[This is aimed at intermediate-level students. It's "big picture" stuff, and doesn't go into all the institutional details. And it's already too long.]
Banks are financial intermediaries that create money. Banks are special because money is special. If we used cows as money, then dairy farms would be special, because dairy farms would create money.
Why is money special (and important)? Money is the medium of exchange (we buy and sell all other goods for money). Money is the medium of account (we quote the prices of all other goods in money). (The store of value function doesn't make money special, because lots of goods are stores of value.)
If there were a fall in the supply of apples, the price of apples would need to rise to clear the market for apples.
But if we used apples as the MOA, apples would not have a price of their own; the price of apples would be the reciprocal of the price of everything else. So the price of everything else would have to fall instead. And that's hard, because some prices are sticky, which means the price of apples would be sticky. And some prices are stickier than others, so a fall in the supply of apples would distort relative prices across the whole economy.
If the price of apples were sticky, a fall in the supply of apples would create an excess demand for apples. Some people would be unable to buy as many apples as they want. But if we used apples as the MOE, apples would not have a market of their own; the market for apples would be the market for every other good. An excess demand for apples would mean people would want to sell more other goods and buy fewer other goods. But if people are buying fewer other goods, they would actually be able to sell fewer other goods than before, and this would mean they would buy even fewer other goods, and so on. If barter is hard, which it is, an excess demand for the MOE disrupts trade of all other goods. It causes a recession.
That's why money is special and important. Changes in the supply and demand for money have macroeconomic consequences.
There are three types of goods that might be used as money: 1. goods that would be valuable in their own right even if they weren't used as money; 2. IOUs, or promises to pay goods that are valuable; 3. none of the above, or goods that are valuable only because they are used as money.
Suppose people like wearing shells as jewelry. That creates a demand for shells. If the supply of shells is not perfectly elastic at a zero price, demand and supply means shells will be valuable. People might use shells as money, as well as for jewelry. That's an example of the first type of money. But notice that the demand to use shells as money adds to the demand to use shells as jewelry, so this monetary demand makes shells more valuable than they would otherwise be.
Suppose now that fashions change, and people stop wearing shells as jewelry, and only the monetary demand remains. People now hold shells in their pockets, and not around their necks. Demand is less than before, but there's still a demand for shells, so they can still be valuable, though not as valuable as before. That's an example of the third type of money.
But notice there's also a second equilibrium, where shells are worthless, and so can't be used as money, and so there is no monetary demand for shells. The demand curve for shells slopes down, but suddenly drops to zero demand when the price of shells drops to zero, so the demand curve crosses the supply curve twice.
What keeps us in the first equilibrium and not the second equilibrium? I think the answer is "network externalities and custom". If everyone else I trade with is using shells as money, I will want to use shells as money too, because it will be easier to buy and sell things for shells. This means it is hard to change what we use as money, because we either all change at once or not at all, and it's hard to coordinate everyone changing at once. Just like languages or computer programs.
Suppose we are using shells as money. If I write "IOU 10 shells, signed Nick Rowe" on a sheet of paper, I might be able to persuade my local supermarket to give me some apples in exchange for my IOU, if they know me and trust me. But if my local supermarket redeems my IOU with me next week, then my IOU is not money. I have simply postponed payment for the apples. But if the supermarket uses my IOU to buy something else, from someone who uses it to buy something else, and it keeps on circulating around my village because everyone in the village trusts my signature, then my IOU is money, because it is being used as a medium of exchange.
I have now created money. But I am not a bank. Because a bank is a financial intermediary that creates money, and I am not a financial intermediary.
"Liquidity" is hard to define precisely, but it means how easy it is to buy or sell an asset quickly at a good price with low transactions costs. Some assets are more liquid than other assets. Other things equal, people prefer holding more liquid assets to less liquid assets. So in equilibrium the more liquid assets will pay lower rates of interest (or lower rates of return). Money is the most liquid asset, so money will pay a lower rate of interest than other assets. (Bank of Canada currency pays 0% nominal interest and usually pays around minus 2% real interest.)
As long as my IOU for 10 shells continues to circulate around the village rather than being returned to me for redemption, and if my IOU pays 0% interest, the village has, in effect, made me an interest-free loan.
If I can lend at a higher interest rate than I can borrow, I might decide to become a financial intermediary, which means I both borrow and lend, and make an income off the spread between my borrowing and lending rates.
If I can create this spread by creating money, I might decide to become a bank, which means a financial intermediary whose liabilities are used as money.
So if you want to borrow 100 shells to buy a bike, you give me your signed IOU for 100 shells plus 5% interest, and I give you my signed IOU for 100 shells plus 0% interest. You take my IOU to the village bike store, and use it to buy a bike. If my IOU continues to circulate around the village, I am now a bank.
My IOUs are paper currency, convertible on demand at a fixed exchange rate into shells. Paper currency is like a "bearer bond", where simple possession of the bit of paper is normally considered sufficient evidence of ownership. There is no central record kept of who owns what bits of paper with which serial numbers. But if I sell you my car, we have to tell the government the VIN and your name as the new owner, and the government keeps records. This helps prevent theft of cars. Cars are not like bearer bonds; if the police stop you, you have to show the ownership paper, otherwise they'll think you stole it.
If I wanted to prevent theft of my IOUs, I might keep a record of the serial numbers on my IOUs and who owned them, and require you to inform me that the ownership of my IOU serial number 12345 had been transferred from you to the bike seller. And the bike seller would want to make sure you had informed me of this. To make it easier for you to do this, I might give you a form letter, where you could fill in the name of the bike seller, then give the letter to the bike seller, who could then mail it to me. That way, I know the bike seller now owns the IOU, and he knows I know this.
