Blue sky money two - money and not banking

Banking is a subset of finance. Money and finance go together.  Money and banking go even more together.

But they don't have to go together. Maybe they didn't ought to go together. Finance is unstable. Banking is even more unstable than the rest of finance. A bank makes promises it knows it might not be able to keep. A bank is an accident waiting to happen.

Unstable money is a very bad thing, much worse than unstable finance or unstable banking. Why do we have an economy in which money is linked to unstable finance? Why do we have an economy in which money is linked to the most unstable part of finance? Isn't this a really stupid sort of monetary system to have?

If we had a monetary system in which money and finance, or money and banking, were separate, then financial instability would just be a spectator sport for monetary economists. You could bail out the banks, or not. Just like you could bail out the carmakers, or not. And us monetary economists would shrug our shoulders and stay out of it, and let the finance guys and industrial economics guys argue it out among themselves. If a financial crash didn't cause an excess demand for money, and the resulting recession, only the microeconomists would care. Sure, there might be some structural unemployment, as workers switched from finance and investment industries to producing consumer goods instead, but monetary policy can't do much about that anyway.

But instead we have a monetary system in which finance, especially banks, and things that work like banks, are heavily interwoven. So if banks go bust money disappears and people can't buy and sell all sorts of things that have nothing to do with finance. If all our cars went bust at the same time (which they don't) we could at least walk to the supermarket.

Money is a medium of exchange and medium of account. It's got nothing to do with borrowing and lending. Forget that "standard of deferred payment" rubbish.

I can imagine an economy with money and no finance. People never borrow and lend. It's tabu, or they all have exactly the same intertemporal preferences and production possibilities so won't want to borrow or lend. But they still use monetary exchange, because barter is such a PITA. They use gold, cowrie shells, or bits of paper for money. But the bits of paper are not IOUs for anything. They are just irredeemable bits of paper.

I can imagine an economy with finance and no money. It's a barter economy, so any good can be traded directly for any other good. But people still borrow and lend apples, cars, land, labour, wheat, whatever. I can even imagine banks in a barter economy (unless you insist on defining "bank" as a financial intermediary whose liabilities are money). A wheat bank lets you withdraw wheat on demand whenever you want some quickly. That doesn't mean that wheat is the medium of exchange or unit of account. Wheat banks would need to keep wheat reserves to cope with fluctuations in the demand for wheat. There could be runs on wheat banks; people would rush to withdraw their wheat first if they thought their bank were illiquid or insolvent, so they might not be able to withdraw their wheat quickly, or ever. Bakers might be worried about runs on wheat banks, but it wouldn't be a macroeconomic problem.

Money and banking don't have to go together. Money and finance don't have to go together. But they do go together. Why?

You can use wheat to bake bread. You can't use a promise to pay wheat to bake bread. You can use a Bank of Canada $20 bill as a medium of exchange. And sometimes, you can use a promise to pay a Bank of Canada $20 bill as a medium of exchange. A promise to pay money can itself sometimes be used as money. If that promise to pay is made by a bank, like the Bank of Montreal, it very often is used as money.

The Bank of Canada issues money that is irredeemable. The Bank of Canada can exchange it for something else if it likes, but it doesn't have to. It doesn't promise to redeem it for anything. It may choose to redeem it, usually for government bonds, but in a quantity or exchange rate that the Bank of Canada chooses to meet its macroeconomic objectives. A $20 Bank of Canada note is not an IOU. It is not a liability of the Bank of Canada. We only say it's a liability to prevent the accountants getting upset when they try to balance the books of the Bank of Canada.

The Bank of Canada is not a bank. It is more like a closed end mutual fund, where the fund managers can choose whether or not to sell assets and redeem units when the price of their units falls. And even that analogy doesn't work well. It's like a closed end mutual fund where the dividends on its assets, minus operating costs, are given to the owners of the mutual fund (the government in this case) and not paid out to the unitholders. The unitholders are willing to hold the units, even though the units depreciate at 2% a year, and pay no dividends, because the units are just really handy for doing the shopping.

A central bank on the gold standard (or with a fixed exchange rate) really is a bank. Its money really is an IOU, redeemable in gold. Its money really is a liability. A central bank on the gold standard is more like a money market mutual fund. (It's not exactly the same, since it still pays any dividends from its assets to its owner, and not to the unitholder.) A central bank on the gold standard can suffer a run, and it can be unable to fulfill its promise of redemption in gold. Its liabilities can "break the (gold) buck", just like a money market mutual fund, if it doesn't hold 100% gold reserves.

When we decided that the gold standard was a bad idea, we decided that central banks should not be banks. Going off gold, and abandoning fixed exchange rates, was a decision to separate money and banking. And it was a really good decision. Banking is unstable. Money ought to be stable. So separating money and banking was a good thing. But we didn't go far enough. Central banks are no longer banks. But commercial banks are still banks.

Deep down, I think that all monetary cranks vaguely understand that there's something wrong with having money and banking go together. Banks with 100% reserves aren't really banks either, because those banks have a perfect match of their assets and liabilities. They have just changed paper money into electronic money. It's like holding $100 bills as assets , and issuing more convenient denominations as liabilities, because the central bank can't be bothered to print $10 bills. Banks with 100% reserves are more like money changers than money producers.

But those monetary cranks are standing in front of the train of the History of Finance, trying to get it to stop. Even if they succeed, Finance will switch to cars, and drive right around them.

People would like to borrow to invest in long, risky, illiquid, and complicated projects. And they would like to lend in short, safe, liquid, and simple assets. They want to hold money, only paying interest. Finance tries to give people what they want. Finance tries to convert all assets into money. Finance wants to join money and finance together, because that's what people want. That's the problem.

What to do?

1. What we are doing now. Regulate banks and finance to try to make them more stable, support them when they fail, and let central banks (which aren't banks) take offsetting action when finance fluctuates.

2. Try to prevent people doing what they want, and try to separate money from finance, especially separate money from banking, which is the most unstable part of finance. Chequable stock market mutual funds, maybe? Your cheque would still be worth a fixed number of dollars, but the balance in your chequing account would rise and fall with stock prices.

3. Dunno.

Comments

[David Pearson]

We did "number 1", as you say, but the problem is that "banking" (supplying credit) has become increasingly divorced from "central banking" (supplying bank reserves). The shadow banking system operates on the basis of daisy-chains of collateral hypothecation. Lehman failed because it ran out of collateral to put up against its trades and Fed swaps; a number of European banks are arguably on the verge of doing the same. Supplying reserves through OMO does not solve this problem. It only gets liquidity to those that already have it (in the form of available collateral), not those that need it. So central banks turn to supporting the collateral value directly through asset purchases. The Fed had some success with this fiscal operation, but it is much more problematic for the ECB to carry it out. In any case, the problem is if you promise to maintain collateral values, markets will lever more and more against that collateral, so that the more stability you have, the more fragility you create.

The solution is to allow low-grade instability to wash away insolvent firms and prevent leverage from becoming the dominant strategy. This requires a discretionary monetary policy with little transparency, a strong deposit insurance scheme to prevent runs, and a firewall between deposits and shadow banks. This roughly describes where we were pre-Greenspan.

[Ralph Musgrave]

Nick, Welcome to the world of full reserve banking.

You say you don’t like a system in which “if banks go bust money disappears”. Quite right. That system is called fractional reserve. Under full reserve the problem does not occur.

You also say “Money is a medium of exchange and medium of account. It's got nothing to do with borrowing and lending.” Quite right. So vote for full reserve!

[Martin]

4. Private note issue. Trade systemic risk for idiosyncratic risk?

Though Nick, I have to ask, don't you mean unstable NGDP rather than unstable money? I asked this question - if I recall correctly - to Scott, and he answered that if NGDP is not allowed to collapse you do not need deposit insurance to prevent bank runs.

As I see it now you need a shock to expected nominal asset returns to get into trouble as a bank. If you have an entity that can buy unlimited assets at will, (solvent) banks will never get into trouble.

[Bill Woolsey]

Nick:

You are treating the gold redemption requirement as something serious and real, but the obligation to maintain the inflation (or nominal GDP) target as just a whim of the central bank.

I can argue either way. The gold standard is a problem? Suspend payments or devalue (or revalue.) Central bank liabilities aren't really liabilities when what they promise to pay can be changed on a whim.

Or...

From the point of view of the monetary authority, the constitutional requirement to keep nominal GDP on target is a binding requirement. It can issue zero interest currency, but it can only issue the amount someone wants to hold and must pull it back out of circulation if they don't want to hold it. Of course, as long as someone wants to hold it, it is a loan. Maybe even interest free, or at a negative real interest rate. But some provision must be made to pay it back when necessary. If we consolidate with the rest of the goverment, then it must issue interest bearing debt as needed or else collect taxes--create a budget surplus. If it operates on banking principles, as an intermediary, it needs to hold assets that can be sold.

