There’s a considerable backlog of interesting material related to the April, 2012 Krugman/Keen blogosphere debate on banking. Market monetarist Bill Woolsey wrote an intriguing piece on banks and monetary policy, somewhat under the radar to the main discussion. I don’t know if it was prodded by the K/K debate, but it may well have been:
http://monetaryfreedom-billwoolsey.blogspot.ca/2012/04/money-multiplier.html
The subject of Woolsey’s post is his explanation of the money multiplier. What caught my eye was not that specific topic, but some of the related description of banking operations he provided as background. The standard textbook multiplier concept as often explained is a highly dubious version of monetary operations and banking. But whether or not one wants to make a case rejecting the concept of the money multiplier is secondary to the main point of my post here, which is the value of Woolsey’s piece for its related descriptive material on banking operations.
Regarding the nature of monetary operations, if I read between the lines, Woolsey may be saying there’s some basic stuff that the post Keynesians think non post Keynesians in general and monetarists in particular don’t know, but in fact they do know:
“Monetarists never argued that the quantity of base money should be fixed. But they did say that it should not be changed to keep any interest rate, including the interbank loan rate, from changing. It was never the case that monetarists were unaware that the Federal Reserve adjusted the quantity of reserves to keep the interest rate on interbank loans at a targeted level. It is rather that they opposed that policy. They favored changing the quantity of base money according to a different principle–keeping some measure of the quantity of money on a targeted growth path…”
No doubt that’s been said before by a monetarist. But it softens the perception that monetarists are somehow unaware that “it’s about price” as far as the Fed is concerned in its modus operandi today. (This is a fact that is not contradicted by the explosion of excess reserves brought about since the financial crisis, since the Fed controls the price of those reserves by deciding what interest rate to pay on balances.)
The most interesting aspect of his post in my view is reflected in the following paragraphs:
“How is it that the quantity of money changes when there is an increase in base money? There is a terribly unrealistic story often told in introductory textbooks. The story starts with a bank having excess reserves. The bank lends those reserves out. The borrower spends the money and the sellers deposit the checks. The checks clear, and the bank that made the loan no longer has the excess reserves, but the bank used by the seller now has the excess reserves. It makes a loan, and so on. Assuming banks’ demand for reserves increases with their supply of deposits, then the amount lent each time is smaller. The initial quantity of excess reserves is “multiplied” into many deposits and many bank loans…
… I often point out to my students that the simple story of banks making loans based upon their existing level of excess reserves is very unrealistic. Do banks put out a sign saying, “excess reserves available today, come get your loans?” In reality, banks set interest rates on both loans and deposits intending to use their deposits to fund their loans. In a growing economy, this generally involves setting interest rates on loans and deposits so that demand for loans from banks is matched by the supply of deposits to banks. The banks then use money market instruments to adjust for any temporary imbalances between the demand for loans and the supply of deposits. Higher loan demand would immediately result in banks selling off government bonds, reducing overnight lending to other banks, or borrowing more overnight from other banks. An increase in the supply of deposits would result in banks buying government bonds, reducing overnight borrowing from other banks, or lending more overnight to other banks…
… It is only when we have a system where banks can use money market instruments to manage their reserves, and banks are funding their loans and securities portfolios with a variety of interest bearing deposits, does a quite different approach to understanding the banking business become reasonable. In particular, changes in the interest rates on money market instruments and the interest rates that banks pay and charge become very important for any reasonable account of the banking business.”
There you have it. Not only does he reject the textbook multiplier story, but he captures the general nature of bank reserve management and asset liability management quite accurately in my view. His high level summary of it compares favorably with similar accounts I can recall seeing on post Keynesian blog posts in general.
He has it right on the general nature of the pricing function across various business lines within a given bank. Commercial bank business lending and deposit business units take their cue on pricing parameters from the bank’s treasury function, which in turn takes its cue from the central bank’s own interest rate management. (Cost of equity capital is also a core input to asset-liability pricing, obviously.) And again he’s right that the textbook story of the multiplier is wrong, or unrealistic, to use his word. I won’t go into Woolsey’s refinement of that story, because it’s secondary to the purpose of this post.
The main point I wanted to emphasize here is reflected in this statement:
“The banks then use money market instruments to adjust for any temporary imbalances between the demand for loans and the supply of deposits”
This is a critical aspect to understanding bank reserve management. It is not a typical focus, at least in much detail, in the post Keynesian blog summaries that I’ve seen, for example. The usual account focuses on the central idea that “loans create deposits”. That idea incorporates the fact that banks don’t require a supply of statutorily required reserves ahead of time in order to create new deposits as described in the textbook multiplier explanation. And that is OK as far as it goes. But that in itself doesn’t really go very far in describing the reserve management process of a commercial bank. So inquiring minds are interested in how this works at an operational level.
For example:
“I’ve read all the numbingly repetitive and identical explanations of capital-constrained bank lending, how deposits aren’t, can’t be, lent out. But I’m still confused. If I write a million-dollar check on my account in Bank A and deposit it in Bank B (or roll a wheelbarrow of cash across the street…), what does Bank B do with that million dollars?”
That question is fundamentally germane to real world banking operations, and I suspect a lot of other people are confused about it as well.
Banks run complex money market operations to respond to the net movements in their reserve accounts that result from asset-liability transaction activity in the rest of the organization, including the main loan and deposit book. (Those flows include voluminous amounts of deposits shifting between banks, without any necessary connection to new loan extensions. Such activity results simply from the ongoing use of existing bank money in regular economic activity. It is inherent in the concept of money “velocity”.) Banks adjust their reserve accounts in response to these effects through offsetting transactions in short-term liquid assets and wholesale deposit liabilities. These money market operations are the balance wheel for reserve account disruptions caused by net flows through the main loan and retail deposit banking operations. Effectively, changes in reserve balances that are the result of client initiated activity in loans and deposits are offset by reversing changes in reserve balances initiated in wholesale money markets by the bank itself.
(Note that in today’s environment of chronic system excess reserves, individual banks with outsized excess reserve positions may simply let those balances decline in response to a net reserve outflow. That is a special case of a bank using the ultimate liquid asset (excess reserves) to absorb a net reserve outflow. When a bank already holds excess reserves, those reserves can be viewed as such a liquid asset that is “sold”, in effect, in order to meet a net reserve outflow. If the bank did not hold excess reserves already, it might sell a more conventional liquid asset such as treasury bills for example, in order to offset the net outflow of reserves that occurs otherwise.)
Reserve management might be imagined as simpler in a reconstituted monetary system where the government assumes a much larger, quasi-nationalizing role in operationally funding the banks, and where individual bank liquidity risk is muted as a result. But that’s not the way it works today, in fact. The deposit business is very competitive and the nature of bank liquidity management much complicated as a result. The idea that “loans create deposits” isn’t sufficient to capture the dynamic of deposit gathering in a competitive banking system. Thus, the nature of deposit origination as in “loans create deposits” does not explain the consequent dynamic that competing banks exhibit in attracting and retaining their own desired share of originally created deposits.
Woolsey’s explanation of all this is obviously high level, but it suffices as an important nexus for a coherent story of banking overall. Separate from that, he offers a good deal of additional explanation regarding the monetarist interpretation of the money multiplier. As noted, he rejects the textbook explanation. But he doesn’t dismiss the idea outright. I think I neither agree nor disagree with his interpretation in total at this point. I’ll leave the detail of that analysis as a separate subject for now. But the multiplier idea as it applies to the point of contact between commercial bank money market operations and the central bank reserve system does suggest a more refined interpretation than the crude loan/deposit form found in textbooks, and Woolsey makes a subtle distinction along these lines. And this distinction reflects the importance of money market operations as they facilitate management of net reserve flows that arise from the rest of the banking book. So the interesting aspect of his post in my view is its emphasis on money market operations in this context. And I think this particular aspect should factor more prominently as a general education point for those who are trying to understand how banks work in fact, because there does seem to be some curiosity and confusion about this point.
Thanks. Let me reframe the question in two possible ways:
1. After my deposit, can Bank B make more loans *without the need* to seek acquire reserves?
2. Why do banks spend so much money (on branches, marketing, etc.) to attract deposits? Does it have anything to do with expanding their loan books (they are borrowing at around 0%, after all…), or is it just about the services, charges, and fees they can get from those customers? Why do deposits matter to banks?
If that million dollars *does* allow bank B to lend another $900K without seeking additional reserves, of course, it doesn’t mean that the banking system as a whole has an extra $900K in lending capacity. (Bank A can lend $900k less, or must seek additional reserves.) But…?
Isn’t one of the reasons banks work so hard to attract deposits due to the fact that they make a lot of money on the fees from deposit account transactions?