But if I'm keeping a record of who owns which of my IOUs, and since the only thing that matters is how many shells' worth of my IOUs different people own, we can dispense with the sheets of paper on which I have written "IOU 10 shells signed Nick Rowe" altogether. As long as you trust me to keep good records. I simply give you a stack of form letters, stapled together for convenience. We call those form letters "cheques". You now have a chequing account instead of currency, and you give the bike seller a cheque instead of currency.
And if I keep my records on a computer rather than in a book, and if the form letter gets sent by email rather than by Canada Post, and if my computer can understand those email letters, we have electronic money. Currency, chequing accounts, e-money: these are all basically the same. Some are just more convenient and trustworthy than others. But on the other hand, money is all about convenience and trust. We wouldn't use money if it weren't more convenient than barter. And if everyone's IOUs were trusted, anyone could create money. Or we wouldn't even need money, except as a convenient way to keep track of the value of goods we have taken from other people and given to other people, so we don't take more than we give, or give more than we take.
If using my IOUs as money is sometimes less and sometimes more convenient than using shells as money, some people will be redeeming my IOUs for shells (withdrawing shells from my bank), and other people will be buying my IOUs for shells (depositing shells in my bank). If a lot of people want to redeem my IOUs at the same time, I might run out of shells. In principle, as long as I am solvent, I could always redeem all of my IOUs by selling the IOUs signed by other people that other people have sold me, when they borrowed from me. But those IOUs will be illiquid. Because the whole point of my being a bank is to make an income from the interest rate spread between my liquid IOUs and your illiquid IOUs. So if I had to sell those illiquid IOUs quickly, I would make a loss, and that loss might be big enough to make me insolvent, so I couldn't redeem all my IOUs quickly. And if people think that might happen, there might be a run on my bank, where everyone tries to redeem before I run out of shells.
If I saw that I would run out of shells, I would suspend convertibility of my IOUs into shells. I would break my promise to redeem my currency on demand at a fixed exchange rate. But if people believed there was some positive probability I would restore convertibility in future, or even if they they expected me to restore convertibility at a lower exchange rate (paying less than 100% of the shells promised), my IOUs would still have some value, even if that value would be lower. But if my IOUs have some value, and they are convenient to use as money, people might still use them as money, so there would still be a demand for my IOUs. And if there is a demand for my IOUs as money, then provided I don't create an unlimited supply, there will be an equilibrium in which my IOUs are valuable even if I renouce any promise of convertibility. (Though there is also a second equilibrium in which my IOUs are worthless and so cannot be used as money and so are not demanded.)
This case of permanent suspension of convertibility is analogous to the case where people use shells as both jewelry and money, and then fashions change so the demand for shells as jewelry disappears, but people keep on using the shells as money.
If I permanently suspend convertibility of my IOUs into shells, my IOUs are now just sheets of paper with ink marks on them. All the number "10" means is that I will convert one "10" into two "5"s or vice versa. If people keep using my sheets of paper as money, I am now a central bank issuing a fiat money with a floating exchange rate. My central bank might get taken over by the government, or it might not. It might depend on whether the government thinks I will do what it wants the central bank to do. Or whether the government wants the profits I can earn.
What happens when there are two (or more) banks? If you bank with Nick, and you buy a bike, and the bike seller uses Mary's bank, then the bike seller sends your cheque to Mary's bank, Mary credits the 100 shells to the bike seller's chequing account, and then brings the cheque to me, and I give Mary 100 shells, and deduct 100 shells from your chequing account. But at the same time, someone else who banks with Mary may have bought something for 90 shells from someone who has an account at my bank. So when Mary and I settle our debts to each other that evening, I only give her 10 shells.
If both banks have lots of customers, and if both banks are doing the same things, the Law of Large Numbers says that most evenings those net transfers of shells will be small in proportion to the gross transfers between me and Mary, and they will average out to zero over time. If Mary and I don't want to keep a lot of shells in reserve to make these payments between us, we could instead just borrow from each other in the overnight market. Instead of me giving Mary 10 shells, she accepts my IOU for 10 shells, but since Mary doesn't need my IOU to use as money, she charges me interest at the overnight rate of interest on loans between banks.
But if Mary succeeds in attracting some of my depositors to switch to her bank, or if I succeed in attracting some of Mary's borrowers to switch to my bank, I will be making net payments of shells to Mary, and unless I want to keep on borrowing from Mary in the overnight market, I won't want this to happen. And I won't want to keep on borrowing from Mary, because Mary charges me a higher rate of interest than I pay my depositors.
This means I won't want to expand my loans without expanding my deposits, and Mary won't want to expand her loans without expanding her deposits. But if we both expand our loans together, and if we are the only two banks, our deposits will expand in line with our loans, and we won't have to worry about this. Unless people withdraw deposits from both our banks because they would rather hold shells instead.
Most countries, like Canada, have several commercial banks that accept deposits from the public and make loans to the public. And a single central bank. The central bank is usually owned by the government, and is an arm of public policy. The commercial banks are usually not owned by the government and can be assumed to maximise profits. The central bank usually has a monopoly on issuing paper currency, but does not normally make loans to the public or accept deposits from the public. It may however, make loans to and accept deposits from the government.
What is the important difference between central banks and commercial banks? What makes a central bank central? What makes central banks powerful? I think there are two important differences.