There are two ways to escape this logic. Irresponsibility--print money and spend it. Or else, a quantity rule. The monetary consititution is fix the quantity of base money or have it grow at a contant rate. It's value depends on supply and demand. As soon as you instead start aiming at something else and have a real commitment to that, it is a liability. When you say, not really, because there is nothing to pay it off in. Well, you can suspend payment with a gold standard too. What is the commitment to fix the price of paper money in gold less of a commitment than to keep the CPI or the growth path of nominal GDP on target.

And, of course, money is a asset--if it is storable. It inevitably is going to be related to finance and banking.

[Nick Rowe]

David: interesting comment. I wish I understood the shadow banking system better. I only alluded to it in my post, because I know I don't understand it well enough.

Ralph: Yep, but if you make banks keep 100% reserves, someone will just set up a bank, and call it something else. Plus, people want their assets to be monetised.

Martin: "Though Nick, I have to ask, don't you mean unstable NGDP rather than unstable money?"

I think the two come very close in practice. What I can't decide is how possible it would be for a central bank to keep NGDP stable if finance were very unstable.

I don't think it really matters whether there's private note issue or just private chequable accounts, if both are redeemable in central bank money. And the natural monopoly aspects of money, where we all want to use the same medium of account, tend to rule out irredeemable competing monies.

[Nick Rowe]

Bill. OK, I see your point. But a central bank that targets the price of gold or foreign exchange and has fractional gold reserves or forex reserves is vulnerable to a run. I can't see an NGDP or inflation targeting central bank being vulnerable to a run in the same way. Sure, if the government is insolvent, the CB won't be able to keep to its target. But it then bends like plastic, rather than breaking like glass. It takes extraordinary incompetence or insolvency like Zimbabwe to fully destroy an irredeemable currency.

[David Pearson]

Nick,
A good primer from Karl Smith on the E. shadow bank collateral problem:

http://modeledbehavior.com/2011/12/01/has-the-ecb-completely-lost-control-of-monetary-policy-ctd/

Izabella Kaminska at FT Alphaville understands this stuff better than most:

http://ftalphaville.ft.com/blog/2011/12/01/775341/draghi-we-are-aware-of-the-scarcity-of-eligible-collateral/

[Ryan]

Now this post is really fascinating.

Nick, you said: "But instead we have a monetary system in which finance, especially banks, and things that work like banks, are heavily interwoven. So if banks go bust money disappears and people can't buy and sell all sorts of things that have nothing to do with finance. If all our cars went bust at the same time (which they don't) we could at least walk to the supermarket."

This fascinates me because it presupposes the necessity of centrally managed currency. Yet money and finance both pre-date banks and centrally managed currency. If all our banks went bust at the same time, we could at least use real money, rather than that which Mises termed "money substitutes." This is important. Competition among currencies is valuable for consumers.

In fact, your entire post presupposes the necessity of central management of money. Why? In absence of central management, we would observe all of the above phenomena: every combination of finance, banking, and money. (Really, you mean "money substitutes," but I won't hold you to my terminology - it is important, though, because we currently do not have money in the Western world, only "money substitutes." In absence of central planning, we would have both money and money substitutes.)

The planners have gotten us in a pickle. Do we now throw up our hands and say, "Oh, well, it's really the best we can do assuming we want a centrally managed currency system"? Or, do we consider legalizing money and allowing it to compete with money substitutes?

[Nick Rowe]

David: thanks, I had read both of those. I mean I understand it at one level, but don't deeply understand it, so I feel comfortable making arguments about it.

[Martin]

Nick,

"I think the two come very close in practice. What I can't decide is how possible it would be for a central bank to keep NGDP stable if finance were very unstable."

What however is the source of the instability in finance? Financial instability is usually discussed, if I recall correctly, when it is caused by those so-called animal spirits. The central bank can break any animal spirit by indirectly targeting the nominal expected asset return when it sets expected NGDP. Any other source of instability has to work from the real economy to asset returns and will then result in 'inflation' due to lower real GDP. The problem then becomes more general and the problem is then how the central bank can keep NGDP stable when real GDP is unstable.

It's late here so perhaps this is neat, plausible and wrong, but so far it seems mostly neat and plausible to me.

[Nick Rowe]

Ryan: Thanks!

Trouble is, money seems to want to centralise itself. Or rather, each of us wants to use the same money as the people around us are using. So it's harder for competing monies to survive, and harder for new entrants to break into the money business. Indeed, that is really why monetary exchange exists in the first place. If we used barter, whichever good was most commonly traded would be even more desirable to accept in exchange for other goods. The monetary race is a "winner takes all" race.

The only competitor for Word would be other programs that mimic Word, by promising to pay Word.

[David Pearson]

Nick,
I'm not sure anyone fully understands it. Perhaps the post-War period was one of uncharacteristically low volatility in velocity, and these shadow banks are just a return to the pre-1934 "information sensitive" bank liability dynamic. Attempting to stabilize velocity through monetary policy suppresses but then amplifies that volatility. I think of stability-seeking regimes as adding floors to an office tower built on a credible foundation. The more credible, the more floors actors add, the higher the center of gravity, the more prone it is to tipping over in moderate winds.

[Mike Sproul]

"A bank is an accident waiting to happen."

Not true. Even if a bank becomes insolvent, the bank can (or should) suspend convertibility. If the bank has assets worth $90 (paper dollars) backing 100 checking account dollars, then after suspension, speculators will value each checking account dollar at $.9 paper dollars. If the real money supply is reduced by this, then the bank can (or should) issue more checking account dollars in exchange for assets worth .9 paper dollars for each checking account dollar issued. This restores the money supply to an adequate level. Bank customers take a 10% loss on their money, but that's life. At least the taxpayers don't get stuck with the bill.

The real problem is not banks, but laws that prevent banks from suspending. This leaves them unable to do anything but shut down, which is a terrible option.

[Nick Rowe]

Martin: even if NGDP were perfectly stable, real shocks, and expectations of future real shocks, would cause asset prices to move around a lot. Just a very small change in the expected growth rate of earnings, or real interest rates, can have very big effects on asset prices. Leverage magnifies those movements. The liquidity of assets can have multiple equilibria (they are easily tradeable because lots of people trade them, and lots of people trade them because they are easily tradeable). And asset prices depend on their liquidity. And that's even before you bring in animal spirits.

[K]

3. Don't prevent people from doing what they want *and* separate credit intermediation from money if *it* wants to separate.

This means

1) Don't force 100% reserve banking. Freedom is good.

2) Provide every citizen/corporation/whomever an account at the CB. Just like the banks (more freedom!)

3) Withdraw deposit insurance. (Less subsidies=less taxes=more freedom)

4) Free banks of regulation (yet more freedom).

5) Let the people choose where they want to keep their money (did I mention freedom?).

Since nobody is forced to use risky money there is no need for deposit insurance as you also discussed. I don't see any way in which this solution doesn't dominate our current setup.

As you know, I fully support the idea of using fully redeemable equity money, rather than credit (deposit/repo) money. But that's a separate decision from democratizing central bank money.

"People would like to borrow to invest in long, risky, illiquid, and complicated projects. And they would like to lend in short, safe, liquid, and simple assets."

People would like to drive Ferraris and pay Ford Fiesta prices. So? That's not where supply meets demand. A Fiesta is not a Ferrari and no amount of lipstick (fractional reserve banking) can make it so. The transformation magic only happens by virtue of  a giant deposit insurance subsidy. If we gave people the choice, they would not deposit money in banks. They would *buy* bank bonds and stocks at market clearing prices and then either sell or repo those instruments at the CB for money. Those repo trades would earn the risk free rate which leaves the investors *way better off* than holding dangerous bank deposits. And it establishes a fair market rate for bank liabilities and frees the banks of regulation. It's total win-win.

"people want their assets to be monetised."

No.

"Finance tries to convert all assets into money."

Yes.

Indeed banks want that since money is a liability of theirs that doesn't pay interest. But nobody in the real economy wants that. People just want to hold capital assets and they exchange the bare minimum of those capital assets for medium of exchange that they require for liquidity (or to hide the proceeds of crime). If they could barter their capital assets for goods, they would want zero money. But since our settlement system for both goods and capital assets typically require several business days (weeks for credit cards) for settlement (yup, no improvement in many decades) that requires us to maintain significant money balances. But that's not what people *want*. That's what they *have* to do given our 19th century system of settlements.

[Ralph Musgrave]

Nick, You say “but if you make banks keep 100% reserves, someone will just set up a bank, and call it something else. Plus, people want their assets to be monetised.”

Re your first sentence, you are saying the shadow bank system will circumvent the rules. I doubt it. First, where is the big problem in passing a law that says any institution which acts like a bank (i.e. takes deposits and makes loans) must make itself known to the authorities, else its prison sentences all round? Second, the shadow bank industry does not do much fractional reserve, as I understand it. They concentrate on connecting large lenders to large borrowers, which is not fractional reserve.