But when they do attract those deposits that should give them excess reserves, right? The deposit is made in the form of either cash or, more likely, a transfer from the reserve account of another bank. So they can make additional loans without acquiring more reserves. But they can also lend the excess reserves to other banks through interbank lending at the FF rate.
If acquiring deposits is cheaper than acquiring reserves from the interbank market/Fed, then it lowers banks’ cost of funding loans. This is part of the reason banks spend so much money attracting deposits.
My use of the word “funding” may be a little misleading there, but the point is acquiring deposits may be cheaper than acquiring reserves.
Dan, wh10: Okay this is great, totally validates my understanding of how things work, but also makes incorrect something that I think a lot of people believe based on the “banks are not reserve-constrained” mantra.
Banks *do* lend their deposits.
But this does *not* validate the “loanable funds”/intermediary concept, because:
1. Banks’ lending (individually and *especially* collectively) is far in excess of what would be supported by (90% of) their deposits. In that sense they are not deposit- or reserve-constrained in their lending. Capital (regulation) is the constraint.
2. Banks don’t *need* to lend their deposits. Per Dan, they “they can make additional loans without acquiring more reserves. But they can also lend the excess reserves to other banks through interbank lending at the FF rate.”
They could also use the deposits buy treasuries, gold, or hell, vault cash from the Fed or other banks. Anything that counts as “external reserves.”
Meanwhile they can borrow more reserves and acquire more capital as needed to make more loans. It’s all a matter of what mix of sources and uses makes them the most money.
So it’s a sources and uses of funds thing. Money being fungible and all that.
Which also validates a conceptual understanding the I’ve bruited in the past: banks lend (90% of) their deposits, then they keep on lending to the extent of their capital constraint, and then acquire more capital if there’s more profitable lending to be done.
The timing of that sourcing doesn’t actually, necessarily, work anything like that, of course. But I think it’s a simple explanation for general thinking by the unwashed like me.
The problem with this representation is that it can be misunderstood: because banks are “lending their deposits,” it seems like they are, in fact, intermediating by borrowing short and lending long. When in fact, at the margin (both individually and collectively), they’re printing new money when they make a loan (“counterfeiting” under license from the Fed). The deposits down in the bottom of the pile are only material in their impact on banks’ sourcing costs — not on their capacity to lend (or in any predictable way, their risk profile or susceptibility to bad economic events).
Seem right?
Just to be totally clear, adding the key item to the list of things that Bank B *could* use the million dollars for: assuming they have sufficient capital, they could lend it out (without having to seek additional reserves).
Hi Steve,
I hope you didn’t mind me making “an example” of your question. It was intended to be just that. The reason I didn’t get back to it on your blog was that I was away during that period and am just now getting up to speed. I’ll be back in a bit to try and add my two cents to your specific question, but it looks like the group is pitching in anyway.
Oh so I can blame this one you, can I? Shoulda known. You’ve now gone beyond the cryptic initials to no byline at all. Whadda ya think you are, the editorial staff of The Economist or something?
Thanks, actually, for using me as an example, and for getting back to this. I was frustrated that none of the team answered that. I suspected at the time it was because of what I suggest above: if we admit that “banks lend deposits,” the whole loanable funds things seems validated. But it’s not.
Banks have not borrowed from the private sector to lend since the days of actual “fractional reserve banking,” when banks accepted deposits of specie and issued certificates against these deposits in excess of their specie holdings in the expectation that demand for redemption wouldn’t exceed supply. If it did, then the bank would have to provide from its own capital and when it no longer could, then it would have to shut its windows and go bust.
Governments that guaranteed exchangeability of its currency for specie at a fixed rate could find themselves in that position, too, but instead of going bust, governments just end convertibility, which is their prerogative as sovereigns.
The “fractional reserve banking” refers to banks telling their deposit customers that they can take out their deposits “on demand” even though all but a fraction has been lent out long term. This is how western banking works. It is brain dead. At some point there will always be too many customers who want to take out their money at the same time. So they try to patch around this by having a central bank that can act as a “lender of last resort” and print as much money as needed. This makes all kinds of new problems. The right thing to do would be to require that banks sell bonds of the same duration as the loans they want to make. So sell 10 year bonds to raise money to make 10 year loans. Then there is no risk of a bank run. Also, banks no longer “make money” with this system. Much more stable financial system if this were done.
http://pair.offshore.ai/38yearcycle/#banksgobust
Steve,
I just don’t see why it’s helpful to think of banks as lending their deposits. From an accounting standpoint, this is certainly inaccurate. You’re not going to see a bank decrease someone’s deposits when it makes a loan.
I think the best way to think of it is simply that a loan and a deposit are created in tandem, and then banks have different sources of funds *if* they are needed to support that type of transaction. In a regulatory regime without reserve requirements, then banks have no need to acquire reserves following a loan, is my rudimentary understanding. We have a tendency to be biased in our thinking because we live in a system where there is a reserve requirement. If we start our model in a world without them, I think the “truth” of the mechanics is less veiled.
Banks compete on price and service.
They compete in asset businesses (lending and investing) and funding businesses (deposits and capital).
They measure gross profit as an interest rate spread in all cases – asset and deposit businesses, plus the margin effect of capital allocation.
They lend at a spread over a benchmark wholesale funding cost.
They raise deposits at a spread below the same benchmark as an interest rate paid.
That benchmark yield curve “splits” the spread between external lending rates and external funding rates.
Surplus funding will show up as surplus reserves. That money can be invested at near risk free rates requiring no capital allocation and no required profit, essentially.
Or it can be used to pay down a money market liability. Either way, the spread earned by the deposit business remains intact.
And if the bank becomes “swamped” with excess funds that it decides it can no longer use without experiencing dysfunction in its money market operations, it can always adjust deposit rates down to turn off the “tap”.
The process works in reverse for loans. The fact that “loans create deposits” is no guarantee the bank will be able to retain the deposits it requires in order to keep its balance sheet “in balance”.
It’s about pricing.
The reserve account at the central bank is an indicator of how much the balance sheet is in or out of balance otherwise. Banks do not want to experience chronic surplus or deficit reserve positions.
(The post 2008 environment is exceptional. Banks must allow for the fact that the system is chronically surplus in reserves. They must have well thought out strategies with a view toward their appropriate or natural or neutral share of such reserves, which are essentially stuck in the system.)
So all asset and funding businesses have their own cost of funds used or return on funds supplied. They are independent to that degree. If there is a business unit that has a mandate to invest in equities, it will pursue that mandate based on its internally assigned cost of funds, without much reference to the actual current experience of the bank in raising deposits at the time.
Banks have centralized risk management committees and asset-liability committees that bring together the disparate business perspectives of all the different business units, in order to track down potential dysfunction risks at the bank-wide portfolio management level. That includes tracking the overall funding profile and any trending liquidity position as reflected in the money market operation.
P.S.
I think the phrase “lending deposits” is not a good one.
The balance sheet objective, simplified, is:
Assets = Liabilities + Capital
Deposits are part of the equation that keeps the balance sheet in balance.
That is fundamental to the business of banking.
Reserves are a reflection of imbalance, very roughly speaking – a mirror image of it.
The monetary authority does not expect chronic imbalance, particularly on the reserve borrowing side.
(Again, post 2008 is very special, with chronic system excess reserves.)
So banks use deposits in large part to match assets in nominal terms.
But they don’t “lend” deposits.
They take money on deposit, and they lend money to borrowers.
The net interbank payment effect shows up as reserves.
Think of bond lending as an analogous counterexample. Banks don’t lend their own deposits the way they can lend bonds issued by others.
yeah meaningless to say lending deposits. Just like I cannot lend my loan to the bank to anyone. I can but nobody will take it.
It’s indicative that banks don’t worry so much about not getting deposits to fund their lending, but instead worry that their liabilities will over time reprice at a higher rate while their assets remain fixed, or that their liabilities remain fixed while their assets over time reprice at lower rates.
“they’re printing new money when they make a loan (“counterfeiting” under license from the Fed).”
I think that the “counterfeiting” analogy falls into the rhetorical bin instead of the descriptive. It’s typical of Libertarian-Austrians. BTW, “Lord Keynes” has a rebuttal of “funny money” over at Social Democracy for the 2021st century.
http://socialdemocracy21stcentury.blogspot.com/2012/05/funny-money-loaded-phrase.html
Banks have a “franchise” from government to create credit money denominated in the government’s unit of account, enjoying an exorbitant privilege that others do not. Banks “pay” for that privilege by submitting to regulation. Is that sufficient payment? Some argue not, in that the government’s exposure to bank losses is inordinate given present institutional arrangements between government and the financial sector, especially considering ballooning systemic risk. The question of reform involves how to revise the franchise agreement in view of recent experiences and current challenges.
