Follow by Email

Monday, May 6, 2013

The Bankers' New Clothes!

After Joseph Ackermann announced (and repeated undeterredly) his 25% ROE-target for Deutsche Bank (incidentally, during his tenure, Deutsche's ROE averaged less than 12%), he was often called upon (in TV shows; by business leaders; by politicians) to justify such a ridiculously high expected return on equity.

Ackermann's standard answer to that question was something like: "Look, that's really not as high as it seems. It's common and acceptable for corporations to have ROE's of 8-10%. Corporations have about 3 times our equity. If we had the same level of equity as such corporations, our ROE wouldn't be higher than 8-10%, either".

Good answer for gullible questioners. The proper reaction would have been the following question:

"Why, then, don't you increase your equity to the same percentage of total assets as corporations have? That would make Deutsche a much safer bank; shareholders would carry much less risk and would be satisfied with a lower ROE. And, by the way, society would not have to ever step in and bail you out!"

I have just finished the fascinating book titled The Bankers' New Clothes. It argues professionally and with enormous empirical evidence what I have argued in this blog on several occasions: the way to control banks and to make them safe is by limiting their leverage; by requiring them to have a greater portion of equity as a source of funding.

Most large European banks (TBTF-banks) fund their assets with less than 3% of equity; i. e. for 97% they rely on debt. Take a bank which has assets of 100, debt of 97 and equity of 3. If the value of its assets declines by 3%, its equity is wiped out (and society has to come in and save the bank because it is too big to fail). If it had had equity of 20 and debt of 80, the 3% decline in asset values would have reduced equity from 20 to 17. Solvency would not have become an issue; liquidity perhaps.

There is only one reason why, historically, large banks have been able to rely so little on equity, and that is: large banks operate with the implicit guarantee of society. The Eurogroup can dream up all the 'liability chains' it wants to dream up; no liability chain will ever be applied to a large multinational bank doing business in multiple jurisdictions with multiple subsidiaries around the world, and with nearly unlimited interdependencies (contagion!). A liability chain simply can't be enforced properly and quickly with such a structure (see the uneven damage which Lehman caused). And no lender will ever lend less money to a, say, Deutsche because of the threat of a liability chain for the simple reason that the threat of a liability chain won't work with Deutsche (or JP Morgan Chase; or...).

Before reading the book, I would have thought that large banks should be forced to return to equity/total assets ratios of 6-8% which they used to have not too long ago. The book argues that there is no reason why banks should have lower equity requirements than the free market typically requires of corporations (20-30%; at least). That argument is made convincingly!

The book is marvellous at taking the smoke screens of banking lobbyists apart. Higher equity requirements would have lower economic growth as a consequence? No evidence to support that! Equity is on the right side of the balance sheet; loans are on the left side. Equity only impacts how the bank funds itself. It does not necessarily impact how the bank applies that funding!

Basel-2 (and its successor Basel-3) is based on the terribly mistaken assumption that 'risk-weighted assets' should drive the capital requirement of banks, which led to terrible distortions: German banks did not lend to the German Mittelstand in the 2000s but, instead, purchased so-called 'risk-free' bonds of the Eurozone which required no capital allocation and AAA-rated paper (i. e. sub-prime) which required little capital allocation. The German Mittelstand still exists; the other papers have shrunk in value.

Banks will reduce loans if they are forced to increase equity? Not if the requirment is to increase equity in percentage AS WELL AS in nominal terms. In fact, with more equity, banks may end up making more loans because they are stronger.

There is not enough equity for banks in the market? Certainly not at book value and possibly not right away and in the full amount. The immediate measure to increase equity is to ban the payment of dividends. Bank share values would evaporate if banks discontinued the pay-out of dividends? Well, check back with Warren Buffett. Berkshire Hathaway, a financial concern in Buffett's own words, has never paid a dividend. It has a high equity ratio and its stock value has increased all the time.

Incidentally, the book shows that in the 3 years following the US government's TARP, the beneficiaries of TARP paid out about half of TARP-funds in dividends. Smart application of tax payers' money!

