Monday, March 11, 2013

Should banks be left as they are?

The Ekathimerini argues in this editorial that banks should be left as they are. After a second reading, I come to the conclusion that they don't mean it the way I understood it the first time around. They mean that if banks are not left the way they are, they would shift back to the model of state-run banks. Well, nobody can want that.

But noone in his right mind can plead for the continuation of a system where tax payers have to bail out banks (sometimes via countries like Greece) and they get absolutely NOTHING in return for doing that!

What they should get in return are temporary ownership stakes in those banks. Yes, as a passionate believer in a market system with a large private sector, I say that there have to be temporary partial/full nationalizations of banks so that tax payers can have the prospect of recouping at least part of their investment when the banks are re-privatized.

Should the state run banks? Heck, no! The state ought to assume partial/full temporary ownership and select a competent bank management to run those banks so that it can later be sold profitably.

Look up the history of the Continental Illinois National Bank and Trust Company of Chicago (CINB). Back in 1984, as the then 6th largest US bank and considered 'too-big-to-fail', it had to be rescued by the government. The government split the bank into a good bank and a bad bank. The old shareholders remained shareholders of the bad bank, which was being liquidated. They were given a share in the good bank to make them go along with the restructuring, but that share was pledged to support the bad bank. At the end of the day, the old shareholders lost everything because the bad bank finished with huge losses and they had to forfeit their share in the good bank, too. The government made a good profit on the sale of the good bank but not good enough to cover all the losses of the bad bank. Still, the final cost to the government was rather manageable.

The government achieved this by putting in place a new CEO (John Swearingen, a retired CEO of Amoco) and a new President (William Ogden, former CFO of Chase). Swearingen was later replaced by Tom Theobald, former No. 2 at Citibank. Within ten years, CINB was again attractive enough to be sold to the Bank of America.

There is a curious situation in Germany these days where Leftists (Oskar Lafontaine, Sahra Wagenknecht) are explaining to the economic establishment the principle of capitalism where shareholders equity is risk and should be treated as such! The capitalists don't seem to believe in that principle and keep bailing out shareholders.

I will show a couple of ways how this could be accomplished (in addition to the CINB model outlined above).

The most brutal way would be for the government to say to a private bank which shows up and 'demands' a bail-out because otherwise they would go bankrupt and market chaos would result the following: "Don't you worry about us or the market. If you can't make it, go ahead and file for bankruptcy. We already have a new company in place which will simultaneously with your bankruptcy declaration take over all the operations of your bank. Your shareholders will be wiped out and we reserve the right to also have some of your professional market participants (i. e. bondholders, etc.) share in the damage. And you will lose your jobs, too." If the bank declares bankruptcy in the morning, its operations will continue smoothly by the same afternoon.

The more elegant way would be what the US government did with GM, AIG, Citibank, etc. --- simply dilute the equity of existing shareholders with new capital, and make a profit on it when it is sold back to the market.

And, finally, there is the 'elegant' way which Chile chose when it bailed out its entire banking system back in 1982: they recapitalized banks against the banks' commitment to not pay dividends for as long as public capital was not repaid with interest. There, the existing shareholders continued to hold shares but they were worthless at the outset. However, the shareholders could hope that their children would be rich again when the bank had recouped its losses out of earnings.

The principle behind the Chilean model is 'earnings power'. There is probably no bank which does not have an earnings power before risk and before taxes (and before bonuses!). Suppose it has an annual earnings power of 1 BEUR and a worthless portolio of 20 BEUR. After 20 years, the bank will have recouped its losses through its own earnings power. All the bank needs is time and the state can provide time more or less infinitely.


  1. Which would you recommend for say Deutsche, SocGen, Bankia, RBS & Lloyds - always the same for each, or maybe different for each?

    And how can we get A) get rid of these too big to fails, B) prevent their reformation and C) reduce systemic risk.

    Its the 2nd anniversary of the Tōhoku event. It resulted in Japan losing 30-40% of its electricity generating capacity - most of the nukes are still down. At one point soon after the event Japan was 60% down on grid power due to interconnect failures and grid collapse. Yet Japan has survived, remarkably well in fact. Could the Global Financial System withstand an event of that magnitude - and be in reasonable shape 2 years later.


    1. A couple of suggestions (off the top of my head):

      First, the pro-cyclical rating system of Basel-2 which relies far too much on historical hard facts should be amended (Basel-3 does not do that!).