Steve, I’m uncomfortable with the terminology of “lending their deposits.” But maybe that’s just because I’m just being too much of a stickler about the accounting terminology and want to avoid confusion. The bank’s deposits are liabilities of the bank, not assets of the bank. So I’m not sure what it means to say that the bank is lending some of its liabilities. But I think I understand what you mean. The bank’s total reserves – it’s vault cash plus its reserve account balance – were built up from various sources, and one of those sources is from funds deposited by customers. In the simplest version, if I walk into a bank with a $1000 bill and deposit it, then the bank creates an account for me (assuming I am a new customer) and credits that account with a $1000 balance. That balance records a liability of the bank, since it represents an amount it owes me. That account balance, the bank’s liability and my asset, is the deposit. In exchange for giving me this asset in the form of an account balance, I relinquish an asset in exchange: the $1000 bill. The bank takes the $1000 bill and puts it in its vault. That $1000 bill is now an asset of the bank – part of it’s total reserves.
The same picture applies if I deposited the initial $1000 by depositing a check on an account at another bank. The check is cleared when a transfer is made from the latter bank’s reserve account to my new bank’s reserve account.
So then, can we at least say that the bank lends its reserves, where some portion of its reserves were built up from customer deposit. It seems better to say that the bank lends against its reserves in the sense that its total reserves are a sort of limit on the bank’s ability to lend without acquiring new reserves. That limit would be meaningful if acquiring new reserves at a reasonable price was hard. But given that the Fed necessarily accommodates the demand for reserves, the quantity of existing reserves doesn’t seem to be much of a limit at all.
I think maybe Krugman’s point was that even if the individual bank doesn’t need to attract additional deposits to make more loans, since it can borrow Fed Funds to acquire the needed additional reserves, the banking system as a whole has to acquire additional deposits in order to make additional loans, because the borrowed Fed Funds had to be some other bank’s reserves in the first place. But that doesn’t seem true. The government is constantly supplying new reserves to the banking system to accommodate growth in loan demand.
I’m also a little confused about the 90% number. Even if the entirety of the bank’s reserves were built up from customer deposits, and if the bank were fully lent out – i.e. the total value of its loans were equal to the full amount allowed by law given its total reserves – then the total value of the loans would, given present rules, be 9 times its reserves. Where does the 90% number come from?
Tom Hickey: “I think that the “counterfeiting” analogy falls into the rhetorical bin instead of the descriptive.”
Oh yeah I totally agree, just using it here among friends. Wink and a nod…
“the total value of its loans were equal to the full amount allowed by law given its total reserves – then the total value of the loans would, given present rules, be 9 times its reserves. Where does the 90% number come from?”
My understanding was that there’s 1) a 10% reserve-to-deposit requirement (allowing lending of 90% of deposits), and 2) a circa 12% capital-to-loans requirement (allowing lending of 8.3 times capital). Roughly; risk adjusted and all that. Am I totally wrong?
My understanding was that there’s 1) a 10% reserve-to-deposit requirement (allowing lending of 90% of deposits),
I believe the first part is correct, in the US, but doesn’t entail the second part. If by “deposits”, you mean deposit balances, which are a bank liability, then I don’t think it makes sense to describe banks as lending their deposits.
But if by “deposits” you mean “the portion of a bank’s total reserves that were deposited by bank customers,” then even if 100% of the bank’s reserves were acquired in that way, the 10% reserve requirement means that the bank can have total deposits equal to 10 times the amount of those reserves. Suppose I am a bank and my total reserves are $1 million, and each dollar of those reserves came to me because some bank customer deposited it with me, then my permissible total deposits are $10 million – not $900 thousand. Since $1 million of those deposits will consist of the $1 million I owe to the depositors who deposited $1 million with me, the other $9 million can consist of loans I have generated.
Kay I’m getting this, maybe this will help others like me. I’ll do this before responding to JKH’s latest.
I think this avoids the confusion of composition (something we stumble on sometimes…) — it’s true both of individual banks and of “the banks” in toto:
I start a new bank, and buy a million shares in it for a million dollars. That’s the bank’s capital.
It can lend $8.3 million (printing new money into deposit accounts under license from the Fed) without exceeding its 12% capital requirement. It does so.
Now somebody deposits a million dollars. Can it lend 90% of that, retaining 10% in reserves? Does the deposit increase its lending capacity by $900K? No. Not without a further $108K of share purchases/capital investment to maintain its capital ratio.
That $108K of extra capital gives it an extra $900K lending capacity regardless of whether it gets the million dollar deposit.
So why does The Economist make such a big deal of loan-to-deposit ratios (LDRs) here:
http://www.economist.com/node/21553015
Do they not get it?
And am I perhaps confused because the entries in the bank’s Fed account are called “reserves,” a paleo-holdover term from the days when banks did in fact lend deposits and were required to hold reserves against those deposits? (Pre-FDIC? Pre-Fed?)
First, you’re working from a required reserve ratio of 10 per cent, which only applies to demand deposits. The weighted average required reserve ratio for the US banking system is less than 1 per cent, due to the proportionately high level of deposits requiring no reserves.
But leave that aside and use the 10 per cent reserve requirement.
In your example, you should assume capital funds of $ 1 million are invested initially in risk free assets such as treasury bills, which require no capital. You then lever up by $ 8.3 million in loans that create $ 8.3 million in deposits. That uses up your full capital allocation according to your example.
And the balance sheet is now $ 9.3 million nominal. Then, working in step logic, you can sell the treasury bills and pay down $ 1 million in deposits, coming back to your balance sheet version. I’ve just separated it out to make the steps clearer starting from an initial capital raise.
This also produces a reserve requirement of $ .83 million according to your version. So the bank raises another $ .83 million in funding and leaves it in the reserve account. If that $ .83 million funding is itself a reservable deposit, then the process repeats again from there according to a sort of inverted multiplier dynamic.
The key to it is that the reserve requirement is determined and imposed ex post to the actual timing of the deposit creation. And the central bank provides the system reserves that are required to satisfy the demand created by new requirements as they are calculated and imposed.
On the capital side, yes the bank can go on to raise more capital through either internal or external generation, and repeat the entire process above.
Capital requirements and reserve requirements involve separate funding dynamics that work along in parallel to each other.
Quick answer on the other – it’s a matter of liability classification. Retail deposits are considered the most stable. Wholesale funding is less stable and may be classified as deposits or not. You have to have a close look at the balance sheet and look at the liabilities that aren’t classified as deposits and figure out exactly what they are – e.g. commercial paper, bonds, perhaps certain forms of interbank borrowing, internal transfers of funds from a parent holding company, etc. etc.
But generally speaking, retail deposits are considered stable and wholesale money is considered potentially “hot” and potentially unstable. The money market operations I was talking about tend to issue “hot” money liabilities. Retaining or losing such funding is very price competitive, but its a necessary function.
In short, that loan/deposit ratio isn’t nearly granular enough to provide good information for a particular bank or even a country of banks without understanding a lot more about the definition of the underlying liability categories. I wouldn’t worry about it a whole lot unless you’re prepared to put a lot of time into the mining of data classifications.
Awesome info. Thanks.
Woolsey’s description of bank operations seems compelling and well-informed. But that description, without further explanation, seems to undermine the case for the the feasibility of the monetarist policy recommendation:
They [monetarists] favored changing the quantity of base money according to a different principle–keeping some measure of the quantity of money on a targeted growth path…”
Right. But my understanding is that this policy was tried, and implemented in a few different ways, before being abandoned, because it turned out that there was no stable relationship between the quantity of base money and the quantity of broad money – no matter how that latter quantity is measured.
It also turned out that, whether or not the central bank is determined to maintain an exact policy rate target value, its scope for policy choice is restricted due to its need to ensure the smooth functioning of the payments system is a system in which the demand for reserves is driven by the demand for credit at the prevailing rate.
I think the fundamental problem that remains is two-sided: (i) even if monetarists now reject the old textbook account on which their policy recommendations used to be based, they still seem convinced that, almost as a matter of definition or principle, the central bank controls the quantity of money, and so if the textbook transmission mechanism fails there must be some other transmission mechanism – expectations management, unconventional securities markets, etc. – that will do the trick; and (ii) a whole host of important real macroeconomic phenomena are “always and everywhere” monetary phenomena, so that if the central bank could manage to control the quantity of money, we would be well on our way to central bank-managed macroeconomic stabilization policy.
But both these contentions seem doubtful to many. The new version of monetarist, in which aggregate nominal spending is targeted and not just the aggregate monetary quantity, seems if anything even more ambitious than the old monetarism, and likely to run into the same types of barriers due to inherent limits on central bank powers in the economic system of 2012.
Two typos in the above:
its scope for policy choice is restricted due to its need to ensure the smooth functioning of the payments system is a system in which the demand for reserves …
should be:
its scope for policy choice is restricted due to its need to ensure the smooth functioning of the payments system in a system in which the demand for reserves …
and
The new version of monetarist, in which aggregate …
should be
The new version of monetarism, in which aggregate …
Excellent post, Dan. Thanks also to JKH for getting back to Steve’s question at Asymptosis!
