Understanding Banks and Two Monetary Policy Instruments

Sometimes I write posts about things I don't understand and want to understand. (Or don't understand very well and want to understand better.) This is one of those posts.

Suppose the government nationalises all the commercial banks, and merges them into the Bank of Canada. And suppose the Bank of Canada abolishes paper currency. So everyone has a chequing account at their local branch of the Bank of Canada, and they use that chequing account to buy and sell everything else. No other financial institution is allowed to offer an account with chequing (or debit card) privileges, just like no other financial institution is currently allowed to print paper currency that could compete with the Bank of Canada's paper currency.

In that imaginary Canada the Bank of Canada has two monetary policy instruments. First, it can vary the quantity of money by buying or selling IOU's (bonds, loans, etc.). Second, it can vary the rate of interest it pays on chequing account balances (and it can make that rate of interest negative if it wishes). [It could also vary the limits placed on overdrafts; and could also vary the spread between the interest rate charged on overdrafts and paid on positive balances. But let's set all those other instruments aside.]

I understand how monetary policy would work in that imaginary Canada (at least, I think I do). Increasing the quantity of money (holding the interest rate paid on money constant) shifts the LM curve to the right/down. Increasing the rate of interest paid on holding money (holding the quantity of money constant) shifts the LM curve left/up. Done.

It's a crude model of an artificial economy. But it's a helluva lot better than a simple New Keynesian model where money (allegedly) does not exist and the central bank (somehow) sets "the" nominal interest rate (on what?).

What I do not understand well enough is how well that thought-experiment translates into the real Canada, where only the commercial banks issue chequing account money for the public, and only the Bank of Canada issues paper currency, and paper currency always pays 0% interest. But the Bank of Canada still has those same two instruments: it can buy and sell IOU's; and it can vary the interest rate paid on the commercial banks' chequing accounts at the Bank of Canada.

[Remember that commercial banks can pay interest in kind, like when they reduce fees for balances above some amount. And remember that when a commercial bank makes a loan it is buying the borrower's IOU.]

Comments

[JKH]

Nick

I think your second paragraph is an excellent starting point for analysis of the real world.

(You've used this construct many times, and I always thought that.)

From there I would model the creation of a commercial banking system as the delegation of a substantial portion of those same mechanics to the competitive, capitalist sphere.

The central bank can do both OMO and interest rate setting in your 2nd paragraph model.

It continues to do those things in the real world.

But the commercial banks do it as well.

In reverse order, they do interest rate setting, guided by the economics of where the central bank sets the policy rate, along with associated nearby market considerations and profit margin requirements.

The rate on short term demand or notice deposits is an obvious example. Banks do now pay a positive rate on interest on such deposits. Spreads required for profit are an obvious consideration.

The rate on variable rate mortgages is another interesting example. There's been some tweaking lately, (notwithstanding a stable bank rate), but the starting reference point is the central bank rate. Many other commercial bank rates on variable rate assets and liabilities are associated with the Bank of Canada rate as the starting reference point.

The key I think is that the BoC rate is an administered rate rather than a market rate (unlike the relationship between market bond or interest rate swap rates and 5 year fixed rate mortgages for example). It's a policy rate anchor in that sense.

And commercial banks do OMO in the fundamental sense of creating deposits when they make loans and in the more comparable sense of creating money when they buy securities from the nonbank sector.

[Roger Sparks]

Nick ends with "And remember that when a commercial bank makes a loan it is buying the borrower's IOU."

We have two implied assumptions here: 1. An object with present value is being traded for an object with future value. 2. Two trades are contemplated: a present trade and a future trade.

I have to ask two questions:

1. Why would a commercial bank want a "borrower's IOU?

2. What is the commercial bank using in trade for the IOU.

I will attempt to answer both questions:

1. The bank will observe that the borrower's IOU could (or could not) be worth more in the future than it is today. If worth more, the bank would like to share in the gain.

2. This is the key question. The bank would like to share in the gain but what is "gain"? Gain is a complex balance calculation comparing what is used for present trade object against what will be received in the future.

For the present trade, to "buy an IOU", any bank would use money. Where would the bank get the money and at what cost to the bank?

