaid and money thus be destroyed.
In the bilateral case, the bank basically emits its own equity to buy the product from the seller. There needn’t be a buyer (of output). Money = bank equity. But equity is always a specific claim, a property right in a company, not the economy as a whole. This view would also imply that money, so defined, is junior to all other bank liabilities. It doesn’t make sense.
Odie: “Nick, other question: Would you say in your picture here that a house is the equivalent of an Odie or Nick Rowe IOU in real life?”
Yes. That was the whole point of my post. To imagine a world in which banks bought houses instead of buying IOUs signed by Odie or Nick Rowe.
What I said about cars in my 9.43 comment probably wasn’t clear. Forget it. I was trying to draw a distinction between money and all other assets, like cars.
If someone offered you a car for sale, but said you were never allowed to sell it again, would you still buy it? Yes, maybe, at a reduced price, because you would still want to drive it.
If someone offered you a $20 banknote for sale, but said you were never allowed to sell it again, would you still buy it? Of course not. Nobody would buy it. It’s worthless.
Thanks Nick, When the houses in your picture will be the same as an Odie IOU should it not also have the same characteristics? I would suggest the following points:
1. I can build a house any time but it will be useless to me. I derive no utility from it. The only thing I can do is to exchange it for money. I can find a non-bank investor but there will be no money creation, only transfer of deposits. That changes when I sell the house to a bank. A bank will expand its balance sheet and create new money.
2. Now I can live in the house. Every rent payment will destroy some of the money the bank gave me while some part of it will be income for the bank.
3. Once I have paid an agreed upon total the bank and I will burn down the house which relieves me of the obligation to pay rent anymore.
4. The bank can sell the house at any time but I will still be the only person who can live in it. To be exact, the bank is not selling the house but the right to collect rent payments from me. As you said above:” …banks would still make their living from the spread between the low yield on their liquid liabilities and the high yield on their illiquid assets”. Any house buyer will not be interested in the house but in the rent payments. A house without a “tenant” is like a defaulting mortgage: a loss to the bank.
Would you disagree with any of the points? Looking at it from this angle: Could banks still buy and sell any quantity of “houses” upon a set price? Or would it not require them to have enough “tenants” available who have a demand for money in exchange for future rent payments in order to be able to create money? So who would control the quantity of money created: solely the bank or also the number of “tenants”?
Tom:
“what do YOU think that Nick meant by “desired velocity?” (see my comment above at October 01, 2013 at 11:09 PM). How about questions 2 and 3? I know that’s not your thing, but in your estimation am I on the right track there re: Nick’s argument?”
“Desired velocity” comes from MV=Py, and the usual idea is that if people start spending their money faster, then V rises, and P will rise as a result, even without any rise in M.
MV=Py is a tautology. It just says ‘money spent’=’money received’. It’s just as “true” of GM stock as it is of money. Let M= # of shares of GM, V=#times/year a share is ‘spent’ buying goods, P=how many shares it takes to buy a unit of goods, and y=goods bought with GM shares. Does that tell you anything about the value of GM shares? Of course not. Value of GM shares is determined by GM’s assets and liabilities (including expectations of future assets and liabilities). Same for paper money.
“2. I’m trying to imagine the perfectly interest-elastic demand curve: x-axis is quantity of money (M), y-axis is interest rate, and the demand curve is horizontal, intersecting the y-axis at some interest rate. Is that correct? So a small change in interest rates causes unbounded changes in the quantity of money demanded?”
When economists estimate the interest-elasticity of desired money-holdings, they get a value close to zero. That’s certainly true of me. I pay no attention at all to the interest rate when I’m deciding how much cash to carry and how much to keep in a checking account. Think of an old-fashioned paper dollar that is convertible into 1 oz of silver. That dollar will be worth 1 oz regardless of the interest rate, and regardless of money demand or money supply. The modern paper dollar is no different. They’ve just suspended metallic convertibility (while maintaining tax convertibility, bond convertibility, possible future metallic convertibility, etc).
“3. So in your Zimbabwe example we don’t see a curve like I’ve sketched in 2.: the quantity of money demanded was not sensitive to small changes in interest rates. Is that it?”
You might have guessed from above that in my world, money demand and interest sensitivity are irrelevant to money’s value.
Mike, thanks. I’ll mull that one over.