The basic premise of the book is that, for making banks safer, the cost to society of prevention is almost infinitely lower than the cost of later repair. If a bank had 20-30% equity, it could lose 20-30% of its assets before it became insolvent. There is no precedent where a bank had to write off 20-30% of its assets. If, despite higher equity, repair would still become necessary in some cases, that repair would not be costly to society because it would have to solve only the liquidity problem and not the solvency problem.

I wish everyone Happy Reading of the book!


  1. Great comment, you have made however a mistake, by having retired, Herr Klaus!
    Of my top of head, definetely equity of 20-30% differentiate the banking functioning, and empirically the book will give important knowledge, however if i may add the equity can be "handled".
    Tier 1 common ratios of selected large banks
    in percent (median) occasionally is changing month by month, an example is the Dexia, were equity was firm, theoritically more than 10% but the bank saved by French and Belgium banks.

    Lloyds Bank, Rabobank, and Credit Suisse using the CoCos increased their equity Tier 1-2

    UBS and Credit Suisse, increase their capital ratios to 19% with up to 9%

    My point is that also tangible common equity (TCE) is something to watch,to explain or particularize that,especially in connection with RWA.

    The book set some crucial questions as you mention.

    Finally, again what is your opinion about full-reserve banking?


  2. Amazing. I'm reading it at the moment too. Yes, they quite methodically explode a lot of common arguments. I particularly like the example they come back to all the time, of a woman taking out a mortgage on a house. How much debt has she? How much equity? How do these factors play out, as prices rise or fall? What does that tell us about the interests of banking industry spokemen/women?

    And I have seen journalists I respect trot out that "more capital requirements means less lending by the banks" argument. Time and again, in this crisis.

    I'm annoyed with myself, that I didn't spot the difference between capital (equity) ratios, and reserve ratios. One needs to know that difference, to see why the argument is false.

    1. Just a little addendum to your comment: if one only regulates the ratios for equity or leverage, it is very likely that a bank pressured to decrease its leverage will do so by reducing assets and not by increasing equity (and, I should add, the assets which the can reduce are likely to be the better and/or more liquid ones; thus, at the end of such an exercise, the quality of the remaining assets is normally worse than before). Why do banks do that? Because it is the far easier route (more equity means having to work harder for the equity holders...).

      This is why the book argues that regulations will not only have to work with ratio requirements but, definitely, with the requirement to increase equity in nominal terms. Take the bank with assets of 100, debt of 97 and equity of 3 (a 3% equity ratio). Regulations would have to stipulate that they increase equity to, say, 20% as a ratio and 20 units as a nominal amount. That way, banks are forced to not only increase equity but to also use it to reduce debt.

      Deutsche is presently making the news with their planned capital increase of 2,5 BEUR. Commentators call that "courageous" and "ambitious". Well, if my memory serves well, if they stopped paying dividends for 2 years, they would retain earnings in more or less the same amount (and thereby inrease equity). The book states that in the most simple terms possible: 'just ban banks from paying out dividends until they have reached satisfactory equity levels'.

  3. In the narrow sense, full-reserve banking means that a bank always has enough cash available (i. e. as overnight deposits with the Central Bank) to pay out all funders if all funders wanted their money back tomorrow morning. That, of course, wouldn't work because the bank would convert into a safety deposit box. It wouldn't have liquidity to make loans.

    In the wider sense, full-resere banking means that the maturity structure of its funding is matched with the maturity structure of its assets. Actually, banks are doing that. I am not sure whether regulations require them to do that or whether they do it because that's the only way to run a bank.

    A 'fully-matched' book means that the maturities on the funding side are fully matched with maturities on the asset side. Example: if a bank has 1 BEUR in loans with a 6-month tenor, it also has 1 BEUR of funding with a 6-month tenor. Should the funding not be renewed, the bank could - theoretically - call back the loans on maturity.

    Typically, banks never have a fully matched book. If they had, they would make a lot less money. In a classic situation, short term rates are lower than long term rates. So if banks fund themselves short-term and lend long-term, they profit from the transformation of tenors. To play the interest rate curve is a major source of bank profits.

