Money Creation in the Modern Economy (Bank of England)

This is an excellent paper:

http://www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2014/qb14q1prereleasemoneycreation.pdf

The paper is written in easy to understand language and helpful balance sheet graphics. It does a particularly good job in simplifying several technical issues. As Cullen Roche notes, the money multiplier is dispatched to the ash can of history, where it belongs. It is reassuring to see this sort of recognition coming out of the central banks as regular production now, although that has been the case to some degree for a while now. And Ramanan  points to some interesting post Keynesian references and influences in the paper.

The paper is impressive in that it also makes a number of insightful points apart from the issue of the money multiplier – points not often seen elsewhere – and treated here matter-of-factly and effectively. Moreover, its scope extends well beyond money creation and into a wider description of the dynamics of money circulation and bank risk management.

The section on quantitative easing in particular is excellent because (unusually) it immediately identifies a fundamental aspect of QE overlooked by most commentators – which is that QE routinely creates broad money as well as bank reserves. I suspect future economic history books will look back on this QE episode and point to a general failure by present day analysts and economists to consider this balance sheet aspect more thoroughly in figuring out just how QE works.

QE is not just about bank reserves. In fact, the bank reserve effect is arguably the least important feature of QE – although that opinion may not exactly warm the cockles of monetarist hearts. The paper appropriately refers to reserves as the “by-product” of QE (a concept I invoked in a back and forth here with Scott Sumner in his early blogging days, roughly 5 years ago).

There are additional points made about QE that are notable – that the central bank isn’t giving the banks “free money” for example. Again, a full appreciation of this aspect requires an understanding of banking system liability effects of QE – including potential or actual pricing effects on the liability side (this may become more of an issue after zero bound lift off has commenced for those systems that are currently parked next to the zero bound).

In all, the paper takes the reader on an illuminating journey through financial system balance sheets as they are affected by money creation and circulation. This includes an interesting section on the so-called hot potato effect (Nick Rowe might appreciate that.) It refers several times to the same 1963 paper by James Tobin as I reviewed here. In that regard it touches on the dynamic balance sheet interaction between banks and non-banks.

I have a different perspective from the paper on one subject previously discussed on this blog. This risks revisiting old ground, but it has to do with the idea of banks as intermediaries. On the one hand, this is semantics. On the other, there is a point of substance here, semantics or not.

It seems that those who want to banish the use of the term ‘intermediary’ claim this because they understand that “loans create deposits” instead. But the “loans creates deposits” idea has to do only with the creation process itself – and the creation process does not freeze bank balance sheets in a permanent birth state. Proactive liability management in particular and vigorous competition for deposits is an essential part of commercial banking. Banks are in constant competition for a share of deposits that have already been created but are now in competitive play. Demand deposits move around among different banks. Time deposits can move around when they come up for renewal. Bank customers can be surprisingly elastic in the choices they make in response to this competitive froth – particularly institutional customers who deal in wholesale money markets, and retail customers at times as well.

Interestingly, the paper does a pretty good job of describing this sort of thing. The composition of individual bank balance sheets changes from the point of loan and deposit creation. And the purpose of the payment clearing system is not only to facilitate payments from one demand deposit to another, but to accommodate the competitive process for bank liability management more generally. The authors touch on this, so the issue is not about an appreciation of such basic banking activity, but rather the connection between that perspective and the relationship between money creation and bank intermediation.

Here is a separate example. Suppose the Bank of Montreal posts an unusually attractive rate for 5 year mortgages – i.e. very attractive in a competitive sense. The immediate elasticity of the associated mortgage borrowing activity can be quite pronounced. The bank may follow some correlated strategy (for example) in the retail GIC funding market in order to help match the liquidity and interest rate risk exposure on the mortgage (a potential margin squeeze if a fixed rate mortgage is funded with a short term deposit and interest rates go up).

