Wednesday, October 02, 2013

A Bank is still not a financial intermediary: Part 2

Lots of wonderful comments on Part 1 of my post on this topic. I want to address them and highlight some that I thought were particularly interesting and though provoking, but before that let me finish what I set out to suggest.

In part 1, I spoke about:
- How the word "intermediary" is vague, and I'm using it in a specific sense -- namely a market maker that matches buyers and sellers and therefore can be safely abstracted away when modeling the system.
- How this understanding of the word "intermediary" is a problem in the specific economic context we find ourselves in today -- 5+ years of an extended recession, which high unemployment, and a host of conventional and less conventional monetary policy responses which have not worked well and which also, frankly, seem weak on theoretical grounds.

So, Tobin, who knows that banks generate loans which in turn create desposits, does on to say that once banks have those liabilities on their balance sheet, they undertake additional steps to reconstitute that balance sheet in ways which make sense for their business and also reflect the portfolio preferences of their non-bank counterparts. For example, they may actively manage the duration of their liabilities to better match their assets. They make take an active role in managing reserve levels to avoid having to use the OIB market. And of course, they stand ready as a counter-party to the non-bank sector -- if a corporation wants to issue paper and collect cash, a bank will assist in that, moving the cash from the paper purchasers accounts to the bank's account etc.

It is this counter-party role which in turn reduces a bank's function to being a safely ignorable match-making style intermediary, and means that the overall composition of bank assets and liabilities reflects the non-bank sector's preferences. To close the argument: focus on buyers and sellers in the market and safely abstract the banks away.

I actually agree with all of the above, but I also think it begs the question and that is: what non-bank sector portfolio desire can banks not be a counter-party to? And the answer to that is: money! I'm being glib, but when most people think of money they think of what's in their savings account without a corresponding non-equity liability (or rather, what's in their savings account that they are not under any obligation to pay back). If you want to move your money from checking to savings, or savings to CDs, or CDs to stocks, the banking system stands ready to move the numbers from one cell in the spreadsheet to another. If you want to take out a loan, which will increase your bank account, and then someone else's bank account when you buy that car or house, the bank will do that as well.

But suppose you want to simply have more money? And suppose you want to have more money as a sector? Well, sector-accounting, or paradox of Thrift if you prefer, means that you cannot get it and this is a portfolio desire, the desire for a bigger portfolio, that banking cannot help with. This is why increasing "reserves" is not "printing money" in the sense that people think and it is this understanding of what banks can and cannot do which is missing in mainstream macro.

In some comment somewhere, I think it was JKH but it may have been someone else, said: "Tobin is not to blame for Paul Krugman's misunderstandings. Krugman is to blame for Krugman's misunderstandings".  I wish that were true but it is not. Krugman did not come up with his misunderstandings all by himself, he inherited it from the norms of his academic discipline, since mainstream macro is all wrong in exactly the same way. Paul points to Tobin as a reference, but seems to claim that Tobin says something different from what he actually said. So the problem is elsewhere, but it is still environmental in origin, not specific to Krugman. Nevertheless, you need to meet people half way, and if Krugman points to Tobin, but gets Tobin wrong, then we can begin there.

23 Comments:

Blogger Ramanan said...

But support you want to simply have more money? And suppose you want to have more money as a sector? Well, sector-accounting, or paradox of Thrift if you prefer, means that you cannot get it and this is a portfolio desire, the desire for a bigger portfolio, that banking cannot help with.

That is precisely wrong because banks stand ready to buy fixed income assets from the private sector which in the process creates deposits.

It's not just loans which create deposits.

And this dynamics happens via both quantity and price adjustments. Price of securities in the financial markets.

It is this mediating role in particular which Tobin is interested in.

He is not looking at banks as a veil.

http://www.jstor.org/stable/1992063

I guess Tobin perhaps didn't word his thoughts properly but I don't think he ever meant that the only function of banks is this mediating role.