    There is probably no larger bank which does not have a daily 'position report' which shows the 'open position' between asset and funding tenors. It shows the profit which the bank makes on the open position as well as the risk inherent in the open position (i. e. what potential losses the bank is exposed to if, suddenly, short rates were to rise). The bank also has the instruments in place to close the open position right away should that become necessary (which would entail costs).

    I am not sure whether the level of the 'open position' is regulated. When a bank is audited, the auditors definitely examine how the bank runs its funding and how much risk it entails.

    HypoRealEstate became suddenly illiquid because of their huge open posititon. The German Head Office, before the crisis of 2007, was one of the best names and could borrow at the best rates in the market. They borrowed very short term and, with that short-term money, they funded the long-term assets of Depfa, their Irish subsidiary. When the crisis hit, 'sudden stop' occurred and even the German Head Office had severe trouble renewing its short-term funding. This was a matter of weeks, if not days. In such situations, it is not a question of putting 'a little rescue money' into the bank. If much of the bank's funding is short term, it becomes a question of replacing much of the bank's funding with tax payers' money (or let the bank fail). HypoRealEstate did not originally fail because of insolvency. Actually, its numbers looked quite good. It failed because of sudden illiquidity. That sudden illiquidity brought the issue to the forefront and once they started taking a closer look at the bank, they discovered that it was also insolvent because the good looking numbers were not as good as they looked.

    Mind you, even a fully-matched position is only theoretically 'secure'. The funders of the bank can call their 6-month funding on maturity and the borrowing bank has to repay if it doesn't want to close doors. It is different on the asset side. The loans may have a 6-month maturity but even the best borrowers may not be in a position to immediately pay all their short-term loans (at least it would cause some major disruption). It that bank still called back all its 6-month loans to repay its 6-month funders, it might as well prepare for closing its doors. Word would get out in the market that the bank is having major funding problems and nothing scares funders away from funding a bank as much as the rumor of trouble at the bank.

  4. Very explanatory comment of yours.

    The "fully matched book" can practically implemented in today's banking structure? Trully i don't get,how to be "implemented", althought you explained it very well.
    I am not sure whether the level of the 'open position' is regulated, too. A "buffer zone" maybe can create a kind of control.Equity 20-30% is wise.
    Assesing fractional reserve banking is creating "a motive" not to control adequately core accounts.
    Sudden illiquidity were it came from in case of Hypo?

    "Typically, banks never have a fully matched book. If they had, they would make a lot less money"

    This might be a wise solution to make less money.


    1. The 'position paper' must be signed off daily by senior management. All it would take to implement a fully-matched book is for senior management to instruct its treasury people to 'match the position'. Before they do that, they will look at the forgone (anticipated) profits from the mismatch and they will think twice about giving such an instruction...

      I cannot emphasize srongly enough how important a source of revenues/profits the tenor transformation is!!! (perhaps over 50% of net interest revenues in some cases!).

      'Sudden stop' at HRE? Well, Lehman had failed and interbank lending had started to dry out. Not just for HRE and with HRE it really hurt badly...

  5. Here's a newpaper article written by Mr. Hans-Werner Sinn, which I find quite interesting. In German, unfortunately. But concerning the questions relevant in this blog:

  6. For Klaus especially - F.A. Hayek: His 114th Birthday from Think Markets, by Mario Rizzo

    @anonymous - Thanks for the Faz link, I read something similar in English authored by Dr Sinn, might have been at Project Syndicate. It was a very good translation, some of his earlier articles there were not so well translated.

    I first heard of Dr Sinn in respect of TARGET-2 via Reuters and FT journalists. As a layman I never understood his arguments on that issue, still don't, Whelan's arguments seem to make more sense then and now. It also annoys me, as a software engineer, that the issue is mystified by the use of the acronym given to a particular suite of computer application programs. As a result I was not a fan of Dr. Sinn.

    However, in more recent times I've changed my views, considerably. A lot of what he's been saying of late makes sense to me, as it takes account of political realities and the idiosyncrasies of the various societies that make up the EU.


    1. 1 of 2
      Re Hayek: since I am not an economist, I have never paid much attention to 'Hayek the economist'. My focus with Hayek is on, as the author calls them, his ‘views on mind, law and ethics’. And, in general, what it is that makes the society a better society.

