Tuesday, October 30, 2012

Keen Disappointment

At Business Spectator, Steve Keen considers The myth of the money multiplier. He begins by reviewing "The standard story about how banks create money, and how reserves work". Of that story, Keen writes

to anyone who’s done empirical research, it’s a myth.

The most recent proof of this is in an excellent discussion paper from Federal Reserve economists Carpenter and Demiralp, entitled “Money, Reserves, and the Transmission of Monetary Policy: Does the Money Multiplier Exist?”. The first clue that it doesn’t exist is given by their abstract, which notes that, “before the financial crisis, reserve balances were roughly $20 billion”. If the textbook model were correct, the total stock of money in the USA would be $200 billion, versus the multi-trillion dollar level of even a narrow definition of the money stock. As the authors note, this makes a mockery of the textbook “Money Multiplier” model

The impression arising from Keen's numbers is that the standard story must be wrong. The numbers are way off, so the money multiplier story must be wrong. Okay, except for one thing: It is Keen's story that is wrong.

If the reserve requirement is "say 10 percent" as in Keen's example, the $20 billion reserve balance is 10% of $200 billion money, and everything works out fine. But the money supply is much, much more than $200 billion, and it doesn't work with the 10% number. So what is wrong?

The 10% is wrong. Keen uses 10% as an example in the setup. But in the denouement he lets us think that 10% is a valid number, and he takes advantage of our confusion to convince us that the money multiplier story is false.

It is Keen's 10% that is false. Only a very small part of U.S. money must abide by the 10% requirement. For most of that money, there is no reserve requirement at all.

If I have it right, for money in circulation the reserve requirement is 10%. And for money in savings the reserve requirement is zero. A reserve requirement of zero means there is no limit to how much bank money can be issued upon each dollar. With a reserve requirement of zero, the money multiplier could easily produce multi-trillion dollars of money. As, in fact, it has.

But I could be wrong. Keen lays out the requirement differently in his post. He says the reserve requirement is "10 per cent of household deposits only in the USA – there is no reserve requirement for corporate deposits". But again, if there is no reserve requirement on some of the money, then there is no limit to how much bank money can be issued upon each dollar of that money.

Either way, the result is that we end up with trillions of dollars of debt, trillions of dollars of money, and an infinitesimal, tiny fraction of the money in reserve.

The impression Keen gives with his numbers is that the story is wrong because the numbers don't work. The numbers do work. The reserve requirement is not universally applied, but the numbers work. The standard story is not wrong.

Steve Keen wants to convince his readers, even if he has to fudge the story to do it. I am frankly disappointed in him.

Oh, and I searched the PDF for the words "mockery" and "mock". The PDF does not contain those words. It is Keen who says "this makes a mockery of the textbook 'Money Multiplier' model". Not the Federal Reserve guys.


Clonal said...

Bank lending is constrained by two factors. First of course is the demand for loans. Second is the capital requirements (note: capital requirements are not the same as reserve requirements!)

For cpaital requirements see Bank Capital Requirements

The BIS rules set requirements on two categories of capital, Tier 1 capital and Total capital:

Tier 1 capital is the book value of its stock plus retained earnings. Tier 2 capital is loan-loss reserves plus subordinated debt.** Total capital is the sum of Tier 1 and Tier 2 capital.

Tier 1 capital must be at least 4% of total risk-weighted assets. Total capital must be at least 8% of total risk-weighted assets.

Joshua Wojnilower said...


Your point about Keen's poor logic regarding the 10% requirement is valid, but I don't think that validates the normal story. In fact, your comment about the lack of reserve requirements should be a signal the normal multiplier doesn't exist. If banks can lend without holding reserves against those loans, then why would an increase in reserves cause banks to increase lending by a given ratio?

From my perspective, Clonal presents a stronger case against the money multiplier and others have done a better job elsewhere too.

The Arthurian said...

If nothing but capital requirements limit bank lending,
And if under capitalism there are no limits to the accumulation of wealth,
Then over the long term there can be no limit to how much banks can lend. No limit other than the sort of financial collapse that undermines the economy for a generation or sometimes for a dark age.

I prefer reserve requirements because they CAN be used to impose limits on the expansion of finance. CAN be and SHOULD be.

I cannot imagine popular support for restricting the growth of capital, but I can imagine support for limiting the ratio of bank money to fiat money. Some people call for 100% reserve (which I think is excessive; but then I think zero is insufficient).

Woj, I think your statement
"In fact, your comment about the lack of reserve requirements should be a signal the normal multiplier doesn't exist."
grows out of the "Money Multiplier is Causal" view -- a view totally alien to me. I can see your point, but only if I adopt the causal view.

I don't really know anything about Bank Capital Requirements...
"Obviously" you say?? :)
Clonal, I will study that link. Thanks.

Greg said...