1. Central banks act as bankers to the commercial banks. [Aside: I'm not satisfied with this section.]
If there are only two banks, mine and Mary's, we don't need a bank to do our settling of accounts every evening. But if there are three or more banks, it gets more complicated. If I owe Mary 100 shells, and Mary owes Harry 100 shells, and Harry owes me 100 shells, we would need all three of us bankers to meet together every evening and do a deal where we agree that none of us owes anything to either of the others. That three-way deal is like a three-way barter deal where a hairdresser gives a haircut to a manicurist, who gives a manicure to a masseuse, who gives a massage to the hairdresser. It's hard for all three to get together and do three-way deals. That's why they use money instead of barter. It's more convenient. And they don't all have to trust the other two to deliver what was agreed on; each needs only to trust one.
It's more convenient, and requires less mutual trust, if each commercial bank has a chequing account at the central bank, and lets the central bank cancel out the circles of obligations. So the central bank acts as banker to the commercial banks, and central bank money is the money the commercial banks use to make payments between themselves. Just as the public uses commercial bank money to buy and sell different goods, so commercial banks use central bank money to buy and sell different commercial bank monies. And just as the public can get an overdraft at a commercial bank to cover unexpected needs for liquidity, so commercial banks can get an overdraft at the central bank to cover unexpected needs for liquidity. The central bank acts as lender of last resort to the commercial banks.
2. Commercial banks promise to redeem their monetary IOUs at fixed exchange rates for central bank monetary IOUs. It's not the other way around. It's the Bank of Montreal's job to make sure a Bank of Montreal dollar is worth the same as a Bank of Canada dollar. Only in an emergency, if there's a run on the Bank of Montreal dollar, will the Bank of Canada act as lender of last resort to make sure the Bank of Montreal dollar doesn't depreciate against the Bank of Canada dollar.
This is analogous to what would happen if the US Fed decided to have a fixed exchange rate between the US dollar and the Canadian dollar, but the Bank of Canada took on no responsibility to keep the exchange rate fixed. The Bank of Canada could do what it liked, and the Fed would have to follow. If the Bank of Canada wanted to create higher inflation and make the Canadian dollar worth less, the Fed would have to make the US dollar worth less too. US monetary policy would be set in Ottawa. Since the Canadian commercial banks fix their exchange rates against the Bank of Canada dollar, Canadian monetary policy is set in Ottawa, by the Bank of Canada. If the Bank of Canada wanted to create higher inflation and make the Bank of Canada dollar worth less, the Bank of Montreal would have to make the Bank of Montreal dollar worth less too.
Right now, the Bank of Canada: charges commercial banks 1.25% interest on overdrafts (negative balances); pays commercial banks 0.75% interest on deposits (positive balances); and has a target of 1.00% for the overnight rate on loans between commercial banks. It nearly always keeps these same 25 basis point spreads between the three rates. Actual positive or negative balances are usually very small. The actual overnight rate cannot be higher than the 1.25% bank rate, because otherwise commercial banks would borrow from the Bank of Canada instead; and cannot be less than the 0.75% deposit rate, because otherwise commercial banks would lend to the Bank of Canada instead. The Bank of Canada keeps the actual overnight rate very close to the target by adjusting, or threatening to adjust, the supply of settlement balances up or down to lower or raise the actual overnight rate. If the Bank of Canada wants to make its dollar more valuable (because it thinks inflation will rise above the 2% target unless it does something different) it raises all three interest rates. That's how the Bank of Canada stops all the commercial banks expanding their total loans and deposits too much. If it wants to make the Canadian dollar less valuable (because it thinks inflation will fall below the 2% target unless it does something different) it lowers all three interest rates. That's how the Bank of Canada stops all the commercial banks expanding their total loans and deposits too little.
You might say the Bank of Canada dollar is an inconvertible fiat money, of the third type. Or you might say that, because the Bank of Canada targets 2% CPI inflation, the Bank of Canada dollar is like an IOU that is indirectly convertible at a slowly depreciating exchange rate into the CPI basket of goods, and is thus a money of the second type.
[Obviously, there's more to say about banks, but for now I'm stopping here.]
"The central bank is usually owned by the government"
Are there any countries in which this isn't the case?
Posted by: Philippe | September 18, 2013 at 11:23 AM
Philippe: I'm trying to remember when the Bank of England was nationalised. I don't know if there are any now. (Some are de facto but not de jure owned by the government, like the Fed, which creates great fun for conspiracy theorists, though I tell them that the Bank of Canada too is actually owned 100% by a little old lady who lives in England). Someone else could give a much better answer than me, because I don't have a memory.
Posted by: Nick Rowe | September 18, 2013 at 11:48 AM
Great explanation, but I don't agree that the store of value function does not make money special.
Let's take again your own examples (where you are the one people trust, while I'm not).
If you want to buy a bike, you issue a 100-shell IOU (interest free). You're the final borrower. The bike seller is the final lender.
If I want to buy a bike, I give you a 5% interest 100-shell IOU, and you give me a 0% interest 100-shell IOU. I'm the final borrower, and the bike seller is the final lender. You are the intermediary.
In both cases, the transaction (buying the bike) is credit-based. The bike seller grants a loan to the buyer. He will be "repaid" when he exchanges the IOU for beer (passing the loan to the brewer).
So, money (the IOU) supposes the loan will in the end be repaid in goods (as long as it is repaid with IOUs, it still is a loan), and there is a stored value that is transferred through time. But it's not an intrinsic value (as would be the case with goods that store value, such as grain, land, etc.). The value depends on the ability of the borrower to repay the loan (i.e. to produce economic goods or services equivalent to a bike). If there is inflation, the IOU will have lost value. And if the borrower is dead, there will be no repayment at all. At the scale of a country, money can lose value if the national economy is in a serious downturn. That certainly makes money special compared to other value-storing goods.