In this example, it is not the mortgage loan itself that created the desired 5 year GIC funding. It is the bank’s strategy to compete for a liability product that is different than the deposit that was created in association with the mortgage. And it is also not likely that the deposit that was created originally will end up itself being a candidate for conversion into the form that the bank actually wants as an offset to its new mortgage loan – because that deposit will probably have left the bank. The mortgage loan will produce a demand deposit that may immediately disappear from the books, with a cheque flying out of the bank to pay for a house sold by someone who now deposits that cheque with the Royal Bank. Meanwhile, the 5 year fixed rate funds that BMO actually wants may come from a Toronto Dominion Bank demand deposit held by a customer who is keeping her eye open to such fixed deposit opportunities. The intermediation effect, to use that word, in this example is in the coming together on the BMO balance sheet of a borrower’s need for money to buy a house and a lender (i.e. depositor)’s desire to lock away some money at a higher rate than she can get on a demand deposit. The intermediation consists of the structure of the balance sheet itself – that the borrower’s liability is not the same as the lender (i.e. depositor)’s asset.

That’s all ‘intermediation’ need mean. It need not contradict “loans create deposits”. And it doesn’t mean that “banks lend deposits”, which is a strange phrase on its own, with some epistemological head scratching associated with it. While “loans creates deposits” ensures a nominal balance in the balance sheet of the entire banking system – quite independently of any bank reserve activity – it doesn’t come close to describing the substance of banking as it occurs at a competitive level, which includes the dynamic management of an asset mix and a liability mix and the two in coordination. And while intermediation may not necessarily be the perfect word to describe this process, it is probably a false argument to exile the word and the idea simply because “loans create deposits” at source. Conversely, the enemy idea of “loans creates deposits” is really not intermediation. It is instead an array of misguided theories such as the money multiplier and loanable funds – but which are also not necessarily logically connected to the idea of intermediation. Let’s not indict an idea that is largely about the dynamics of money circulation already in existence when the actual concept in contention is the nature of original money creation.

An intermediary matches up supply and demand in some way – which is what banks are attempting to do in their proactive asset liability management. For example, they are attempting to create the mortgage that the borrowing customer wants and the deposit that the lending customer wants – both within the constraints of what the bank itself can accommodate and wants according to prudent risk management, product design and pricing constraints. That seems to satisfy a sensible definition of intermediation. This idea need not imply anything untoward about how money is originally created or how the bank reserve system works as banks compete for those assets and liabilities.

I think that this perspective was the theme – broadly speaking – of Tobin’s 1963 essay already referenced. Banks do not lend deposits into existence with nothing more to do from that point on. For one thing, a bank that makes loans can’t operate for long if the demand deposits created along with those loans flee to another bank and nothing else happens. The central bank won’t permit that kind of bank to survive indefinitely – even if it provides some temporary relief through the last resort facility. And there are other risk management considerations as already noted.

“One common misconception is that banks act simply as intermediaries, lending out the deposits that savers place with them. In this view deposits are typically ‘created’ by the saving decisions of households, and banks then ‘lend out’ those existing deposits to borrowers, for example to companies looking to finance investment or individuals wanting to purchase houses. In fact, when households choose to save more money in bank accounts, those deposits come simply at the expense of deposits that would have otherwise gone to companies in payment for goods and services. Saving does not by itself increase the deposits or ‘funds available’ for banks to lend.”

Again, I’m not quite sure this is the best definition of what is meant by ‘intermediary’. But the more interesting thing about this quote is that it correctly identifies what is in effect a confusion between NIPA type accounting (for saving) and flow of funds accounting (for loan and deposit creation). These are separate but connected, and the connection has to be right in accounting terms. It is the absence of such a connection along with a thorough treatment of the banking sector that has been missing in mainstream economics. That is a problem post Keynesian economics has responded to by effectively integrating NIPA and flow of funds accounting and analysis in a coherent framework for understanding monetary economics.

But here is an interesting exception where the authors actually depend on the idea of bank intermediation:

“As the pension fund does not hold a reserves account with the Bank of England, the commercial bank with whom they hold a bank account is used as an intermediary”

The authors are referring to a QE transaction in which the Bank of England buys a bond originally held by a pension fund. The pension fund deposits the money in a commercial bank deposit account and the bank is credited with reserves by the Bank of England. This illustrates the broad money effect referred to above. I suppose the authors use the word ‘intermediary’ in this case because money has been created by the central bank. The commercial bank intermediates this process between the depositor and the central bank. But it is not immediately obvious why this should preclude a similar meaning in the case where one bank attracts a newly created deposit from another bank – and leaves the clearing result in the form of reserves. My question would be – why is that situation not intermediation as well? The commercial bank balance sheet profile is the same in both cases – incremental deposits and incremental reserves.

