Sunday, June 23, 2013

Stopping Bank Crises Before They Start

This is a Wall Street Journal Oped 6/24/2013

Regulating the riskiness of bank assets is a dead end. Instead, fix the run-prone nature of bank liabilities.

In recent months the realization has sunk in across the country that the 2010 Dodd-Frank financial-reform legislation is a colossal mess. Yet we obviously can't go back to the status quo that produced a financial catastrophe in 2007-08. Fortunately, there is an alternative.

At its core, the recent financial crisis was a run. The run was concentrated in the "shadow banking system" of overnight repurchase agreements, asset-backed securities, broker-dealers and investment banks, but it was a classic run nonetheless.

The run made the crisis. In the 2000 tech bust, people lost a lot of money, but there was no crisis. Why not? Because tech firms were funded by stock. When stock values fall you can't run to get your money out first, and you can't take a company to bankruptcy court.

This is a vital and liberating insight: To stop future crises, the financial system needs to be reformed so that it is not prone to runs. Americans do not have to trust newly wise regulators to fix Fannie Mae and Freddie Mac, end rating-agency shenanigans, clairvoyantly spot and prick "bubbles," and address every other real or perceived shortcoming of our financial system.

Runs are a pathology of financial contracts, such as bank deposits, that promise investors a fixed amount of money and the right to withdraw that amount at any time. A run also requires that the issuing institution can't raise cash by selling assets, borrowing or issuing equity. If I see you taking your money out, then I have an incentive to take my money out too. When a run at one institution causes people to question the finances of others, the run becomes "systemic," which is practically the definition of a crisis.

By the time they failed in 2008, Lehman Brothers and Bear Stearns were funding portfolios of mortgage-backed securities with overnight debt leveraged 30 to 1. For each $1 of equity capital, the banks borrowed $30. Then, every single day, they had to borrow 30 new dollars to pay off the previous day's loans.

When investors sniffed trouble, they refused to roll over the loans. The bank's broker-dealer customers and derivatives counterparties also pulled their money out, each also having the right to money immediately, but each contract also serving as a source of short-term funding for the banks. When this short-term funding evaporated, the banks instantly failed.

Clearly, overnight debt is the problem. The solution is just as clear: Don't let financial institutions issue run-prone liabilities. Run-prone contracts generate an externality, like pollution, and merit severe regulation on that basis.

Institutions that want to take deposits, borrow overnight, issue fixed-value money-market shares or any similar runnable contract must back those liabilities 100% by short-term Treasurys or reserves at the Fed. Institutions that want to invest in risky or illiquid assets, like loans or mortgage-backed securities, have to fund those investments with equity and long-term debt. Then they can invest as they please, as their problems cannot start a crisis.

Money-market funds that want to offer better returns by investing in riskier securities must let their values float, rather than promise a fixed value of $1 per share. Mortgage-backed securities also belong in floating-value funds, like equity mutual funds or exchange-traded funds. The run-prone nature of broker-dealer and derivatives contracts can also be reformed at small cost by fixing the terms of those contracts and their treatment in bankruptcy.

The bottom line: People who want better returns must transparently shoulder additional risk.

Some people will argue: Don't we need banks to "transform maturity" and provide abundant "safe and liquid" assets for people to invest in? Not anymore.

First, $16 trillion of government debt is enough to back any conceivable demand for fixed-value liquid assets. Money-market funds that hold Treasurys can expand to enormous size. The Federal Reserve should continue to provide abundant reserves to banks, paying market interest. The Treasury could offer reserves to the rest of us—floating-rate, fixed-value, electronically-transferable debt. There is no reason that the Fed and Treasury should artificially starve the economy of completely safe, interest-paying cash.

Second, financial and technical innovations can deliver the liquidity that once only banks could provide. Today, you can pay your monthly credit-card bill from your exchange-traded stock fund. Tomorrow, your ATM could sell $100 of that fund if you want cash, or you could bump your smartphone on a cash register to buy coffee with that fund. Liquidity no longer requires that anyone hold risk-free or fixed-value assets.

Others will object: Won't eliminating short-term funding for long-term investments drive up rates for borrowers? Not much. Floating-value investments such as equity and long-term debt that go unlevered into loans are very safe and need to pay correspondingly low returns. If borrowers pay a bit more than now, it is only because banks lose their government guarantees and subsidies.

In the 19th century, private banks issued currency. A few crises later, we stopped that and gave the federal government a monopoly on currency issue. Now that short-term debt is our money, we should treat it the same way, and for exactly the same reasons.