And that specificity of money may matter when you consider savings. You can save by investing in securities or capital goods, or you can save money (i.e. lending to other people, through the financial intermediary of banks). But then, you need to be confident that those other people (the next generation) will be able to repay, which might not be the case if the economy is a wreck, if there is a decline in demography or if that generation faces costs you delayed to it (such as climate change for example).
Money is fascinating. By mostly ignoring money, neoclassical economics not only lost track of the most important economic variable, it also made the discipline boring.
Posted by: Zorblog | September 18, 2013 at 11:54 AM
The Bank of England was nationalised in 1946. The Fed describes itself as follows:
"The Federal Reserve System fulfills its public mission as an independent entity within government. It is not "owned" by anyone and is not a private, profit-making institution."... "the Federal Reserve can be more accurately described as "independent within the government" rather than "independent of government."... "The Federal Reserve, like many other central banks, is an independent government agency but also one that is ultimately accountable to the public and the Congress."
The Fed describes the Reserve Banks as follows:
"The 12 regional Federal Reserve Banks, which were established by the Congress as the operating arms of the nation's central banking system, are organized similarly to private corporations--possibly leading to some confusion about "ownership." For example, the Reserve Banks issue shares of stock to member banks. However, owning Reserve Bank stock is quite different from owning stock in a private company. The Reserve Banks are not operated for profit, and ownership of a certain amount of stock is, by law, a condition of membership in the System. The stock may not be sold, traded, or pledged as security for a loan; dividends are, by law, 6 percent per year."
http://www.federalreserve.gov/faqs/faq.htm
Posted by: Philippe | September 18, 2013 at 12:13 PM
It wouldn't make sense for a government/bank to allow its currency to become worthless when it can effortlessly prevent it. So the only cases where a money is permenantly inconvertible (as opposed to inconvertible for an indefinite period of time) is when the government/bank is broke or ceases to exist. These are the same cases where convertible money becomes worthless.
Posted by: Max | September 18, 2013 at 12:35 PM
Nice. Teaching notes on banks and money... or, put differently, the last 5 years of monetary blog debates in a nutshell.
Posted by: JP Koning | September 18, 2013 at 01:46 PM
before countries went off the gold/silver standards, central bank currency was normally redeemable for Treasury Coin, or bullion, right? (unless specie payment was suspended by the government)
Posted by: Philippe | September 18, 2013 at 02:34 PM
Zorblog: Thanks!
But all IOUs, like stocks and bonds, and not just money (which may or may not be an IOU) may fluctuate in value according to the fortunes of the debtor.
Canoes are a store of value too, and if they weren't stores of value (if they fell to pieces 5 minutes after you put them in the lake) they wouldn't function as canoes. But being a store of value doesn't make canoes special, and isn't what defines them as canoes.
"Money is fascinating. By mostly ignoring money, neoclassical economics not only lost track of the most important economic variable, it also made the discipline boring."
I find money fascinating too. And weird. But I don't think it's fully accurate to say that neoclassical economists ignore money, even though some neoclassical models, like Walrasian General Equilibrium models, do ignore money. many economists from the neoclassical tradition have spent a lot of time thinking about money, and building it into their models, with varying degrees of success.
Philippe: Aha! My memory is not so totally bad after all!
Max: well, Zimbabwe allowed its currency to become worthless. Maybe it wasn't easy to prevent it, all things considered? But you can have a money that is not convertible into anything *at any fixed exchange rate* and yet not worthless.
JP: Thanks! LOL, yep, it does read like that. Now I can tell my Dean that my blogging really does help my teaching. Because I couldn't have written this if I hadn't been writing and reading blogs and comments over the last few years.
Posted by: Nick Rowe | September 18, 2013 at 02:41 PM
"many economists from the neoclassical tradition have spent a lot of time thinking about money, and building it into their models, with varying degrees of success"
Do you have any specific good examples?
Posted by: Philippe | September 18, 2013 at 02:47 PM
"you can have a money that is not convertible into anything *at any fixed exchange rate* and yet not worthless"
I'm not sure its correct to include 'fiat money' in that category, as it's always 'convertible' into a reduction in liabilities owed to the state, at the very least. i.e. you can always use it to pay public debts (and also private debts, if it's legal tender for them).
Posted by: Philippe | September 18, 2013 at 02:52 PM
Philippe: "before countries went off the gold/silver standards, central bank currency was normally redeemable for Treasury Coin, or bullion, right? (unless specie payment was suspended by the government)"
That's basically right. I think the precise details varied a bit, but I forget them.
One example that springs to mind: Don Patinkin "Money Interest and Prices". Though you could say he didn't properly integrate money into the model.
Yes, but the value of those public and private nominal debts then rises or declines in proportion to the value of money, so they don't really provide any sort of real anchor. It's like two ships anchoring to each other.
Posted by: Nick Rowe | September 18, 2013 at 03:07 PM
Highly liquid assets (eg in the form of securities) serve as MOE substitutes -- but the supply and value of these assets is contingent on constantly shifting production processes with rival time schedules and alternative output potentials. In other words, these highly liquid assets have rubber-like quality as a measure of value -- their valuation can shift as the usefulness of longer or shorter production processes shift or as the valuation of superior or inferior output shifts.
As these asset valuations shifts, so to does the liquidity of these assets.
Think out yo-yos in housing construction and the valuation and liquidity of securitized mortgages, as just one example.
Now plug in liquid assets of flexible valuation and liquidity where Nick discusses "apples".