Re your second sentence, I don’t see the relevance of the fact that “people want their assets monetised”. The large majority of bank loans are backed by collateral. I.e. the “collateral is monetised”. That applies under both full and fractional reserve.

[Ralph Musgrave]

Nick,

I fully agree that having 100% safe instant access accounts fund long term and less than 100% safe investments is a “problem”, as you put it. Mervyn King, governor of the Bank of England described this as “alchemy”.

The essential nonsense is that if depositors are insulated from the risks inherent in having their money put into less than 100% safe investments, the risk does not disappear. In practice the risk to date has simply been loaded onto taxpayers. This largely explains the implicit too big to fail bank subsidy which was estimated by the UK’s Independent Banking Commission as being worth over £10bn a year (about £150 a year for each UK resident).

Your solution, namely “Chequable stock market mutual funds…” is a poor solution, I suggest, because it gives instant access to money which has been invested long term: a nonsense.

I suggest the solution is to force depositors to come clean and choose between two sorts of accounts. 1. Checkable accounts, which are 100% safe and instant access, and which because the money is not invested, would earn little or no interest. Perhaps the money could be deposited at the central bank.

And 2, accounts where money is invested, and which as a result would earn interest, but which because of the inherent risk, would not be guaranteed by the state. Plus the money would have to be locked up for months or years.

This two account solution is advocated in this paper:

http://www.positivemoney.org.uk/wp-content/uploads/2010/11/NEF-Southampton-Positive-Money-ICB-Submission.pdf

Amongst the advantages of the two account system is that bank runs would be much more rare. As to “safe” accounts, there’d be no point in depositors doing a run because the money is guaranteed to be there (absent blatant criminality). As to investment accounts, depositors just can’t withdraw their money in a hurry.

[K]

Nick,

Spam filter?

K

[K]

Ralph: "it gives instant access to money which has been invested long term: a nonsense."

What does that mean? Why can't I trade stocks for cucumbers? It's an excellent solution.

[Nick Rowe]

Mike: I'm wondering: is there any fundamental difference between your bank and a stock mutual fund where you can write cheques on your balance? Is your bank just halfway between mine and a regular bank? Half plastic, half glass, like a preferred share?

Ralph: Suppose you had 10 people, each of whom was not a bank, but all 10 were a bank, and yet the 10 people didn't even know each other, and didn't know what the others were doing. Maybe David Pearson will chime in on this.

With 100% reserve banking, the stock of capital is totally independent of the stock of money. At the opposite extreme, with 0% reserves, and when banks own every asset, the stock of capital equals the stock of money. The real world is halfway between those two extremes, but "wants" to go to the K=M equilibrium.

[Nick Rowe]

K: Found it in the spam filter!

[Nick Rowe]

K: "2) Provide every citizen/corporation/whomever an account at the CB. Just like the banks (more freedom!)"

OK, now suppose the CB subcontracts out the operation of its banking services, and lets people choose between subcontractors. How is that different from 100% reserve banking?

"People just want to hold capital assets and they exchange the bare minimum of those capital assets for medium of exchange that they require for liquidity (or to hide the proceeds of crime)."

Most people don't understand stocks and bonds and stuff, and just want to park a lot of their savings somewhere simple, and safe, where they can get it out if they need it. So banks spring up, even without deposit insurance. Then the CB faces the Samaritan's dilemma of having to bail them out.

Ralph: "I fully agree that having 100% safe instant access accounts fund long term and less than 100% safe investments is a “problem”, as you put it. Mervyn King, governor of the Bank of England described this as “alchemy”."

And I called it "magic"!

http://worthwhile.typepad.com/worthwhile_canadian_initi/2010/01/finance-as-magic.html

"Your solution, namely “Chequable stock market mutual funds…” is a poor solution, I suggest, because it gives instant access to money which has been invested long term: a nonsense."

Nope. It means you can only take your money out if someone else is willing to put his money in. And if everyone tries to take money out at once, the price falls until they stop trying.

[K]

Nick: "And if everyone tries to take money out at once, the price falls until they stop trying."

No it doesn't. Imagine that a hundred units of money are backed by 100 shares of stock. If I demand back my 1 share in return for 1 unit of account, then 99 units will be backed by 99 shares. So each share is always worth 1 unit. There can't be a run. Ever.

[Max]

It seems like everyone with an opinion is divided into two equally wrong camps: either "100% reserve" banking is a panacea, or else it's the work of the devil. It's neither. It's just a sensible (though politically impossible) minor tweak which doesn't fundamentally change anything.

[Nick Rowe]

K: sorry. I wasn't thinking/writing clearly. If you pull your money out, the mutual fund has to sell shares, which means someone else must be persuaded to put his money into this shares, not into the mutual fund itself. Depends whether we are talking closed or open mutual fund. The shares themselves are like a closed end mutual fund in some real investment.

[Scott Sumner]

Nick, Good post as usual. Quick story. I only met Bernanke once, about 15 years ago. I asked him just one question: What happens if a banking panic occurs while a central bank is successfully targeting NGDP? I don't recall exactly what he said, but it was something to the effect that it wouldn't be good, but not as bad as if they allowed NGDP to fall.

BTW, I love monetary thought experiments with no banks.

[Determinant]

Speaking of deposit insurance, in Canada deposit insurance was not introduced until 1967 and when it was it was meant as a competition-enhancing measure to allow small banks and trust companies to solicit funds from the public against the Big Five chartered banks. I can't understand K's antipathy to deposit insurance in Canada when it practice it has been a non-factor in our banking system.

The worst case that ever happened, the twin failures of the Canadian Commercial Bank and the Northland Bank in Alberta in 1985 were a result of same factors present today in other parts of the world: declining real estate values resulting in a declining mortgage portfolio and a reliance on wholesale funding.

I don't see the evidence of CDIC having ever had much an impact on the operations of the Big Five banks which have 90% of Canada's deposits, though. The OSFI, yes, CDIC, no.

[M. Ricks]

Reminiscent of your "bike banks" post ... I would offer this interpretation. We need an efficient way to get money into circulation. Not all methods are equal. One natural way for the government to get purchasing power to those who want it is to lend it out. Why is this natural? Because borrowers by definition want more purchasing power than they have on hand. So it makes sense to distribute the money supply through the financing markets -- you're getting it to agents who want more of it. The problem is that the government isn't too good at underwriting credit. So it outsources this role. Under this view, banks are the government's licensed distributors of the money supply. They're agents of the state for the purpose of getting money into circulation efficiently. My sense is you don't see it quite this way ...

[Nick Rowe]

Scott: Thanks!

Part of banks' problems come from bad monetary policy, which screws up everything, finance and banking included. But sometimes causality works the other way. And most of the time it just spirals, with bad banks causing bad money which causes bad banks, etc.

Determinant: I didn't know we didn't have deposit insurance before 1967. Wow!

Yep, Canada's banking history has been very good. But I have never heard a really convincing story why. Smart banks, smart regulation, or did we just get lucky? In either case, I just can't be confident that we will always be this smart or this lucky.

M Ricks: Yes, I don't see it that way. We buy and sell apples for money. We buy and sell IOUs for money. Because money is the medium of exchange, so everything gets bought and sold for money, whether it's apples or IOUs. But that doesn't mean IOU's (loans) are an intrinsically monetary phenomenon any more than apples are an intrinsically monetary phenomenon. We can imagine IOUs are for apples, not money.

[K]

Nick,

One thing at a time. I thought we were talking about one central bank backed by a some fraction of all capital assets. So there is no other unit of account than that bank's liabilities. Then those liabilities *cannot* have a run in principle. Let's say you the bank owns 10% of the capital assets and there are 100 units of account outstanding. Now let's say holders of 99 units want to redeem their units all at the same time. They will each get 0.1% of all the capital assets in the economy. Will that change the relative value of money and capital assets? No. The last unit holder will *still* get 0.1% of the capital assets if he comes for redemption.

And the scenario of the CB having to "sell" assets to raise money for redemptions doesn't make sense. If the unit holders want money they already have it. All the can want is to have their capital assets back.

Now you could imagine that the capital assets would trade in the market at a price that was not equal to 1 unit of account per 0.1% of capital assets. But then an arbitrageur would immediately fix that price discrepancy by exchanging capital assets for units of account with the central bank at the fair price of 1 unit per 0.1% of capital assets. Absolutely no runs. And the whole market can never change price in nominal terms. It's always worth 1000 so the market doesn't move. (Obviously since the numeraire is the market itself.)

[M. Ricks]

Understood everything gets bought and sold for money -- no disagreement there. I'm afraid I failed to get my point across. How about this thought experiment. Imagine a sovereign state with a barter economy. Say the government decides it wants to transition to a monetary economy. You're appointed head of the new monetary authority and tasked with effectuating the transition. What do you do? What's your plan, your institutional design? It strikes me that this is a major administrative challenge. You print special bits of paper, but how do you get them into circulation, precisely? And then how do you modulate the supply going forward? I assume you don't start buying apples. Seems to me like buying IOUs (making loans) would be a better idea.