11:35 AM  
Blogger Unknown said...

IMO opinion in some sense banks arent intermediairies and in a sense they are. For example they intermediate in reserves or asset markets which they dont create. But assets they create like deposits they originate instead of intermediating.

8:03 PM  
Blogger Neil Wilson said...

"That is precisely wrong because banks stand ready to buy fixed income assets from the private sector which in the process creates deposits. "

I see that as just loan creation by another name. The 'fixed income asset' will pay interest and have a redemption date.

Only the loan originator mechanism has changed.

There is no operational difference to loans create deposits, or frankly buying a tranche of loans from a different financial institution - particularly a shadow one.

The bank gets paid an income stream from some private entity and holds an asset which will get repaid in the bank's liabilities at some point.


11:51 PM  
Blogger Игры рынка said...

Neil, sure, the general rule stands that whatever banks buy from the non-bank sector, they create deposits. But it is also wrong to strech it too far and say that operational details are the same. The same to who? The same accross which dimension? Can there be other dimensions that are actually more important or not? And there are plenty of dimensions and details to think of than a simple fact that the balance sheet has increased and that means both assets and liabilities have increased.

1:22 AM  
Blogger Nick Edmonds said...

Neil,

"There is no operational difference to loans create deposits, or frankly buying a tranche of loans from a different financial institution - particularly a shadow one."

I'm not sure what you mean by operational difference, but are you suggesting that there is no difference in terms of macroeconomic effect?

Would that then imply that no matter how fast banks are pushing out new loans, if they're selling them off to the shadow banking sector at the same rate, then one cancels out the other and there is no overall impact on the economy?

2:23 AM  
Blogger Neil Wilson said...

"Would that then imply that no matter how fast banks are pushing out new loans, if they're selling them off to the shadow banking sector at the same rate, then one cancels out the other and there is no overall impact on the economy?"

That's not what I'm getting at.

What I'm saying is that there is no difference between a bank originating a loan (on the same terms as 'fixed income asset') to a third party, and that third party originating a 'fixed income asset' and selling it to the bank. The accounting result is identical between the two processes - for both the bank and the third party.

2:40 AM  
Blogger Neil Wilson said...

"And there are plenty of dimensions and details to think of "

I disagree. It's the same fundamental operation between the two - particularly when you throw in securitisation processes.

You can end up over-egging details that are irrelevant to the fundamental processes going on - particularly on the aggregate.

It's about getting to the nut of what is going on - which is what the original post is all about.

2:46 AM  
Blogger Ralph Musgrave said...

Winter says “But suppose you want to simply have more money? Well, sector-accounting…. means that you cannot get it and this is a portfolio desire… that banking cannot help with.” I don’t agree.

If I want more money, I just go along to my bank with some collateral and ask to have my account credited. That’s it. I have more money. Put another way, my collateral has been monetised.

2:49 AM  
Blogger Ramanan said...

"I see that as just loan creation by another name. The 'fixed income asset' will pay interest and have a redemption date. "

Well a bond is not a loan exactly because loans are as much traded as bonds.

A loan is made by the bank. A bond which is bought by the bank is in the secondary markets in my context. The borrower who issued the funds went to the markets initially.

This accommodative behaviour of the banking system is what is important to understand how the portfolio allocation decisions of the non-bank private sector.

6:28 AM  
Blogger Ramanan said...

"Well a bond is not a loan exactly because loans are as much traded as bonds."

Typo... meant loans are NOT as much traded as bonds in secondary markets.

7:38 AM  
Blogger winterspeak said...

Ramanan:

Good point, banks create deposits through means other than making loans. Buying fixed income assets is one way, as you say, but they will also purchase other assets as well.

However, I don't see how this matters. You've stressed how a bank can stand as a counterparty to the non-bank sector so that sector can adjust it's portfolio, but you haven't made it clear why that matters in the context at hand. So in this regards, I'm in agreement with Neil.