Or simply plug in securitized mortgages where Nick discusses apples.
That is the real monetary world we live in.
Changes in the supply and demand for production processes of shorter and longer periods and with variable superior or inferior output -- and the supply and demand for securities directly linked to these production processes -- have macroeconomic consequences.
Posted by: Greg Ransom | September 18, 2013 at 03:31 PM
Nick,
With bank money, negative balances can function as the medium of exchange. For example, if you and I both have overdrafts and I buy something from you, no money (as conventionally defined) changes hands. How do you see that? Do you extend your definition of money to cover negative balances? Is it something else other than money acting as the medium of exchange? Or do you treat it as if there were some implicit money involved?
Posted by: Nick Edmonds | September 18, 2013 at 04:20 PM
Nick, did you catch that David Laidler on EconTalk has adopted the Hayekian explanation for the yo-yo of the economy over the last 10 years as his own?
http://www.econtalk.org/archives/2013/09/david_laidler_o.html
Posted by: Greg Ransom | September 18, 2013 at 04:20 PM
"If I permanently suspend convertibility of my IOUs into shells, my IOUs are now just sheets of paper with ink marks on them. All the number "10" means is that I will convert one "10" into two "5"s or vice versa."
The important thing is that the convertibility to shells wasn't needed. It is the convertibility to anything available to being sold at any time that is important with money.
I like to view holding money as a promise from society at large to give you something valuable. When you have a certain amount of money, it means users of money owe you a certain amount of goods or services that you can redeem whenever you want. Carrying money is effectively, the most flexible, shortest maturity, least risky bond you can have. The fact that money is redeemable for stuff at anytime from anyone has huge benefits but also a huge downside.
With all other forms of savings, the claim on wealth is backed by specific people, things or activities (let's forget complex contracts and derivatives for the sake of this argument). For example, you can save by buying something you will need in the future. You own the thing immediately so its value is obvious and immediate. You can also save by buying stocks in a company. In this case you have a claim on the value and profits of this particular company and you can evaluate the risk in that claim by looking at the business worthiness of that company. The stock value is backed by this company. You can also buy bonds or loan money. In this case you have a claim on future wealth of a particular party and you can evaluate the creditworthiness of these specific parties. These parties back your investment.
When it comes to saving by accumulating money on the other hand, it is not immediately clear who or what 'backs' its value. I don't mean this in a legal way. This is not about who can legally convert or tax or produce money. I mean this in purely economic way. When there is a stock of money that people want to spend, who is responsible for producing the things this money will be traded against?
The answer with money is that everybody that trades in money are collectively responsible to "back" money, by which I mean they are responsible to produce the goods that trade for money. This has a very important consequence. It means that if there is too large a stock of idle money out there compared to the ability of the economy to produce things, then this money is not backed by enough capacity and will likely either lose its value when people collectively try to spend it or otherwise require market distorting central bank interventions to maintain its value.
This is why, as much as possible, money or money like instruments (such as safe short term government bonds) shouldn't be overly used as a long term saving instrument. Money is best used as an intermediate medium so that people don't have to resort to bartering.
People who want to save longer term should save in things that are backed by real economic activity. Things that have a defined role in the economy which can be evaluated for worthiness.
In order to prevent people from stockpiling money, it is critical that money has a return that is lower than real private safe investments. This means that in low growth conditions where the safest investments in the private sector temporarily have real returns of less than -3% or -4%, money shouldn't artificially produce a return of -2%. That is, in these conditions inflation should be higher than 2%.
Otherwise money crowds out private safe investments and the -2% being a few percent above what the private free market can generate is effectively a subsidy to holders of money and money like instruments. When banks and businesses hoard cash and government bonds, instead of investing in instruments that are tied to more intrinsic real world value, we know that electrons or pieces of paper have been artificially turned into vehicles that funnel towards its holders more value than the free market can provide. Accumulations of idle money is not backed by real economic activity unless there is someone out there waiting with stockpiles of surplus stuff or spare production capacity.
In conditions of persistent high unemployment and low interest rates, low inflation is a central bank subsidy to cash savers and established businesses that don’t depend on loans. The wealth is extracted from small entrepreneurs and innovators that rely on loans to buy new equipment to jump start their businesses and create new wealth and new jobs.
What's more, in these conditions, when central banks don't create enough inflation, the flow of money slowing down and the businesses and banks stockpiling of money creates a higher risk of inflation in the future when money velocity will be restored but production can't be ramped up quickly enough.
It is a bit of a paradox but overshooting inflation early in a slump, to maintain the velocity of money, actually decrease the overall risk of inflation by preventing people from accumulating stockpiles of money that could later flow back too quickly in the economy.
Posted by: Benoit Essiambre | September 18, 2013 at 04:20 PM
Professor Rowe --
I have a totally unrelated question, forgive me. I'm slowly making my way through the archive of your posts, so I hope this isn't something you've already discussed in the past.
Disequilibrium macro (notional vs effective demand, etc.) peeks through in a lot of your writing. What do you think of the mathematical modeling that was developed in the disequilibrium literature, specifically the Barro/Grossman model? Do you think that model could be used usefully to analyze contemporary issues?
Thanks.
Posted by: BH | September 18, 2013 at 05:35 PM
Nick Edmonds: I haven't thought about the negative money question in some time. I can't remember what conclusion I came to when I last thought about it! I will run it over in my mind again.
Greg: yes, I did read the transcript of David's talk. He did acknowledge the Hayekian perspective.