For the future trade, the bank would expect to get the money back. What value of money would it receive back?

The answers to these questions (and more) would allow the bank to form a favorable/unfavorable opinion about the transaction.

[JKH]

Maybe this will help a bit:

Commercial banks run complex transfer pricing systems in order to manage interest margins on their assets and liabilities as distinct portfolios.

For example, the demand deposit manager will "sell" those funds to an internal clearing house that pays a rate and a positive margin to that portfolio.

The variable rate mortgage manager will "buy" funds from the clearing house in order to fund his portfolio.

Pricing signals abound in this system - including market bond and interest rate swap market, as well as the Bank of Canada policy rate (and/or closely correlated very short term market rates).

Any change in the Bank of Canada rate is a significant event for ensuring the updating of such price signals.

[Ralph Musgrave]

Nick, I find your question imprecise. You say you don’t fully understand something. But that something is fairly complicated, thus without you saying EXACTLY what you aren’t clear on, it’s difficult to know where to start with an answer. Still, I’ll try.

Also I find this passage puzzling, “What I do not understand well enough is how well that thought-experiment translates into the real Canada, where… the Bank of Canada…. can vary the interest rate paid on the commercial banks' chequing accounts at the BoC.” Surely in the “real world” commercial banks do not have chequing accounts at the BoC – at least HOUSEHOLDS’ accounts are not there. They’re in the books of commercial banks.

Anyway, in the real world (at least before the 2007/8 crisis) central banks manipulated interest rates by keeping commercial banks short of (or slightly less short of) reserves (which they need for settling up purposes). To illustrate, to cut interest rates, the BoC would purchase government debt, which supplies commercial banks with more reserves (aka more base money).

However, with vastly more base money now sloshing around thanks to QE, I’m interested to know how central banks propose raising interest rates: presumably by paying more interest on reserves. But I don’t care for that: it involves rewarding people for simply hoarding money.

[Majromax]

> In that imaginary Canada the Bank of Canada has two monetary policy instruments. First, it can vary the quantity of money by buying or selling IOU's (bonds, loans, etc.). Second, it can vary the rate of interest it pays on chequing account balances (and it can make that rate of interest negative if it wishes).

The problem with the analysis is that these aren't fully independent tools. If the net money supply is to remain constant, paying interest on balances now commits the central bank to selling IOUs later. Actually making that sale (to a willing buyer) will require yields on those IOUs in excess of the rate of interest on balances.

You've invented a yield curve. The rate of interest on balances is the short-term rate, the rate of interest on IOUs is the long-term rate. This adds an extra dimension to IS/LM, and it complicates New Keynesian analysis by providing two distinct "the" interest rates.

[Majromax]

As an addendum to the above, coming at this form a math but not economics standpoint I have a hard time of thinking of a constant equilibrium. It seems very strange to me that we can talk about overdrafts, interest, and IOUs without also talking in great detail about time.

The dynamics are important for my intuition. More concretely, my intuition tells me that the Fed's 'Operation Twist' should have failed precisely because it sent Neo-Fisherian signals.

[Ralph Musgrave]

Having said above that one way of upping interest rates given commercial banks' current very large stock of reserves is to pay more interest on reserves, another is to impose or increase minimum reserve requirements. I think the Chinese do that more often than is normal in the West.

[Antti Jokinen]

Nick: Your second paragraph reminded me of Wicksell. Change "Bank of Canada" to "Norges Bank" and we will go nearly 100 years back in time. Below is a second-hand report on what Wicksell said in "The Scandinavian Monetary System After The War" in 1917. I think this was originally a lecture given at Norges Bank and published in a Norwegian journal, but I haven't yet found the original online (I'm fairly fluent in Norwegian and Swedish).

I rely on what I believe is a quite obscure paper by Boyanowsky ja Erreygers: ”Social Comptabilism and Pure Credit Systems: Solvay and Wicksell on Monetary Reform”, 1999. (The subject is interesting! Unfortunately I'm no good in French and so I don't get the many quotes they haven't translated. If I remember correctly, there's for instance Walras commenting on Solvay's ideas.)