[Determinant]

Nick:

The usual explanation is that Canadian banks are modelled on Scottish banking: The Royal Bank of Scotland, the Bank of Scotland and the British Linen Company (actually a bank). A heavy concentration in a few institutions, banks that see themselves as pillars of the economy and/or rent extractors, their job is to maintain their top dog status, not grow recklessly and thereby shoot the golden goose.

Put another way, the profits of Canadian banks parallel the health of the Canadian economy, with a slight premium for the fact that they have less risk in their business models than a widget factory.

RBS and BoS tossed their cultural background over the side in the last 15 years and it cost them their solvency. Canadian bankers and regulators never forgot their culture.

I listened to a Bank of Montreal program for small businesses two years ago where a banker explained that "the role of the Chartered banks is to provide senior, secured financing to business, as laid out in the Bank Act."

I think smart banks and smart regulators boils down to the fact that both banks and regulators in this country still have a healthy appreciation and fear of risk. Risk is dangerous and too much risk is deadly. The market here is small enough that you can't paper over every transaction with an insurance-like derivative, not that Canadian banks think that way nor do regulators let them.

In practice that means that Canadian banks have had rock-solid capital ratios for years. The minimum in Canada even during the boom years was 7% Tier 1 and banks routinely ran 9%. First that means you aren't taking silly risks, you can't afford to. Second, even if you do do something stupid it will not ruin the bank. It will hurt, but it won't kill you. That's why Canadian banks are so resilient.

Everything that has happened since 2008 should show everyone that when it comes to banking there is no substitute for lots of paid-up capital.

[K]

Nick: "OK, now suppose the CB subcontracts out the operation of its banking services, and lets people choose between subcontractors. How is that different from 100% reserve banking?"

1) Those aren't safe unless *regulated*. And why provide the CB service only to some people? What is so objectionable about everyone getting access to the CB on the same terms? Why the desire to extend privileged access and then have to regulate operations and monitor anticompetitive behaviour? Can we not just get out of each others way?

2) 100% reserve banking forces banks to behave in a certain manner. Why can't we let the market decide what kind of commercial banks we want? Maybe some people want a fractional reserve bank that pays higher interest. Or guarantees privacy.

"Most people don't understand stocks and bonds and stuff, and just want to park a lot of their savings somewhere simple, and safe, where they can get it out if they need it. So banks spring up, even without deposit insurance."

Or... We transform capital assets *into* money by turning the CB into a giant open ended mutual fund holding a representative portfolio of capital assets. Then we no longer need to worry about stocks and bonds. Just hold money and we won't have to worry about investments either unless we have a compulsive desire to really fine tune our exposures (and to get involved with asset management firms). And what's even better, our capital asset holdings never fluctuate in nominal terms because, well, they *are* the numeraire. Will money banks spring up? Not to provide safety or simplicity. You can't beat the CB for that. But I guess tax evaders, organized crime, libertarians and others concerned about governments might find a use for them.

[K]

But even with repo-based rather than equity-based money we still don't need commercial banks to supply money. Like I said above, the entire money supply can be created by investors repoing stocks and bond (including bank liabilities) at the CB. So still nobody *has* to worry about having to deal with financial instruments. They still get their (CB) money by borrowing it at the bank (who gets it by selling bonds to investors who get it via repo loans from the CB). It solves all of the agency problems of our current system, but not the problem of giving money a nominal anchor (which is solved by the CB mutual fund approach).

[K]

M Ricks: "The problem is that the government isn't too good at underwriting credit. So it outsources this role."

To banks who take lots of risk and need to be insured. But consider the CB repo method in my previous comment. Produces all the money you need and as long as you use liquid collateral with a big haircut you don't need to worry about credit analysis. In the US for example there are at least $30Tn of liquid, price transparent assets and less than $10Tn of M2 most of which is really short term bonds rather than money. If all of that M2 were bank bonds there'd be another $10Tn of bank bonds available in the market. So with $40Tn of liquid assets available the CB wouldn't be taking any risk by repo lending $10Tn.

[Ralph Musgrave]

K and Nick, K: You ask me what is wrong with selling stocks for cucumbers (i.e. giving depositors instant access to money that has been locked up in long term investments). My answer is “nothing wrong there, as long as the stocks lose value when a significant volume of stocks are sold”. That’s No.2 solution in Nick’s post above.

Re K’s suggestion (5.59pm) that stocks won’t lose value because a bank can always obtain cash for a unit of stock by selling the unit, no doubt that works at the micro economic level. But what happens given a general decline in stock and property prices, as occurred over the last few years? The taxpayer is on the hook for several trillion to bail out banks. So there is still a risk there.

Next, I think the problem with Nick’s No 2 solution (“the balance in your chequing account would rise and fall”) is that there is a strong desire by many people to have bank accounts where when they deposit $X, they can get $X out a year later. Indeed, I suggest that is a fundamental human right, particularly for the financially unsophisticated. In fact it’s politically impossible to withdraw that right, isn’t it?

Hence the clear separation I suggested above between, 1, accounts which are 100% safe and involve no investment, and 2, what you might call “Nick No 2 accounts” i.e. accounts where funds ARE invested, but where account holders clearly carry the risk.

K: you ask (10.07) “Why can't we let the market decide what kind of commercial banks we want?”. My answer is: look at the 1800s. That was a real wild west free market: banks going under left right and centre – small savers losing years’ worth of saving. I don’t think that is acceptable.

[K]

But Ralph, the small savers can just keep their money at the CB. Nobody will *need* to use a commercial bank any more. That is the key to everything I'm saying.

[K]

And with mutual fund money backed by general capital assets there's no such thing as a "general decline" in nominal terms. That was my point at 9:06pm.

[Nick Rowe]

K: I'm not sure if you are missing my point, or if I am missing yours.

Suppose the BoC lets ordinary people have a chequing account at the BoC. We have $100 in our chequing accounts, and the BoC has $100 in Tbills to match. Then the BoC sets up a subsidiary branch, to handle this business. The subsidiary holds $100 assets (our chequing accounts), and has $100 on reserve at the BoC, which is an asset for the subsidiary and a liability for the BoC, and the BoC still has $100 Tbills. It's a wash. Then the BoC privatises the subsidiary, with the stipulation that the subsidiary must keep 100% reserves at the BoC.

A 100% reserve bank cannot extend loans. There's no regulation needed, except to check that it really isn't making loans, and does have 100% reserves.

So, I'm arguing that there is no essential difference between: letting ordinary people have chequing accounts at the BoC; and 100% reserve banks.

Can anyone else explain this more clearly than I am able to?

Now, if the BoC started making loans to ordinary people, that *would* be a difference.

[K]

"There's no regulation needed, except to check that it really isn't making loans, and does have 100% reserves."

No I agree. It's still regulation though. And now you have to manage the money oligopoly for anticompetitive behaviour.
And fight with the 100% banks over what exactly it is that they are allowed to do. If the answer is nothing apart from intermediating access to the CB, then why do we need them?

So the question remains, why don't we just give everyone acces to the CB?

[K]

Also, unless you ban use of fractional reserves, you'll blur the distinction between 100% reserve banks, and fractional reserve banks. It's really important that people understand where their money is safe and where it's not. Otherwise we'll end up back exactly where we started bailing people out to save the money supply.

[jonny bakho]

Money flows through multiple channels. One channel is the lending channel from banks. Another channel is "wages" or more broadly, money paid for goods and services. Another channel is transfer payments.

Productive economies make efficient use of materials, labor and capital to produce the desired level of goods and services. The banks serve a function to allocate money in a way that contributes to economic balance. In our current economy, we have excess labor AND and excess capacity AND unmet demand. We had a downward spiral that simultaneously reduced demand for labor, decreased capacity utilization and increased the gap between demand and unmet demand. This combination causes the risk premium for lenders to skyrocket. The key is reducing the gap between demand and unmet demand. This simply means that more money must flow to those in the economy with unmet demand. Banks must have return on investment, so they will not lend to those with high risk premium who happen to have the greatest unmet demand. Any effective strategy must lower the risk premium. The risk premium for the unemployed is lowered by creating jobs. The risk premium for underwater borrowers is lowered by cramdown. The risk premium for many can be lowered by transfer payments. The risk premium cannot be lowered by giving money to banks. That channel is at capacity and will not open until the risk premium problem is fixed. This means that the banks must be bypassed with fiscal stimulus or transfer payments until the risk premiums are reduced and the lending channel unclogs.

[Nick Rowe]

K: "So the question remains, why don't we just give everyone acces to the CB?"

OK. We are on the same page.

For the same reason the BoC doesn't sell coffee and doughnuts on Sparks street, and leaves it to Tim Hortons instead. Because the BoC isn't a bank, and isn't a coffee and doughnut shop.

[K]

Nick,

Now it's me who doesn't follow. The CB doesn't make donuts so it shouldn't open donut shops. But it *does* provide the medium of exchange. It already provides it electronically to all the banks and to everyone else in paper form. Why is it *bad* to also provide it electronically to everyone else? You still haven't explained what it is about some people that entitles them to special access. What is that special property that they have?