In the wake of Great Depression bank runs, the U.S. government chose to guarantee bank deposits, so that people no longer had the incentive to get out first. But guaranteeing a bank's deposits gives bank managers a huge incentive to take risks.

So we tried to regulate the banks from taking risks. The banks got around the regulations, and "shadow banks" grew around the regulated system. Since then we have been on a treadmill of ever-larger bailouts, ever-expanding government guarantees, ever-expanding attempts to regulate risks, ever-more powerful regulators and ever-larger crises.

This approach will never work. Rather than try to regulate the riskiness of bank assets, we should fix the run-prone nature of their liabilities. Fortunately, modern financial technology surmounts the economic obstacles that impeded this approach in the 1930s. Now we only have to surmount the obstacle of entrenched interests that profit from the current dysfunctional system.


  1. Treating government bonds and money as interchangeable undercuts the independence of the central bank. So I would modify the proposal by not allowing treasuries to substitute for reserves, and also forbid the Fed from purchasing treasuries.

  2. Dodd-Frank a colossal mess? The UK’s equivalent, the so called Vickers report is just as big a mess. Laurence Kotlikoff tore it to shreds here:

    Anyway, a great article in the WSJ.

  3. "Institutions that want to take deposits, borrow overnight, issue fixed-value money-market shares or any similar runnable contract must back those liabilities 100% by short-term Treasurys or reserves at the Fed."

    What about backing them with 3-month triple A commercial paper?

    How do you scale this idea to non-US situations? Should banking institutions in nations whose governments are prone to default back their liabilities with government paper?

    1. Backing with “triple A commercial paper” is doubtless OK in theory. Problem is that’s a slippery slope: banks then start lobbying for “B paper” and then “C paper”. And politicians being the suckers that they are, would fall for it.

    2. Overnight reverse repurchase agreements with haircut proportional to price-volatility on any collateral. Government Bond collateral makes the most sense but if there isn't enough of it, no reason not to extend to other asset classes.

      That's the safest asset there is and that's what I've recommended should back deposits (here:

      AAA commercial paper is pure unsecured credit risk to corps. It's kinda safe because generally the corps get downgraded progressively prior to default but relying on that is dangerous. Plus you have risk to accounting fraud on the part of the corps (small, I agree).

  4. The shadow banking system was a bastion of pure unregulated free enterprise. The Libertarians were free to negotiate contracts with each other and order their own affairs free from the "dead" hand of the regulators. The government should have allowed the shadow banks and all who worked there or invested there go bankrupt.

    The difference between the dot com burst and the housing collapse was that dot com represented huge paper gains followed by huge paper loses. The housing crisis represented huge real resources spent in over building housing and wasted as surely as if they had been thrown in the ocean.

    1. Wrong, the ratio of housing to population didn't change much. It was housing PRICES that went far outside the historical norm.

    2. TGGP:

      There was a lot of new housing construction.

  5. You seem to be missing the point as to what banks do.

    Banks trade on their high-quality name to borrow at low costs of funds, and put that borrowed money at risk, earning a spread on the difference between their cost of funds and the rates at which they lend.

    In the 500 year history of banking, this has not changed.

    Banks borrow in the sort-term and lend over the intermediate-term. This exposes the bank to interest rate risk and liquidity risks.

    The function of banks is to provide liquidity.

    The entire banking industry is built on the assumption that only a fraction of the banks creditors will call their loans simultaneously.

    1. Yes, banks engage in “borrow short and lend long” i.e. maturity transformation which as you say “creates credit”. But the problem is that over and over again, they borrow too short or lend too long.

      Plus “credit” or money can be supplied by the state at zero cost, so why let commercial banks do it? As Milton Friedman put it, “It need cost society essentially nothing in real resources to provide the individual with the current services of an additional dollar in cash balances.” (Ch 3 of “Program for Monetary Stability”).

    2. Either they are prone to borrowing to short and lending too long, or they are borrowing too much, Or they are not being charged enough to borrow money.

      I would say that there is economic value in the risk allocation function. If banks don't provide it, it would be possible to dump more responsibility onto the markets. Although, I am not sure that that is necessarily a good idea. I would not like to see the state to become the primary commerical bank.

      From a regulatory point of veiw, what I would really like to see is a mechanism for banks to fail without triggering fears of systemic failure.

  6. A very specific point: I have an old account in the Vanguard Treasury Money Market Fund. The fund has been closed for years. Even with Vanguard's low costs, the T-bills don't pay enough to keep the fund open. Thus, the low risk all-Treasury money market fund that your op-ed postulates cannot exist in the financially repressed real world of today -- unless the government subsidizes it.

