Saturday, April 16, 2016

A better living will

"US rejects 'living wills' of 5 banks," from FTWSJ puts this event in the larger story of Dodd Frank unraveling. Juicy quotes:
WSJ: “living wills,” ... are supposed to show in detail how these banking titans, in the event of failure, could be placed into bankruptcy without wrecking the financial system.

FT:...the shortcomings varied by bank but included flawed computer models; inadequate estimates of liquidity needs; questionable assumptions about the capital required to be wound up; and unacceptable judgments on when to enter banktruptcy.

FT: David Hirschmann of the US Chamber of Commerce, the biggest business lobby, said the living wills process was “broken”. “When you can’t comply no matter how much money you put into legitimately trying to comply, maybe it’s time to ask: did we get the test wrong?” he said.

WSJ: Six years after the law was passed, and eight years since the financial crisis, regulators given broad authority to remake American finance, with thousands of regulatory officials on their payroll, cannot figure out a system to allow financial giants to fail, even in theory. What are we paying these people for?
It seems like a good moment to revisit an idea buried deep in "Toward a run-free financial system."  How could we structure banks to fail transparently?

Picture of bank structure

Recall, here is how banks are structured now (extremely simplified). Banks hold assets like loans, mortgages and securities. Banks get money to fund these assets by selling a tiny amount of equity, i.e. stock, and by a huge amount of borrowing, including deposits, long-term bonds, and short-term debt.

The trouble with this system is, if the value of the assets falls by more than $10 in my example, the equity is wiped out, and the bank can't pay its debts. If short-term debt holders worry about this event, they all clamor to get paid first, so a run can happen. That's not really a problem either; bankruptcy is set up exactly to handle this situation. The creditors who lent money to the bank split up the assets. Yes, they don't get their full money back, but if you lend to a bank that's leveraged like this, that's the risk you take.

The trouble is the widespread feeling that big banks are too big, too complex, too illiquid, to utterly muddy, to carve up this way. If it takes years in court, and if all the value of the assets is drained away by lawyers, you have a real problem. Furthermore, we often want the profitable parts of the bank to remain in operation while the creditors squabble about assets. (Ben Bernanke's classic paper on banking in the great depression makes this point beautifully.) The ATM machines should not go dark, the offices where people know their customers and can keep things going should stay in operation.

Hence, big banks become too big -- or too something -- to fail. In that situation, the government is mighty tempted to bail out the creditors and keep the thing limping along. Given that temptation, a lot of large, politically well connected creditors also scream that there will be ``systemic dangers'' if they don't get their cash now, adding to the bailout pressure. A "living will" is supposed to stop this chain, by allowing  bank assets to very quickly get divvied up among creditors.

But the large banks are, apparently, so large and complex that nobody can figure out a living will. That's debateable, for example Kenneth Scott and John Taylor argue bankruptcy can work.  But let's go with the idea. Is there an alternative to Bernie Sanders' bust up the banks? Here's one.

picture of altered bank structure that is easy to resolve

Starting from the left, suppose the bank holds all the same assets it does today. But, it issues 100% equity to finance its assets. Now, a 100% equity financed bank cannot fail. If you don't have any debt, you can't fail to pay debts. Yes, the bank can lose money and slowly go out of business. But it cannot go bankrupt. As it loses money, the value of its equity declines, until shareholders get mad and liquidate the carcass. Nobody can run to get their money out ahead of the other person. End of bankruptcy, end of bank runs, end of financial crises.

(Technical note. Yes, that's a bit overstated. A bank can potentially invest in derivatives and other securities where it can lose more than all of the investment. The amount of monitoring needed to make sure this doesn't happen is trivial next to the Basel sort of thing required to make sure a bank never loses more than a few percent of its value.)

OK, gulp, you say. But don't people "need" to have bank accounts? Isn't "transformation" of debt into loans the crucial feature of the financial system? Don't equity holders "require" high risk, high-return stock? No, argues the "run-free financial system" essay. But let's not go there. Let's just restructure things so that the bank can hold exactly the same assets it has today, and its investors can hold exactly the same assets they hold today.