[Ryan]

Nick,

Does money really want to centralize itself? Do you think the advent of central banking was a natural market phenomenon, and if so, why do you think it is illegal to use gold as currency (at least, so it is in the USA)?

Like all goods, I think there is a tendency for businesses to expand and grow in order to capture economies of scale, but is that the same thing as currency tending to centralize itself? Perhaps. I don't think I'm convinced. The 19th century saw plenty of private-bank-issued currency, and the occasional bank-run, too. One could argue that the market was self-regulating in the sense described by e.g. Mises, punishing credit-expanding banks and clearing currency values at the point of exchange.

I don't see the need or the tendency for centrally managed currency, but maybe I'm missing something?

[Nick Rowe]

K: The BoC used to handle Canada Savings Bonds. They got rid of that business. If we really wanted the government to get into branch banking, Canada Post already has the network of branches. That's how some countries (Austria? Japan?) do it.

Back much later. Gotta set an exam.

[Nick Rowe]

Or, for starters: because it would be like giving CUPE/CUPW direct control of the BoC's printing press! No way!

[Martin]

Nick,

"Martin: even if NGDP were perfectly stable, real shocks, and expectations of future real shocks, would cause asset prices to move around a lot. Just a very small change in the expected growth rate of earnings, or real interest rates, can have very big effects on asset prices. Leverage magnifies those movements. The liquidity of assets can have multiple equilibria (they are easily tradeable because lots of people trade them, and lots of people trade them because they are easily tradeable). And asset prices depend on their liquidity. And that's even before you bring in animal spirits."

1. But if NGDP is assumed to be stable then keeping NGDP stable in the face of financial instability should not be a problem for the Central Bank, it's stable by assumption ;).

2. I don't think asset prices moving a lot is a sufficient condition for financial instability. You need at the very least asset prices to move a lot together. And I have a hard time seeing how that is likely to happen without animal spirits.

3. Regarding multiple equilibria for the liquidity of assets see #2. I agree with you, however I do not see how it can be financial instability just because particular assets are less liquid. Let's assume that all financial assets are suddenly (a lot) less liquid due to self-fulfilling expectations, how is that possible without a general increase in the demand for money? And how are those events possible without animal spirits? Furthermore how is it possible for all those assets to remain less liquid in the face of the assumption of stable NGDP? Shouldn't there be a massive drop in Real GDP then?

I am reminded of your post on 'hares' and I wonder whether I am using 'animal spirits' and 'unstable finance' in a different way from you.

PS. I tried to post this yesterday but the filter on this website was preventing it.

[K]

Nick: "CUPE"

Now you've totally lost me. Are you talking about CB lending? I'm just talking about *deposits*. Transaction services! LVTS for everybody. No lending except with huge haircuts (e.g. 75% margin = a 25% loan) against super-liquid collateral. Or no lending *at all* if we go the mutual fund route.

And what is it about private 100% reserve banks that prevents whatever the problem is that you refer to as "CUPE." Those banks wouldn't lend either. (Sorry. I'm not trying to be coy here. I've truly lost the tread.)

[K]

But it's not like I'm imagining the average retail investor going to see the BOC for a margin loan (though I guess it's possible). For the vast majority it would just be a place to keep their money and do transactions. Period. Large investors (and securities brokers) are the the ones who would create new money by borrowing at the CB against their collateral.

And as far as using the Postal network as the mortar and bricks... I hadn't really imagined branches. I basically never go to the branch. Maybe for FX and I used to go more when I borrowed from them. But I don't do transactions there. All *I* want from the BOC is a transactions website and a debit card.

[Max]

"A 100% reserve bank cannot extend loans. There's no regulation needed, except to check that it really isn't making loans, and does have 100% reserves."

Banks can use deposit-like (short term) funding whether or not they take deposits. Actually the main problem with deposits is not the asset-liability mismatch. It's that the customer is exposed to the risk of a single bank failing. If everyone prudently spread their money across many banks, there would be no need for deposit insurance (except as an emergency measure during a financial crisis).

[Nick Rowe]

K: Canadian Union of Public Employees. Canadian Union of Postal Workers. Sorry.

Baksun, as Owl said.

[Nick Rowe]

Martin: I checked the spam filter, and it's only got spam. Did you remember to answer the anti-spam question before posting?

[Jacques René Giguère]

Nick: on stability of CDN banks:
IN the 19th century, the Bank Act provided for "double responsibility". If a bank went bankrupt, the shareholders lost their shares plus an equal amount pledged on theirmpersonnal assets. Given that banks had few shareholders, ( the good old cdn oligo-kleptocracy), it put the fear of whatever you fear into the bankers culture. The meme still transmits itself.

Maybe have I lost the thread but:
100% reserve banking has the following consequences: any deposit is a withdrawal from the spending flow and is contractionnary...Unless the deposits are only chequing accounts for transactions purposes, each check immediately balanced by the check being deposited ( and it is not the intent here, constitutionnal rights to get back your money and all), this bank is a permanent recession machine. Unless the gunmint send back the spending power via countervailing deficit , in effect borrowing from the bank.
Am I lost?

[Steve Roth]

"A $20 Bank of Canada note is not an IOU. It is not a liability of the Bank of Canada."

Well it is, but only if BOC and the Canadian Treasury are seen as a single entity. (As you say: "the owners of the mutual fund (the government in this case)")

In that case, they *must* accept the $20 note in payment of taxes.

A bit of a stretch to call it a "liability" perhaps, but...

And yes per other comments: full-reserve banking is pretty darned interesting. Remove banks' license to counterfeit, make government the monopoly supplier of money.

[K]

Jacques: "Am I lost?"

I think so :-)

We probably need to stop throwing around the word "bank" carelessly in this conversation. We should use "bank" for deposit taking institution, and credit broker for consumer and commercial lenders.

In a 100% reserve economy, *credit brokers* lend money that they raise in the capital markets by selling their bonds and stocks. The money supply can originate in a number of ways. The one that is most similar to our current system would be for the owners of capital assets to borrow it via extremely well collateralized loans from the CB, using liquid, transparently priced collateral. The CB can regulate the money supply by varying the rate on those repo loans, just like it does right now. This would feed into bank lending via competition between CB repo and the money market for short term credit broker liabilities (uncollateralized lending).

Now, I agree that credit brokers could also take deposits on a 100% reserve basis. But there is no necessary relationship between these two businesses so it's better, at least for the purposes of clarity of discussion to imagine them as separate.

[K]

Oops! I meant "This would feed into *credit broker" lending via competition..."

Can't even follow my own nomenclature proposal for 2 minutes :-(

[Phil Koop]

Let's call K's CB mutual fund a CB ETF instead; ETFs already work the way he wants, as a unit of ETF is redeemable for a basket of the underlying stocks, not money. (You can also create a unit of ETF by delivering the same basket; I guess the CB could let you create money that way too.)

OK, so now the market is the numeraire. Do we think we will be living in a M-M world in this economy? If so, it seems to me that finance (I mean, credit-broker money) will be expensive, since the rate of return will have to exceed that of the market. That's fine for financing investment projects; no different from the situation that obtains today. But other types of loans, in particular mortgages, will experience a sea change. I think it's a debatable point, meaning that I'm confident that K can dream up a reason it favour, if his prior is commitment to equity money. But he will need to convince people who don't hold the same prior, and that will be no slam-dunk.

It occurs to me that our CB ETF will earn negative seigniorage, since if money is to represent a fixed holding in the aggregate equity market, the bank's holding will need to be continually rebalanced as relative stock prices change, and this will incur frictional costs. But this could be handled by inflation: steadily reducing the percentage of the market represented by each unit of money. It amounts to a few bp a year in real ETFs, so if people are willing to hold today's money at -2%, that should be no insuperable obstacle.

I wonder, though, about the foreign exchange effects in a small, open economy like Canada's. The equity market is pretty volatile in today's nominal terms, and if Canada's CB is the only one that switches to this system, that would presumably translate into FX volatility, possibly imposing trading costs.

[123]

Nick,

I agree with Bill Woolsey that money is a part of finace. But even if it was possible to separate money from finance, it would not possible to separate finance from AD. Collapse of bubbles will always have consequences for the distribution of AD. Sometimes the consequences of bursting bubbles are manageable (dot com crash), sometimes they are not (collapse of Greek bond bubble and its consequences for Greek AD).

[Bill Woolsey]

I agree that have money based on nomimal GDP is better than basing it on gold. But just think of redeemability as a way of enforcing the rule. I agree you can have runs with a gold standard, but you just have to suspend when you run out of reserves.

If nominal GDP is above target, how fast to you get it back down? If you run out of gold reseves, you suspend, gold goes to a premium. How fast to you get it back down?

How is it different?

I am more and more thinking that when anyone borrows short and plans to borrow again to repay, they should have an option clause like the old Scottish banks. If you can't pay, the short bonds become higher interest long bonds until you can pay them.

