Saturday, May 30, 2015

Zoltan Jakab and Michael Kumhof — Banks are not intermediaries of loanable funds - and why this matters

In the intermediation of loanable funds model of banking, banks accept deposits of pre-existing real resources from savers and then lend them to borrowers. In the real world, banks provide financing through money creation. That is they create deposits of new money through lending, and in doing so are mainly constrained by profitability and solvency considerations. This paper contrasts simple intermediation and financing models of banking. Compared to otherwise identical intermediation models, and following identical shocks, financing models predict changes in bank lending that are far larger, happen much faster, and have much greater effects on the real economy.
Putting another myth to rest.

Bank of England
Working Paper No. 529: Banks are not intermediaries of loanable funds - and why this matters
ht Stephanie Kelton

36 comments:

The Arthurian said...

From page iii: "The DM [deposit multiplier] model however does not recognise that modern central banks target interest rates, and are committed to supplying as many reserves (and cash) as banks demand at that rate."

I see that argument all the time. It seems not to notice that modern central banks change interest rates, in order to change the demand for reserves.

Unknown said...

Art-

Changes in interest rates do not change the demand for reserves as demand for reserves is based on the bank's number of demand deposit accounts that have a reserve requirement attached.

The two concepts are totally independent of each other.

Kristjan said...

It is a good read. The profession must be embarrassed. But the show goes on I guess.

NeilW said...

Reserve requirement is a moot point. The central bank must supply sufficient reserves to meet demand or it loses control of its interest rate instantly.

And it's hardly an issue in a QE world where the central bank is paying interest on massive quantities of reserves rather than juggling around OMOs to so that Treasury pays the interest on reserves.

A large reserve requirement, or a large amount of QE on which interest is paid floods the system with bank deposits which may or may not be a cost to banks depending upon the dynamics of the deposit market.

Ryan Harris said...

When will the text books be updated to reflect reality? That is all that matters.

Economists should pressure Krugman, Mankiw and the other introductory text authors to make new editions because those are the texts that most future regulators, leaders and congressmen will read. They will create a basic economic framework in their minds of most that will eventually formulate policy.
Once people are taught ideas, flawed or not, it is nearly impossible to change minds and we've seen the ignorance from economic orthodoxy is nearly limitless in the wrath and destruction it unleashes on the world.

Kristjan said...

The DSGE is garbage of course but in this paper they just want to integrate the right framework into DSGE, they don't do it actually.
It is crazy, nobel prize winners don't understand basic accounting. How could banks lend their own liabilities, yet they think that's what they do.

The Arthurian said...

Auburn Parks: "The two concepts [interest rates and the demand for reserves] are totally independent of each other."

Jakab and Kumhof: "... modern central banks target interest rates, and are committed to supplying as many reserves (and cash) as banks demand at that rate."

Auburn, it sure looks like Jakab and Kumhof disagree with your claim that interest rates and the demand for reserves are totally independent of each other.

And yet, J&K still manage to ignore the Fed's longstanding policy of changing rates in order to change the growth of lending (and reserves).

MRW said...

@Neil Wilson,

Dummy alert. What does OMO mean? Thx.

MRW said...

Arthurian, call the Fed and ask. Auburn Parks is correct. "Changes in interest rates do not change the demand for reserves as demand for reserves is based on the bank's number of demand deposit accounts that have a reserve requirement attached."

I think you have the causation backwards.

NeilW said...

OMO = Open Market Operations.

It is where the central bank buys and sells government bonds until the interest rate in the inter bank market hits their target window.

They get the government bonds they sell by 'borrowing' them from the government...

(No seriously).

NeilW said...

"it sure looks like Jakab and Kumhof disagree with your claim that interest rates and the demand for reserves are totally independent of each other."

Only if you squint through a fog of pre-conception.

What they are saying is that the quantity demanded won't *alter* the rate the central bank is set - as it would in any market where quantity is restricted.

The rate is set irrespective of quantity demanded, because the quantity of reserves is totally irrelevant to anything. It's just an accounting fig leaf to make the payment system clear properly.

Nick Edmonds said...

The problem with papers like this is that they imply that the deficiencies of loanable funds explanations are all to do with the operation of banks. This is just not the case. In a monetary economy, investment decisions drive the quantity of savings, not visa-versa. This has nothing to do with banks, beyond the fact that banks as we know them (creating deposits by lending) could not exist in a non-monetary economy. Even if regulation was introduced to make banks operate as depicted in deposit multiplier stories, investment / saving causality would remain unchanged. Say's Law would still fail.

Anonymous said...