      Second, the valuation of assets must be reviewed. The large banks have huge asset volumes which are not valued on the basis of a notional claim but, instead, based on the Black-Sholes formula. What if someone sometime finds a mistake in that formula? Also, there have to be much clearer rules regarding 'mark-to-market'. Sovereign bonds (or bonds in general) should only be valued at nominal if (a) the clear intention is to hold them until maturity and if (b) there is no reasonable doubt that they can be paid at maturity. Otherwise, immediate mark-to-market.

      Third, the focus on capital adequacy ratios (excluding the so-called 'risk-free assets') must be expanded. The ultimate control over a bank's size is 'leverage' (total liabilities to equity). I think if leverage were capped at 20:1, it would already be a very high leverage. 15:1 would be a lot better.

      Those are a couple of general guiding principles. The specific application may vary a bit from bank to bank to fit each bank. The idea is not to stubbornly enforce a rule but to achieve desired results.

      How would Deutsche go from a leverage of 40:1 down to 15:1? Presumably not as one bank. But if JP Morgan and Citibank get by with a leverage of 9-11:1, why shouldn't Deutsche get by with 15:1? After all, unduly high leverage overstates profits (particularly when the risk of that leverage is carried by tax payers).

      I don't think that one can get rid of systemic risk in a globalized world, but one certainly can and should make the systemic risk transparent. If derivatives are weapons of mass destruction, trading them over the counter is like opening a bazar for WMDs. Which reminds me that I should have added another point above: the transaction volume below a bank's balance sheet (contingent assets/liabilities) must be contained in a relation to the balance sheet total. Many banks have more risk below than above 'the line'.

    2. Trully great comments mr Kastner, one of the best i ve read.
      You might know the terms of recapitalitation of Greek Banks. Generally the basics here

      Its the Contingent convertibles CoCos

      But the scheme planned does not give not only for current stockholders the motive to participate or even for new potential,. In an environment of low interest rates the coupon which is paid (7%) does not create a motive because the participants ignore the panic or stock market volatily or even manipulation. Even those with the warrants. The warrants price is calculated around P/BV 1,5 when foreign banks are 0.75
      The warrants should have as a trigger price a price in accordance with convertibility price (of the warrants)

      Also the proposal does not give the right to current shareholders to buy warrants at conversion price, or does not give a period of maturation.
      The potential investors do not want to participate because of the market possible instability and the valuation of investment.

      Do you have an idea- proposal?

      I have read-a bit analyse COERC as a well designed more practical


    3. There is a blog called Independent Insight Views which has written extensively about the recap program. I have read everything and still, I have a bit of trouble understanding it. My sense is that existing shareholders are given the chance to keep a share of their bank if they can come up with capital. You seem to confirm that. What I don't understand is what if they don't come up with that capital. Will that nationalize the banks 100%?

      My understanding is that behind some of the banks are some of the wealthiest Greeks. By that, I don't mean people who count their wealth in single-digit BEUR figures but in double, if not triple-digit BEUR figures. Why they wouldn't immediately come up with the 'peanuts' of a few hundred million to keep control of their banks is a mystery to me.

      At the same time, one would have to ask the question to what extent tax payers' monies should be used to help some of the richest people in Europe. I don't have a clear answer to that.

      The key question is how good/bad these banks really are. If they are, as some suggest, totally bankrupt, then existing shareholders should be thankful that someone else is saving them so that they can eventually become good banks again (which would make the shares of the existing shareholders valuable again). It's only fair to ask that, to realize this chance, the existing shareholders should come up with some money on their own.

    4. @anonymous - so that's what CoCos are - nothing to do with the Keeling Islands that the Clunies Ross' sold to the Aussies in the '70s - no wonder I get confused ;)

      @Richard Bourke - thanks for the link to Peston's very informative piece. Whilst I was there I listened to his Sep 2009 interview of Adair Turner. I'm not convinced much has been done in the ensuing 3.5 years that would prevent a replay of 2008 in say 2014.

      Maybe that's the reason the MSM have so little to say about Banking Regulation. Firstly its couched in arcane language that even the so-called experts don't seem to understand, and whilst there's a lot of talk, nothing seems to get done - unless you count banker bonus caps.

      Paul Volker proposed his rule Rule in 2009, Obama signed up to it in Jan 2010 - current date for implementation is July 2014. So there's at least another 15 months of watering (down) to be done. Then there will be more delay to argue the merits of the watering. One has to ask - will he live long enough to see his Rule implemented. By now it must be so full of get-outs, loopholes and exceptions that its only effect will be... an increase the wealth of attorneys.

      Thought : Greeks are top of the Planning First Division, but towards the bottom of the Execution Fourth Division. Some commentators seem to think that the solution is - more Planning (presumably 'cos we're good at that).