I also think rapid reorganization is important. Like, if an instituion is insolvent, the former creditors get new debt, worth much less that the total assets of the insolvent firm, and then equity claims to the difference and the current management can carry on. If the all of the new stockholders want to replace them, then they can.

I don't think that works with government. But generally, this idea that we have borrowed short and have to borrow to pay back those who lent to us, and so, somebody has to bail us out or it is the end of the world is just a disaster.

If this discourages people form lending short, then so be it. If it means that people save less and consume more, then so be it.

[Mike Sproul]

Nick:

This comment apparently didn't make it through, so here it is again. If a checkable stock fund suffers from a drop in stock prices, then there are fewer checking account dollars in the account. If a normal bank suffers a loss of assets, and it suspends convertibility, then each checking account dollar becomes worth (say) .90 paper dollars.

In both cases the bank is not an accident (bank run) waiting to happen. A run only happens if the stock fund fails to reduce the number of dollars in the account, or if the normal bank fails to suspend convertibility. Historically, failure to suspend usually happens because government regulations outlaw suspensions. So blame the regulators, not the banks.

[K]

Phil: "Let's call K's CB mutual fund a CB ETF instead"

Good idea. I agree.

" If so, it seems to me that finance (I mean, credit-broker money) will be expensive, since the rate of return will have to exceed that of the market."

I think that fixed rate bonds would not exist with this choice of numeraire anymore than we have instruments today that pay the return of capital markets plus a spread. Bonds exist because the serve as a rough proxy for the ability to lock in a future rate of consumption (modulo surprise inflation) or because some people suffer from serious money illusion. But since there is no real asset that accomplishes this trick we need central banks to inflation target in order to achieve it. But this is an illusion. Since there *is* no such capital asset, any instrument that promises a future quantity of defined consumption must in fact be a derivative. Ie someone must be shorting the consumption basket for someone else to buy it.  But that same derivative, ie an instrument that pays the return some desirable index could still exist. And maybe some people's risk preference is such that that instrument would have a lower expected return than the general market. Or maybe the current desire for bonds is just entirely driven by money illusion and corporate tax avoidance and they would just disappear with ETF money in which case we'd all just equity finance which Modigliani-Miller tells us makes no difference anyways, and which as Bill points out would save us enormous dead weight bankruptcy losses. 

One other important point about ETF money. The units don't  *have* to be in fixed quantity. If we really wanted the unit of account to track some imaginary fixed consumption basket we could do that by splitting or reverse splitting the ETF units. This could be used to deal with nominal rigidities if we so desired (and then we wouldn't make any progress on the money illusion issues and we'd still get lots of bonds so maybe it's a bad idea). A better idea might be to adjust the number units so as to keep NGDP (which is a very different beast with our new definition of "nominal") growing at a constant rate.

"the bank's holding will need to be continually rebalanced as relative stock prices change"
I don't think this is correct. If the market is made up of two equally weighted stocks and and then one of them doubles in price we will be holding twice as much of that stock. Which is still the correct proportion of the whole market. Relative price changes don't change anything. But you will need some rules to add and delete assets based on liquidity and price transparency. 

I don't have a "prior" desire to have ETF money. I see lots of good reasons, but the biggest one by far is that it makes it way more difficult to have a recession since you can't flee from capital assets to money. You can only sell capital assets for consumption. Which isn't going to cause a recession.

[K]

Phil: Your FX comment made me think that you could use the ETF money unit splits to keep FX rates fixed at unity. We could keep each medium of exchange backed by domestic capital assets, but there might be some convenience in having fixed exchange rates (like the pre Euro EMU).

[Nick Rowe]

Jacques: If banks hold 100% reserves against their chequable demand deposits, then changes in public preferences between currency and demand deposits have no effect on the size of M1, given the monetary base. The switch to 100% reserve banking would presumably require the central bank to make a one-time increase in the monetary base. But thereafter, the system would be much more stable, and you would need the CB to act as lender of last resort to banks with 100% reserves, or insure deposits.

[Nick Rowe]

Bill and Mike: your proposals are really very similar. When the borrower cannot repay the debt, then the debt converts to equity, and the institution (bank) carries on, possibly under new ownership. Instant automatic chapter 11. I like these sort of proposals. There was some talk about them in the blogosphere in 2008/9, but they seem to have gone quiet.

And I see your proposals as being a sort of halfway house to the sort of proposals I and K are making, with money being a form of equity rather than debt.

[Nick Rowe]

Ryan: private monies like LETS do exist in Canada and other countries, but they never seem to take off into wider acceptance. They seem to fail the test of the market.

[Nick Rowe]

I rather like the proposal K has been arguing for, with money backed by a market ETF. But why does it have to be the central bank running that money? Why can't it be a private bank, rather than a state-owned bank? These seem to me to be two separate issues.

[Nick Rowe]

BTW, who was the finance economist who proposed a monetary system very much like we are talking about here, with money being a stock mutual fund? Was it Fama, or Black? My memory is going.

[Nick Rowe]

123: (TMDB): Yes, any financial crash or bursting bubble will change the *distribution* of AD, both between people and between goods. That's probably unavoidable, not necessarily a bad thing, and, in any case, isn't really a money/macro problem anyway. It's when it causes fluctuations in AD itself it's a bad thing and a money/macro problem. And yes, the 1987 stock market crash and the 2000 dot.com stocks falling didn't create a money/macro problem, and that fact seems important to me. First it shows how equity finance is more resilient than debt finance. Second it shows that volatility in finance doesn't have to cause volatility in AD, and it won't if the volatility is in a part of finance that has little to do with money.

[Nick Rowe]

K: "I don't have a "prior" desire to have ETF money. I see lots of good reasons, but the biggest one by far is that it makes it way more difficult to have a recession since you can't flee from capital assets to money. You can only sell capital assets for consumption. Which isn't going to cause a recession."

This, to me, is very important.

Let me try to restate what K is saying (or, maybe, should be saying, just in case I've misunderstood him).

Say's Law is false. But it doesn't have to be false. We want a monetary system in which Say's Law is true. We want a monetary system where an increased demand to hoard money does *not* mean an excess supply of goods. We want a monetary system where an increased demand for money *is* an increased demand for Kapital goods (investment, in the macro sense of the word).

This point deserves a post on its own. It's sort of where I've been going with these two Blue Sky Money posts. In the first post, an excess demand for money created an excess demand for labour to work panning gold. Theoretically nice, but unworkable since there isn't a gold stream in everyone's back yard. And gold is a silly form of capital anyway. A more sensible and workable approach is where Kapital is itself money (or where a claim on Kapital like an ETF is itself money).

Blue Sky Money three - making Say's Law true?

[Nick Rowe]

Or, to put it another way, in Keynesian terms, we want an economy where desired saving *is* desired investment. And it's doable.

Did I understand you right, K?

[Martin]

Nick, it isn't just money that is the problem in your view of Say's law. It's debt. When an entrepreneur invests into supply, she does so because she expects to be able to sell the additional product from that investment to consumers. She expects so because there are people buying those products. What she does not see however is how those people are financing their consumption, she does not see what constitutes their power to demand. If it is supply, then they will be able to sustain their demand. If it is debt, that is not necessarily the case. And if it is debt, that entrepreneur's demand will also not come from supply, but from debt.

Making money income follow a stable path deals with but one friction. Regulating debt and credit is necessary to deal with this friction.

[Martin]

Or at the very least change the entitlement between debtors and creditors so that they have the incentive not to impose those 'externalities' upon unwitting entrepreneurs.

[Nick Rowe]

Martin: I disagree. I've done a number of posts arguing that *only* an excess demand for the medium of exchange can create an excess supply of newly-produced goods in a monetary exchange economy.

For example:
http://worthwhile.typepad.com/worthwhile_canadian_initi/2010/10/the-paradox-of-thrift-vs-the-paradox-of-hoarding.html

http://worthwhile.typepad.com/worthwhile_canadian_initi/2011/04/walras-law-vs-monetary-disequilibrium-theory.html

[K]

Exactly, Nick.

The old monetary system proposal you are thinking about may be Greenfield and Yeager's BFH. I think it involved convertibility into consumption goods, which I think is a flawed idea. Bill Woolsey has written about in the past.

[K]

Nick: "Exactly" was a reply to your Say's law comment. As far as Martin's comment on debt goes, well *I* think it depends on the choice of numeraire. Debt denominated in future units of present consumption basket is trouble because it creates the illusion of a collective ability to transport consumption in time. And you can tell that this illusion runs very deep because people get so angry when they suspect that the buying power of their money/bonds might be eroded. That's what so wrong with inflation indexing: money illusion should be redirected towards something that actually *can* be transported in time, ie capital assets which are (roughly) a claim on a fixed fraction (*not* a fixed quantity) of all future consumption.

[Ralph Musgrave]

Nick, You say (8.09am) “there is no essential difference between: letting ordinary people have chequing accounts at the BoC; and 100% reserve banks.” Strikes me that everything there hinges on the word “LETTING”.