So, moving to the right in my little picture, suppose bank stock is held in a mutual fund, exchange traded fund, or a special-purpose "bank." Bank stock is the only asset these companies hold, and that stock is also traded on exchanges. These banks fund themselves by the same mix of debt, equity, deposits, and heck even overnight wholesale debt, commercial paper, and so forth.

Now, if the value of the bank stock falls, these holding companies fail, just as my original bank failed. But there is a huge difference. You can resolve the holding company in a morning and still make it to play golf in the afternoon.  The only asset is common stock, commonly traded! There are no derivatives positions to unwind, no strange positions in offshore investment trusts, or whatever.  The "living will" simply specifies how much common equity each debtholder gets in the event of bankruptcy. There is never any need to break up, liquidate, assess, or transfer bits and pieces of the big bank.

Furthermore, there is no more obscurity over the value of  the holding company assets. We see the value of bank assets, marked to market, on a millisecond basis.

The holding companies can provide all the retail deposit services banks now provide. In fact, they could contract out to the banks to provide those on a fee basis, so the customer might not even need to know.

In addition, any sane holding company would hold the stock of several banks, diversifying the risk, and thus reducing the chances of ever needing to be wound up. Come to think of it, any sane holding company would also diversify out of banking, but now we're back to my larger vision of equity-financed banking and sensible small changes in financial structure to achieve it.

In the meantime, there you have it. 100% equity financed banks can still give bank creditors exactly the same assets they hold today, and allow failures of those debts to be resolved in a morning.


  1. Thanks for the article, professor. Let them fail and go to regular process as any other business. No "too big to fail" policy. Brgds

  2. I think (not sure) you're saying that it's ok to have demandable/short-term debt liabilities so long as your asset portfolio is mark-to-market. And I think (not sure) in your system any entity that owns non-mark-to-market financial (or nonfinancial?) assets must fund itself with 100% equity. Is that basically right?

    BTW, I have a new book out (U. Chicago Press) that includes a discussion of your run-free financial system piece.

  3. "Banks hold assets like loans, mortgages and securities. Banks get money to fund these assets by selling a tiny amount of equity, i.e. stock, and by a huge amount of borrowing, including deposits, long-term bonds, and short-term debt."

    You seem to be forgetting that banks can issue far more money than they get from the public (from deposits, bonds, etc). A bank that has received $100 (in Federal Reserve Notes, say) on deposit can proceed to issue and lend $900 of its own checking account dollars to its customers, in exchange for "deposits" of $900 worth of those customers' IOU's.

    1. Indeed monetary financing of investments is a defining property of banks, and Cochrane's diagram is somewhat confusing or perhaps misleading where it labels an investment company, financed entirely by equity, as a bank.

      I don't think that he sees what his structure is doing to the possibility of monetary financing. This is not a sensible small change in financial structure. Carolyn Sissoko (@csissoko) comments here sometimes to point out the problems caused by cash-in-advance constraints, which are being re-introduced in the modification depicted above. Such constraints and resulting problems are also discussed in academic work that looks critically at narrow-banking proposals.

  4. Who does derivatives in this system? If it's the bank subsidiaries, then you will have another claimant besides the stockholders. If it's the holdcos, then they will have assets besides the stock.

  5. John Cochrane’s article “Towards a run free financial system” was better than the above post in that in the “run free” article he advocated splitting the bank industry into two distinct halves (as did Milton Friedman and as does Positive Money and other present day advocates of full reserve banking like Laurence Kotlikoff).

    One half simply accepts deposits which depositors want to be totally safe. To reflect that, money is simply lodged at the central bank and/or invested in short term government debt.

    The second half lends to mortgagors, businesses, etc, but that half is funded just by equity (or something similar, like bonds that can be bailed in). The chance of either half failing is vanishingly small.

    In contrast, in the above article, JC seems to advocate having THE SAME entity take deposits and make loans to less than entirely safe borrowers, like mortgagors. In that case, if the assets fall far enough in value, then the promise made to depositors, i.e. that they’ll get their money back, can’t be fulfilled. Enter the taxpayer stage left with truck loads of $100 bills to save the day.