While banks' aggregate demand for reserves is indeed determined by the aggregate level of demand deposits, the level of demand deposits is determined by their volume of lending, and the volume of lending is sensitive to the cost of funds. How sensitive? That's an empirical question that can't be answered on theoretical grounds. But we know the answer isn't "not sensitive at all." After all, we know Paul Volcker was able to decrease the rate of bank lending and cause a recession by jacking interest rates way up.

Anonymous said...

Clearly the level of national saving does not fully determine the level of national investment; but it would be juts as wrong to say that the level of investment determines the level of saving. One thing we can say is that saving is a necessary condition for investment. A society that is consuming 100% of its output and saving 0% can't invest.

Kristjan said...

"but it would be juts as wrong to say that the level of investment determines the level of saving"

I don't agree. It is right to say that the level of investment determines the level of saving. What you don't get is that once you've invested, you have saved. Opposite is not true.

Anonymous said...

Yes, Kristjan, investment is part of saving and so once you have invested you have saved. But saving always exceeds investment, and the level of investment doesn't determine what the level of non-investment saving will be.

Kristjan said...

Saving equals investment (real)

Tom Hickey said...

While banks' aggregate demand for reserves is indeed determined by the aggregate level of demand deposits, the level of demand deposits is determined by their volume of lending, and the volume of lending is sensitive to the cost of funds. How sensitive? That's an empirical question that can't be answered on theoretical grounds. But we know the answer isn't "not sensitive at all."

Right. Some assume that the sensitivity is for all practical purposes close enough to uniform to be disregard in developing a rationale for money policy. Keynes showed how and why that is wring and why monetary policy works toward the middle of the curve but not in the same way at the tails.

Tom Hickey said...

It's not quite true that saving equals investment (function, causal relationship). Rather, saving is identical with investment (accounting identity, tautology).

It's not the case the 100% consumption in any period prevents investment in the next period when funding and real resources are available.

Tom Hickey said...

Saving (amount saved) is always identical investment (amount invested) in a close system because accounting. In a system with more than one sector, there can only be net saving in one sector if its matched with net borrowing in other sectors. There can be net saving in national accounts because open system which includes an external sector. There cannot be net saving that is not identical with net investment in the global economy, which is closed system. Right now, the global economy is more significant than national, regional and local economies because interdependence. No subsystem can buck the system for long. That's what floating rates and liberal economics are based on.

Kristjan said...

"It's not the case the 100% consumption in any period prevents investment in the next period when funding and real resources are available."

When we are talking about gdp then we are always talking about one period, investment, consumption etc in that period. Be it one year, one quarter or one day.


"It's not quite true that saving equals investment (function, causal relationship). Rather, saving is identical with investment (accounting identity, tautology)."


It is certainly true. No matter how you slice it, it is true.

Tom Hickey said...



There a difference between an equation and an identity. The symbol for an equation is to horizontal bars and the symbol for an identity is three horizontal bars. Accounting identities like saving is identical to investment is an identity, not an equation.

Keeping this straight in economics is important and many mistakes are made when it is not observed.

For example, "y = f(x)" says that varying the independent variable "x" produces an invariant change in the dependent variable "y" as a result of applying the operation denoted by "f". "x" is the input and "y" is the output.

Many make the mistake of thinking that "saving equals investment" means that saving is an input and investment is the output of a function based on the operation of using saving to fund investment. That is not true based on the accounting.

The reality is in a one sector closed system of households and firms, the accounting residual after subtracting consumption from income on both side must be the same for both sides. It could be given the same symbol to make this less confusing but instead different symbols are used that are equivalent. In accounting, saving doesn't create investment and investment doesn't create saving. These are two totals than must be equivalent if the accounting is correct and all books balance.

However, it is also possible to interpret accounting identities causally with respect to how the entries are created. Then the line between accounting identities and theory is crossed. That line is often blurred.

Observing what happens in the relation of banking, finance, and economics, firm spending generates the income for household consumption. The amount spent on consumptions must equal the amount received according to the rules of accounting. Saving being the residual from income not spent. investment is the residual of income that firms received less the amount received from sale consumption goods.

Obviously, the two residuals after consumption are the same. That is, saving is identical with accounting.

But then the question arises, where the investment came from. Some may have come from cumulative saving, which is called "savings". But some comes from creation of new funding from borrowing. The point of the analysis of endogenous money creation is that lending does not presuppose prior saving, i.e., banks create deposits by lending, that is, they do not intermediate among savers and borrowers as many still mistakenly think.

Then one can see from actual operations that firm spending results in total income, where income is equivalent to expenditure in the income-expenditure accounting model. Total income over a period is spread over consumption and saving as a residual. So the direction runs from firm spending through consumption to household saving as deferred consumption, just as borrowing is income pulled forward. What firms received from income not related to the sale of consumption goods is classified as investment.