  7. "Run-prone contracts generate an externality, like pollution, and merit severe regulation on that basis. "

    This is the most troubling thing I've read in a long time. When friends of free markets like John Cochrane start leading the charge toward more banking regulation, we are in deep trouble. Are banks incapable of matching the maturities of their assets and liabilities without the wise guidance of government officials? Are there no arbitragers to take over badly run banks? Did John ever learn the difference between pecuniary externalities and physical externalities?

    1. "When friends of free markets like John Cochrane start leading the charge toward more banking regulation, we are in deep trouble."

      Bottom line, should you let banks fail? A free marketer would say that they are like any other business and should be allowed to fail. A non-free marketer would say they perform a vital service and should be heavily regulated with government back stops.

      Here is a question for you - how do you feel about electric utilities? Should they be allowed to fail even if it meant that you might be without power for months on end?

    2. "Are banks incapable of matching the maturities of their assets and liabilities without the wise guidance of government officials?"

      Yes. Cochrane is candid that the run started in the shadow banking sector where regulation was least (or non-existent). Lehman Brothers and Bear Stearns were at 30:1 leverage.

    3. Frank:

      Yes; let bad banks fail. Good banks will step into the void. Utilities, unlike banks, are natural monopolies, so a (not very strong) case can be made for bailing them out.


      If a firm fails, the losses can be borne by its shareholders or by the taxpayer. That's an easy choice. The story I find most convincing is that the crisis started because the government was forcing banks to lend to customers who couldn't repay their loans, so regulation caused the crisis.

    4. Mike

      If a financial firm fails the losses can be borne by the creditors.

      You say: "The story I find most convincing is that the crisis started because the government was forcing banks to lend to customers who couldn't repay their loans, so regulation caused the crisis."

      I suspect you find this convincing because it is consistent with your belief system. I know this explanation is widely cited on the "right". The explanation makes no sense to me because:
      1) The crisis went far beyond any group of borrowers the regulators wanted to favor;
      2) Even if the regulators leaned on retail banks to make unsafe loans no one forced the ratings agencies to give the resulting paper (when the mortgages were pooled in MBSs) high ratings or forced Lehman Brothers or the rest to buy the resulting paper. Without the complete breakdown of due diligence in the secondary market there would have been no housing bubble or financial crisis.

      I know the explanation you cite was adopted in the "White Paper" put out by the Romney campaign (it was signed by Mankiw and three others). Given my views on the matter that became a reason for rejecting the so-called "White Paper" as a whole.

  8. 1. If this had been written 10/15 years ago I would have paid attention. Being late to the part is impeaching, goes to the favor of the witness

    2. Here is also why, part

  9. John,

    "Regulating the riskiness of bank assets is a dead end. Instead, fix the run-prone nature of bank liabilities."

    "To stop future crises, the financial system needs to be reformed so that it is not prone to runs."

    "Runs are a pathology of financial contracts, such as bank deposits, that promise investors a fixed amount of money and the right to withdraw that amount at any time."

    The way you fix the run prone nature of any financial contract is:

    1. Make that contract non-marketable (A loan made by a bank is retained by the bank until maturity). Dodd-Frank tries to address this by forcing banks to retain a percentage in any loans that it makes.

    2. Make that contract non-guaranteed (equity in lieu of debt). And that also applies to the federal government.

  10. John,

    "Clearly, overnight debt is the problem. The solution is just as clear: Don't let financial institutions issue run-prone liabilities."

    From a run-prone aspect, the problem is not that the debt is short term. The problem is that a good portion of the debt is simultaneously rolled over.

    The same thing could happen if all of a bank's debt was 15 year debt and it was all coming due next Tuesday.

    This is one reason why equity is important. It is a perpetual security, rather than a fixed duration security.

  11. "In the 19th century, private banks issued currency"
    It continued into the 20th century in Canada. They didn't have a central bank during the Great Depression (or "unit banking" laws), which may have something to do with why they weathered it better. George Selgin & Charles Calomiris have written more about that history which doesn't fit the whiggish story Cochrane tells about the period.

    Here's a contrarian thought: the 1987 stock crash didn't cause any macroeconomic harm because Greenspan ensured there was sufficient liquidity. Housing markets turned south a while before the rest of the economy did, it was when the Fed got spooked by oil prices that inflation might be increasing that markets realized they weren't going to offset any declines in housing & finance. And that's when things fell apart. If our central bank, like Australia's, kept things steady we wouldn't have to worry about any of this. There would be no "too big to fail" because even if the largest institution went under, the worst result would be a lot of inflation that year. And yes, I'm basically just paraphrasing another blogging economist from Chicago.