Yes. John Carney had an excellent post on how it was tried by Volcker.
I put my marker (i) in the wrong place above. The sentence would make more sense if the (i) came after “they still seem convinced that”. So (i) and (ii) are supposed to be two monetarist views that are problematic.
JKH,
Yes even Friedman realized this. What’s equally important was that the Monetarist adamance was tried in the UK leading to a huge slump and was called the “Scourge of Monetarism” by Nicholas Kaldor.
In “How Monetarism Failed” http://www.jstor.org/stable/40720321, Kaldor wrote – on targeting a growth of monetary aggregates as was forced on the Bank of England – which obviously didn’t work:
“The U.S. and U.K. experiments in monetarism have thus left Friedman and the monetarists in an intellectually highly embarrassing position. Friedman has admitted that as far as the United Kingdom is concerned, the money supply is not exogenously determined by the monetary authorities, but he attributed this to the “gross incompetence” of the Bank of England. Later he implied the same about his own country. However, this puts an entirely new complexion on monetarism. It was nowhere stated in the writings of Friedman or any of his followers that the quantity theory of money only holds in countries where the monetary authorities are sufficiently “competent” to regulate the money supply. If the Bank of England is so incompetent that it cannot do so, how can we be sure that the Bank of Chile or of Argentina or Mexico- to take only the highly inflationary countries- is so competent, or rather so competently incompetent, as to make it possible to assert that the inflation of these countries was the consequence of their central banks’ deliberate action in flooding them with money? How, indeed, can we be sure that any of the central banks- not excluding even the German Bundesbank or the Swiss National Bank- are sufficiently competent to be able to treat their money supplies as exogenously determined? And what happens if they are not? Surely we need a general theory of money and prices that is capable of embracing the cases of countries with “incompetent” central banks, such as Britain and the United States.”
Friedman also admitted other things:
Monetarism and UK monetary policy http://cje.oxfordjournals.org/content/4/4/293.extract
“The basic contention of monetarists is that there is a stable function of money in relation to income (which comes to the same as saying that there is a stable velocity of circulation, invariant to changes in the quantity of money in circulation). This assertion, first put forward by the early followers of the quantity theory of money in the 18th and 19th centuries, was denied by Keynes and reasserted by Friedman on the basis of statistical evidence which shows a high correlation between changes in the amount of money in circulation and changes in the money value of the national income. Friedman admitted, however, that there is nothing in his findings which logically excluded an interpretation diametrically opposite to his own—i.e., that the change in the money supply may be the consequence, not the cause, of the change in money incomes (and prices), and that the mere existence of time-lag—that changes in the money supply precede changes in money incomes, is not in itself sufficient to settle the question of causality—one cannot rule out the possibility of an event A which occurred subsequent to B being nevertheless the cause of B (the simplest analogy is the rumblings of a volcano which frequently precede an eruption). Apart from that it is notoriously difficult to establish the existence of a lead of one factor over another, when both move in the same direction in time and the whole question of the existence of a ‘lag’ is by no means established.”
Woolsey:
“It is rather that they opposed that policy. They favored changing the quantity of base money according to a different principle–keeping some measure of the quantity of money on a targeted growth path…”
But that’s just a minor tweak to Monetarism, isn’t it? It’s the same old.
Thanks, Ramanan.
As I noted, the purpose of my post really was not to explore monetarism to any great degree, notwithstanding the title. It was to point to the nature of money market operations as the front end function of bank reserve management – much more so than lending. And I think Woolsey’s reference to this aspect was rare in its insight.
Yes JKH understood. I actually become all ears when some Monetarist tries to say something about causalities. Actually I was going to write something on my blog something related: New Keynesian endogeneity versus PKE.
Btw, a decent account of the money markets exists in Moore’s Horizontalists and Verticalists. Sadly the book is out of print. I will try to see if there are papers of his which can be posted online . Moore has good discussion and good history of Fed’s attempt to control credit via money markets. Also Moore would phrase “Banks compete on price and service.” as price competitiveness and non-price competitiveness.
But the importance of money markets is definitely missing in blogosphere around “loans make deposits”.
Steve,
Try this way of looking at it:
Suppose, in a simplified banking system:
Loans = Deposits + Capital
By initial construction, in order to be in balance, this system has zero required reserves and zero excess reserves (either of those would be an additional asset in the equation)
Now impose required reserves of RR and assume excess reserves of ER.
Then:
ER + RR + Loans = Deposits + Capital
Then the following relationships hold:
- Loans create deposits at origination
- Some deposits are converted to capital as the system grows and more capital is required
- RR are generated after the fact as banks expand deposits
- The CB supplies adequate RR by acquiring assets
- Banks compete for deposits in order to sustain funding that nominally matches RR + Loans
- The CB also supplies ER through OMO, in order to stabilize the policy interest rate and payment system clearing; the OMO create the deposits that produce the ER
NOTE:
A positive ER means RR is satisfied
A negative ER means RR is not satisfied (e.g. the Fed has occasionally run negative ER positions for the system in the past when tightening aggressively)
THEN:
Banks do NOT necessarily require positive ER in order to make new loans and create new deposits. Loans create deposits without any immediate impact on system reserves.
BUT:
Positive ER positions may well encourage banks to acquire money market assets, which can create new deposits in the same way that new loans create them. That is one of the technical messages of the post, regarding the money market dynamic.
@JKH:
- Loans create deposits at origination
For “the banks,” but not necessarily for the bank. The borrower may/will move the money out of the bank instantly.
- Some deposits are converted to capital as the system grows and more capital is required
Really?? How are deposits converted to capital? And if they can be, don’t deposits increase an individual bank’s lending capacity by relieving its capital constraint? Or is this at the aggregate level? If so, how does it come about via individual banks’ operations (with each other and the Fed)?
- Banks compete for deposits in order to sustain funding that nominally matches RR + Loans
I need to understand “deposits are converted to capital” before I understand this.
- Loans create deposits without any immediate impact on system reserves.
Right, with “immediate” being a key word. Over time the Fed *must* supply adequate additional reserves relative to deposits, or lose control of the interbank rate.
- Positive ER positions may well encourage banks to acquire money market assets, which can create new deposits in the same way that new loans create them. That is one of the technical messages of the post, regarding the money market dynamic.
That’s an aha that I’m not totally getting. Say there are ten banks, and five of them have ERs. What do those five banks do, and how does it create new deposits?
“For “the banks,” but not necessarily for the bank.”
- Yes, obviously, I hope.
“I need to understand “deposits are converted to capital”
- For the system, internally generated capital is a “slice of net interest margin” that comes out of the hides of the borrowers net of what the depositors are paid. I.e. borrowers pay the interest through debit to their deposit accounts. Depositors get credited a lesser rate of interest. So that net interest margin is an effective conversion of deposits into internally generated capital (net of all the other associated expenses of banking of course). Externally raised capital is paid for by cheques written on deposits, which again effectively converts deposits to capital.
“Right, with “immediate” being a key word.”
- Yes. The time order is the key. Banks don’t need to react to newly imposed reserve requirements until the Fed has already supplied them on a system basis.
“What do those five banks do”
- For example, they buy treasury bills and commercial paper from non banks (using investment dealers as conduits). The non banks (e.g. pension funds, investment funds) deposit the cheques into their bank accounts, which creates higher deposit levels.
Great work JKH. Keep it up, and thank you for the free education.
thanks Erik
JKH, intriguingly post, though you may be a little too kind to monetarists and a little too unkind to Post Keynesians.
Marc Lavoie (pp. 12-17) observes that some voices in officialdom do have a sophisticated view of banking. The section where he holds Fed economists Keister and McAndrews to account for modifying the standard “money multiplier” story gets to the crux for me. There would be reason for celebration if orthodoxy just dispensed with that notion altogether (instead of trying to modify it in some way or another) and freely admitted that they have caused great confusion there.
http://www.levyinstitute.org/pubs/wp_606.pdf
In the bigger scheme of things, the rise of Friedman’s monetarism went hand-in-hand with the downfall of neo-Keynesian, and ill-fated commencement of the neoliberal era. When I think of monetarism I get mental images of the US Fed bankrupting developing economies in the 1980s in its quest to quell inflation; the elevation of inflation-targeting over full employment; and the silly idea that the “market” is some sort of all-seeing equilibrium-seeking perfectly-rational God-like-entity.
So, I’m not sure I’d ever give monetarists a compliment, let alone a backhanded one. Woolsey’s version of monetarism may be a little more sophisticated but the matter is one of balance; and, I do not think that any of the orthodox schools teaches or understands money and banking that well whereas I think Post Keynesians do. Yes, the broad focus in the endogenous money literature is on ‘loans create deposits’, though my reading of the intramural “accommodationist versus structuralist” debate starting in the late 1980s is that Post Keynesians have thoroughly gone through bank management practices on both the asset and liability side.