Benoit: to my mind, the question is: what does the money supply curve/function look like? A promise of convertibility by the issuer of that money tells us something about that supply function. Put that supply function together with the demand function, and you're getting somewhere.
We *can* think of money as a responsibility of the whole community, but I'm not sure how far that gets us.
BH: I liked that old Barro/Grossman model. It was a big influence on me. I think it contains some insights that help us understand all monetary disequilibria.
Posted by: Nick Rowe | September 18, 2013 at 09:21 PM
Nick, very nice story. Someone should animate it to replace this one:
https://www.youtube.com/watch?v=jqvKjsIxT_8
My very dull blog (in comparison to this exciting story about apples, shells, bikes and IOUs) illustrates some of this, but only a small part. This is a great summary of several of your posts put together into a unified whole. I await part II.
Posted by: Tom Brown | September 18, 2013 at 09:59 PM
"But you can have a money that is not convertible into anything *at any fixed exchange rate* and yet not worthless."
Yes, but why drag in *permanent* non-convertibility, as if that is a central feature of modern money, as opposed to an exotic feature of a small number of relatively unpopular monies (orphaned bank notes, bitcoin)?
Posted by: Max | September 19, 2013 at 04:53 AM
Very nice post. I have to applaud you because I tried to explain "banks just as financial intermediaries long term" vs "banks as money creators short term" and it was hard. And still you managed much more with a far greater clarity. I agree with JP Koning - this post is 5 years of debates in a nutshell. Putting some link to this or that article here and there and it may be an ultimate monetary economics WIKI entry :D
Posted by: J.V. Dubois | September 19, 2013 at 05:50 AM
"If I can create this spread by creating money, I might decide to become a bank, which means a financial intermediary whose liabilities are used as money."
The main reason people use commercial bank liabilities as money is because they think they are US dollars. Deposits are just claims on central bank money. They are being misled.
If the central bank provided the possibility of depositing and transacting in US dollars (base money)then the demand for commercial bank deposits would disappear.
Posted by: lxdr1f7 | September 19, 2013 at 05:53 AM
Tom: thanks!
Max: I'm not sure I follow. Some central banks issue money that is directly convertible at a fixed (rather, pre-specified, because it could be a crawling peg) exchange rate, either into real goods (like gold) or into a foreign money; and some don't. And there are intermediate cases, like inflation targeting, with indirect convertibility at a rather fuzzy crawling peg with random walk errors into the CPI basket of goods. We need to consider both cases, and the intermediate cases as well. Permanent suspension of convertibility is one possibility, at one extreme.
Does it read differently?
(Though, to my mind, a central bank that has a fixed exchange rate acts more like a commercial bank. Or rather, a commercial bank that has a de facto monopoly on money within its own neighbourhood.)
JV: thanks! (Yeh, reading it through again, I am rather pleased with how this post turned out. Except for that one section on why central banks exist, which is not 100% clear in my own mind.)
lxdr: "If the central bank provided the possibility of depositing and transacting in US dollars (base money)then the demand for commercial bank deposits would disappear."
Suppose the Bank of Canada offered chequing accounts and made loans to the public. It's like if the government gets into any business. If it offers a better product at lower prices it will presumably drive its private competitors out of business. It might do that because it is a more efficient producer, or it might do that because it is subsidising its business. In this case it would be difficult to tell. The Bank of Canada has a monopoly on note-issue, and could use those profits to cross-subsidise its chequing account and loan business.
Would we want the government to be the one that decided who did and didn't get a loan?
Why not go the other way: let commercial banks issue notes and compete with the Bank of Canada?
The answers to those questions aren't obvious to me.
Posted by: Nick Rowe | September 19, 2013 at 07:28 AM
Nick Rowe
The bank of Canada can simply allow people to directly hold reserves that doesnt mean it should engage in lending though. If the bank of Canada needs to expand the money supply because inflation is too low or NGDP is too low it can simply expand the money supply evenly to the broad public in a non debt basis in order to attain its target.
Commercial banks can issue whatever financial securities they want except for Canadian dollars which are issued by the Canadian monetary authority. I can manufacture cars but I cant manufacture a Ford.
Posted by: Mike | September 19, 2013 at 09:02 AM
I understand what you are saying. If the supply of money isn't constrained by shells, it then becomes about whatever monetary policy people trusts will be followed. However, since monetary policy is usually set relative to prices or sales of everything, moving from shells to a target dependent on a basket of everything or their amount produced is what makes fiat money special.
I still think the main difference between money and other saving instruments is that money is "backed by" the aggregate of everything produced instead of being a claim on specific things. If there is a mismatch between the amount of things produced and the amount and flow of money in the economy we get problems.
Posted by: Benoit Essiambre | September 19, 2013 at 09:04 AM
Mike: "I can manufacture cars but I cant manufacture a Ford."
I can't make a car and put a "Ford" badge on it, because that would be counterfeitng, just like money. But I can make a car, that works better than a Ford in some cases, worse in others, and promise to exchange it for a Ford whenever the buyer wants.
If my IOUs were the only thing that Canadians used as money, would it make sense to ban me from making loans? Would it be good for me? Would it be good for other Canadians?
Posted by: Nick Rowe | September 19, 2013 at 09:59 AM
Benoit: "I still think the main difference between money and other saving instruments is that money is "backed by" the aggregate of everything produced instead of being a claim on specific things."
Gotta be careful about the "backed by" bit, because it usually means the assets of the issuer of money. But who is responsible to ensure the implicit claim on all goods (both produced and non-produced goods like land and old houses), and who is actually able to take action to ensure that implicit claim is honoured? The Bank of Canada.