    In short, the above “split in two” element is absolutely essential.

    1. Ralph,

      "In short, the above split in two element is absolutely essential."

      Treasury Secretary Robert Rubin, Treasury Under-Secretary Larry Summers, Federal Reserve Chairman Alan Greenspan, Democratic President Bill Clinton, and a host of politicians from both parties all disagree with you. Once upon a time we had separate commercial and investment banking. And then along came Graham-Leach-Bliley:–Leach–Bliley_Act

      What JC would like is for depositors to be treated as equity share holders, which is fine, as long as they are given a voting stake in the operations of the bank like every other equity shareholder.

    2. Frank, Re politicians, I have little time for them: they don’t know much about economics. Also splitting banks into those investment and commercial categories is quite different from splitting them into deposit taking and lending categories. The latter split is better.

      Re your claim that JC wants depositors to be treated like shareholders, that’s what he says in the ABOVE article (if only implicitly) but not in his earlier “Towards a run-free financial system” article. The latter article (which as I said above, I agree with) advocates the same as Milton Friedman, Positive Money and others, namely that depositors have a choice. First, they can remain depositors in the traditional sense, i.e. plonk money in a bank or similar entity and that money is totally safe because it’s simply lodged at the central bank. Or second, they can have their money loaned on to mortgagors, businesses etc, but in that case they must in effect become shareholders – or something similar, like bondholders who can be bailed in.

      As to the second lot having voting rights, that’s fine by me.

  6. Is it that easy to split up the asset and liability side of a bank? Take commercial lending: a single bank will typically be a creditor (medium-term commercial loans) and a debtor (managing payments services, holding liquid assets) to the same company. This might be because acting as the latter gives the bank information relevant to the credit-worthiness of the company, allowing them to lend at lower interest rates than the market.

    In your proposal the "bank" would not have the information from acting as a provider of liquidity services (that would be delegated to the "bank holding company") neutering this key advantage of bank-based finance (vs. market-based finance).

    1. Interesting point. The solution to that problem is to have payment services and lending offered by the same bank under the same roof, but with the two activities (lending and payment services) kept separate. Positive Money’s system allows that. Obviously that increases the risk of supposedly safe money being used in a risky way. Perhaps the solution to that is for FDIC to charge banks extra insurance where they want to engage in that riskier “under the same roof” option.

  7. If I get the essence of this idea, it seems to be that a separate holding corporation is assuming and guaranteeing the risks of the lending bank. How does that differ in substance from simply requiring every bank to buy private insurance from regular insurance carriers? Surely Lloyds of London could do better than FDIC and the Frankly Doddering regulators combined? Am I missing something?

  8. a structured finance solution which is too clever by half:

    all you have done is move the situs of the "run" to the mutual fund company whose depositors will demand their money back at par when the bank's equity is impaired.

    as long as there are short term liabilities payable on demand at par funding long term assets subject to impairment, instability is inherent in the system.

    1. Jim in this post he is only showing how to get a better living will. His way to eliminate runs is in the narrow banking paper he linked, which removes the mutual fund holding company, eliminates banks altogether, and just leaves us with investment companies financed entirely with equity.

  9. I am confused about one point: When the bank gets into trouble, and a depositor goes to withdraw their cash, do they get actual cash, or are they handed shares of the bank? If the former, how much cash does the mutual fund keep on hand? If a small amount, then they have the same bank run problem. If a large amount, how can they keep sufficiently invested in the bank?

    If the mutual fund hands out shares instead, then this is a significantly worse prospect for the depositors. The whole point of call deposits is you can get cash at any time. Not possibly illiquid and definitely uncertainly priced shares.

    1. Foolish Jordan,

      "When the bank gets into trouble, and a depositor goes to withdraw their cash, do they get actual cash, or are they handed shares of the bank?"

      Presumably they already own shares in the bank (as depositors) so they can attempt to sell that stake in the bank on the open market or as voting shareholders they can vote their displeasure at the next bank shareholder's meeting.

      It's my personal belief that depositors as equity shareholders would never work without granting shareholder voting rights to those depositors. With voting depositors, they have a say in the amount of risk a bank takes on.


Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.