I do not doubt that issues such as how banks use money market instruments to manage their balance sheets (or where equity issuance fits into the picture) get much run in Post Keynesian blogs. Academic papers are another matter and that is where nitty-gritty technical discussion occurs. Orthodoxy does typically run with its ‘deposit/loan multiplier’ story so Post Keynesians emphasise ‘loans create deposits’. The basic operations of the banking system are worth reiterating until they are accepted: Krugman’s ignorance of banking is a case in point.
I will take your inference that in some academic debates the protagonists talk past each other: it is much easier to critique ‘straw men’ caricatures. As one example, Steve Keen tends to present US Fed Chairman Ben Bernanke as well, basically as a an idiot for not understanding Minsky and Fisher. I have read at least one other blog from Steve where he pulls out the following quote as proof that “Before the crisis, Bernanke dismissed Fisher’s debt-deflation explanation for the Great Depression”:
“because of the counterargument that debt-deflation represented no more than a redistribution from one group (debtors) to another (creditors). Absent implausibly large differences in marginal spending propensities among the groups, it was suggested, pure redistributions should have no significant macro-economic effects. (Bernanke 2000, p. 24)”
http://www.debtdeflation.com/blogs/2012/04/16/inet-presentation-minskian-perspective-on-instability-in-financial-markets/
That Ben must be idiot… but wait… immediately after he adds something important and I will quote Bernanke (1995) from the original paper continuing on from the above:
“However, the debt-deflation idea has recently experienced a revival, which has drawn its inspiration from the burgeoning literature on imperfect information and agency costs in capital markets… From the agency perspective, a debt-deflation that unexpectedly redistributes wealth away from borrowers is not a macroeconomically neutral event: To the extent that potential borrowers have unique or lower-cost access to particular investment projects or spending opportunities, the loss of borrower net worth effectively cuts off these opportunities from the economy… If the extent of debt-deflation is sufficiently severe, it can also threaten the health of banks and other financial intermediaries (the second channel).”
http://fraser.stlouisfed.org/docs/meltzer/bermac95.pdf
Bernanke did not pursue how differences in marginal propensities to spend between debtor and creditor agents (and flows of “inside debt”) impact on aggregate demand (which Tom Palley calls the ‘debt service transfer’ mechanism) but he DID NOT reject Fisher’s theory. His so-called “financial disruption hypothesis” of the Great Depression has a Minskyan flavour insomuch as the focus is on the effects of balance sheet congestion. Why let nuance get in the way of a good story? Then again, Bernanke took a fancy to the ‘Great Moderation’ fable (where financial “innovation” and adoption of inflation-targeting monetary policy frameworks had permanently dampened the business cycle), so on balance he gets a big thumbs down for that. Think I’ve said enough here
When I wrote monetarism makes me think of “the elevation of inflation-targeting over full employment” I should have written the “the elevation of price stability over full employment”. The early 1980s monetarist experiment with money-supply targeting was a dismal failure though the focus on price stability above all else lived on in inflation-targeting regimes.
Also, JKH, I forgot to note that the purpose of your post was “not to explore monetarism to any great degree”. So take my comments on the right ‘balance’ when discussing monetarism and Post Keynesian as general remarks directed to everyone.
My first comment is not yet up (being moderated), nevertheless, had some other thoughts.
JKH: “The deposit business is very competitive and the nature of bank liquidity management much complicated as a result. The idea that “loans create deposits” isn’t sufficient to capture the dynamic of deposit gathering in a competitive banking system. Thus, the nature of deposit origination as in “loans create deposits” does not explain the consequent dynamic that competing banks exhibit in attracting and retaining their own desired share of originally created deposits.”
I think Fullwiler (2012) explains matters fairly well:
“The key here is to understand the business model of banking—which is to earn more on assets than is paid on liabilities, and to hold as little capital (equity) as possible (since that’s generally more expensive than assets). The most profitable way to do this is to make loans (that are paid back, obviously, so credit analysis is an important part of this) that are offset by deposits, since deposits are the cheapest liability; borrowings in money markets would be more expensive, generally. So, Bank A, if it is not able to acquire deposits is not operationally constrained in making the loan, but it will find that this loan is less profitable than if it could acquire deposits to replace the borrowings.”
http://www.nakedcapitalism.com/2012/04/scott-fullwiler-krugmans-flashing-neon-sign.html
JKH you liked Woolsey’s remarks: “The banks then use money market instruments to adjust for any temporary imbalances between the demand for loans and the supply of deposits.”
Following quote from Palley is interesting:
“The money multiplier approach embodies “portfolio endogeneity” whereby the money supply depends on agents’ portfolio preferences for money relative to other assets… Changes in the money multiplier arise from changes in the composition of portfolios. In this regard there are two sets of portfolios — those of the non-bank public and those of banks. The non-bank public consists of households and non-bank firms, both of which demand cash, checkable deposits, and time deposits. Variations in these demands are the traditional focus of money multiplier analysis. Banks also demand liquidity in the form of excess reserves, so that changes in bank preferences also affect the money multiplier. These portfolio composition effects are an integral part of the Post Keynesian structuralist approach to the money supply. They provide a mechanism whereby banks are able to generate additional liquidity to support deposits created by increased bank lending. Bank asset and liability management activities are central. Banks vary their own portfolio demands (asset management), and induce variations in the portfolio demands of the non-bank public (liability management). Together, these variations allow banks to accommodate increased lending independent of the monetary authority’s stance.”
http://www.thomaspalley.com/docs/articles/macro_theory/endogenous_money.pdf
In basic terms Palley is saying that ‘loans create deposits’ is not enough to understand the business of banking. Hope that helps
Thanks. Those are very good quotes.
The Fullwiler piece looks at net interest margin analysis more than liquidity management. Those are connected but separate. Also, I would not agree that the business model aims to minimize equity. There are certainly bank CEOs that like to have some excess capital on hand as a rule. But I think STF was probably just simplifying there.
The Palley piece focuses on liquidity, but is still slightly loan centric. Liquidity management must deal with reserve management demands from potential deposit/liability outflows that may be quite separate from loan demand.
Yes, the Palley quote is really quite good. Fairly balanced actually.
This is key:
“Bank asset and liability management activities are central.”
I’ve seen it before, but can somebody offer a definition of “Post Keynesian structuralist approach to the money supply”?
Calling Ramanan?
5, 4, 3, ….
Giuseppe Fontana has written a paper summarizing the accommodationist v. the structuralist approaches: “Post Keynesian Approaches to Endogenous Money: a time framework explanation” (2003)
http://cas.umkc.edu/econ/economics/faculty/fontana/winter%202005/rope.pdf
“Lord Keynes,” Endogenous Money: A Bibliography
http://neweconomicperspectives.org/2012/05/quantitative-easing-and-commodity-prices-an-mmt-approach.html for the principal lit on PKE views regarding endogenous money.
“I’ve seen it before, but can somebody offer a definition of “Post Keynesian structuralist approach to the money supply”?
Calling Ramanan?”
There was a group of economists such as Kaldor, Lavoie, Moore who came to be known as Horizontalists (because of Moore’s book) who argued xyz and were taken unsuccessfully to task by another group called structuralists. Briefly, the argument of the latter was that interest rates rise on loans start rising. The structuralists were arguing the same old: central bank controls the monetary base but money endogeneity was a result and a function of banks’ innovations rather than the central bank acting defensively. So in their approach the central bank can create more reserves and this can lead to more lending. Although Moore has nice descriptions of general banking/liability management, the stress was more on the central bank acting defensively. But one can find good stuff on articles trying to dispute Moore.
I final point addressed here to Ramanan. Some structuralists did argue that the central bank can target the money supply and use the monetary base as a policy instrument though that was not the only issue: the horizontalist / structuralist debate has a lot of ‘grey’. I refer you to the three paragraphs after the following quote from Palley (2008, p. 9):
“For horizontalists, banks are unconstrained by the supply of finance which is infinitely elastic at a price set by the monetary authority. This horizontalist position is wrong on three counts. First, it ignores the effects of regulation. Second, it ignores the effects of increasing risk on individual banks. Third, it ignores bank portfolio diversification concerns.”
http://www.peri.umass.edu/fileadmin/pdf/working_papers/working_papers_151-200/WP178.pdf
Fontana’s (2003, p. 297) relaying on Davidson’s views suggests shades of grey: “Indeed, Davidson moved from the position of one who ‘neither accepted nor denied the “extreme form” of endogenous money, namely a “full accommodation” at a given interest rate’ (Davidson, 1989, pp. 489–490), to the acceptance that ‘an increase in the demand for money induces an endogenous increase in supply if bankers are willing and able to expand under the rules of the game that regulate banking operations’ (Davidson, 1994, p. 136).”
http://cas.umkc.edu/econ/economics/faculty/fontana/winter%202005/rope.pdf
Both camps make solid points on certain issues. When Lavoie (2010, p. 13) writes: “Unfortunately, this idea of idle reserves, associated with credit-rationing banks, is also rampant among some Post Keynesian economists (Palley 2010)” I agree with him in black and white terms.
http://www.levyinstitute.org/pubs/wp_606.pdf
BF,
Don’t agree with Palley, although I couldn’t catch your general point addressed to me. Moore called it a storm in the teacup
Ramanan, my general point was that your 12:26pm was a little too thrift, and there are many issues in the debate which are a little ‘grey’. I quoted Davidson not so much as support for structuralists but to underline that if a man of his knowledge is little tentative on the debate, well, that speaks volumes for me that there are no clear overall winners and losers. Is Moore right on certain issues and wrong on others; arguably, yes. The same goes for structuralists.