    5. @ CK

      In an extremely complicated world, there aren't any unaltered or hidden "truths". In open markets when banks fail, and people pay gov's have responsibilities to implement some rules. In Gr its irrelevant if the bankers had the major responcibily or the political elites-people individually for what is happening.As an elite they had a major responsibility. Definetely there were many examples of mismanagement but these - explained in numbers- do not cause the major systemic imbalance.
      So for the "truth" from the numbers it is decided to "give" a 10% as a reward to old shareholders to participate, not to destroyed totally from dilution and to keep the private character of the bank.
      But CoCos as planned do not constitute a resonable opportunity for current and potential shareholders because of market instability etc i mention to mr Kastner.
      Its much more clear to decide that because the billions needed for recap are many, the banks stay under the control of gov. End.
      But when some hundred people "discover" CoCos, and all the rest do not understand a clue its not all this a solution. When y do something do it well.


  2. Very interesting column. And your remarks about leverage ratios and some of the perverse incentives ("zero risk weighting" for Greek Sovereign Bonds ist still in effect, I believe) of Basel-II (and presumably Basel-III too) seem right on target.

    One of the really frustrating things I find, in my reading around this whole issue, is just how little coverage there is on banking regulation. What's gone wrong, what might conceivably be changed. There are almost no journalists in the mainstream press writing about it, since the subject is so technical and dry.

    One honourable exception - and the only exception I'm aware of: Robert Peston at the BBC makes a brave attempt to explain leverage ratios, and the trade-offs concerned, in layman's language.

    1. This is a good article; thank you! Obviously, a lot more specific than my superficial comments.

      A bank which has zero risk (for the sake of theoretical argument) can't help but make money. Why? Because banks get paid for just keeping assets on their books (interest). A corporation, in contrast, cannot afford such luxury; it always has to turn assets from cash to inventory to work in process to finished products to accounts receivable and then back to cash.

      Thus, a bank is as good (or as bad) as the quality of its assets, the quality of its loans. There is no way that an outsider can assess the quality of a banks' loans. Basel-2 sets the model for classifying assets according to their perceived/calcualted risk. If that model is not adequate, the whole result can be wrong (as you point out).

      Restricted leverage will not prevent bank failures. The only thing it does is to limit the potential damage of a bank failure.

      Take a leverage of 15:1. It means that a bank has, say, total liabilities of 1,5 BEUR against an equity of 100 MEUR. If that bank has to write-off 10% of its assets, its entire equity is wiped out (and 50 MEUR more!).

      Writing-off 10% of one's assets used to be unimaginable if a bank was only half-way seriously managed. How does 10% sound in today's world?

      Try the same calcualtion for a bank which has a leverage of 40:1 and you start shivering!

      When the Continental Bank failed, the potential damage was easy to calculate. Virtually all its assets represented a nominal claim (i. e. no derivatives which are valued based on a formula and not a nominal claim). One only had to calculate how many of those nominal claims would have to be written off. Still, even back in 1984, the interbank business of that bank was so inertwined with the system that a systemic chaos was feared if the bank had been allowed to fail outright. This is why it was rescued in the way I desribed. But compared to today, it was a fairly simple and manageable process.

      Having been in US banking during the Paul Volcker years (and having once met Paul Volcker in a small private luncheon), I would suggest the following: make Paul Volcker in charge of the worldwide financial system and give him total authority. The financial world would be a better place!

      Incidentally, Paul Volcker is the man who once said, after being asked what all the financial innovation had given the world by way of benefits: "I think the invention of the ATM did more good to the world than all financial innovation taken together". All clear!

  3. Thanks Klaus - I'll need to sleep on that and reread in the morning...

    Did you see this in today's EKathi - The Byzantine empire's own 'eurozone' crisis... - quote

    Those responsible for the crisis were shown no mercy. The Herman Van Rompuy of the time, a eunuch named Nikephoritzes, was lambasted by an angry population faced with price rises and a fall in the standard of living, and was eventually tortured to death.