But my intention is not to reduce this issue to semantic angels on the head of a pin – rather to illustrate the dynamics whereby bank balance sheets are routinely and proactively managed well beyond the original point at which “loans create deposits”.

Finally, the paper does quite a good job with a brief overview of how risk management factors into lending and deposit creation. Capital management is referenced somewhat obliquely, and this is an area that could be strengthened in the overall explanation, since risk management is inextricably linked with capital management and decisions on capital allocation. It is a holistic risk capital system in that sense.

But this is a very good paper – and relatively easy to read for such a technical subject. Once again, please have a look at it:

http://www.bankofengland.co.uk/publications/Documents/quarterlybulletin/2014/qb14q1prereleasemoneycreation.pdf

About

View all posts by

Leave a Reply

126 Comments on "Money Creation in the Modern Economy (Bank of England)"

Notify of

Guest
Nick Rowe
3 years 3 months ago
“In no way does the aggregate quantity of reserves directly constrain the amount of bank lending or deposit creation.” Oh yes it does. The BoE must make damn sure it does, if it wants to keep inflation on target. “A related misconception is that banks can lend out their reserves. Reserves can only be lent between banks, since consumers do not have access to reserves accounts at the Bank of England.” Oh yes they can. There is nothing preventing a bank swapping reserves for cash, and lending that cash to consumers. “Deposit creation or destruction will also occur any time the banking sector (including the central bank) buys or sells existing assets from or to consumers, or, more often, from companies or the government.” Yes! Money is created whenever a creator of money buys something. “Making a loan” is just buying a non-monetary IOU. But it is also true that money can be created without buying anything, by just giving it away, as a charitable donation. In fact, the simplest way of putting it is that money is created by creating money! “The limits to money creation by the banking system were discussed in a paper by Nobel Prize winning economist James Tobin and this topic has recently been the subject of debate among a number of economic commentators and bloggers.(4)” I wish they had said *which* bloggers! But this proves that at least some of us are read. Yipee! “The ultimate constraint on money creation is monetary policy. By influencing the level of interest rates in the economy, the Bank of England’s monetary policy affects how much households and companies want to borrow.” Finally! They (implicitly) recognise that reserves are not just “supplied on demand”! “And it is because there is demand for central bank money — the ultimate… Read more »
Guest
3 years 3 months ago

“….what determines the interest rates the BoE will set? And how would it change if a rise in the price level (above target inflation) would cause an increase in loans and deposits and broad money, and a proportionate rise in demand for central bank money? It would increase, of course, to prevent that increase in the quantity of central bank money, which would not be “supplied on demand”.”

I think it’s a stretch to say that the reason for raising rates is to prevent the increase in base money. The level of base money is not seen as important in its own right. It’s quite possible that the amount of base money might have fallen, despite the rise in the price level, but that is unlikely to have any effect on the interest rate decision.

I know it’s all just different ways of looking at the same thing, but the practical reality is that when they set the rate for the next period, the expected outcome for the level of reserves doesn’t figure into it.

Guest
3 years 3 months ago

I would say the increase in the interest rate is more directly to lower loans and deposits, not to lower reserves. What happens to the quantity of reserves is more a product of the nuances of the interbank rate system that the central is running, as I touched on above.

Guest
3 years 3 months ago

I agree. The exchange rate effect is also important in the UK. And it’s also possible that higher debt service cost has more of an impact on consumer spending than the corresponding higher interest income for depositors.

Guest
3 years 3 months ago

C’mon Nick. When will economists learn to accept their mistakes?

Both Samuelson – the previous best-seller before Mankiw became #1 and Mankiw have the money multiplier description where the central bank fixes the monetary base and the money stock because of the multiplier relation.

Guest
Nick Rowe
3 years 3 months ago

Oh dear!