Take, for example, that Palley feared that the Fed would no longer be able to set the fed funds rate and influence other interest rates in the economy due to the rise of e-money. Fullwiler (2006) argued that: (1) there will always be demand for reserves because banks need them to settle transactions between the public and Treasury; and, (2) the idea that the Fed’s ability to set interest rates pivots on quantity changes in reserves is not right. For some time the Fed has attained the fed funds rate by ‘open mouth’ policy rather than engineering quantity changes in reserves. And as Fullwiler (2006, p. 510) argued: “the Fed’s influence on other rates occurs via arbitrage in other markets against the federal funds rate, what matters is that a demand for reserve balances exist that is significant enough—call it a nontrivial demand for reserve balances—for such arbitrage to continue into the future.”
S. Fullwiler, ‘Setting interest rates in the modern money era’, Journal of Post Keynesian Economics, 2006, Vol. 28, No. 3.
Then again, with the rise of nonbank lenders (the ‘shadow banking system’) the Fed is limited in its capacity to provide liquidity support to nonbank lenders (i.e. balance sheet substitutions) because these financial firms are outside the ‘traditional’ transmission belt of monetary policy (which means that the Fed’s ability to influence broader interests rates can be weak when a ‘shock’ occurs as the US monetary system was designed to suit an institutional setting/structure where depository banks occupied centre stage); hence the alphabet soup of Fed lending facilities created in the crisis.
Whether all structuralist positions are ‘a storm in a teacup’ is a matter for debate. If horizontalism implies that “liquidity pressures” and “risk” do not affect bank behaviour, well, I have difficulty with that. Minsky is sometimes viewed as a structuralist. Palley (2008) depicts Lavoie and Wray as ‘weak’ horizontalists as opposed to the ‘strong’ horizontalism of Moore. In one of the three paragraphs that I referred you to Palley (2008, p. 11) offered in a footnote: “Wray (2007) also points out that capital requirements mean that banks cannot simply accommodate all loan demand as suggested by Moore (1988b), even if that loan demand is creditworthy.”
When JKH 1:52pm writes: “the finance supply curve should be horizontal on a risk adjusted basis” I think the position of structuralists was that, from their perspective and interpretation of Moore, his argument was the finance supply curve is horizontal [insert full stop] such that “liquidity pressures” and “risk” do not matter. Whether Moore was taken out of context or oversimplified his position I do not know. Two decades of debate suggests that there must be merits to the views of both sides; and, I do not think that the issues can not be covered here let alone resolved.
Given that I opened this can of worms and the issues are technical can I suggest that there may be more to gain from going through the Krugman/Keen debates (as JKH sought to do): there are fundamental differences between orthodoxy and Post Keynesians on banking
BF,
Thanks. What I was trying to say was that there was a general tendency to argue that Moore was holding extreme positions when it was perfectly normal. I will hold off commenting on this – maybe sometime later – because it’s not the main topic here. Have to be careful as well. Recently on a social networking group I was trying to argue that Minsky held the loanable funds intuition as late as 1994 and was met with major attacks! Anyways, battles in economics rarely end. Having said, yes there are fundamental differences between PKE and orthodoxy. Woolsey seems to suggest “he always knew” but errs when he suggests that the central bank can target the growth of money stock. That sort of thinking led to a disaster in the UK in the late 70s and early 80s.
But yeah there is a lot lot in literature that’s not in the blogs.
I’ve been struggling to get some intuition from the Palley paper about what “structural” is supposed to mean, in the context of horizontal, vertical, and diagonal (
).
The best I can do is associate horizontal with the central bank determining the interest rate policy benchmark, combined with endogenous money creation by banks, and structural with the balance sheet complexity of banks, including asset/liability mix with risk adjusted pricing spreads relative to the central bank benchmark rate.
Part of it seems to be a debate about the slope of the “finance supply curve” for banks, but that shouldn’t be an issue if risk is measured consistently and a growing bank is sufficiently capitalized to bear the risk. In other words, the finance supply curve should be horizontal on a risk adjusted basis.
All in all, it strikes me as awkward analysis. My instinct is to leave the horizontal concept clean in terms of the central bank setting the benchmark rate and the banks creating money endogenously through lending, with appropriate risk analysis and pricing. It seems to me that the entire structural conversation is a bit of a muddle of what should be more properly categorized as bank risk analysis.
JKH,
Here’s a good paper on this. It’s kind of a convoluted debate and the “strong” and “weak” horizontalists appear to have converged to some degree with the structuralists. It’s further complicated by MMT’s language regarding vertical and horizontal as vertical is used to describe transactions between the govt and non-govt and horizontal is often used to describe the banking transactions (even though the Moore definition of horizontal implies something totally different). I much prefer the MMT definition as it clearly defines which entities create NFA and which do not. I am not sure that Palley, Moore or Tobin made this distinction as clearly as MMT does….Further complicating this is that MMT’s prescriptive components are almost entirely vertical in their nature giving MMT a verticalist perspective (even though they’re “weak” horizontalists by Palley’s description).
http://scholarworks.umass.edu/cgi/viewcontent.cgi?article=1149&context=peri_workingpapers
Thanks, Cullen.
I note that Hudson in that recent article on Krugman we discussed contested the use of the word “structural”, although again in a completely different context.
Words and their meaning.
Eh, they’re just words. Who needs ‘em when we’ve got metaphors!!! Wait….
that’s right – you gotta watch your step around here
I quoted Scott as he explains in simple terms why banks compete for deposits: it is a lower cost liability relative (at least) to borrowing in money markets. Presumably, banks also compete for deposits as that may result in additional clients, to whom they hope to lend to.
When Scott wrote banks want to “hold as little capital (equity) as possible” I understood it as meaning banks are leveraged entities; and, as a general rule keep capital adequacy near but ‘prudently’ above the limit as that usually means more profitability. Hence, bankers do not want to minimise equity per se, but they surely want to leverage out their balance sheets close to the extent allowed by regulators (and one would hope restrained by prudent estimations of credit risks in order to protect the value of shareholder claims and client deposits but as the recent crisis attests that is not always so).
“Hence, bankers do not want to minimise equity per se, but …”
Well said and agreed.
“In basic terms Palley is saying that ‘loans create deposits’ is not enough to understand the business of banking. Hope that helps ”
True, its not enough to tell the whole story, nor do I think it was intended to by those few who use the phrase. It is, however, absolutely true, in contrast to the deposits “create” loans idea which is prominent……………………. AND false.
JKH
Overall a very interesting post. Thanks
I think we are on the same page here Greg. In my 1:20am I wrote: “Orthodoxy does typically run with its ‘deposit/loan multiplier’ story so Post Keynesians emphasise ‘loans create deposits’. The basic operations of the banking system are worth reiterating until they are accepted: Krugman’s ignorance of banking is a case in point.”
My 4:38am remarks that “‘loans create deposits’ is not enough to understand the business of banking” was not to deny the veracity of the idea but to place it within a context that Post Keynesians get that “Bank asset and liability management activities are central.”
BF
Saw that after looking back. Sorry I missed it . As a friend of mine used to say
“Sometimes I doesnt read good”
Thanks, Greg.
Also on the same page, although that might not have been clear from my approach to the post.
Sorry, I just now saw your 1:20 a.m.
I’ll be back.
Hi BF,
I won’t interrupt your discussion with JKH on the banking specifics, but I am more generally curious about your thoughts on MMR as a development – assuming you have some thoughts there….
Cullen
I’ll have get back to you on MMR sometime. What I can is that the forum appears to be one where the contributors seek cordial exchanges. No one or any theory has a stranglehold on truth when it comes to economics. So in my view the dare I say humble and tentative pursuit of knowledge that could be said of many posts and comments here is refreshing relative to some other forums (including many exchanges between academics) where dismissive language and egos can be counterproductive to learning and participation.
