Monday, 1 June 2015

Jakab and Kumhof on Banks and Loanable Funds

Zoltan Jakab and Michael Kumhof  (JK) have produced a working paper for the Bank of England on whether banks should be viewed as money creators or simple intermediaries.  One of their main claims is that the mainstream modelling of banks is flawed because it views banks as the latter, when in reality they are the former.

I don't find their reasoning very convincing.  They produce some detailed models in which they compare results where banks are intermediaries of loanable funds (ILF) with those where there is finance through money creation (FMC).  By introducing various frictions into the ILF versions, they dampen the impact of shocks so that the FMC versions display greater volatility.

JK's argument is that most mainstream models with banks implicit assume that they those banks operate under the ILF model.  Presumably this means that those models are somehow reflecting an unrealistic dampening of shocks due to implicit frictions.  I find this line of argument odd.  On the whole, mainstream models are constructed to exclude all frictions other than those under consideration.  It is not clear to me what frictions JK would remove in those models to make them more realistic.

Part of the problem is that when JK compare the results of their different models, they are not comparing like with like.  There are some important structural differences between their ILF models and their FMC models to do with which agents do what, and it is this that is giving their results rather than anything to do with bank operation.

Take for example their ILF Model 1 and FMC Model 1.  In ILF Model 1, they have two types of household - borrowers and lenders.  Borrowers have real capital assets which they use as collateral to get loans from banks.  Lenders just hold bank deposits.  Lenders are then assumed to face a transaction cost friction which depends on their holding of deposits.  The greater the level of deposits the less this friction.

In FMC Model 1, these two types of household are folded into one.  The representative household has loans, deposits and real capital.  It still faces the same transaction costs, but is now in a position to mitigate this.  It has no need to borrow to fund a holding of capital assets, but it can use the collateral to borrow to raise its holding of deposits and reduce the transaction cost friction.  In the ILF Model 1, this cannot happen because lenders need the deposits but do not hold the collateral.

Now this may reflect a genuine friction that arises in the real world, but it seems to me to be all about heterogeneity and distribution and nothing to do with the operation of banks.  Their entire result here depends on there being two classes of agent in one model, but only one in the other.  JK seemed to have compared two different structural set-ups and concluded rather arbitrarily that the different results are all to do with loanable funds.

Even from the start, when they give a simple overview of what they see as the different models, they make this mistake of not comparing like with like (pp 11, 12 and figures 2 and 3).  To make a proper comparison, we need to use the same assumptions about who is involved and what they are trying to achieve and then examine how the results might differ if we restrict what steps can be involved in getting to the result.  

It's helpful to illustrate this by rejigging their examples slightly.  So, we start with three non-bank entities (which I'll call A, B and C) and the Bank.

- A holds gravel (as per their example), which it does not wish to consume currently - it wishes to hold a bank deposit instead.

- B wants to invest in machinery, but has no current resources so needs to borrow.

- C has machinery to sell and would like to acquire gravel

- Each of A, B and C will take credit risk on the Bank, but only the Bank is willing to take credit risk on any of A,B or C.

We then have two different ways in which these objectives can be reconciled, the ILF route or the FMC route.  These are illustrated in the diagrams below.

In the ILF model deposits and loans are contracted and settled in real goods rather than monetary payments (or ledger entries).  (Odd as this sounds, it is a common assumption in mainstream models of banking.  However, it is usually possible to reconstruct these into monetary models that are structurally equivalent - see here for example.)  So, here, the steps are:

ILF1  -  A deposits gravel with Bank
ILF2  -  Bank lends gravel to B
ILF3  -  B trades gravel with C in exchange for machinery

This scenario is the same as the ILF one that JK describe in the paper.

In the FMC scenario, we have monetary deposits so the steps are:

FMC1  - Bank lends to B (creating a deposit for B in doing so).
FMC2  - B buys machinery from C (by "transferring" the deposit).
FMC3  - C buys gravel from A (by "transferring" the deposit)

This is not equivalent to the FMC scenario in JK's paper, which lacks the final step here.

As I have set them out, we can see that in both scenarios we get the same end result.  The issue then becomes whether the process allows these steps to take place as set out or not.  In JK's alternative scenarios, they end up with completely different outcomes, so they don't even get near this issue.

I think there are some legitimate points to be made about how the operation of banks might impact on transaction frictions and what this means for volatility and response to shocks.  Unfortunately, I do not think this paper has very much of interest to say on the topic.