So the RHS, C+S is equivalent to the LHS, C+I in a one sector closed model. These same relationships hold in multi-sector models over all the sectors as a closed system. The global economy is closed system in which saving and investment are identical by accounting identity as long as the same accounting rules are applied across the system.

Tom Hickey said...

Oops, LHS and RHS should be reversed in "So the RHS, C+S is equivalent to the LHS, C+I in a one sector closed model. These same relationships hold in multi-sector models over all the sectors as a closed system. The global economy is closed system in which saving and investment are identical by accounting identity as long as the same accounting rules are applied across the system." The household sector is conventional written on the left and the firm sector on the right.

This is a matter of convention that should not be confused with putting the dependent variable "y" on the left and the independent variable "x" on the right in the function "y-f(x).

As an operational fact, saving, which appears on the left, is actually an outcome of firm spending that appears on the right. But that's just a coincidence of convention.

MRW said...

@Neil Wilson,

"They get the government bonds they sell by 'borrowing' them from the government...

(No seriously)."


You mean it's a book entry? Because OMO (thanks, btw) are done through primary dealers, aren't they?

Roger Erickson said...

Sorry Tom,
We must have been writing simultaneously.

Banks are not intermediaries of loanable funds? - Well Duh! Didn't John Law Say This ~1715, And Beardsley Ruml In 1946? We Can PRACTICE Doing Better.
http://mikenormaneconomics.blogspot.com/2015/05/well-duh-didnt-john-law-say-this-1715.html

Tom Hickey said...

Just explaining the difference between equation and identity, and how it applies to saving and investment, where confusion still abounds.

A lot of confusion results from not understanding the underlying logic. This is what philosophical logic and analytic philosophy is about. I wrote my dissertation on the logic of justification in ordinary language by looking at how "certainty" is used in different contexts.

Anonymous said...

Saving only equals investment in economic models and accounting frameworks in which "investment" is defined as including inventory accumulation. That is, if even accumulations of unconsumed consumables count as "investment", then all saving is investment.

This is one of those big brazen double-thinking lies macroeconomists tell routinely, and have incorporated into their basic definitional rhetoric to sell ideology as a set of unchallengeable definitional truisms.

Imagine a society consisting entirely of self-supporting farmers. Every year they eat 90% of what they grow, and they can the rest in Bell jars to save for an emergency. Their farming methods, area under cultivation and total annual output never changes from year to year. So clearly, the annual savings rate is 10%. But according to macroeconomists and national income accountants, the rate of investment is also 10%, even though there is zero capital formation in the society!

Why do macroeconomists tell this lie? Because it helps in propounding an old myth that it is their business to hawk: that the rich are necessary to society because only the rich save and all of the the savings of the rich make the economy grow. This myth faces the embarrassing common-sense knowledge that not all of the output that is not consumed consists of capital equipment or other factor inputs that are invested in the creation of more capital.

No problem, say the economists! We'll just redefine "investment" so that includes absolutely everything that was produced but not consumed, no matter how that stuff is actually used.

The Arthurian said...

Dan Kervick: "Saving only equals investment in economic models and accounting frameworks in which "investment" is defined as including inventory accumulation."

Yes. And it's an important point.

"Why do macroeconomists tell this lie?"

Not sure, Dan, but I think it originates in the accounting.

Anonymous said...

"Not sure, Dan, but I think it originates in the accounting."

I would say instead that the accounting originates in the lie. People who devise accounting frameworks have choices about how to define the fundamental terms they use. They can define them in accordance with previously well-established conventional usage so as to minimize misleading associations, or they can concoct crazy BS definitions that torture established terms into new shapes so as to uphold lies and bad theory by definitional legerdemain.

NeilW said...

This is the problem with the terms 'investment' and 'savings'.

They are massively overloaded.

This comment thread started with *financial* investment and savings, and then ended up incorporating real investment and savings, including inventory.

It's important in the analysis to separate out the financial circuit from the real circuit. Conflating the two causes havoc (deliberately so as Dan rightly points out).

Because real investment is in the eye of the beholder. One person's investment is another person's waste of money. And one person's consumption is another person's investment (training courses for example!).



Nick Edmonds said...

Tom Hickey - "But then the question arises, where the investment came from. Some may have come from cumulative saving, which is called "savings". But some comes from creation of new funding from borrowing. The point of the analysis of endogenous money creation is that lending does not presuppose prior saving, i.e., banks create deposits by lending, that is, they do not intermediate among savers and borrowers as many still mistakenly think."