Posted by: Nick Rowe | September 19, 2013 at 10:04 AM
"if my IOU pays 0% interest, the village has, in effect, made me an interest-free loan."
As soon as I see you getting that interest free loan, I'll offer Mike dollars that pay 1% interest. You'd probably counter with Nick dollars that pay 2%, and we'd keep bidding until we both pay 5% interest on our dollars. (Unless printing/handling costs add up to %3, in which case we'd pay 2%.)
Posted by: Mike Sproul | September 19, 2013 at 10:54 AM
Nick Rowe
"I can't make a car and put a "Ford" badge on it, because that would be counterfeitng, just like money. But I can make a car, that works better than a Ford in some cases, worse in others, and promise to exchange it for a Ford whenever the buyer wants."
The commercial banks tell people they have Canadian dollars in their account. They dont tell people that their account holds claims on currency or commercial bank liabs. Thats like me saying I will sell you a Ford not you can exchange my cars for a Ford.
"If my IOUs were the only thing that Canadians used as money, would it make sense to ban me from making loans? Would it be good for me? Would it be good for other Canadians? "
Monetary policy doesnt need to be tied to lending (unless for emergency lending). It would be superior to conduct monetary policy on a non lending basis becuase the money supply can be adjusted without necesitating greater debt.
Canadians can use anything as currency. The central bank can create the best medium of exchange though if it manages MP effectively so therefore most people will demand Canadian dollars over other currencies. Plus it is the only way of paying tax.
Posted by: Mike | September 19, 2013 at 11:11 AM
Mike Sproul: of course you will. But I've already got you in my (implicit) model. Only I called you "Mary", or "Harry". Given free entry, new banks will enter the industry until profits are zero, which means the spread equals the transactions cost. I expect I should have said that explicitly, but this is just where banks fit into the big macro/money picture. The micro aspects of banks come later.
Posted by: Nick Rowe | September 19, 2013 at 11:12 AM
Nick Rowe
"Of course you will"
I'm sorry I lost you there what are you referring to?
Posted by: Mike | September 19, 2013 at 11:36 AM
Yeah you are right. "back by" leads to confusion. I'm not sure what's the word I'm looking for.
Posted by: Benoit Essiambre | September 19, 2013 at 11:39 AM
Mike (not Sproul): There are now two Mikes commenting! I have edited my previous comment to show I was responding to Mike Sproul.
Posted by: Nick Rowe | September 19, 2013 at 11:52 AM
"But if people believed there was some positive probability I would restore convertibility in future, ... my IOUs would still have some value"
Suppose you took all your assets and spent them on candy. No fair keeping any assets, since that creates the possibility that you might pull a Bank of England and restore convertibility later. Well, once again you have gotten a free lunch, which inspires me to produce rival Mike dollars to get a piece of that free lunch. That would reduce the demand for your Nick dollars, with no stable solution short of both of our dollars being worth zero. (No surprise since that's the value of the assets backing them.)
It sure would be easy to get confused here. Like, for instance, you might suspend convertibility but keep all your assets, and people would therefore value Nick dollars at par with the assets backing them (just like stocks and bonds). Then some misguided souls would think "inconvertible=unbacked", and invent the crazy idea of fiat money.
Posted by: Mike Sproul | September 19, 2013 at 07:41 PM
Nick, do you accept the concept of a degree of "moneyness" to financial or other assets? If there were a continuum of goods with all conceivable degrees of moneyness in your examples here, how would that change things? See JP Koning or Roche for a couple of ideas on this moneyness concept:
http://jpkoning.blogspot.com/2012/12/why-moneyness.html
http://pragcap.com/understanding-moneyness
Posted by: Tom Brown | September 19, 2013 at 08:01 PM
Mike Sproul: How would you think of the case of shells, where fashions change, so people stop wearing shells as jewelry, but keep on using them as money. Something that can't happen?
Tom Brown: I'm in two minds. One mind likes JP's ideas of moneyness. The other mind says: OK, liquidity is only a difference in degree, but there's something special about the most liquid good on a spectrum. The winner always gets the prize, even if he's only a couple of seconds faster than whoever comes second. We wouldn't use the second place finisher when we could use the winner. All the other goods get bought and sold for the winner of the liquidity race.
Posted by: Nick Rowe | September 20, 2013 at 06:21 AM
Nick:
You've got me there. I have a hard time explaining shells, Somali currency, Iraqi Swiss Dinars, and bitcoin. You have a hard time explaining why all central banks hold assets, why the dollar should be called "fiat money" when the Fed is clearly able to buy back every federal reserve note at par, how so-called fiat money could hold its value in the face of competition from rival moneys, why money should be valued on principles that are completely different from those governing stocks and bonds, etc.
It seems to me that you have more explaining to do than I do.
For what it's worth, I explain shells etc. with the "Baseball card" theory, which says that curiosity value can create a demand for some things, and as long as they are limited in supply, they can have positive value. Once some small value has been created in this way, then monetary demand can increase the value even more. Gold and silver are the best cases in point.
Posted by: Mike Sproul | September 20, 2013 at 10:14 AM
Mike Sproul: Yep, we circle around, searching for weak spots, making each other think things through and explain things more clearly!
"You have a hard time explaining why all central banks hold assets,.."
1. To keep accountants and backing thoerists off their ..ummm..backs?
2. To preserve some vestige of independence from government?
3. Those assets are all a...umm..shell game anyway, if the government owns the central bank, and the central bank owns mostly government bonds.
".. why the dollar should be called "fiat money" when the Fed is clearly able to buy back every federal reserve note at par,.."