  1. The last step, step 3 , in both models is a red herring , its really an addition put onto the models so that A arrives as the lender at the end of both models. It is stage 2 which truly represents the economy with Banks, and its at this stage the difference in the models occurs. In model ILF, A has knowingly and voluntarily lent his gravel to B. In the FMC model C has sold his machinery in return for what he considers to be a liquid asset, which is infact fact debt from B. Which requires future production from B for C to realise his wealth.
    And so the difference is that at that stage, in ILF, A is the lender.
    In FMC , at that stage, C is the lender.
    A proceeds cautiously knowing that they hold a debt that relies on A being able to honour it for A to realise his wealth. In FMC model, C proceeds unaware of this. In ILF the deposit represents the gravel. In FMC the deposit represents future production that has not been realised yet. In ILF model, A is a concious patient saver, who has saved past production. In FMC model, C is not a saver, they are only a deposit holder and possibly an impatient spender. So that with the observation that the first model requires an agent who deliberate sets aside the use of a real commodity and the other does not , the status of money does make a difference.

    1. I put in all the steps to show what is necessary to fulfil the same set of objectives for the same set of people. Alternatively, we can tell two different stories which achieve different sets of objectives for different sets of people. If (as JK do) you assume different sets of people and different objectives, you will get different results, regardless of what you have to say about banks.

      You say stage 3 is a red herring, but the point is this. If stage 3 does happen you end up with the same result. If stage 3 doesn't happen, you have to explain why. You can get rid of stage 3 (in the FMC model), by changing your assumptions in that model so that (say) A does not want to hold a deposit instead of gravel. This is effectively what JK do. But then, the reason you are getting a different result is only because you have changed the assumptions.

      What we need to be asking is about the sort of questions you suggest in your last paragraph. Will these agents be more cautious about taking each of these individual steps if they have to be taken in a particular order? Unfortunately JK do not address this and I don't think DSGE is the right tool for doing so anyway.

    2. This comment has been removed by the author.

    3. You can simplify it to two steps

      ILF step I A lends gravel to bank in return for a deposit
      ILF step 2 Bank lends gravel to B

      FMC step I Bank lends to B by creating a deposit for B
      FMC step 2 B buys gravel from A by "transfering" depsit to A

      The same set of objectives are satified , as you require. You have A with the same objectives , the deposit, but A is a different agent in the economy.. In ILF A is a saver in FMC he is not a saver , he is a seller.
      In ILF A is a saver and B is a buyer.
      In FMC you have twice the demand as A can be a buyer and B is a buyer. FMC fascilitates commerce , creates new buying power and demand. Big difference. In FMC the funds are not restricted to the amount of gravel and machinary or savers.

    4. "So that with the observation that the first model requires an agent who deliberate sets aside the use of a real commodity and the other does not , the status of money does make a difference."

      I can see where you are coming from but certainly in both models the same amount of real savings have to happen. I think it might happen either voluntarily or due to inflation.

      The models might be similar given Nick's exposition but isn't FMC a richer one - B might consume without A's permission which is not possible in ILF model? If one's assumption is that B's net worth (viewed by the bank) allows more debt and consumption it will happen regardless of A's (saving) preferences.

    5. Quite true. ILF requires a saver before a borrower can consume. As you say in FMC more debt and consumption occurs regardless of A's saving preference. Likewise Deposit holders spending and consumption occurs regardless of borrowers repayments and so there is no increase in deposit holders spending and consumption when borrowers are reducing spending and repaying credit. And so FMC applies when credit is growing and when credit is shrinking.

    6. Dinero,

      I agree that you could do a simplified version like that. It would be a good alternative to what I've done. I did it with three parties to follow what JK did in their ILF example, but yours works as well. You can then ask important questions like the ones you raise. JK however did not do this - instead they compared two models which were not equivalent and which produced different results only because they were not equivalent.

  2. I think this is one of the most interesting topic on macroeconomics, and too rarely discussed (thanks).

    Any comment on Keen's work (, would your post applies to his modelling as well? Have anyone else tried to do the comparison between ILF / FMC, esp. the way you are suggesting?

    Can you comment on velocity of money - it clearly drops in recession. Does this suggest that banking might be an integral part of the economy and should be explicitly modelled? It seems _obvious_ to me that the velocity and boom is feed by banking and its shadows? Is there a counterargument I cannot see?

    1. Yes - I find this very interesting as well. You could see it as a question of how to reconcile two things. 1. That the amount of funding for investment is constrained to always be equal to the amount of savings (this is just an accounting identity). 2. That decisions to invest are not constrained by decisions to save (except in simple barter economies, where the only way to save is by holding real goods).

      On Steve Keen, I applaud his focus on the importance of debt, but I think he is confused on the mechanics.

      I'm not very interested in the velocity of money. I don't like the concept, because I don't see money as something that circulates anyway. I also don't see one measure of money as being particularly important and prefer to consider general liquidity, which includes undrawn facilities.

    2. I find his work interesting, even if not the easiest to follow. I have read and agreed with critical comments on his unconventional accounting and mixing up flows and stocks etc. Yet I think Keen has changed at least some of it.