    1. Well, a lot of inflation results from a drop in real growth since you are talking about keeping NGDP steady. The "worst result" is not the increase in inflation (which is good because it keeps NGDP steady) but rather the decrease in real output. Even with an excellent central bank, which we clearly don't have, it is still important to limit financial crises that hurt the real economy as well as the nominal economy. Managing externalities is part of creating an efficient real economy.

  12. Professor Cochrane,

    Is this really different from Krugman's point (from years ago) that we need to regulate the assets of anything that is effectively a bank? Both of you are saying that we need to prevent risky assets and short-term liquid liabilities from going together. Krugman said that the difficulty is in designing regulations that can determine what is effectively a bank. Your idea has the same difficulty; we have to design regulations that can prevent institutions with risky assets from taking on short-term liquid liabilities without knowing the exact nature of these liabilities ex-ante.

  13. I'm somewhat puzzled by your post. Wouldn't capital requirements a la Hellwig and Admati solve the same problems in pretty much the same way without imposing restrictions on the type of debt that financial institutions issue?
    Regulating what liabilities constitute short term and what type of capital needs to be held when short term liabilities are issued seems to open a whole new can of worms similar to the risk-weights in the Basel accord.
    Would love to hear your thoughts on capital requirements vs regulating maturity!

  14. Excellent article - I couldn't agree more. In the article, you ask the question: "Won't eliminating short-term funding for long-term investments drive up rates for borrowers?" i.e. isn't maturity transformation critical to supporting long-term lending? Let me add a few points to support your claim that it is not.

    In modern developed economies, life insurers and pension funds manage a large pool of funds that explicitly prefers long-term investments to short-term investments. Most banks in fact hedge their long-term interest rate risk with insurers and pension funds so in many cases, the duration risk of long-term assets has already been shifted away from banks onto insurers and pension funds. This means that "structural changes in the economy have drastically reduced and even possibly eliminated the need for society to promote and subsidise maturity transformation" as I argue in this post Questioning the Benefits of Maturity Transformation

  15. I like this intelligent and thoughtful approach of this post.

    But still, it is a call for regulation, and there are usually unintended consequences---and industry end runs around---to regulations.

    In general, I add this: regardless of what is the very best solution to feeble banks, in good government the KISS solution is often preferred, as in Keep It Simple, Stupid.

    I prefer a simple, easy-to-understand solution to feeble banks. Maybe higher reserves, a maximum size, and the sticky deposits is the answer.

    After all, the real purpose of banks to is pool savings and put them to productive use, through lending.

    How much of this overnight banking and derivative whiz-bang stuff is connected to that basic purpose?

    Will our economy suffer if banks go back to the basics?

    1. Dodd-Frank currently stands at ten thousand pages: which is nothing more than a lawyer’s paradise. Finding loopholes in that lot will be easy.

      In contrast, the basic principles of a Cochrane type system has got KISS written all over it. In fact I can set it out in one sentence:

      All bank creditors must be loss absorbers, and must be legally bound to absorb the biggest loss a bank can possibly make, the only exception being creditors who want their bank to return the EXACT SUM they’ve deposited, and those deposits must be backed by 100% safe assets.

    2. OK - there is no such thing as a 100% safe asset. Instead of saying "100% safe assets", I should have said "99.5% safe assets" or perhaps "assets which are in the top 5% safety-wise".

    3. What I was getting at was how do you measure the "safeness" of an asset?

      1. Liquidity - how easily can you convert the asset to currency
      2. Durability - how well does it hold up over time
      3. Supply / Demand - are there constraints on either that would negatively affect the value of the asset over time
      4. Legal protections / limitations - what recourse do you have if the seller of the asset misrepresents what he / she sold to you
      5. Volatility - how much do the other four measures of "safeness" change over time through technological or legal / political change

    4. Ralph,

      What you should realize is that there is no such thing as a 100% safe asset. Currency is the ultimate in liquidity, is fairly durable, is legally protected as a means of exchange, and is non-volatile, but has a tendency to lose purchasing power over time.

      All other assets involve a trade off.

  16. Thanks, Prof. Cochrane. You have correctly identified the cause of and solution to the most important economic & policy issue of our time.

    I understand the dynamics of an N x N matrix of derivative exposures, and how net exposures devolve to gross as the participants start to fail. There is no need, little benefit, and considerable risk, as we have seen, to having moral hazard fund the book. Or the many other variations of it.

