Thursday, 30 October 2014

The mistake made by Martin Wolf and bank regulators.

Summary: Their basic mistake is that they think that upping bank capital requirements involves costs. It doesn’t: as the Modigliani Miller theory explains, changing the way a bank is funded (e.g. changing the mix of capital versus depositors versus bond-holders, etc) has NO EFFECT ON the cost of funding a bank.

It is now widely recognised that that amount of bank capital was too low prior to the recent crisis, and after millions of hours of haggling those capital requirements are being raised a small amount.
Of course banks have resisted that change with a variety of totally dishonest and not even desperately clever arguments. I wouldn’t expect bankster / criminals and psychotics to do otherwise.
However, regulators and several household name economics commentators like Martin Wolf have been happy to concede that banks DO HAVE SOME SORT OF POINT. That is, they’ve gone along to some extent with the idea that increasing capital requirements comes at a price. Thus we supposedly cannot go wild and impose a 50% or 75+% capital requirement. For example Martin Wolf advocates 25%, which is way above the percentage advocated by Vickers, Dodd-Frank, etc.
However he says, “I accept that leverage of 33 to one, as now officially proposed, is frighteningly high. But I cannot see why the right answer should be no leverage at all. An intermediary that can never fail is surely also far too safe.”
Well the simple answer to that is that if it costs nothing to make something totally fail safe, why not do that? I.e. the CRUCIAL QUESTION is whether raised capital requirements cost anything.
Here’s why they don’t.

Deposits versus capital.
Take two hypothetical banks: one is funded ENTIRELY BY shares, and the other ENTIRELY BY depositors.  But in other respects the banks are the same: in particular, the risks stemming from the type of loans they grant are the same. It follows that the REWARD that both depositors and shareholders will want for covering that risk will be the same.
Ergo the cost of funding the two banks is the same!!
Ergo it makes no difference what the mix of capital and deposits is that funds a bank: the cost of covering relevant risks is the same!
Which is pretty much a re-statement of the Modigliani Miller theory.

Deposits, bonds and wholesale money markets.
Having rather suggested above that there are only two ways of funding a bank, capital and deposits, there are of course other ways: e.g. bonds and loans from wholesale money markets. However, the latter are essentially deposits of a sort. Indeed loans from wholesale money markets sometimes have to be repaid in one or two weeks, which makes them the same as retail term accounts where one or two week’s notice of withdrawal is required.
The word “deposit” will be used in a loose sense from now on: it covers conventional retail deposits and bonds and loans from the wholesale money market.

 The real world.
Of course in the real world, shareholders demand a higher return than depositors but that’s for the simple reason that depositors enjoy a cast iron guarantee provided to a greater or lesser extent by taxpayers that deposits are safe, whereas, while bank shareholders are subsidised TO SOME EXTENT by taxpayers, the “cast ironess” of the subsidy / guarantee is not the same. E.G. in the UK during the recent crisis bank shareholders took a hair-cut. Depositors did not.
In short, FOR A FAIR COMPARISON between the costs of funding a bank via shares and via deposits the comparison must be done on a strictly “no subsidy of any sort is available” assumption. And on that assumption, to repeat, there is no difference between the cost of funding a bank via capital as opposed to via deposits.

There is however an important difference between the two scenarios. The difference comes where it suddenly turns out that incompetent loans have been made (as occurred with many banks at the start of the recent crisis).
In the case of a bank funded just by deposits, ANY FALL in the value of the bank’s assets (i.e. the loans it has made) means the bank is technically insolvent. And if faith in the bank vanishes, a run starts, and the bank is ACTUALLY INSOLVENT.
Alternatively, if the bank is funded about 3% by capital and about 97% by deposits as was common prior to the recent crisis, then assets have to fall at least 3% before the bank becomes technically insolvent. But that’s frankly not much different the scenario just above where ANY FALL in asset prices means technical insolvency.
In contrast, where a bank is funded just by capital / shares, asset values can fall 50%, even 75% (which is practically unheard of) and the bank still isn't insolvent! All that happens is that the value of the shares fall to about 50% or 75% of initial value. In that sense, the bank cannot fail.
Now in what sense is Martin Wolf right to say such a bank is “too safe”? Exactly what is wrong with that fail safe characteristic? Absolutely nothing!
To rephrase that, where a bank is funded just by deposits and it makes disastrous loans, the bank collapses. In contrast, where it is funded just by shares / capital, it soldiers on.
The “fund by deposits” option doesn’t have a leg to stand on!

