Monday, April 02, 2012

Keen vs Krugman

In general, I'm not a fan of Steve Keen's work for two reasons:

1. He ignores the vertical component of money which, I think, is important and also necessary in understanding the monetary system.
2. I find his computer models a needlessly complicated way to do something that (to me) is simpler with basic double entry book keeping and t-tables.

However, in his back-and-forth with Paul Krugman, Keen is absolutely right. Paul says
If I decide to cut back on my spending and stash the funds in a bank, which lends them out to someone else, this doesn’t have to represent a net increase in demand.
This used to be my model of banking as well -- what does it matter if people save more by putting more money in the bank? That just gives banks more money to lend out, right? Wrong. Academic macroeconomics does not understand how banks work. Loans create deposits. Bank lending is not reserve constrained.

21 Comments:

Blogger Neil Wilson said...

"He ignores the vertical component of money which, I think, is important and also necessary in understanding the monetary system."

I can promise you he doesn't ignore it. He hasn't resolved it to his chosen model format as yet.

I'm still waiting for the evolution of the model for the 'very strange debt indeed'. Should be fun when it arrives.

"I find his computer models a needlessly complicated way to do something that (to me) is simpler with basic double entry book keeping and t-tables."

Running it on a computer to get the dynamic interactions is a good evolution. As the models get more sophisticated it will be near impossible to do them on paper.

However shortly after that point the flowchart modelling and Ordinary Differential Equation approach will similarly hit a cognitive limit.

Then we might be able to break out the Information Science toolbox and start modelling properly.

12:22 PM  
Blogger The Arthurian said...

"what does it matter if people save more by putting more money in the bank? That just gives banks more money to lend out, right? Wrong. Academic macroeconomics does not understand how banks work. Loans create deposits. Bank lending is not reserve constrained."

So, what do banks DO with deposits? Are you saying they refuse to lend them out? It seems to me a piece of the story is missing.

4:41 PM  
Blogger Dan Kervick said...

The Arthurian,

Deposits are liabilities, not assets. They are money owed, not money owned. So the banks don't do anything with them. Asking why a bank doesn't lend out its deposits is like asking why I don't lend out my mortgage balance.

If the customer's deposit balance comes about as a result of a walk-in deposit instead of a loan, then the depositor might give the bank a hunk of actual cash in exchange for the deposit balance. The cash goes into the vault and becomes part of the bank's total reserves. And that cash is an asset. But it is offset by the deposit balance it creates in exchange, which is a liability. And it just means that the next day as part of its liquidity management operations, the bank will have less need of acquiring additional reserve balances.

The point is that the bank doesn't need to get additional deposits to make additional loans.

5:09 PM  
Blogger Greg said...

Art

Scott Fullwiler explains it all here;

http://www.nakedcapitalism.com/2012/04/scott-fullwiler-krugmans-flashing-neon-sign.html

5:12 PM  
Blogger Greg said...

I think this is the best part of the whole post Fullwiler makes;

"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 borrowing

5:19 PM  
Blogger The Arthurian said...

Good quote, Greg. I want to shorten it and say the idea is that banks want to make loans that are offset by deposits.

I'm not sure what "offset by" means:

When I deposit my paycheck in the bank, the bank has a new liability (owing me that much money) AND a new asset (the money). I think "offset by" means the bank wants to use my money, by lending it out.

I think banks DO lend that money out, and I think it is incorrect to say, as Dan says, "Deposits are liabilities, not assets. They are money owed, not money owned. So the banks don't do anything with them."

??

11:29 PM  
Blogger Greg said...

Art

I think you have to look at it this way;

A banks balance sheet must balance. Liabilities and assets being equal. So if you never take in deposits you can still loan but it would be much more expensive to acquire another type of liability. Its simply less profitable to not acquire deposits.

I've come to take my own "Greg-ian" view of banks. They do not ever loan me anyone elses money. They are only loaning me my own future potential income stream and it is usually offset (collateralized) by some present "saving" vehicle. They simply turn potential future flows into a present stock.

2:47 AM  
Blogger The Arthurian said...

I don't do accounting, so tell me where I go wrong.

Because of the accounting, liabilities and assets are equal. Okay. So when I deposit my paycheck, that creates a liability for the bank AND an asset for the bank.

The liability is the money the bank owes me. The asset is the money I gave them to hold.