Thanks BF. We try over here to be fair.
In the end, there is a machine. We (as a civilization) need to understand how the machine works.
Yes, I am not sure how new you are to the site, but we’ve had some fairly substantial disagreements with MMTers although we agree with them that the post-keynesian framework is a far superior understanding than the neoclassical framework. Ultimately, as Michael alluded to, we’re just trying to describe how the “machine” works so we can establish a better foundation from which to build policy. We believe MMT and most of economics begins with a policy approach and works their understanding of the system around this (generally ideological) approach. Some claim it’s impossible to extract the politics from economics, but I am not so sure that’s true….
Anyhow, I’d love more of your insights as yes, we’re very open to interpretation. We don’t pretend to have all the answers and any insights that can contribute to a better understanding are certainly positive.
“I’m not sure I’d ever give monetarists a compliment”
This post was a deliberate stretch in that direction, with a purpose in mind. It’s intended to be a catalyst to get something else going. As a sidebar only, I would like to understand better what it is the monetarists think they’re trying to do in terms of formulating and visualizing the execution of their recommended monetary policy. I don’t think they explain their operational vision particularly well. Maybe it’s unrealistic to wish for more from them. At the same time, I’m not entirely satisfied with some of the post Keynesian descriptions that I see of the monetary system more broadly. I believe there is some intertwining of factual and counterfactual lines of reasoning going on in cases where one would be more effectively separated from the other, although perhaps some of this is attributable to blog simplified translation requirements. I agree they have banking down reasonably well. But this particular area of bank reserve management is one that could use some mutual understanding from all quarters. It’s actually not that easy to explain the way it works now, because it is such a central function from a commercial bank perspective, while interfacing with a similarly central operational function of the monetary authority. Finally, I’m hoping this post is one of several in a series that gradually explores what one might try and define as the crux of the Krugman/Keen difference in views of the importance of banking. That could be a bit ambitious, but it’s what I have in mind.
“I do not doubt that issues such as how banks use money market instruments to manage their balance sheets (or where equity issuance fits into the picture) get much run in Post Keynesian blogs. Academic papers are another matter.”
Yes. I do try to be sensitive to that in general, and qualified it here as:
“His high level summary of it compares favorably with similar accounts I can recall seeing on post Keynesian blog posts in general.”
Thanks for your interest!
This was intended to be addressed to BF.
JKH: my 2:37am “take my comments… as general remarks directed to everyone” was to acknowledge that my 1:20am reflected less what you wrote and more about my own general dislike of monetarism. (By the way, I’m in a quite different time zone to most of you, so you’ll have to forgive me if my replies are a little delayed).
I think one of your points, JKH, was that Post Keynesians typically discuss the banking system in the context of a newly-created loan when the business of banking involves active management (including reserve positions) in response to the entire range of client initiated activities (e.g. the payment and receipt of wages, consumer spending, buying/selling assets, etc.) and their own trading accounts. So why the loan centric focus? By analogy, the economy is a forest, and how bank credit impacts on the macro economy (e.g. the money supply, aggregate demand and business cycle) is a big tree in that forest.
The theory of endogenous money arose in opposition to monetarism; hence, one reason for the loan centric focus is because it makes much more sense to look at the flow of lending as an indicator of economic activity than changes in the stock of money. Looking at the economy though an endogenous money lens is vastly different to looking at the economy entirely through the orthodox ‘loanable funds’. (That said, I do think nonbank credit is an awfully big tree, which deserves a little more spotlight).
Going through the Krugman/Keen differences should be a worthwhile venture. There is a great deal of confusion surrounding the topic of banking (including some ‘specialists’).
I do hope you get round to Steve’s argument that the positive (or negative) change in private debt volumes measured as a percentage of GDP equals the addition (or deletion) to aggregate demand supplied by private debt. Surely a net increase (or decrease) in debt by a financial intermediary has very different effects on aggregate demand than say consumer finance or nonfinancial business debt? Do net increases (or decreases) in mortgage finance to buy existing homes as opposed to that to build new homes also add to or take away from aggregate demand on a one-to-one basis? One thinks no not at all.
The issues are first getting the description right and then establishing causation. Neither are simple in complex system, especially those involving human beings. Intense arguments develop that are difficult to settle due to lack of formalization such as applies in math and harder sciences. Here’s a guest post at Steve Keen’s on this by a mathematician and economist. He’s critiquing Krugman-Keen but the same issues apply pretty much across the board in econ.
http://www.debtdeflation.com/blogs/2012/05/16/an-attack-on-paul-krugman/
We get all worked up over who is right and who is wrong, but the big picture is that we are all trying to get a better grip on an enormously slippery subject that has huge implications social, politically and economically through both institutions and policy. No one has the ocean in their bucket.
Tom
I think you are right, it is a complex system. So neo-classical economist create theories and models, with simple but outrageous assumptions about the economy such as it is:
• non-monetary and simple barter
• equilibrium-fixated
• uncertainty-free
• institutionally barren (no banks) with hyper-rational individual-based reductionist
Now as an educated individual who has worked with complex system I can accept that maybe you cannot get an accurate vision to see a small recession and those can sneak up on you, but to miss the largest financial disaster since the last Great Depression to me suggest maybe the idea that capitalism is:
• Monetary
• inherently cyclical
• embedded in time with a fundamentally unknowable future
• institution-rich and holistic
• And we should considered the interactions of its four defining social entities: industrial capitalists, bankers, workers and the government.
Would get a better reception at the table of ideas, since for some reason the post-Keynesian and Austrian schools who explicitly consider some or all of the things above in their models of the economy, expected a crisis to some degree.
http://mpra.ub.uni-muenchen.de/15892/1/MPRA_paper_15892.pdf
Paul Krugman neoclassical theories in which macroeconomic problems arise only if there are microeconomic ‘imperfections,’ cannot IMO ever be reconciled to Keens views that you cannot derive models of the macro economy from any microeconomic models of the behavior of individuals.
I do not think MMR can truly understand the economic machine unless they choose the correct path on how it is put together. Is it heterodox economics or neoclassical economics; for me the choices are simple, you can work on trying figure out why Krugman and others failed to see the largest housing asset bubble in history, or why Keen and others were able to see it coming.
@Oilfield Trash:
Fully agree. Try modeling the weather system with supply and demand curves.
Most economists, particularly the Neo classical, I think are now virtually the only ‘scientists’ who attempt to model a real-world system using static, equilibrium tools.
As a result of this isolation, economists have been shielded from developments in mathematics and other sciences which have revolutionized how scientists perceive the world.
As Keen put it “The attempt to conduct a critical dialogue within the profession of academic economics has therefore failed, not because neoclassical economics has no flaws, but because – figuratively speaking – neoclassical economists have no ears.”
I do not think I would care much if they would stop playing with the dials on the real economy or as MMR would see it, stop trying to tune a machine they do not understand. At best they currently are as skilled as physicians who would practice bloodletting to cure sickness in medieval times.
If the patient gets better the cure worked, if it get worse or stays the same then we just need to do more bloodletting. If the patient dies then it is god’s will.
If the economy get better they take credit, if it get worse or stays the same then we just need more of the cure until the economy get better. If they kill the economy then it is the will of the Black swains.
Tom,
Neutral question:
Do you think that it’s actually possible that Paul Krugman doesn’t understand that when you walk into a bank branch and have a loan approved and funds advanced, that one of your options is to have those funds immediately credited to your deposit account with that bank, and that if you know something about basic accounting, that the result will be an increase in the bank’s assets and an increase in the bank’s deposits, at that very moment in time? Is it actually possible he doesn’t understand that? If not, and if he does understand that (which I think he does), what do you think accounts for his apparently strange explanation of something that appears to contradict that fact?
JKH, I don’t see how PK cannot understand. But he often says that he doesn’t understand heterodox economist when they talk about money & banking and finance. I think he just accepts major orthodox positions such as money neutrality, loanable funds, etc. that underlie the monetarist position. It doesn’t seem that he has taken the time to get into it. He certainly buys into IGBC, and before he changed his position, he was on the side of those claiming that the US could become insolvent.
I have a terminal degree in another field (pun intended), and I know well that one cannot know everything about a field, let along about the specialized aspects. I admit to superficial understanding about some important areas in my discipline, since it is not physically possible to get up to date and remain up to date in the whole discipline.
It may be that PK has just not pursued mone & banking and finance and is just going with the orthodox understanding, presuming that the supposed experts know what they are talking about. Which is strange, since he often criticizes supposed experts for not knowing what they are talking about.
Add t to that Warren’s report of having explained MMT principles for a couple of hours. That is a lot time considering how parsimonious Warren is in explanation. Maybe PK just nodded assent to be polite even though it went over his head.
The short answer is that PK is an enigma on this issue.