This is what Boyanowsky ja Erreygers say about Wicksell's proposition:

-------------------------------------------------------------------------------------------

He explained the transition to pure credit using as an example the balance sheet of the central bank of Norway, which is supposed to be, for now, the only bank of the economy. For the sake of illustration, Wicksell (p. 179) imagines that the bank has before the reform assets consisting of gold reserves (100 million krona) and claims (200 million krona), and liabilities in the form of deposits (100 million krona, which also includes the capital of the bank) and notes in circulation (200 million krona). With the permission of the Norwegian parliament, the central bank calls in its notes; half is paid out to note holders in gold and the other half is put down as deposit at the individual’s bank account, on which he or she has the right to draw cheques. Cheques cannot be paid out in cash (gold or notes), but only be transferred to other accounts. At the same time, the bank stops issuing notes and accepting gold as deposit or in payment. The bank’s total assets have now been reduced by 100 million krona (the gold reserves that were given out) and comprise, on one side, claims (200 million krona, as before) and, on the other, deposits (200 million krona, including the bank’s capital). That sum can increase through new loans (which bring about directly or indirectly corresponding deposits in the bank) and can be reduced through payment of debts to the bank. Wicksell acknowledges that at the beginning gold coins would probably be used as means of payment between individuals, but - since they are no longer received by the bank, nor (“let us assume”) held as cash balance – they would soon cease to be used as means of payment and, instead, be melted down and absorbed by industry. In such an economy, “money means henceforth only the unit in which the bank’s accounts are kept.“

--------------------------------------------------------------------------

It is interesting that Wicksell thought of this as a "pure credit" system (and I more or less agree with him!), while you see 'money' on the accounts? Your view is also at odds with Schumpeter's, whom I quoted in a footnote to my latest post: "… in a pure account-settling system the concept of money supply would correspond to nothing at all."

I feel that Wicksell was greatly influenced by Jevons, and I dare to say that he, too, would view that kind of system as a "complicated and perfected system of barter" (see chapters from XIX onwards in "Money and the Mechanism of Exchange"). I'm under the impression that even John Hicks, later on in his life, expressed similar thoughts (anyone?). (I recall him talking about "pure credit" and how he thought we were close to that kind of system.)

Wicksell's “money means henceforth only the unit in which the bank’s accounts are kept“ above could be interpreted as compatible with my "abstract unit of account" view.

Did the "old guard" understand something important which is now mostly forgotten? I think so.

JKH: Thanks for enlightening comments, once again!

[Nick Rowe]

Thanks JKH. Glad to hear you think it might work.

Let me ask you: assume we are well above the ZLB. If the Bank of Canada cuts the deposit rate and Bank rate by 1%, and does not do OMO (holds the size of its balance sheet at least roughly constant), how do you think would commercial banks respond?

Cut deposit rates across the board? By 1%?

Do Open Market Operations (increase loans, by cutting loan rates to increase demand)?

Bit of both?

[JKH]

Nick

If well above the zero bound, the banks would likely cut the prime rate, the variable mortgage rate, and administered variable deposit rates by one per cent.

"Well above" is relative, but 5 per cent or more is a pretty safe bet for smoothly marching administered rate adjustments.

Moving below that is a bit dicier in terms of precisely matching Bank of Canada rate declines. Deposit rates would already be well below a 5 per cent bank rate and heading toward the zero bound. The idea of cutting those rates with full force becomes painful from a marketing perspective.

There is a more general issue of rate compression that has to do with the fact that no interest is paid on the book equity position of a bank. This means that as the general level of interest rates declines, the net interest margin attributable to book equity as a source of funding gets squeezed. This is a complicated and critical part of bank interest rate management, exacerbating the problem that occurs when interest rates paid on deposits also start approaching the zero bound.

I'd say there is an increasing caution about matching Bank of Canada rate cuts 1:1 as rates approach the zero bound, and that caution kicks in before rates actually hit the zero bound.

On the loan side, the same issue holds. If there is resistance in cutting deposit rates 1:1, then the same resistance occurs with loan rates - in an attempt to preserve margins.

Demand will be what demand will be at the new interest rate level. Supply will be driven by credit assessment and return on equity pricing criteria, just as it is at any general level of interest rates.