      I lately watched Minsky videos second time and Minsky modelling itself seemed to be coherent and logical - written code is pretty formalized anyway. I think with Minsky one can play with the reconciliation assumptions. Reconciliation frontier is the most interesting and I think it is very much about monetary implications (which ILF misses).

  3. Welcome back to blog world after a hiatus.

    I am only on page 13 of this paper. So far, it seems to describe the concept that I hold: banking must be viewed from two simultaneous perspectives. The two perspectives, ILF and FMC, seem to be adequate descriptions, at least to page 13.

    As I read this paper, I watch to see if it conforms to my view that money is always "property" owned by someone. FMC conforms to this view by assuming that banks can instantly create new property (that is 'financial property'). Instant creation of property is banned by microeconomics (it is counterfeiting) but is observable in the macro-economy (debt held by central banks increases year-on-year).

    I would abandon the "money is property" view if I found convincing evidence that it was incorrect.

    As I continue to read this paper, I will keep your comments in mind. Thanks for the post.

  4. I completed a reading of the JK paper this afternoon. I must not pretend that I could now translate their equations into spreadsheet or code; I did not gain that kind of grasp of their work.

    However, in general, their work and results correspond with my much less mathematical framework of bank operations.

    Going to your posted comments, and more particularly to your FMC diagram where gravel ends up in the hands of actor A, I agree that JK do not include that final step of transferring gravel (synonym for money) in their FMC model. I think the reason for omission is that the initial deposit is never needed for the ongoing banking FMC based system to work. But this brings it's own problem:

    If an initial deposit is not needed for a bank to make a loan, we could have any actor set up their own bank and borrow exclusively from it. For example, General Motors could set up a GM bank and self finance. All with no initial deposit. JK acknowledges that restraints exist other than initial deposits and I would certainly agree. It is the other restraints that prevent GM from self-financing.

    I think Zimbabwe was trying to self-finance when the country went into hyperinflation.

    On the matter of model frictions, you point out that the models generally remove undesired frictions. The ILF model has time as a friction. Time is required to build the deposits before lending. The FMC model removes this time friction by way of changing spending/savings constraints. The exact mathematical method of doing this is one of the features I do not yet understand.

    The frictions found in ILF models could be injected into FMC models if needed to accommodate government borrowing from the private sector.

    I will close by saying that Central Banks have no money obtained as the result of labor or sale of resources. They have a lot of money obtained by stroke-of-the-key money creation. I think JK are trying to model the effects of stroke-of-the-key money creation.

    1. You did well to stick with that paper to the end - it's not an easy read!

      In my opinion, much of the stuff in their model is irrelevant. You could demonstrate the same results with a much simpler model. This sort of model should be about making it easier to see why things happen. To make things easier to see, the models need to be as simple as possible - all the complicating factors need to be abstracted away. Putting extra stuff in just obscures what is going on. And I can't believe they think it makes the model more realistic.

      "Time is required to build the deposits before lending. The FMC model removes this time friction by way of changing spending/savings constraints." This is correct. In the FMC model, the households can build up their deposits by borrowing. So when they need more deposits they just borrow and the level of deposits (and loans) jumps to whatever is needed. It's easy, because in this model the borrower and depositer are the same person. In the ILF model, they are different people. The people needing deposits cannot just borrow, because JK only allow people holding real collateral to borrow. So they have to build up the deposits over time.

    2. I haven't read the paper, I very much like the way Nick summarized/simplified it in the post.

      Ï'm not sure I follow but isn't this quite important passage which describes what the ILF implies. I mean in ILF the bank doesn't consider the credit/project worthiness but the collateral/capital. I think there is a difference. In order to get collateral one needs to have preceding asset to fully cover the loan. The collateral can be gained only by having assets slowly appreciating in value due to economic growth. The implication is like in the passage: the collateral needs to be built (slowly) over time. Deposits cannot be created in addition to this slowly building collateral.

      That's of course not what happens in reality. Borrowers mostly loans against the cash flows expected from a project. Collateral/equity is needed to cover only a fraction, a haircut, of the total amount borrowed. When the economy booms profitability is high, haircuts are low and thus it allows rapid loan growth and deposit creation. This is confirmed by the data.

    3. I skimmed the paper and it's very good. This may be the much delayed break-thru of endogenous money concepts in mainstream economics. Credibility and collateral of borrowers are critical to practical banking and the actual limits are there. This study, however, clearly tells that lack of money is not a reason for a bank to reject loan proposals. To get a comprehensive and realistic picture it's nesessary to use double entry accounting. My own text in employs the same. It is somewhat simpler and a bit more concrete.

    4. I don't agree that this is a good paper. I don't think it demonstrates at all what it purports to. The authors may have a good grasp of the mechanics of DSGE models, but I don't think they have a good grasp of the mechanics of money.