“As with the relationship between deposits and
loans, the relationship between reserves and loans typically
operates in the reverse way to that described in some
economics textbooks. Banks first decide how much to lend
depending on the profitable lending opportunities available to
them — which will, crucially, depend on the interest rate set
by the Bank of England. It is these lending decisions that
determine how many bank deposits are created by the banking
system. The amount of bank deposits in turn influences how
much central bank money banks want to hold in reserve (to
meet withdrawals by the public, make payments to other
banks, or meet regulatory liquidity requirements), which is
then, in normal times, supplied on demand by the Bank of
England.”

No! Have these guys never heard of inflation targeting? The BoE does not sit idly by, happily “supplying” whatever is demanded at a fixed rate of interest. If banks decide to lend more, and this increases spending and pushes inflation up above target, the BoE will raise that rate of interest, precisely because the BoE cannot keep inflation on target if it simply lets reserves be “supplied on demand” at a fixed rate of interest.

They have an ant’s eye-view of the economy. They really need to step back and see the big picture. Any increased demand for reserves that is a result of actions that would lead to inflation rising above target will be met with a refusal to supply *any* additional reserves. The BoE will raise the rate of interest to make the supply curve of reserves perfectly inelastic in that case.

Guest
3 years 3 months ago
Nick, I’m going to expound upon your broader view of what the CB does, which is correct, but argue that you’re still focusing too singularly on quantity of reserves. To simplify matters, I’m going to assume the central bank targets a fixed price level, rather than growth rate in prices. Suppose to attain the price level the CB desires, they decide the target rate needs to be 1.5%. This will lead to a level of bank lending activity consistent with their price level target. Once in equilibrium, the CB maintains the target rate by setting the floor rate (IOR) at 1% and the ceiling rate is 2%. Let’s say the reserve demand curve flattens out to 2% at 100 reserves and flattens to 1% at 200 reserves. Assume that setting reserves at 150 achieves the 1.5% interest rate. Let’s say this also corresponds to a level of deposits and loans around ~1500, where the reserve requirement ratio is 10% (thus 150 reserves). Now say all the sudden the price level increases beyond the central bank’s target. Say the central bank thinks that if deposits/loans were brought down to 1000, then they would achieve their price level target again. Assume they think that if the interest rate is at 2%, then loans/deposits will decline from 1500 to 1000. So how can the CB achieve this? In the short-run, assume that reserve demand won’t change very much because it takes a while for the economy to respond to the new interest rate of 2% (meaning loans/deposits stays around 1500, so demand for reserves stays around 150). In this case, the CB can simply change the floor and ceiling to 1.5% and 2.5% respectively, to hit its target of 2%. Of course, since this doesn’t impact demand for loans in the short run,… Read more »
Guest
3 years 3 months ago

So yes – the central bank sets rates based on their price target. But no – the central does not necessarily have to change the quantity of reserves. The more fundamental point is that if inflation is connected to the quantity of bank lending, then we need to consider what bank lending is a function of. Bank lending is a function of what JKH said: “mostly of capital and pricing of risk.” The CB ultimately affects this by changing the interest rate, and changing the interest may or may not involve changing the quantity of reserves.

Guest
JKH
3 years 3 months ago

“If banks decide to lend more, and this increases spending and pushes inflation up above target, the BoE will raise that rate of interest, precisely because the BoE cannot keep inflation on target”

I don’t think anything they’ve written contradicts that or that they would disagree with it. But they would disagree with your characterization of how the BOE would manage the reserve system in conjunction with that interest rate increase. Unfortunately, so would I. But this is what the debate is all about.