The fact of “letting” people have chequing accounts at the BOC does not preclude fractional reserve: i.e. it does not stop commercial banks creating money or “lending money into existence” as the saying goes. (I think I’m saying much the same as K just after your 8.09 comment.)

In contrast, if no bank accounts were allowed other than accounts at the CB, that’s the same as full reserve. Alternatively, if every dollar deposited at each private bank must be backed by a dollar at the CB, that is also full reserve.

Ryan (8.53), I agree, to quote you, that money does not “want to centralize itself”. And I agree that central banking is not a “natural market phenomenon”. As to what the arguments are for a centrally managed currency, strikes me the arguments are as follows.

1. Free markets clearly have a tendency to boom and bust. Ideally those managing a centrally managed currency can mitigate these instabilities, though they are not at the moment not 100% competent in this regard, to put it politely.

2. A central bank can offer 100% safe accounts. Commercial banks can’t. Moreover, commercial banks are motivated to take the maximum possible risk with depositors’ money: and have taxpayers foot the bill when it all goes wrong.

Jacques René Giguère, That’s an interesting point about “double responsibility”. What’s your source for that?

Re 100% reserve banking being contractionary, I agree. But that is not a problem because the central bank can easily increase the monetary base to compensate. Indeed, this is just what central banks have done over the last couple of years in response to the credit crunch (not that I agree with the ACTUAL WAY they’ve done it: stuffing the pockets of the rich and ignoring Main Street).

Steve Roth, You say it is “a bit of a stretch” to call a $20 bill a liability of the central bank. Agreed. As Willem Buiter (former member of the Bank of England Monetary Policy Committee) put it, “These monetary (base money) ‘liabilities’ of the central bank are not in any meaningful sense liabilities, because they are irredeemable.”

[Martin]

Nick,

assume there is a two goods, one firm, and two households economy. This economy exists for two periods. The firm can produce two goods, exploding and non-exploding toilet-paper. The first period one household (A) lends everything to the second household (B). The second household proceeds to purchase exploding toilet-paper.

Second period the firm has closed down production for the non-exploding toilet paper based on the purchases of B and has invested in the exploding variety. B has to pay back A and cannot demand the exploding toilet-paper. A does not want the exploding variety, therefore does not buy and both A and B are unemployed.

What is the cause of this recession and excess demand for money or a mal-investment of capital based on the first period demand of B? I'd argue both, you argue that it is only the first. I agree that raising nominal income will fix it by wiping out debt relative to income, I disagree that therefore an excess demand for the medium of exchange is the sole cause. They're fungible. I'll give you that from the perspective of policy they're not after the crash, but before the crash they are.

If the debt contract is expressed in terms of nominal income, however...

[Martin]

Apologies for the typos and punctuation at the end.

[Jacques René Giguère]

Ralph Musgrave: from
"Banking en français"
http://www.amazon.ca/Banking-en-fran%C3%A7ais-Ronald-Rudin/dp/2890522423

http://artsandscience.concordia.ca/news/pdfs/rudin.pdf a history of early franophones banks in Canada
(available in either the original english or in french translation)

by
Ronald Rudin ( history professor at Concordia)

Nick: Am I lost? All is ok . I have seen the light...

http://artsandscience.concordia.ca/news/pdfs/rudin.pdf

[Ryan]

Nick, fair enough, but everyone knows you can't compete with the government. Even Porter Airlines needed special treatment to break into the Canadian airline market. I think we both know that's not a real market test.

[Mandos]

Nick: on another thread I mentioned Parecon (which stands for participatory economy). On the subject of "blue sky", although still kind of OT, I just thought I'd drop of a link to a FAQ on it. In a nutshell, it's an attempt (a VERY detailed attempt by its promulgators) to outline from the ground up a social/economic order based on a sort of anarcho-syndicalism. I'm not 100% convinced myself about it (I'm not 100% convinced about *any* worked out solution, as a rule) but it does have a *sort* of monetary policy, I guess, with a rather different aim.

[dlr]

K-

Let's assume bonds do not disappear (I don't think you want your idea to hinge on this possibility actualizing since it requires strong assumptions about preferences). Your MOE and MOA are not necessarily the lowest risk asset. So let's say you have a decline in the natural risky real rate. Stock prices, real estate prices, and thus representative ETF prices rise (or want to rise) relative to output prices and bond prices. So the demand for ETF-backed money increases, and output prices want to fall. This is like an increase in the demand for gold under a gold standard, but maybe a lot more common given the volatility of risk asset prices, even as a whole. Presumably the price level unexpectedly wanting to fall is still potentially problematic.

I assume in this case your CB would conduct monetary policy by accommodating the demand for money to keep prices or nominal spending along some stable path. The CB seemingly has to do this by changing its backing terms somehow. Maybe by doing a stock (money) split. But wait. Suddenly money is not really "backed" by capital assets at all. In the case of a stock split to prevent unexpected deflation, a holder of money is suddenly worse off than someone who held the ETF basket directly. Money is actually backed by the CB reaction function with respect to some nominal variable. They didn't hold a representative portfolio after all, they held a claim whose purchasing power was determined by nominal CB target. That sounds like standard CB fiat money. People would then actually treat money like a risk-free asset (ultimately determined by the CB reaction function and not by backing -- in the alternative case the CB would not permit unwanted inflation due to a large decline in risk asset prices), leaving intact the liquidity segmentation you are trying to extract from the monetary system. The backing would be as irrelevant as under a gold standard that was trumped by a nominal reaction function that operating to regularly adjust the price of gold. Or am I misunderstanding something from your proposal?

[K]

dlr: "Let's assume bonds do not disappear (I don't think you want your idea to hinge on this possibility actualizing since it requires strong assumptions about preferences)"

I agree.

"So let's say you have a decline in the natural risky real rate. Stock prices, real estate prices, and thus representative ETF prices rise (or want to rise) relative to output prices and bond prices."

That depends. The relationship between expected future growth and real yields is really complicated, and we can debate all the details if necessary. But to first order real rates rise when the expected rate of GDP growth rises. In that case capital assets do not respond to changes in the growth outlook because changes in risk free real yields exactly offset the change in expected growth. Other issues relate to expected central bank failure, the rate at which future innovators usurp the claim on growth from current owners of capital, and above all "risk premium". Most of our past bubbles appear to have been related to declines in risk premium. Certainly real yields never rose during either the dot com or housing bubbles. (It would have been great if we'd had an NGDP futures market to see what was happening to expected GDP but unfortunately that's missing data.) The important point though, is that the value of the whole market doesn't depend directly on the outlook for future capacity growth.

"In the case of a stock split to prevent unexpected deflation, a holder of money is suddenly worse off than someone who held the ETF basket directly."

No. Everyone gets their unit holdings split. Ie. they all get more units. So their claim on capital assets is unchanged like in a regular stock split.

[dlr]

The important point though, is that the value of the whole market doesn't depend directly on the outlook for future capacity growth.

I'm not sure I agree that this is the important point. I think the important point is that the price of future risky consumption (your ETF) will surely be volatile in terms of current output and safe future consumption. It isn't enough to say that this volatility doesn't depend directly on the outlook for future capacity growth. The very fact that it will surely vary for myriad reasons (likely more even than (say) the demand for gold in terms of output might), presents problems if the price level is sticky. It is a nice feature that a riskier medium of account doesn't invite reflexive searches for safety, but the price level volatility still seems like a serious problem.

No. Everyone gets their unit holdings split. Ie. they all get more units. So their claim on capital assets is unchanged like in a regular stock split.

Wait a minute. If this is the kind of monetary policy split you have in mind under a backed regime, you are powerless to stabilize nominal spending if the demand for the backing changes relative to current consumption. So what do you do if time preference and/or risk aversion declines and (and there is no perfect offset) and people want to bid up the price of risky future consumption? A unit split like you imply above does nothing. You have to change the backing terms to avoid pressure on the price level, right? What do you envision as the nominal anchor in your world? Isn't it the price of whatever risk-level of future consumption the CB is able to represent via a "representative" ETF in terms of goods and services?

[jean]

If money is not based any longer on debt, how can its value be credibly maintained? If you want the money to have a stable value, you need to be able to provide service and goods (gold, CPI basket, GDP fraction) to which money is pegged or to be able to withdraw an important part of the money supply. Of course, like in the Chuck Norris posts, these capabilities don't need to be really used, but some credibility is needed. For the second capability, I don't see any other way than holding short term debt or benefiting from the fiscal support of the state.

[Too Much Fed]

"Unstable money is a very bad thing"

Is it unstable medium of exchange because demand deposits from loans can be defaulted on and paid off?

"Why do we have an economy in which money is linked to unstable finance?"

You need to able to answer that question correctly.

"If a financial crash didn't cause an excess demand for money, and the resulting recession, only the microeconomists would care."

How about if a too much currency denominated debt crash didn't cause an excess demand for medium of exchange?