    Brilliant - ROFL


    1. Several points for several of the above:
      a) The Black_Scholes is wrong. That can be easily proved: the formula comes out of something called the Ito-Stratonovich formulation that allowed the rigorous treatment of stochastic differential equations, This, together with Ito's lemma and Ito's integral allowed the creation of diff. equations that described the financial derivatives markets. Using Ito's integral these stochastic diff. equations were solved producing Black-Scholes. However for the integral to be possible the stochastic process must not get information from the future-ie it must satisfy causality. Unfortunately orange juice markets at least are known to predict Florida weather better than the Weather Service(papers that do that are very difficult mathematically and it is difficult to make sure that data used are reliable so I will not quote the papers-if you feel inclined to study the subject I will go through my files)Also markets are known to predict announcements (also known as insider trading). With two counter examples, for the formula to be proved, causality must be proved, something impossible.Why is it used? For two reasons: A) it is by far the best we can do b) it works most of the time. I also suspect many people find it very useful for their pockets or political careers. But as an anchor of the system? NO WAY JOSE. I am sorry if the above is incomprehensible, but it is the best I can do. Remember what I said in a previous comment: economics done with rigorous maths is very difficult.
      b) Marked to market is meaningless. We need to be precise. OTC markets often seize up in strange ways. This is because if a player does something fishy the only way to stop him is by a civil suit, as we are dealing with a contract dispute. If the player is big then the market part that said player serves freezes. An example from personal experience: the Hedge fund I worked for was making money by exploiting pricing anomalies caused by the big banks bond pricing algorithms. When we were caught out (after some years) we and I suppose the other funds that did the same, had the Bloomberg feed messed up and our money blocked. We had to go begging and bribing as suing Barclays, Deutsche etc was out of the question. It was restored, but this proves that there is no single price in an OTC market, it is just the price that maximizes big bank bonuses. Just like the LIBOR mess.If marked to market is to be used it must be based on a well regulated market, along the lines of a stock market. Not perfect, but better.
      c) The esteemed Oxford professor must be taking something illegal. If he was a Greek I would say he is προγονοπληκτος-in awe of one's ancestors, a common Greek affliction. Maybe he is trying to give us Greeks a friendly hand. I don't know. What I do know is that his description of the Eastern Empire is highly sanitized, to the point of sophistry. The history of Byzantium is a fascinating subject, especially its duration and robustness. Two things though stand out about its longevity: a) the Empire was the only stable island in a very unstable war torn continent b) it was essentially a rent seeker, taxing the trade along the ends of the Silk Route and exporting its currency, the only good quality gold coin in Europe at the time. Once the deplorable 10th Century came to a close and stability started to return to W. Europe trade patterns changed. The first to move was Venice under the Orseollos. Its ships challenged the Byzantine trade monopolies, eventually causing the 1070 events, a crisis from which the Empire never really recovered: it was downhill from then on economically.
      d) Having being in the US at the same time I perfectly agree with P. Volcker comment.The guy is phenomenal.

    2. Three anecdotes about risk management/controlling.

      In the 1980s, Continental Bank bought First Options, an options trading company in Chicago known for its outstanding risk management/controlling. In October 1987, First Options went bust after the crash. I queried a colleague how that could have happened when they had the best risk management/controlling around. His answer: "They did indeed have the best risk management/controlling. After the crash, they knew immediately that they had gone bust. Their model was built on 1929 as the worst case. October 1987 was worse than 1929. So, their risk contolling worked exactly as it should have..."

      The book "How Genius Failed" describes beautifully the demise of the legendary hedge fund LTCM in 1999. Over a period of 4 months, the fund lost steadily its capital base of about 5 BUSD. A continuous daily downward trend. One exception. On a Friday, the staff had taken the day off and the office was vacant. That was the only day during the whole period where LTCMs computers recorded a profit.

      And, finally, Myron Sholes, Nobel Prize winner, was a partner of LTCM. During their initial fund raising, they called on a Midwest institutional investor and explained how they would make money by doing arbitrage with minimal margins but with huge leverage. A CFO in the audience blurted out: "With that simple strategy, you will never make real money!" Sholes' response was: "Because of fools like you, we will make a lot of money". As it turned out, the Nobel Prize winner was wrong and the fool was right.

    3. The last anecdote is hubris, which, this being an about Greece blog, we know is inevitably followed by Nemesis. The other two though encapsulate my view of the intractability of economic maths. Closing down LTCM made a modest profit for the banks involved. Therefore it was bankrupted by a liquidity crisis that was irrational. A liquidity crisis is, mathematically, a discontinuity, a concept not really manageable by mathematics.
      In the case of First Option I can imagine what happened: they built a good mathematical model that considered the worst case scenario and known sources of instability plus a margin of safety. This method works in engineering bridges and such because our assumptions about the working environment can be reliably tested and the future can be reliably projected from the past (ergodicity). But it doesn't work in finance, where LTCM like events prove that the past is not a reliable way to predict the future (non ergodic) and very low probability or even previously unthinkable parameters become dominant. Be very suspicious of any financial mathematics.
      In order to avoid being called a refusegnic my best guess about financial maths is another little known Samuelson suggestion from the 60's: the Lotka-Volterra equations, a set of highly non-linear equations from biology. They are currently intensely researched and the mathematicians describe them with their telling understatement: they have rich behavior.