I think they must want to DO something with the asset. Other than kill Jason Bourne I mean.

From what Fullweiler said, the bank wants to use the asset to make additional loans, because that is cheaper than...

Oh, does Fulwiler mean the bank wants to make a loan and THEN go find deposits to "offset" it on the books, that being the cheapest way to perform the offset?

Okay, I don't care. But if the bank should happen to get the deposits BEFORE they make the loan, it seems they should still be able to use those deposits to "offset" the loan that comes later.

If a loan is created at one bank and (after the money is spent) the deposit shows up at another bank, then the first bank can borrow reserves from the second? Wow, that's convenient. But the first bank only needs 10% of the loan for reserves I think, so the second still has excess. And will want to lend it out.

Banks don't throw deposits away, I'm sure of that.

3:20 AM  
Blogger Neil Wilson said...

"The liability is the money the bank owes me. The asset is the money I gave them to hold."

No. The asset is the Central bank reserves purchased with the cash if you deposited cash, or it may be the loan the bank gave your employer to pay you if the deposit was electronic.

Or it may be the loan the person took out to buy the stuff from your employer who then paid you.

Or it may be... You get the picture.

3:55 AM  
Blogger circuit said...

WS, notwithstanding Krugman's faulty understanding of reverse causation, I tend to think his point is well taken as far as being a criticism of Keen's definition of aggregate demand, which Keen views as total spending plus change in debt.

4:36 AM  
Blogger Neil Wilson said...

" I tend to think his point is well taken as far as being a criticism of Keen's definition of aggregate demand, which Keen views as total spending plus change in debt."

You need to understand the assumptions you and he are making with regards to time periods. It's not really an accounting statement.

Keen's work is using continuous time and that definition is a simplification - starting that it is aims to be generally right rather than precisely wrong.

Bear in mind that accounting uses an accruals system. Transactions are moved from the time they actually happen to the period they are related to.

It works like a credit card. When you are issued with a credit card you are given more potential spending power, yet no spending has taken place. But knowing that you now have that spending power can alter behaviour so that more orders are made.

The accounting however handles the actual spending and the credit card transaction that paid for it (moved by accruals accounting to the correct period). And that is how you get your standard 'after the event' calculation of aggregate demand.

So the two are consistent - its just a different temporal viewpoint.

Steve probably needs to explain these concepts better.

4:55 AM  
Blogger winterspeak said...

circuit: yes, that was inaccurate for Keen to say.

But the thrust of Keen's argument is that deposits do not fund loans, in fact the opposite is true, loans create deposits.

Reserves are adjusted after the fact to handle settlement balances. Also, in non-OIR regimes, reserves are also drained (or topped up) to try and have the overnight interbank loan rate match the Fed's FFR target.

The notion that reserves somehow enable lending is a terrible confusion of how the "monetary base" works, bank lending works, and reserve works. It is also Krugman's position. And that was what Keen is (rightfully) attacking.

NEIL:

9:56 AM  
Blogger winterspeak said...

arthur:

When a bank makes a loan it wants to be sure that the loan will be paid back, and that it will not run afoul of regulatory capital requirements, which govern how much money a bank can loan out relative to how much capital it has on its balance sheet.

Note that neither reserves nor deposits come into the picture.

If a bank has a lot of deposits, then its cost of capital is cheaper. This helps the bank operate more profitably.

Banks may do things to try and attract deposits, like putting expensive branches in nice locations, or selling CDs at competitive interest rate. They are looking to lower their cost of capital.

10:05 AM  
Blogger Ramanan said...

Arthurian,

The thing is a new loan immediately makes new deposits.

It is true that the bank is attracting deposits and that it may have found itself with a lot of reserves before making the loan. However it doesn't "cause" the bank to make more loans because it doesn't make the borrower more creditworthy.

I partly agree with your attitude about deposits. Bankers do look for deposits.

It's not just profitability which drives banks to look for deposits. It's a bonus of course but deposits are a stable source of funds for banks. Non-deposit sources of funds can be tricky because a deterioration of market conditions (due to factors outside the bank's control) may lead to a lot of redemptions leading to *potential* liquidity troubles for the bank.

Prevention is better than cure.

1:58 PM  
Blogger The Arthurian said...

Neil: "The asset is the Central bank reserves purchased with the cash if you deposited cash, or it may be the loan the bank gave your employer to pay you if the deposit was electronic."