Houses and money differ in many ways. If you want to focus on the special features of money, perhaps it would help to imagine that money is created out of an asset that has the same investment characteristics as money (e.g. overnight repo), except for not being a medium of exchange. Such a bank wouldn’t even need capital, because there’s no mismatch between its assets and liabilities.
Then you could go on to discuss financial intermediation, without any reference to banking or money.
Make sense or not?
(yet another delayed response)
“[ME] “But as noted above (by somebody?) the equity of homeowners is critical for banks in the real world.”
{Nick Rowe] I wonder. If homeowners put their equity into bank shares instead of houses, banks could then buy the houses because they would have the capital cushion to offset the riskier asset. (There’s some sort of Modigliani-Miller theorem at the back of my mind).”
My point why it matters is the part about behaviour of renters versus owners. If you trash your rented house, that’s a small impact for the hypothetical bank-landlord, and so it not really be much of a cost, even if you own shares in the bank. Of course, if the “Representative Household” did that, things would be different…
“[Nick] My guess is that’s the important thing. But then renters do exist (because there are various offsetting benefits, like if you move around a lot, or don’t have much equity), so why don’t banks merge with companies that rent apartments?”
Theoretically they could, but the convention is that banks generally hold assets that they can keep at par on the books (or else they have to hold massive equity reserves against those assets). You would need to change that practice.
In the real world, investors tend to dislike conglomerates (or at least they should). The idea is that renting apartments and banking are very separate; one is “financial intermediation” with “limited” capital needs, the other is taking a capital-intensive long-term real asset positions and running an operating company. (OK, banks need to build up a payments systems and bank branches, but that is more of a barrier to entry and eventually becomes a very small part of the balance sheet.)
Investors would probably distrust the earnings of the property arm (there’s no way of telling if it is being subsidised by the parent bank), and so it’s earnings would have a low multiple. This means that the effective cost of capital is very high for that line of business.
That said, the Canadian banks are now financial conglomerates, but I think that they generally are in areas which are much less capital intensive than owning real estate.
Max: I think that makes sense. Presumably though, banks can make higher returns if they can hold very illiquid assets, because that would give them a higher spread? And illiquidity and risk and longer maturities tend to go together? And maybe this is why banks tend to be the way they are?
Brian: I think we are roughly in agreement.
Tom Brown, how about this?
So if the risk weighting is 1 for houses/capital requirement is 100% for houses, then people can still borrow money from the bank to help buy a house. The bank would be limited by the amount of money saved as bank capital.
This would be similar to loanable funds.
Too Much: yes, that sounds right.
“People would pay rent to the bank, instead of paying interest to the bank. So banks would earn rental income from their assets instead of interest income. The revenue side of their income statements would look a little different, but banks would still make their living from the spread between the low yield on their liquid liabilities and the high yield on their illiquid assets.
So not much would change.
But it would be a lot simpler to teach money and banking.
Instead of: 1. me promising to repay a loan from the bank; 2. the bank crediting my chequing account for the amount of the loan; 3. the bank debiting my chequing account when I buy the house; 4. and the bank crediting the house seller’s chequing account, — we can skip all but the last step. The bank buys the house by crediting the house seller’s chequing account. Period. What happens to the money after the house seller gets it is exactly the same in both cases. And that’s the important part. That money keeps on circulating around the economy. Until the bank sells a house, which destroys the money it had previously created.”
1) I believe the rental income paid to the bank would destroy demand deposits.
2) I might sell the house because I believe the bank is overpaying for the house. I may just hold the demand deposit and wait for the price to come back down. The demand deposit will have zero velocity in the real economy.
3) I am pretty sure mortgages are more liquid than the houses themselves.
test
Lookup
zyxzyxooxwo
http://economix.blogs.nytimes.com/2013/10/14/shiller-on-boom-bust-and-human-nature/?_r=0
Robert J. Shiller
“I see no signs that home buyers have learned the lesson I tried to convey in the second edition of my book “Irrational Exuberance” in 2005. That message was that existing-home prices have shown virtually no tendency to trend upward in real, inflation-corrected terms over the last century. While land is limited, it’s only a small component of home value in most places. New construction often brings down the value of older homes, which wear out and go out of fashion, dragging down prices.”
zyxzyxooxwo
Comment(s) in spam?
Comment(s) not displaying???
Thanks for getting the comments to display!