[Nick Rowe]

Ralph: the commercial banks have reserve accounts (that can have either a positive or negative balance) on the books of the Bank of Canada. Those accounts work for the commercial banks in very much the same way that chequing accounts work for people. So I call them "chequing accounts". The Bank of Canada sets two interest rates: the "deposit rate" that it pays on positive balances; and the "bank rate" that it charges on negative balances. They are always 50 basis points apart. That creates a floor and ceiling for the overnight rate (the target for which is halfway between those two rates). The Bank of Canada fine-tunes by adding or subtracting tiny amounts of reserves.

Majro: remember the central bank can vary the amount of profits it hands over to the government.

Economists no longer think in terms of a constant equilibrium either. "Equilibrium" just means "the solution to a system of equations", and it can be moving over time.

Antti: very interesting quote there. Yes, it's very much the same as the banking system in my 2nd paragraph.

For a long time economists thought of chequing account balances as "credit", and not really money, like paper currency. (And, I suspect, for a long time before that they maybe thought of paper currency as just another form of credit, and not really money, like gold coins.) And I think I read the LETS people as often saying that LETS is just a sophisticated system of barter, not money. I disagree with them all. If there's a central clearing house, so offsetting crefits and debits get cancelled around the triangle/circle, it's money to my mind.

Money and the Clearing House

Money, barter, the clearing house, and balance sheet recessions

There is an essay by John Hicks I read decades ago, and have never been able to find since. He talked about an accounting system to make sure people didn't keep on buying more than they sold. It was right up your street, if I remember correctly. But I don't remember the title, or book it was in. But I think he wrote it later in life. Are you maybe remembering the same one? My memory is crap (but then I'm older than you, so I have an excuse).

[Antti Jokinen]

Nick said: "For a long time economists thought of chequing account balances as "credit", and not really money, like paper currency. (And, I suspect, for a long time before that they maybe thought of paper currency as just another form of credit, and not really money, like gold coins.)"

I expected you take this up. I think you have identified the central issue. But there are two ways to approach it. Yours is the conventional one:

Paper currency serves as a replacement for gold coins. Thus, paper currency is money. Positive checking account balances at commercial banks serve as a replacement for paper currency. So they are money, too. And when you say 'money', you see the 'stock' and you see 'velocity' -- you haven't abandoned the Quantity Theory.

Mine is this:

Positive checking account balances are credits. Paper currency is a "portable, bearer credit balance". It is clear that the Quantity Theory doesn't apply here; it's invalid.(I'll explain my view on the link to gold later. Trying to do it here would probably just create unnecessary controversy. It's a bit longer story.)

Nick said: "And I think I read the LETS people as often saying that LETS is just a sophisticated system of barter, not money. I disagree with them all. If there's a central clearing house, so offsetting crefits and debits get cancelled around the triangle/circle, it's money to my mind."

If you want to call it 'money', fine. But why would there be in that kind of system, in the Quantity-theoretical sense, something we should recognize as the money stock, and how do we define velocity?

A simple example:

We have 100 pairs of agents ready to trade (one sells, the other buys). These agents' checking account balances are all zero (we can assume "a new monetary system from scratch"). The goods about to be bought and sold are all priced at $1,000. All the buyers accept this price.

The trades are executed and the bank accountant is instructed to make the entries on all the accounts. He makes them.

We have witnessed 100 fairly large trades -- important activity in the economy. The "gross money stock", as you define it, is now 2 x 100 x $1,000 = $200,000 (Oliver and many others would say $100,000). What is velocity? And if the velocity, as you define it, is measurable, then what does it tell us?

Can you see how it makes at least some sense to see no money serving as a medium of exchange in this case? How Schumpeter can conclude that "the concept of money supply would correspond to nothing at all"?

[Antti Jokinen]

I said: "And if the velocity, as you define it, is measurable, then what does it tell us?"

Should read: "And if the velocity, as you define it, is measurable, then what does that number tell us?"

[Johan Meriluoto]

Nick wrote: "For a long time economists thought of chequing account balances as "credit", and not really money, like paper currency. (And, I suspect, for a long time before that they maybe thought of paper currency as just another form of credit, and not really money, like gold coins.)"