This is where it gets tricky. What does it mean to ask whether the investment came from saving or something else? And the saving can never be prior; it must be simultaneous. I don't think that sort of question is very helpful.

Instead, we should be asking things like what happens if there is a sudden increase in the propensity to save? Does this lead to an increase in the actual volume of saving? In a crude loanable funds model it will. In a realistic monetary model, the answer is more complicated. The volume of savings may even fall. In any event, we probably want to distinguish between the immediate impact and steady state equilibriums.

In relation to this Jakab Kumhof paper, the issue is whether our answer here depends on how banks are structured. If banks operated as passive intermediaries, would that mean the crude loanable funds model would then apply?

Jose Guilherme said...

If banks operated as passive intermediaries, would that mean the crude loanable funds model would then apply?

As the authors explain on page 10, the loanable funds model of banks as intermediaries would be valid only for a barter economy where banks are modeled as institutions of barter. As they rightly conclude, "such institutions simply do not exist".

Nick Edmonds said...

Jose Guilherme

I agree with that. That is why I say that the issue is really about the difference between monetary economies and barter economies and nothing to do with the operation of banks.

Tom Hickey said...

The story goes that banking is founded on the idea of gold certificates, where the goldsmiths become bankers held gold on deposit in their vaults and issued transferable certificates for amounts. So initially, gold was being bartered using the medium of a promissory note that was backed by physical gold 100%.

This is the story told in Money Masters, for instance which is widely viewed on Youtube.

According to the narrative, the proto-banks realized that there was enough less demand for conversion to allow them to issue more promissory notes than there was backing available. Eventually, the goldsmiths dropped the smithing and just focused on taking gold deposits and issuing promissory notes payable to the bearer. That was the beginning of bank credit money, that is, the bank issuing its own unbacked liabilities out of the thin air because it could.

That led subsequently to bank runs, which led to panics and that eventually led to modifications in banking until we arrive at the present system.

Now bank reserves issued by the central banks are supposedly the replacement for gold in a fiat system. Since there is no metal backing reserves, there is no (real) saving backing them.

IN this view, real saving in a monetary economy that is sound is gold and silver that are minded and saved and certificates issued against metal reserves. A completely sound system is 100% backed by metal, and this is "full reserve."

So paper money is only a veil for the exchange of metals for goods and services. It's essentially a barter system and banks are intermediaries that cannot create bank money be issuing bank credit that is not 100% backed.

So under this view, issuance of central bank money that is not metal backed at the top of the pyramid is just a scheme to shore up an unsound system.

Regardless of the degree to which this true historically, a lot of people believe it and it's what a lot of the debate is founded upon. (Notice in bitcoin, "coins" are "discovered" by "mining." In this sense Bitcoin is a "sound" replacement for fractional reserve banking based on unbacked central bank money. No matter that Bitcoin is not backed either. The amount discoverable is finite, whereas the amount of fractional reserve bank credit "backed by" central bank reserves is infinite. So no inflation with bitcoin, whereas fractional reserve is fraught with impending inflation or at least biased signaling in markets that upset the natural operation of market forces.)

In this view fractional reserve bank created credit money is backed by central bank created money, so the system is exogenous and limited to the amount of reserves that the central bank creates. Ergo, the amount of loanable funds is based on reserve balances.

I think that this is the predominant view but that is only anecdotal.On the other hand, most people also seem to think that banks lend out the funds that customers deposit and the government guarantees deposits through deposit insurance. They believe that government can do this because money comes ultimately from the government.

But this don't have a clear idea of how government gets the money. Most think that government either taxes or borrows in order to get it. The problem here, of course, is infinite regress in money creation because they are not clear on the starting point.

So a satisfying explanation has to cover these points showing which are correct and way, and which are incorrect and why. An accessible book that is a commentary on Money Masters could be good seller in this environment. Of course, its purpose would really be as a teaching guide to impart a correct view.

Tom Hickey said...

@ Nick



Here i had the common belief in mind that entrepreneurs put up some of their prior saving and also borrow from banks or other creditors, often family. So it is true that some savings enter the picture but it is not true in the case of endogenous money generated by banks from keystrokes.

I don't think that most people get this. They think that money borrowed from banks is depositors' savings that the banks lends out and charges a spread on in order to make a profit.

Tom Hickey said...

Oops. The reference didn't print.

But then the question arises, where the investment came from. Some may have come from cumulative saving, which is called "savings". But some comes from creation of new funding from borrowing. The point of the analysis of endogenous money creation is that lending does not presuppose prior saving, i.e., banks create deposits by lending, that is, they do not intermediate among savers and borrowers as many still mistakenly think.