There is no limit to how much a bank can lend but there is a limit to what they can be paid back. The key is for banks to accurately forecast your ability/willingness to part with you future income flows.
Of course the value of the assets you purchase plays a role in ability to pay back as well

As I have started thinking about it, what banks are actually lending me is MY future income. How much of that future income I wish to spend today is a decision I have to make. In good times the bank will make it look easy to me to pay it back, in bad times it will make it look harder.

Im going through a refi right now. I am getting a loan that will drop my monthly payments by almost half (for a longer term) My personal balance sheet has never been better. This will be my only source of debt when this is finished and the total loan is less than half the value of my house. Ive had the same job for 25+ years and this loan payment will be about 6% of my take home pay. It is taking almost 6 months to close this loan!! This has nothing to do with whether the bank has this money and they are going to lend it at 2.25%!. They are simply looking at ME differently. They are checking out MY vault.

I got the original loan in about 3 weeks.

The Arthurian said...

Greg: There is no limit to how much a bank can lend but there is a limit to what they can be paid back. The key is for banks to accurately forecast your ability/willingness to part with you future income flows.

Sure. But I don't think it is such a great plan, for our whole financial system to be just as close to the edge as it can be. There is no reason for it.

People seem to think we need lots and lots of credit, because that is how the economy has been run for as long as anyone can remember. That doesn't make it right.

Credit money should be in proportion to fiat money, as economic growth is to the size of the economy. We want what, 5% growth? In a perfect world we could get by with 5% of total economic activity being on credit. Since ours is not a perfect world, double it: 10% of our transactions on credit. We need no more credit than this.

The rest of our transactions should be fiat-based. But fiat money has been restricted to fight inflation, and we have nothing to spend but credit money.

Hey, good luck with that loan Greg. Heh, 6%, you're right where I said we should be!!

Greg said...


I'm not making a statement about whether or not the way the credit system works is good or bad (You know I think its bad but this particular post is not a morality play), Im simply wanting to understand the dynamics involved.

" But I don't think it is such a great plan, for our whole financial system to be just as close to the edge as it can be."

Okay, agreed. But accepting my description puts the lie to the money multiplier because there is never a lack of money within the banking system to lend there are only a lack of customers they deem capable of paying it back.

"There is no reason for it"

Sure there is. The reason the banking system works as it does is that its purpose is to make a profit and it makes it profits by taking 150,000$ of our income over 10 years and only giving us 100,000$ and a house. We are betting that the house will be worth at least 150,000$ when we are finished. If its not we "lost" and the bank won. If it is we both win.

I totally support your arguments to get us off credit money but there is no way to make the case against the banking system and its deleterious affects if we use the fractional reserve lending models taught in textbooks. Those models ALWAYS make reserves (the portion the govt controls via the CB) the culprit in inflationary times. They see reserves as a cause rather than an affect.

Luke Smith said...

What relevance are capital requirements when a special purpose vehicle (SPV) can be used to shift liabilities off the bank's balance sheet? That is what Citi and other banks had been doing after Basel II, initially created by the 1986 Tax Reform act.

Even non-banks like GE use SPVs to fund the credit they gave to people to buy their toaster-ovens. Heck, even Federal Reserve bailout programs used SPV structures, which is why their balance sheet showed a net decrease in government securities (I call the "the divot") - they shifted the securities into programs like TALF and TAF. The ECB does the same thing, and through a wide range of rehypothecated transactions it is essentially trying to purchase the PIIGS's debt, collateralize it and then issuing bonds on the collateralized debt - which is supposed to reduce risk. Which is why a Greek default will cause a cash call from the other PIIGS.

The Arthurian said...

It's saturday morning & I finally sat down to read that "Bank Capital Requirements" link. It's wfhummel! That guy is great.

What is "loan-loss reserves"?

Thanks for the link.

nanute said...

Art. I can't speak for Clonal. Here's a relevant quote on what loan loss reserves are: A relatively large accrual for commercial banks, loan
loss provisions have a signifi cant eff ect on earnings
and regulatory capital. Because loan loss provisions
are at the discretion of bank managers, there is the
potential for banks to provision more or less than
necessary as a way to smooth their income. From an
accounting perspective, this could introduce discretionary
modifi cations to banks’ earnings and reduce
the comparability of results across firms. http://www.richmondfed.org/publications/research/economic_brief/2012/pdf/eb_12-03.pdf Note the fact that loan loss reserves are at the discretion of bank managers. (What could possibly go wrong?)

The Arthurian said...

"What could possibly go wrong?" :)

Hi Nute. As it happens, we closed on a refinancing this morning and among all the papers we had to initial and/or sign, on one page the words "loss reserve" caught my eye. In this case (for me) it would be an alternative to mortgage insurance. Money set aside for emergency, I suppose.

From your link: A strengthening of accounting priorities in the decade prior to the financial crisis was associated with a decrease in the level of loan loss reserves in the banking system.

Everywhere you look, things have been done to facilitate finance.

Thanks Nanute.