Indeed, for example, in a ruling from 1702 (Ward v. Evans) on paper notes, although specifically concerning the goldsmith-bankers' notes, and contrary to the merchants' practices, Chief Justice Holt declared: "I am of opinion, and always was (notwithstanding the noise and cry, that is the use of Lombard street, as if the contrary opinion would blow up Lombard Street) that the acceptance of such a note is not actual payment."

However, Holt's decree was already losing ground; the notes constituting 'conditional payment' for precedent debts but 'final payment' for simultaneously contracted ones.

[Nick Rowe]

Antti: "Positive checking account balances are credits. Paper currency is a "portable, bearer credit balance". It is clear that the Quantity Theory doesn't apply here; it's invalid."

I don't understand that. What's the difference between positive checking balances and paper currency? Or are you saying they are both credit balances? And what's the "here" in "the Quantity Theory doesn't apply here"? Checking balances, or both checking balances and currency?

Johan: interesting! I'm trying to think of a modern analogy:

Consider two cases:

1. Both buyer and seller bank at BMO. Buyer pays by cheque, which the seller deposits in his account, then BMO suspends convertibility into BoC currency.

2. Buyer banks at BMO, seller banks at TD. Buyer pays by cheque, tries to deposit in his account at TD, but TD refuses to accept it because BMO has suspended convertibility.

In case 1 I think we would say the buyer has paid. In case 2 I think we would say the buyer has not paid, and the seller reclaims his car.

[Majromax]

@Nick Rowe:

> Majro: remember the central bank can vary the amount of profits it hands over to the government.

I don't think this is an independent tool, since the amount of profits are entirely defined by the interest paid on money and the sequence of OMO actions. Besides, I think we'd think very differently of a central bank that persistently ran a loss, such that the fiscal authority had to run a primary surplus to finance money-destruction – it wouldn't "feel" like monetary policy.

Perhaps we could say that the NPV of the government's future primary surplus is a third monetary policy variable? If we hold that constant for the sake of argument, then varying the level of CB profits is an accounting wash from other spending/tax changes.

> Economists no longer think in terms of a constant equilibrium either. "Equilibrium" just means "the solution to a system of equations", and it can be moving over time.

My biggest difficulty is with the dynamics of it. As in this post, the combination of OMOs and interest on money to me sets out a yield curve, and that should have a real impact on the economy. The idea that OMOs and interest on money both shift the LM curve (and that's all) doesn't really capture that impact.

> Let me ask you: assume we are well above the ZLB. If the Bank of Canada cuts the deposit rate and Bank rate by 1%, and does not do OMO (holds the size of its balance sheet at least roughly constant), how do you think would commercial banks respond?

Commercial banks would seek to borrow reserves to purchase short-term government bonds until the secondary-market yields on them fell to approximately the reduced bank rate. If the central bank (consolidated with the government) conducts no OMOs and the loans are collateralized by these bonds, then there might be a shortage in the relevant bond market. If the loans are uncollateralized, then I'm not sure we can fairly say the central bank's balance sheet is unchanged.

@Antti:

> The trades are executed and the bank accountant is instructed to make the entries on all the accounts. He makes them.

You've taken the limit as settlement time goes to zero, then looked at behaviour. You need to do it the other way around.

Imagine all the retail banks are cautious: they withdraw funds immediately, but there's a 24-hour hold on deposits before it hits the chequing account. If that applies across your example, then every agent needs to pay one day's overdraft interest. There's your liquidity premium.

In the Red+Green world, we need to look carefully at who can create money.

If the central bank creates red and green money only by its own whim, then we have a well defined red and green monetary base but the transaction sequencing you describe might not be possible. Additionally, we can observe a shortage of money, where two people without money are completely unable to trade no matter how willing they are.

If the central bank (or commercial bank) creates offsetting red+green money on demand, then we don't have a well-defined red/green monetary base. The world is more analogous to ours, where if I use a credit card to pay a merchant the quantity of green money in circulation actually increases, and QTM looks better with M2 than MB. In this world, we can't observe a strict shortage of money – two people without money can trade if one of them creates red+green ex nihilo to facilitate the trade. In this world, the interest rate paid/charged on money causes the liquidity premium, as with a sufficiently large interest rate the aforementioned trade might be possible but the buyer would not be willing.