      My favorite Hayek-book is The Constitution of Liberty. One may or may not agree with him but, still, I think that's a book which everyone should have read before he/she graduates from High School. Particularly his views not only on the State of Law but, equally importantly, on a state-of-law-culture with the adequate institutions are timeless, as far as I am concerned.

      Re Sinn: his problem is similar to what Thatcher's problem was - he provokes, antagonizes and polarizes. Sadly, that overshadows the substance of his views much of the time. People don't like to agree with someone by whom they feel provoked, antagonized and polarized.

      Sinn has been right more often than not. After German reunification, he wrote a book where he casticized the policy decisions of the West and warned that they would fail (that was the book which put Sinn on the map). Obviously, that was not 'politically correct' at the time, but he turned out to have been right.

      In the early 2000s, he wrote the book 'Can Germany still be saved?'. Again, he was casticized for similar reasons as above. He suggested in the book what Germany should do to save itself and, under Gerhard Schroeder, Germany did most of that.

      Not only are his books provocative, when he speaks publicly and particularly in TV-discussions, he leaves the impression that he has all the right answers and that everyone who doesn’t agree with him is dumb. One doesn’t make friends that way.

      Re Eurozone: back in the spring of 2010, Sinn did NOT stand out with the proper solutions (now, in retrospect, he has them, as shown in the article). Everyone then only saw two choices: either bail-out Greece of let it default. Sinn was for the latter. No one that I can remember suggested the third choice which would have been the right one, in my opinion: have Greece reschedule its entire foreign debt with existing creditors and have officialdom only finance the Fresh Money requirement. Hiring people like Bill Rhodes and Lee Buchheit for that job would have accomplished everything. Remember that the Chief Economist of Citibank said a couple of years ago: “The Europeans did not know that sovereign reschedulings have come a dime a dozen in recent decades”.

    2. 2 of 2
      Re Target2: Sinn deserves credit for bringing that most important subject to public attention and for showing what an important factor it is in the Eurozone’s functioning.

      Target2 is nothing other than the Eurozone’s cash management system and the letters stands for the name of that software. Any Assistant Treasurer of a multinational company with subsidiaries knows what that is because they have their own group-internal cash management and netting systems (albeit with different names).

      The ECB is the equivalent of the multinational’s HQ (or rather: the multinational’s central Treasury Unit). Take a multinational with 20 subsidiaries, each with their own local banking relationships. 18 out of the 20 have credit balances with banks totaling 20 MEUR and 2 are borrowers with debt totalling 18 MEUR. The consolidated statements of the multinational show net credit balances with banks of 2 MEUR. That’s why one doesn’t see the Target-2 claims/liabilities in the ECB’s statements; one has to look at the statements of the individual national Central Banks.

      Now, under Target2, the ‘subdidiaries’ don’t have their own banking relationships and debts. The borrowing subsidiaries borrow from those subsidiaries which have cash. And that works automatically via the software.

      Take a German company which has a production subsidiary in Romania which sells to Head Office. The subsidiary does not have a local banking relationship. All its payments are routed through the central Treasury’s cash management system. If their deliveries to the German HQ are not paid, the receivables due from HQ go up. If they pay their bills through the central Treasury, their receivables go down. That’s netting.

      In the consolidated statements, this results in a wash (the Romanian accounts receivable from HQ are netted against accounts payable to the Romanian subsidiary at HQ). A reader of the consolidated statements would never guess that HQ has a lot of debt and the subsidiary a lot of due-from’s. However, should the German HQ go bankrupt, a problem occurs: the Romanian subsidiary may lose all its accounts receivable from the German HQ.

      The principal difference between a multinational and Target2 is that with a multinational, most payments are ‘for value received’ (products being shipped from one place to the other). Under Target2, there is no value received from Greece in exchange for Target-2 ‘funding’ which Greece gets because that funding is not used to buy products from the ECB but, instead, from third parties (imports).

      All clear?


    The above article indirectly relates to what I have said about The Bankers' New Clothes (but takes a different view). The comments to the article are equally readable (and I don't say this because I have also made comments...).