I have a bit of a theory about that Krugman/Keen discussion. Part of it is that a great deal of what PK rolled out in sequence was a reflexive attack on Keen’s definition of aggregate demand, which is an unusual one. In other words, a lot of the money stuff he came up with was incidental to trying to illustrate that Keen’s particular definition of aggregate demand was problematic. His insistence that banks don’t create money out of thin air may have been a way of saying that banks do need to ensure that loans are matched by funding at the end of the day – literally at the end of the day – apart from LLR advances. So there is a proactive element of deposit gathering that’s necessary in order to ensure that for the individual bank “the balance sheet balances” as expected by regulators, without undue reliance on LLR, and notwithstanding the origination of new money with new lending. The final part of it relates to the idea that a good deal of bank financing is bridge financing by nature – e.g. a firm that borrows bank funds temporarily prior to a debt issue. That doesn’t contradict the fact that loans create deposits, but it modifies the description of where loans fit into the broader financial system pattern in the long run. That’s all defending him, I suppose. But that said, I think he made one error that I would identify more clearly as an error when he started describing the system effect of currency withdrawals, where I think he missed entirely the central bank requirement to pump more reserves back out into the system in order to replace those used by banks to “buy” currency. Anyway, I hope to have more to say on this sometime later.
i.e. I think he missed the point that “the central bank requirement to pump more reserves back out into the system in order to replace those used by banks to buy currency” would typically result in the creation of new deposits to be held by the non bank end sellers of assets to the CB in that case – when the point he was attempting to make apparently was that deposits are “net lost” to the system when currency is issued; not so in normal reserve times
Seem to me that you are crediting PK with a lot more sophisticated understanding of money & banking and finance than he has displayed. Do you have any evidence that he has this level of understanding. Seems to me rather that he has made a lot of freshman errors.
No evidence. Just hope, I guess.
I find it impossible to understand how somebody in that position doesn’t understand what happens when you go into a bank branch for a loan. It’s a matter of observation plus high school accounting. It’s not sophisticated, is it?
As matter of fact, it is. Very few people understand what happens behind the window. Add to that all the erroneous assumption of orthodox economics, and it is pretty impenetrable without actually studying up on it. I think that PK may be getting forced into doing just that to remain credible. He is taking a lot of flack over this.
I think the circuitists call it initial finance versus final finance.
There’s more to banks. It is difficult to make a private sector model with just households and production firms because inconsistencies arise in definition of profits and interest payments. So one is forced to introduce banks in a description/model due to self-consistency requirements.
On “loans create deposits” (I prefer loans make deposits because loans create deposits gives the impression that only loans create deposits) – I think the more natural starting point is the Keynesian principle of effective demand.
@Ramanan: “It is difficult to make a private sector model with just households and production firms because inconsistencies arise in definition of profits and interest payments.”
This is the conclusion I came to in the course of the whole S = I + (S – I) discussion, and I think it’s a key place that my confusion (which I think ignited that whole business in the first place) derived from. In a two-sector model it seems like you have to think about business profits devolving to both businesses and households at the same time. Or more accurately/alternately, you need to choose one, and neither is accurate. Adding banks makes it more complicated, but also more complete in a way that allows a coherent, internally consistent understanding.
(Though perhaps the only problem is the lack of capital accounts for firms and households in most two-sector representations?)
Which is one reason so much of “textbook” economics doesn’t seem coherent to me. They’re thinking inside the NIPAs (essentially a barter economy with money for convenience), absent Flow of Funds.
“So one is forced to introduce banks in a description/model due to self-consistency requirements.”
So JKH, what do you think? Is it impossible to present an internally coherent two-sector (firms and households) model of a modern monetary economy? There have to be dangling terms, right? — rows and/or columns that don’t sum to zero — unlike here for instance:
http://www.concertedaction.com/wp-content/uploads/2012/03/Transactions-Flow-Matrix.bmp
Or does the expedient of adding capital accounts solve that in a two-sector model? I don’t think so, because the “capital” has to appear from and disappear to somewhere… ?
“I think the more natural starting point is the Keynesian principle of effective demand.”
Yeah. It was a great relief and aha to me to find that he’d already thought long and hard about what had (eventually) seemed obvious to me: at least when you’re talking about big market sectors like “labor” and “goods,” supply and demand within each is necessarily contingent upon supply and demand within all the others — not just on the price within one. The shapes and positions of all the curves are mutually determined (in a way that I think leads to the need for something like Keen’s differential equations for solution). Or that’s how I think about effective supply and demand. Correctly?
Not sure which issue you’re driving at – two sector or barter or both.
I see no problem conceptually. The accounting can be handled for two sector barter with income, balance sheet, and flow of “funds” statements.
Financial claims become real claims. E.g. a household balance sheet with a car and a house as assets and food payable claim (equal to the value of the car) and equity on the funding side. Flow of funds becomes flow of real claims connecting balance sheets. Income becomes real revenue saved or spent in barter.
Again, not sure which issue you’re focusing on. I may be missing the point.
Two big problems in articulating Y = C + I + G +(X-M), ignoring the external sector for simplicity. First, firms are owned by households, so there has to be care to avoid matter like double counting. Secondly, the formal financial sector is really public-private institutionally and therefore overlaps with government, while the extensive and growing informal (“shadow”) financial sector is somewhat of a wild card that few people have a handle on.
Steve Keen acknowledges much of this and it trying to get it into a formalizable box that can account for dynamism and complexity. But it’s not clear that he has the understanding of accounting, money & banking and finance to do so. It’s probably more than a one-man project to design an adequate macro model that takes all the factors into account.
I was wondering today about how the repo markets fit into Godley’s matrixeseses.
Steve/JKH,
Just by looking at one kind of flows such as consumption, sales, etc it is difficult to conclude I imagine. One needs other kinds of flows which have a stock equivalent such as change in bonds etc. I think when one includes them, it may be possible to show that the flow of funds accounts doesn’t make sense without banks. I will try to find a simple proof for this. I vaguely remember Dirk Bezemer saying something about this. Maybe I am completely wrong and dreaming.
“The difficulty lies, not in the new ideas, but in escaping from the old ones, which ramify, for those brought up as most of us have been, into every corner of our minds.”
- JMK
JKH,
I think the general tendency to say that its all just accounting and monetary stuff but IMO there’s more than that. If it were true then all bankers would know it all and we would have been in a different world. Economics is a lot harder than that.
e.g:
Warren Mosler repeatedly says xyz (e.g, IS/LM) is fixed exchange rate model – but he misses the point. The model is wrong to begin with and doesn’t apply to fixed exchange rates to begin with. Does the banker check if the exchange rate is fixed or floating when making a loan?
“ramify”
love that word
you should too!
funny you should bring up the fixed/floating “clean bifurcation” (my characterization)
I have an issue with that too; for later
Yes Ramanan shall like “ramify”.
Yes no bifurcation. To me none at all. In one its fixed and in the other floating. Even in the fixed case – in any arrangement – it floats within a band.
IIRC, Warren connects ISLM to the assumption of loanable funds, which he connects with gold standard thinking. PKE and MMT critique Krugman based on his seeming presumption of loanable funds and crowding out.
@JKH: “My instinct is to leave the horizontal concept clean in terms of the central bank setting the benchmark rate and the banks creating money endogenously through lending, with appropriate risk analysis and pricing.”
Is this suggesting that the driving force of creation is (almost?) purely horizontal, with the Fed necessarily being forced post facto to create reserves in response, then trading some of those in and out to adjust IBL rates, nothing more? Not sure if you’re staking that strong a position.
I think the Fed in setting the fed funds rate is the driving force of interest rate creation starting at the short end of the yield curve and working out toward the long. And I think the market is the driving force of interest rate creation starting at the long end of the yield curve and working back toward the short. The two meet somewhere in the middle. The Fed rate sends out a signal that factors into longer term interest rate expectations and the yield curve. The yield curve itself sends back a signal to the Fed about how the market is thinking about expectations, which may factor into the Fed’s short rate policy setting at the next stage. And so on.
That sounds right SR – not a strong position.
If you see some comments above (on “structuralists”) … This is exactly why structuralists incorrectly thought such positions were extreme.
I think most people who read this first time tend to ask
“do you mean the central bank is completely powerless?, surely this is an extreme position”.
While for a given rate, the central bank necessarily accommodates, but it itself can hike interest rates in response and this could be damaging (for example late 70s/early 80s). Even in the recent crisis, the increases by the Fed caused a lot of effects because it really increased the interest burden of household debt.
People also tend to associate “full accommodation” by banks themselves to such views – as if banks lend to anyone and everyone as Palley seems to suggest about such views.
There are however situations one can think of easily in which the central bank however is less powerful. For example there could be a boom due to strong exports or a fiscal expansion – in which case, the increase of interest rates by the central bank may not work at all and may be coincident with higher borrowing.