Guest
Nick Rowe
3 years 3 months ago
JKH: Interesting read. A couple of comments on the paper: 1. “In this view deposits are typically ‘created’ by the saving decisions of households, and banks then ‘lend out’ those existing deposits to borrowers, for example to companies looking to finance investment or individuals wanting to purchase houses.” That is certainly not the “textbook” view. In the textbook view, “saving” simply means not consuming part of your disposable income. It has no necessary connection with deposit creation at all (except indirectly via how the central bank may respond to increased saving). 2. “This article explains how, rather than banks lending out deposits that are placed with them, the act of lending creates deposits — the reverse of the sequence typically described in textbooks.” That is a total mischaracterisation of the “textbook” view. There are two questions: what starts the ball rolling; and what happens when the ball does start rolling? The “textbook” view assumes the banking system is initially in equilibrium. (Textbooks almost always assume the economy is initially in equilibrium, and then hit it with a shock.) What causes the banks to expand loans, and starts the ball rolling? It could be almost anything. The textbook view typically assumes it is a decision by the central bank to do QE (which the textbook calls “open market operations”). The central bank buys a bond for $100 from a member of the public, who then deposits that $100 at the Bank of Montreal. That is what starts the ball rolling. This first step in the “textbook” story is exactly the same as the authors’ description of the first step in QE! The “textbook” story then says that Bank of Montreal then expands loans and deposits, because it is now out of equilibrium and holds more reserves than it desires. It simply… Read more »
Guest
JKH
3 years 3 months ago
Hi Nick, I think your # 1 is a very good point. I made that same point in my post, at least regarding potential confusion between the concepts of NIPA saving and that of non-NIPA flow of funds such as occurs in LCD. I agree the article is too loose in the use of the term ‘saving’ when discussing flow of funds in banking. And I’ve alluded to similar concerns about the authors’ characterization of intermediation – which is roughly analogous to your concern about their representation of the textbook story. Regarding the textbook story itself: “The “textbook” story then says that Bank of Montreal then expands loans and deposits, because it is now out of equilibrium and holds more reserves than it desires.” That story is only fundamentally correct for bank money market operations (I know you hate the term ‘money market’). Money market operations are the generalized response to OMO rate impulses – but MM operations work at the level of risk free assets – or very near risk free assets. The capital requirement and required incremental risk analysis is minimal. It is essentially a response when the CB wants to tweak the policy rate trading level. Your equilibrium concept works pretty well there. But that context is really quite limited in the framework of banking as whole. It is not how banks work at level of risk lending, which is a function mostly of capital and pricing of risk. Banks do not increase their lending risk profile in any meaningful way due to that particular operational equilibrium concept. It is not a good idea IMO to associate strategic QE with tactical OMO under non-QE conditions. Yes the transactions are the same. But the constraints on reserves in terms of interest payments and the use of reserves as… Read more »
Guest
Ian Lippert
3 years 3 months ago

I do not understand this idea of “loans create deposits”. Aren’t the banks lending out money that they have previously received? A loan releases money into the money supply but that seems to differ from the concept of money creation. Under the concept of LCD are there no supply side constraints for the banks? The article listed only 3 constraints and they were all demand side constraints.

Guest
Fed Up
3 years 3 months ago

“Aren’t the banks lending out money that they have previously received?”

Not usually.

I save $1,000 in currency. Someone else starts a bank and sells me $1,000 in new bank stock. Assume a 10% reserve requirement and a 10% capital requirement for everything (unrealistic). The capital requirement applies to the assets.

The bank creates $10,000 in new demand deposits (reserve requirement applies to them) and buys $10,000 of new loans (capital requirement applies to them). The bank keeps the $1,000 in currency for the reserve requirement.

Assets: $10,000 in loans plus $1,000 in currency
Liabilities: $10,000 in demand deposits
Equity: $1,000

$1,000 in currency (MOA and MOE) is no longer circulating in the real economy. $10,000 in demand deposits have been created out of thin air (not previously received) and are circulating in the real economy. Demand deposits and currency are fixed convertible at 1 to 1. That makes demand deposits MOA. Demand deposits can also be used to settle transactions. That makes them MOE. $10,000 in MOA/MOE has been created and are circulating.

Overall, $1,000 of MOA/MOE is saved, and $10,000 of MOA/MOE is “dissaved”. The difference is an increase of $9,000 of MOA/MOE.

Does that help?