Full reserve.
And what do you know? A system in which lending entities are funded just by shares is what full reserve banking involves (at least that’s what Laurence Kotlikoff’s version of full reserve consists of). As to deposits, they are of course needed for day to day transactions, but under full reserve, that is done with accounts which are totally safe and involve no lending: those accounts are backed by central bank issued money, i.e. base money.

Modigliani Miller.
Of course MM has been criticised. But the criticisms are feeble. See this paper of mine, section 1.4 under the heading “Flawed Criticisms of Modigliani Miller.” Also Sir John Vickers devotes several paragraphs to MM in a paper published in 2012 (well after the Vickers commission final report) and suggests a few possible weaknesses in MM, but does not seriously question it.


P.S. (same day):  Another reason why bank capital may well cost more than deposits at the moment is that bank shareholders in the US have recently had to pay around $100bn in fines for crimes they didn’t commit: that’s Libor manipulation, laundering Mexican drug money, etc. (Yes that’s billion, not million) The actual perpetrators of those crimes (specific bank employees and executives) have got off Scott free. No doubt something similar applies in the UK and elsewhere.

If people with red cars have to pay the speeding and parking fines incurred by owners of blue and green cars, then the cost of running a red car will be higher than the cost of running a blue or green car. But that’s not a valid or fair comparison of the costs of running red, blue and green cars, is it? (Sorry about the change in font size there, if you see one: this blogging system goes mad sometimes.)


  1. Very sweet reasoning!!

    Is it also OK to have full reserve for instant/quick access deposit banks, with lending entities financed by both equities and bonds (not just equities)?

    1. Good question. My answer is “no”. However I’ve got some opposition. Under Milton Friedman’s version of full reserve he says lending entities can be funded by shares or “debentures” as he calls them (that’s bonds).

      And under Positive Money’s version of full reserve, lending entities or “investment accounts” as I think they call them, are funded by depositors to some extent. (I mention depositors because they have something in common with bonds: both represent a liability of a bank which is fixed in value (inflation apart) – as distinct from shares which have no set value.

      I don’t think deposits or bonds should be allowed. Reason is that the mere fact of allowing them makes it possible for a lending entity to go insolvent. In contrast, and entity funded by shares just can’t go insolvent (as George Selgin pointed out). Plus as regards depositors, if they want totally safe deposits, they can go for PM’s safe accounts, or the equivalent under Friedman or Kotlikoff’s full reserve systems.

      But I’ll be expanding on that in a paper I am publishing in a week or so, and which is a response to a discussion paper published by Sir John Vickers in 2012.

  2. Banking is one of the sectors where Miller-Modigliani is known not to apply. It doesn't really work anywhere because of the differential tax treatment of equity and debt.

    1. I dealt with that point in section 1.4 of the paper / book featured in the left hand column: “The Answer is Full Reserve….”. Basically my answer is that tax is an entirely artificial imposition, thus for the purposes of working out REAL costs and benefits, it should be ignored.

  3. The cost of debt is lower than the cost of equity to a bank. The reason for this is that depositors do not expect or receive the the full risk premium connected with their loans (either because they misunderstand the nature of a deposit or they think the government will bail out the bank). Equity investors, however, understand the true risk premium with investing in banks - and demand a higher return in order to induce their investment. They know that when banks go down, equity tends to end up with less of the cake than the debt (depositors).

    1. Yes, obviously where a bank is funded by shareholders and depositors, the former carry a larger risk. But as Modigliani and Miller rightly pointed out, if the shareholder to depositor ratio is changed, there is no change to the OVERALL risks involved in running a bank. Ergo raising capital requirements should not have any effect on the total charge made for carrying those risks.

    2. I think what I am saying is that government intervention in the banking market has caused debt funding to be relatively cheaper to banks than equity funding, such that MM does not apply.

    3. Yes, clearly if government offers protection to depositors, protection which is not offered to shareholders, that skews the market. It means that MM does not work in the real world, or at least it does not work as theory predicts. But if you “skew a market”, then APPARENT costs (i.e. prices) do not equate to REAL costs. And if we’re going to maximise GDP, we should consider REAL costs. I.e. we should ignore the above protection offered to depositors.

    4. So, if we ignore the protection offered to depositors, and raise the capital requirements on banks, real costs to society are the same but the apparent cost will increase to particular consumers (borrowers). I agree with that.

      I think the question is: how can you ignore the apparent cost, when that is the vehicle for delivering monetary policy (rightly or wrongly)?


Post a comment.