[Antti Jokinen]

Nick said: "What's the difference between positive checking balances and paper currency? Or are you saying they are both credit balances? And what's the "here" in "the Quantity Theory doesn't apply here"? Checking balances, or both checking balances and currency?"

I just followed your example where you said that in the past people thought that checking account (credit) balances are not money, like paper currency. For me, both should be viewed as credit balances. Quantity Theory applies to neither.

Bear in mind that when I say 'credit', it doesn't mean that its holder is owed money. If paper currency is a 'credit' (to its holder; liability to no one in particular), then there is no money in existence that could be owed to the holder of that credit. I guess this separates me from most of the people who used to say that credit balances on checking accounts are not money, but credit? They saw those as "promises to pay money".

[Nick Edmonds]

A few points on the impact of OMOs on commercial bank rates.

If the central bank carries out OMOs, the marginal impact is on the level of reserves. If people end up holding too much cash, they exchange it at their commercial bank for a deposit. Unless there's an actual cash shortage, the amount of cash in circulation is just what people want to hold; the balance gets held as reserves. So the fact that cash pays 0% makes little difference when reserves are interest bearing.

The impact of OMOs depends on whether the bonds involved are short or long dated. If they are short dated, then the rate on those bills is not going to change much if the rate on reserves is pegged. A certain amount of reserves is needed to meet clearing and possibly regulatory requirements, but beyond that they have no special liquidity benefits to banks compared with bills. The main requirement is for banks to meet liquidity coverage ratios, but bills can serve this purpose as well as reserves can.

There are some minor differences. As non-banks cannot hold reserves directly, there is a wider investor base for bills, so in a situation with large amounts of reserves, you can easily find bills trading through the reserve rate. Going the other way, the BoE recently suspended the inclusion of reserves in the calculation of the leverage ratio, which could make reserves more attractive than bills for some banks. But the impact of all this is pretty small. So deposit rates and loan rates are unlikely to change much.

If OMOs are conducted at the long end, then there is much more scope for the rate on long dated bonds to change. This may then effect the cost of funds for corporate bond issuers or equity issuers or people taking out fixed rate loans. But again the effect on (short dated) deposit and loan rates is likely to be limited. Banks do hold long dated bonds, but in general the exposure to long rates is hedged.

[Antti Jokinen]

Nick said: "And remember that when a commercial bank makes a loan it is buying the borrower's IOU."

I'm mostly off-topic, as always, but this got me thinking:

How would you describe this in case of an overdraft?

There is no loan, no borrower?

There is no IOU? (For you, an overdraft is a liability, not a debt.)

But in both cases, a negative (debit) balance in the bank's books is involved.

[Antti Jokinen]

Nick: Even I don't have time to answer all comments directed at me, so I understand it's impossible for you. I also don't like when people push me to answer their comment.

That said, I want you to know that I felt that this comment of mine might lead to a chance to break some new ground. I put a lot of thought behind it. I felt that you might not have any ready-made answer to it. Partly because of that, I don't expect you to answer it now. But it would be nice if it existed as a stain on your brain, so that it would be easier to get back to the subject at some point in future :-)

[Too Much Fed]

"And remember that when a commercial bank makes a loan it is buying the borrower's IOU."

I would say that it is new demand deposits for a new borrower's IOU, not existing in either case.

"In that imaginary Canada the Bank of Canada has two monetary policy instruments. First, it can vary the quantity of money by buying or selling IOU's (bonds, loans, etc.)."

JKH said: "Demand will be what demand will be at the new interest rate level. Supply will be driven by credit assessment and return on equity pricing criteria, just as it is at any general level of interest rates."

I don't think the "BOC" here has any discretion with new loans. If creditworthy borrowers show up and the spreads are good, then the "BOC" has to buy new loans/bonds. Nothing like the loan/bond is denied because that will cause too much "money" to be created leading to too much price inflation.

I think the "BOC" has some discretion with buying existing IOU's/bonds.