And banks cannot help at all, at a sectoral level, when the problem is the size of the portfolio, not the composition. Moreover, I don't think that time preference is a good way to model this behavior either. That may be worth a post on it's own : )

12:43 PM  
Blogger Ramanan said...

Winterspeak,

I guess you will see this if you write sectoral models. If you do not use some of the works of Tobin, you will be left with ad-hoc assumptions about how much money the private sector wishes to hold as part of its wealth.

Suppose the private sector wishes to hold less money. If you do not invoke some crucial points of Tobin's asset allocation theory, you will be forced to assume that if the non-bank sector wishes to hold less money, then consumption should rise and so on.

Loan issuance and bond issuance are different at origin. Loan issuance increases the stock of money but bond issuance doesn't.

But banks purchasing or selling bonds changes the money supply.

So while loans and bonds seem same at some level, there are differences both at the static and dynamic level.

Another difference is that households don't hold loans directly but can hold bonds of corporations.

So one crucial role of the banking system is the accommodation of the changes in portfolio preferences of the non-bank sector.

My point is that the Monetarist intuition is not so easy to dismiss and can be dismissed/done systematically by embracing Tobin's ideas.

More generally let us forget Tobin and Krugman. The banking system in the United States has done one of the biggest act of intermediation in finance with securitization. It is consistent with some of your definitions. I don't know why you have to say "bank is not an intermediary".

If someone is wearing a pink T-shirt and blue jeans would you say "the person is not wearing pink"?

As to your question on why it matters: Suppose the stock of money is $100 at the end of one period and banks make loans worth $2. However if households wish to hold $103 of their wealth in deposits, where does the extra $1 come from? Or what adjusts so that the desired holding and supply are equal?

A macroeconomics study using flow of funds is not just doing the records but to understand why something happens.

I don't think I have managed to articulate well what I am trying to say but will try some other time.

1:11 AM  
Blogger Nick Edmonds said...

Winterspeak,

I think you are saying that the important feature here is that banks' desire to provide money (through making new loans) is not dependent on the non-bank desire to hold money. So if people try to hold more money, they can't. Whereas, if they were simple intermediaries, then the non-bank desire to hold more money would automatically lead to banks lending more and providing more money to meet that demand.

Have I got that right?

1:51 AM  
Blogger winterspeak said...

Ramanan: Please feel free to come back and try again! In all sincerity, I'm listening and want to understand.

Unfortunately, while what you say makes sense, I don't see why it's relevant to the context at hand. If the private sector wants to hold less money, either a different sector needs to hold more, or the private sector pays down debts (depending on what you mean by "money"). I don't need Tobin for this, I can get there with t-tables. Similarly loan issuance and bond issuances are distinctions without a difference. And no one's claiming that banks are limited to extending credit to create deposits. But again, so what? Why is the fact (not in dispute) that banks manage their liabilities and stand ready to be a counterpart to some non-bank portfolio preferences, "crucial" to the topic at hand?

Nick: Economists ignore banks in their models because they see them essentially as match-makers between buyers and sellers, with interest rate being the market clearing price. They have more or less sophisticated reasons for this, but it is this matching function which is the "intermediation" which can be safely abstracted away in models.

However, this is not how banks operate and not a good way to model the economy. First, because the assumption "for every borrower there must be a lender therefore it's a wash in the aggregate" ignores the credit dynamic at the heart of banking. And second because it ignores the difference between gross and net financial assets. This stuff isn't that complicated, it's just considered unimportant and therefore the thinking about it is very unsophisticated. After hearing JKH and Ramanan stress how banks also actively manage their liabilities, I have some sympathy for the economists!

10:03 PM  
Blogger Nick Edmonds said...

Winterspeak,

I know that banks do not often appear in economic models, although I think that the consequences of this omission are sometimes overstated.

The important point, as I see it, is recognising that a decision to save is not the same as a decision to lend. Rather than simply matching savers and borrowers, we therefore have a three way process with savers, lenders and borrowers. The public may wish to save more by accumulating more bank deposits, but unless banks wish to lend more, it's not going to happen.