But my bank does not know whether my employer took out a loan to pay me. Cannot know.

Maybe this is wrong, but I thought that if I OWE money it is a liability and if I HAVE money it is an asset.

Winter: "Note that neither reserves nor deposits come into the picture."

Noted. But both exist in the real world, and are only absent from your picture. And this is exactly my complaint: the picture is incomplete!

Ramanan: "It is true... However it doesn't "cause" the bank to make more loans..."

Understood. But I think it saves the bank the trouble of having to scurry around looking for reserves, after the loans are issued. This must be part of the whole 'loans create reserves' thing, or else I really don't understand it at all.

4:11 PM  
Blogger Neil Wilson said...

"But my bank does not know whether my employer took out a loan to pay me. Cannot know."

Why would it need to know? The asset exists in the system and that created the same amount of liabilities as a matter of accounting

The payment system is nothing more than shuffling the ownership tags on those liabilities.

So you have central bank assets and bank created assets owned by the average commercial bank. That ensures there is the same amount of liabilities at that bank. Those liabilities have owners that change.

When the owner want to move those liabilities to a different bank they are first offset by liabilities coming in the opposite direction and the ownership tags are just switched.

If there is a net gain or loss then generally an amount of central bank assets are moved as well - but that's only because of the rules in place.

11:37 PM  
Blogger Unlearningecon said...

FWIW, Keen's accounting appears to be correct:

http://rwer.wordpress.com/2012/03/29/keen-krugman-and-national-accounting/

5:54 AM  
Blogger winterspeak said...

Arthur:

Nice explanation here at pragcap:
http://pragcap.com/banks-are-not-mystical

Say you want to buy a car. You take out a car loan. The bank loans the money, which appears in the car dealer's account -- at the bank!

So, you now have a car (asset) plus a loan (liability). The bank now has a receivable (asset) plus a deposit (liability). The car dealer now has less inventory (asset) and more cash (asset). He might have more retained earnings as well which get booked on the liability side.

If the car dealer and you are all at the same bank, reserve positions are unchanged. If you are at different banks, then reserves change to make sure all the checks clear. Reserves are primarily about settlements, and they are also used by the Fed to set the FFR (in normal times, we're in OIR right now).

The Fed must supply reserves as needed, or the payment settlement system will break, or the ONIB rate will miss the Fed's own FFR target. If the Fed wants payments to clear, or hit its own target, it must provide (or drain) reserves as needed.

12:56 PM  
OpenID Dhimmisoftheworldunite said...

re: The Arthurian's question of what banks do with customer deposits:

Think of customer deposits as a cheaper source of reserves than the overnight interbank market or the Central Bank's discount window. Demand deposits never (AFAIK) pay more than these. And there are costs to a bank associated with customer deposits. They need to be insured and the bank (if it has a retail business) needs to staff to deal with retail customers who want to access their retail accounts in person. Those costs are deducted, more or less, from the rate the bank pays for demand deposits, further reducing the rate a depositor earns compared with what the bank will pay for an overnight loan of reserves from another bank.

Banks need reserves to make payments and aren't allowed to persistently overdraw with the CB. So they "source" reserves from wherever is cheapest. "Walk in" customer deposits are one of those sources. But they are by no means the only source, which is why it is not true that banks need to receive deposits before they can lend.

7:28 AM  
OpenID Dhimmisoftheworldunite said...

And someone may object with the question, "where do the reserves come from that are lent between banks in the overnight interbank market"? They don't come from bank customer deposits either. The total level of reserves in the banking system is equal to the cumulative from inception deficit of the sovereign currency issuer less the sovereign debt issued (which drains reserves) plus the Central Bank's balance sheet (which is adjusted via open market operations to hit the CB's overnight interest rate target and which may also be inflated by extraordinary operations, such as QE or bank bailouts/supports -- and doubtless I'm overlooking things).

8:01 AM  
Blogger Michelle said...

While banks and their debt may be important in the greater scheme of things, it matters not what they do or how they lend because a disciplined Fed will always have control of the monetary base.

Bank Lending Criteria
http://businessloans.doobizz.com/bank-loans-2/2011/12/bank-lending-criteria/544/

3:34 AM  

Post a Comment

Subscribe to Post Comments [Atom]

Links to this post:

Create a Link

<< Home