"Money is a medium of exchange and medium of account. It's got nothing to do with borrowing and lending."

I don't believe that is correct the way the system is set up now.

"I can imagine an economy with money and no finance. People never borrow and lend."

I can too. I believe it would be a lot better. *** One entity would probably need to dissave in the correct way ***

"The Bank of Canada is not a bank."

The way JKH and others have explained it to me is that the central bank is like a bank even if it is not on the gold standard or doesn't have a fixed exchange rate.

[Too Much Fed]

"A 100% reserve bank cannot extend loans. There's no regulation needed, except to check that it really isn't making loans, and does have 100% reserves.

So, I'm arguing that there is no essential difference between: letting ordinary people have chequing accounts at the BoC; and 100% reserve banks.

Can anyone else explain this more clearly than I am able to?

Now, if the BoC started making loans to ordinary people, that *would* be a difference."

That is not what Bill Mitchell says.

http://bilbo.economicoutlook.net/blog/?p=16407

Saturday Quiz – October 8, 2011 – answers and discussion

Question 1:

Some “Occupy Wall Street” protesters are demanding that bank lending should be more closely regulated to ensure that all bank loans were backed by reserves held at the bank. However, this would unnecessarily reduce the capacity of the banks to lend.

The answer is False.

Explanation follows at the link.

[K]

dlr: "You have to change the backing terms to avoid pressure on the price level, right?"

Which you are. If you split the unit of account in two, each unit will only be backed by half as much capital, right? So goods whose prices are stuck in nominal terms will drop in real terms by a factor of two. So you can both target whatever you want for the value of the unit of account *and* deny people a fictitious monetary safe harbour.

The effectiveness of that depends on your theory of what the CB is doing. But if, like many (most?) modern monetary theorists, you believe that nominal rigidities (disequelibrium sticky prices/wages) are a crucial part of the mechanism of macro disequilibrium then a purely nominal solution seems like the ideal solution.

As far as the real value of capital assets being naturally volatile, I'm not so sure. As far as I know, the history of market volatility is very difficult to explain in terms of the realized volatility of profits. I think money illusion is a significant driver of risk premia and that it is arbitrary changes in risk premia *themselves* (essentially changes in our "animal spirits") that drive most of volatility. And if AD itself is well regulated at the LRAS then we will also lose divergences between real interest rates and the natural rate as a source of capital asset volatility. I simply don't think that what's left is very significant. But no, that certainly doesn't constitute proof.

[nemi]

Nick:
I don´t understand your distinction between thrift and hoarding. The only property I can think of that money has, and other stores of value does not, is, to some extent, extreme liquidity.

Is your theory that people are not able to sell their assets, and thus cannot convert their assets to demand for goods? I mean, why would they want to hold money except for the reason that they want to save (in an asset with relatively low risk)?

I can see why a “tax” on money would decrease your willingness to hold money, and for equilibrium to be restored, the payoff of all kind of savings would have to decrease – but why would it amount to something special if you taxed one type of asset, e.g. money, compared to another, e.g. stocks (given a simplistic neoclassical “always and everywhere equilibrium model”). Isn’t the thing you really want to tax “saving/thrift”?

Of course it is very easy to increase the amount of the particular asset “money” compared to other assets, and that is relevant with respect to how to combat e.g. a recession, but that feature cannot be used to explain why there was an excess demand for money to start with (I think).

[Nick Rowe]

nemi: line up all the assets in a row, from most liquid to least liquid. Money is at one extreme end of the liquidity spectrum. But even if it's only slightly more liquid than the next asset in line, that small difference makes all the difference in the world. Money is the medium of exchange. All other goods, and assets, are bought and sold for money.

If there's an excess demand for any other asset, too bad. Everyone wants to buy some more, but can't, because nobody wants to sell. End of story. But if there's an excess demand for money, then each of us, as individuals can always get more money by buying less other things. And that's what causes a recession.

There's only one way to get more land -- buy some more. There are two ways to get more money -- buy more, or sell less. If people sell less money (i.e. buy less other stuff) there's a recession.

[Ralph Musgrave]

Too much Fed,

I’ll take your Dec 4 (11.11pm) comment bit by bit.

“A 100% reserve bank cannot extend loans.” My answer: yes it can. It just has to find a saver to fund the loan rather than create the money needed out of thin air.

“There's no regulation needed, except to check that it really isn't making loans, and does have 100% reserves.” My answer: it CAN lend for reasons just given, but regulation is certainly needed to make sure private banks are not creating thin air money. The way to do this, far as I can see, is to ensure that private banks settle up at the end of each day’s trading at the central bank (which they currently do anyway). It would be obvious from this “settling up” system if any private bank was surreptitiously creating money.

However, private banks could always get round this by settling up between themselves, though doing so is inefficient compared to doing it the central bank way. But I’m sure these regulations would never be 100% foolproof: regulations never are.

I don’t agree with Bill Mitchell. His answer to the accusation that full reserve would constrain bank lending is to point out that banks don’t need reserves to lend. Well all he is saying is that banks can get round the constraints of full reserve by engaging in fractional reserve!!!! That is an illogical or “begging the question” kind of answer.

Full reserve certainly constrains bank lending – ALL ELSE EQUAL. In particular, if the monetary base is left constant, then lending is constrained. But of course there is nothing to stop central banks expanding the monetary base: the US and British base have been expanded by unprecedented amounts over the last two years. (Canada as well, presumably.)

[Min]

Nick Rowe: "But if there's an excess demand for money, then each of us, as individuals can always get more money by buying less other things."

Well, some of us have a zero lower bound. ;)

[Min]

jean: "If money is not based any longer on debt, how can its value be credibly maintained?"

The experience of the American colonies in the 18th century showed that the value of fiat money can be maintained in two ways. One, as you say, is by debt. Pennsylvania created money by lending. Another is by taxation, as the MMT people point out. During the Revolution Benjamin Franklin urged the Continental Congress to support the Continental Dollar by taxation, but it did not, and the result was hyperinflation.

[Rob]

"There's only one way to get more land -- buy some more. There are two ways to get more money -- buy more, or sell less"

Surely for both money and lend the same rules applies "There is only one way to end up with more money/land - buy more than you sell ? I believe the difference may be that if everyone's demand for land increases this affects the land-market. If everyone's demand for money increases then this affects every market.

[Becky Hargrove]

Given the news out of Europe now, it's easy to see how people might really want to separate their money from the whims of the financial world. "No haircuts" may have taken things a bit too far.

[Bob Murphy]

Nick, I'm echoing one of the other commenters above, but I too kept thinking while reading your post, "100% reserve banking with a commodity money solves all of these problems."

I'm not saying that rule needs to be imposed by the government, either. I think if the government stepped back and let nature run its course--with depositors getting wiped out if their bank gambles with "their money" etc.--that there would be an option for very safe demand deposits backed close to 100%. The system as a whole (I claim) would be a lot less fragile than what we have now. You are saying "we" decided to leave gold etc. and you think it was a good thing, but I disagree.

[K]

Bob: I don't know if I was the commenter you were referring to, but I'm not in favour of a gold standard. Gold is as bad as unbacked money because it's actually fairly worthless at least compared to the global money demand. The principal feature of the classical gold standard wasn't that it was backed by something. It was the fact that the unit of account under that system wasn't allowed to be split which resulted in severe nominal rigidity problems. My principal point was actually that 1) money should be backed by something intrinsically valuable since it is inefficient to force people to surrender valuable capital assets in order to satisfy their liquidity requirements; 2) that something should be, as broadly as possible, representative of the market portfolio since a) that's what we'll be holding on average anyways and b) that's what modern portfolio theory tells us we should be holding and 3) if money *is* capital assets then we can't flee capital assets to hold money thus causing recession. None of that can be said for gold.

[K]

But I think it's pretty remarkable that despite some disagreement in the above comments, no one has actually come out to make a determined case in favour of state sponsored fractional reserves. Maybe this is one for Mike Moffatt's Economist's Party platform.

[Sergei]

Nick, you make a lot of right half-steps but still end up with "I dunno" leaning towards a standard but wrong conclusion about 100% reserve banking.

You say: Money is a medium of exchange and medium of account. It's got nothing to do with borrowing and lending

And you say: Money and banking don't have to go together

And finally: Dunno

Your "dunno" solution is to separate payment system (exchange medium) and banking (borrowing and lending). It is an easy solution given current technology. Any central bank is effectively a payment system and everybody can have an account for payments and an account for savings. Any central bank acts as a banking agent for government and often as debt portfolio manager. So such central bank will directly sell government bonds as savings accounts. Everybody in the economy will have 3 distinct choices: risk free payment account with no interest, interest bearing but credit risk free savings account, and interest bearing but also subject to credit risk bank account. Banks will still be banks but they will not be allowed to run payments and will have to back all lending with term deposits. Such term deposits will have no guarantee.

Dunno is to separate money and banking system which is to separate payments system from banking system.