This essential dynamic can be picked up without actually mentioning banks (see for example my comment on your last post regarding the Krugman paper) although I'd agree that it is preferable to recognise the role of banks in the process, as it helps us relate our models to the real world.

In any event, what I am interested in is whether you think that the thing that makes banks more than mere passive intermediaries is exclusively a feature of banks, or whether it also applies to other financial institutions. It is true that some non-bank financials are fairly passive - they are merely acting as a conduit for their end investors. However, there are others that are no more simple match makers than banks are.

The question is, if we think that we can't ignore banks, can we ignore other financial institutions?

7:52 AM  
Blogger winterspeak said...

Nick: A bank's willingness to lend more has nothing to do with a public succeeding in increasing its bank deposits. You need to think of the public as a sectoral aggregate. Taking on more debt is different (in fact, opposite) from building its nest egg.

5:15 PM  
Blogger Nick Edmonds said...

Winterspeak,

In the simplest models, bank balance sheets consist only of loan assets and deposit liabilities. In this scenario, not only do loans create deposits, but deposits must be equal to loans at all times. The quantity of deposits in existence is therefore strictly determined by bank lending and the willingness of banks to lend has everything to do with the public succeeding in increasing their bank deposits.

But I think you know this. The more interesting question is what the model looks like when banks have more complex (and realistic) balance sheets and where they compete for funds with other financial institutions, who do not issue monetary liabilities. Is there a clear distinction between banks as active decision takers and other FIs as simple match-makers?

3:16 AM  
Blogger Neil Wilson said...

"and where they compete for funds with other financial institutions, who do not issue monetary liabilities."

Where does the 'other financial institution' deposit the money it has competed for?

Clue: not in its sock drawer.

4:26 AM  
Blogger Nick Edmonds said...

Neil,

I'm sure you don't need me to spell that out for you, so I'm guessing you're simply pointing out that a payment from a non-financial to a non-bank financial doesn't itself reduce overall claims on banks.

The important issue though is what you conclude from that. Some people think it means that the actions of other financial institutions have no macroeconomic consequences - only those of banks matter. What I interested in finding out is what Winterspeak's view on this is (although I'd be interested in your view too).

6:13 AM  
Blogger winterspeak said...

Nick: Look at your model from the non-bank sector's perspective. To that sector there is a big difference between an asset with an associated non-equity liability vs an asset with an associated equity liability. Banks can generate one sort of financial asset, but not the other.

7:57 AM  
Blogger Nick Edmonds said...

Winterspeak,

I'm not sure I entirely understand what you're saying here. I'll have a go at interpreting it, but I doubt I've got it right.

The non-equity liability of the non-bank sector is the bank loan, right? And the asset is the deposit? So when you talk about an equity liability, are you thinking of that as being the sort of asset (issued by the non-financial sector) that a non-bank financial institution would hold? If so, do you think that is the key distinction between banks and non-bank FIs, i.e. that banks hold loans and non-bank FIs hold equity?

9:35 AM  
Blogger winterspeak said...

Nick:

Yes, a non-equity liability held by the non-bank sector would be booked as a loan and the associated financial asset is the deposit somewhere else in that sector. The bank sector would hold the deposit as a liability, and have a receivable as the asset.

But the non-bank sector includes both households and governments, and it's worth splitting those out. My point is that the household sector desires financial assets which cannot be provided by the banking sector and can only be provided by the Govt.

9:55 AM  
Blogger Greg said...

"My point is that the household sector desires financial assets which cannot be provided by the banking sector and can only be provided by the Govt."

Exactly!

Additionally I think that ultimately this is what the banking sector desires as well. They dont just want their own "deposit created from loan" back, they want NFAs which the households hold.

5:03 AM  

Post a Comment

Subscribe to Post Comments [Atom]

Links to this post:

Create a Link

<< Home