I don't like the words "fiat money" either. In the bible, God said "Fiat lux!", and there was light. But governments are not God. If people don't want to use those bits of paper as media of exchange, the government can't create money.
"... how so-called fiat money could hold its value in the face of competition from rival moneys,..."
Network externalities and custom, sometimes backed up by legal restrictions too. Whoever moves first, and gets people speaking his language/using his OS/using his money gets a "natural" monopoly.
".. why money should be valued on principles that are completely different from those governing stocks and bonds, etc."
It's not. But we have to recognise that more liquid assets pay a lower rate of return in equilibrium, and that the marginal value of that liquidity premium may be a decreasing function of the real quantity of that asset, because demand curves slope down.
Posted by: Nick Rowe | September 20, 2013 at 10:40 AM
In my reading of this post, Nick seems to be a little more accommodating of the Sproulian view than at times past, but not as accommodating as Mike would like. Dunno if you two agree.
Incidentally I was listening to David Laidler on Econtalk yesterday and he used the word "moneyishness". I'm not going to rush off and rename my blog or anything, but it highlights that money-as-adjective be a useful approach, especially when we talk about the marginal value of the liquidity premium on various assets. I've always thought that thinking in terms of liquidity premia/monetary convenience yields/moneyness is a way for both Nick and Mike to be right about central bank money. Nick's QTM works via the liquidity premium, and the Mike's RBD takes care of everything other than the premium.
Posted by: JP Koning | September 20, 2013 at 02:06 PM
"...a way for both Nick and Mike to be right about central bank money." Can't we think of them as both wrong instead? :D
Posted by: Tom Brown | September 20, 2013 at 02:49 PM
The liquidity premium clearly can raise the value of gold or silver, since they are commodities with rising supply curves. But if the supply curves of gold and silver were horizontal, the liquidity premium could push the demand curves for gold and silver all over the place without affecting their value at all. Now, bits of paper and computer blips aren't commodities, and it's not kosher to speak of demand and supply curves for paper and blips. But if I were to hold my nose and talk that way anyhow, I would say that the supply of paper and blips is horizontal. So here again, the liquidity premium is irrelevant to money's value.
Posted by: Mike Sproul | September 20, 2013 at 08:06 PM
Before I forget, let me throw in a little whining. Whenever publishers' sales reps send me a new macro textbook, I always scan the book for any mention of the real bills doctrine or the backing theory. In the 20+ years that I've been checking, I have never seen those words appear in any macro book. It's so unfair!!
Posted by: Mike Sproul | September 20, 2013 at 08:12 PM
Mike: inflation rates of (say) 10% are not uncommon, with central banks paying 0% interest on at least some of their monetary liabilities, and yet people still hold those monetary liabilities, even at minus 10% real yield. Central banks' seigniorage profits do not seem to get competed away. Or, to say the same thing a different way, competition from commercial banks does not seem to force central banks to supply Friedman's Optimal Quantity of Money.
Posted by: Nick Rowe | September 21, 2013 at 03:23 AM
Central banks' seignorage profits are probably an illusion created by inflation and inconvertibility. If a central bank does operate at a loss (say, because printing and handling cost exceeded its interest earnings), then their currency loses value and erases their loss. A better test of profitability would be to look at the central banks that maintain a fixed rate of convertibility into gold, US dollars, etc. Those banks operate at a loss, and sometimes crash and burn, pretty commonly.
I don't follow you about the importance of the 10% thing or the Optimal Quantity of Money. Classes start this week, so I'll let those two go for now.
Posted by: Mike Sproul | September 21, 2013 at 04:03 PM
What I think monetarist analysis of money lacks, is a proper analysis of the moment of emission (and that of destruction) of money. Money is created to finance something. Considering it always creates a corresponding debt liability, money is also always an investment, and investments have to pay off. Monetary emission is the moment at which a price is set. So we have a point where money inevitably connects with the real economy by giving some good, asset or service a price tag and hoping more will come back. When I read things like But if we used apples as the MOA, apples would not have a price of their own; the price of apples would be the reciprocal of the price of everything else it sounds like you are treating money as an independently manipulable object that, for reasons unknown but nicely pictured in cute Genesis fables, just happens to sit on the other side of the fraction of all other goods. I don't think this is a particularly useful way to look at things. And it's probably the origin of all those fantasies of central bank omnipotence that monetarists can't seem to let go of. As soon as one looks at money as the numerical investment counterpart to the real world, one immediately arrives at more than crude quantitative answers to monetary phenomena such as inflation etc.. I would say e.g. that the stability in the value of money is directly related to the perceived QUALITY over time of the investments it finances. Hope that makes sense...
Posted by: Oliver | September 25, 2013 at 04:18 PM
I don't agree with you Prof. Rowe. I don't see how network externalities and customs are self-enforcing only in one direction but not the other. If the government ceased to collect taxes, perhaps people may still exchange with one another. This is because people accept fiat on expectation that other people will accept fiat (who are only accepting it because they assume further others are accepting it). It is a regressive loop. I can see how in this direction fiat becomes self-fulfilling as money.
But shocks in the money market happen all the time due to market fundamentals. Money demand will eventually fall. If people begin to stop accepting fiat, other people will stop accepting fiat on expectation that other people will not accept it (who are not accepting because they assume further others are not accepting it). This too is self-enforcing and self-fulfilling. I don't see how network externalities and customs are one directional. I believe that when this happens, people would get fed up with the volatility of the fiat (since after all there is no fundamental demand establishing the intrinsic value of the 'money') and exchange something such as gold or gold notes.
Posted by: Ian | September 30, 2013 at 05:07 PM