Guest
JKH
3 years 3 months ago
It’s a good question, because LCD is in fact more complex in context than some give the idea credit for. There are two basic ways of thinking about what can happen at an operational level. For example, a bank can attract a deposit from another bank and increase its reserves that way. It can then make a loan where the bank writes a check to the borrower and the borrower takes that cheque and deposits it in another bank. That drains the reserves from the lending bank and increases the reserves of the next bank. The lending bank ends up with a deposit and a loan and is net flat on reserves. But the deposit in the final link in that chain represents a new deposit for the banking system as a whole. So the loan has created a deposit in that sense – even though it hasn’t created the deposit in the lending bank. The more typical way of thinking about LCD is that a bank can lend money by first crediting the borrower’s deposit account. That’s the first thing that happens. But that borrower may then write a check on his deposit account and the deposit balance ends up being transferred (in effect) to another bank. That leaves the lending bank short of reserves, which it must cover by attracting more deposits. The knock on effect in this case is the same but in the reverse time order of the first case. In either case, a new deposit is created. It’s just a matter of what order and where. Regarding constraints, the first constraint they mention is really supply side – in the sense that the degree to which the banks will supply LOANS is restricted by credit worthiness of the borrower, other risk criteria that the bank… Read more »
Admin
3 years 3 months ago
Part of the problem comes from the way we want to think about cause and effect. I have been meaning to write a post for over a year called “Banking and the Theory of Constraints” http://en.wikipedia.org/wiki/Theory_of_constraints We want to think about there being some sort of cause/effect relationship between input variables and the output. But with banks and money creation, it seems all mashed up when you try to identify the final determinants of what is causing the money to be created. I strongly suspect this is happening because money creation is primarily based on constraints rather then permissions. At any given moment, there is some overriding constraint on money creation. This can easily shift from price to demand to balance sheets to bank capital to any one of a long list of possible inputs in the banking system. The idea “loans create deposits” is mostly trueish, but as you point out, it’s pretty complex in real life when you consider all of the details. It’s not like you can point to any single part of this and say “Here is the key, and loans always create deposits” However, if you think of loans being created in response to a series of binding constraints on the overall quantity, then I think the picture gets a bit more clear. In this case, loans create deposits, full stop. Still, there is some quantity of deposits which exist in the world, and as you point out JKH, it’s entirely possible there could be a scenario where the most important constraint on loan creation is “available deposits”, and there is little the world can do but find a way to get more deposits if we want more loans. This is all part of the general idea of “Banking and the Theory of Constraints”. What… Read more »
Guest
jt26
3 years 3 months ago

Agree. The interest rate story is a story of “average” interest rates and average returns. Prices are a story of average return on money vs. goods/assets. But, we know people don’t think like that. In the real world, returns can be viewed as options, momentum, disaster insurance, etc. In the real world, banks and businesses think of network effect returns, commissions, razors & razor blades, My guess is the constraints on bank lending is now the government. As Banks will slowly abdicate their “true” money creation to the government and essentially become transactional fee pigs, waiting for the government or government subsidies to spin off securitized product. The Fed will become the largest MMF sponsor (reusing the “Reserve Primary’ fund name it kept from MaidenLane purchases of Lehman assets).

Guest
JKH
3 years 3 months ago

Write that post, Mike.

But please have a look at this first – from five years ago – my first encounter with Bill Mitchell (backed up by Scott Fullwiler and Warren Mosler):

http://bilbo.economicoutlook.net/blog/?p=3921

“JKH: My general impression however is that in making the case you and the others are combining the correct explanation of operational capability with a near-rejection of the real world importance of balance sheet “policy” restraints or constraints.”

Admin
3 years 3 months ago

Ho man thats a good post. How can they be shocked you ended up over here instead of over there?

Guest
JKH
3 years 3 months ago

🙂

Guest
3 years 3 months ago

Not just the interest rate. Lenders essentially liquify collateral and apply a haircut in doing so. The size of the haircut also factors into demand and it varies in the cycle. When credit standards are loose, lenders relax the haircut, lending more against the perceived value of the collateral, which may also be richly appraised. When credit is tight the opposite applies. Standards are usually loosened gradually pro-cyclically and tighten suddenly counter-cyclically.

As you have point out before, Mike, credit money in the US is backed largely by real estate as the underlying collateral. All we have to do is look at recent experience to see how this worked in the lead up to the bubble, the crash and now the recovery with tight credit even tough rates are at historical lows. Much of the activity in the housing markets has not been through credit but historically high cash sales.

So it seems that other factors are operative in addition to the interest rate. The Fed as chief regulator could affect this, for example, by regulating collateral, as Warren Mosler has recommended.

1 3 4 5
wpDiscuz