    The same phenomena pertains to other domains including the money markets, repo, and deposit insurance. Let the participants who create the risk adequately capitalize it, and let the investors who by the asset wear the risk.

    Your comments about providing low/no risk investment alternatives to retail investors are also appropriate and actionable. Distribution of Treasury notes and bills could easily accommodate the scale. Banks and intermediaries could compete on the basis of whatever risk/return profile they deem workable in the market. Let them compete without a call option on the taxpayer's pockets.

    Well done and an important piece of work.
    Hunter Brown

  17. Runs are a pathology of financial contracts, such as bank deposits, that promise investors a fixed amount of money and the right to withdraw that amount at any time.

    i wouldnt say that (but you know more)..

    if they did not use fractional reserves that were too high, then the pathology would also have a cure (also downstream re-fractioning too. can someone tell me the right term please? i would be grateful)

    if all they did was store my money, and charge me a fee..
    such a pathology could not happen, but we also agree to let them play with it while we aren't using it (as long as they give us a share). no?

    so the pathology is that the money is in play, but the owners of it are not risking it. however, the bank is giving them a micro share in exchange for that guarantee.

    ie. the minute the state sets an amount, all banks go to it. if a good marginal rate would be lower, too bad... higher, also too bad..

    i have not thought of the latter enough, but i have thought of the former.

    its sets up a destructive situation in which none can avoid it (and sadly, those that are publicly owned can avoid it less as its impossible to justify a more prudent course given outcomes i point out below).

    lets say that absent a dictated amount, banks set their own. larger banks have an advantage, in that they can get away with a larger amount. but all in all, each bank sets their own. those going to high, end up failing by overextending and losing. those going too low, end up losing market to those dancing next to the edge.

    and clients would look at that the way we look at risk in stocks, and so that may offset the seeming confusion that would come. [ie. banking would be more like stock, which is your point. no? except that now the depositor can lose but also has a choice of which risk is acceptable. no?]

    but now, in comes the state and it sets an amount.
    now what happens?

    well, the meek banks set it to that amount, the others set to that amount and its simple game theory that plays that script out.

    if set low, the system wont make maximum...
    if set just right, balancing on a razor blade, not in averages across many, goldilocks will be very happy.

    however, the system and those setting favor setting it too high, and riding the outcome for as long as possible hoping that they get to get theirs before the hot potato is in their hands and the music stops.

    a lesser bank cant afford to set a more prudent rate and be wrong (they dont know if that rate they pick is right, or the state rate is right).

    a larger bank can make a lot before that wave hits shore...

    there is no way to stop this train as part of the deal of letting the state do this, would be that the state also offers the banks stability insurances in exchange for not fighting the rate setting. one big happy family, united under state insurance, even if its a promise.

    in shorthand: if your going to tell us a rate, then your going to have to insure that what you pick is not the bad choice as we will not leave the outcome of our banks to your choices without such a guarantee. otherwise we will choose... (say the banks) [not a bad deal if one side is smart and we should point out, its not the political side]

    this whole situation is a gold mine for large banks as they could ride the wave and when the inevitable hits, get a free ride to kick off on the next wave coming in as the agreements kick in (even if just promises).

    so if you go back to them setting it, and you get to choose which reserve you put your money against, you could choose no reserve, which will cost you money for them to hold it for you (rent), smaller fraction(which is much less likely as a system to collapse if sufficient), large fraction (you can lose your money).

    the point is that the fixes created the pathology...

    some of these arrangements and things would always lead to a bad end, just because they set up a self organized system that converges on that in a kind of monetary Abilene paradox.. though thats not quite right, either.

  18. I think the lending regulations were counterproductive because they were so complex. Instead, all lending institutions should be treated the same, with a few simple rules regarding loan requirements, and reserve requirements, no bailouts, and most of all total transparency.

    And BIG penalties for breaking any of these simple rules.

  19. BTW---there is a fad now to say that legislation has a certain number of pages, therefore must be bad legislation.

    Every year, we get about a cumulative 8,000 pages of legislation from the Housed Armed Services Committee, and the relevant House Appropriations Committee, and the pertinent two Senate Committees, on the Department of Defense.

    That's every year!

    So, one could say "there are 8,000 pages of legislation on this year's Defense budget---what an example of boondogglerey!"

    Dodd-Frank has 10,000? Only 2,000 more than what comes up every year on Defense.

    1. OK, but I think the relevant point here is that science respects simple laws which explain a lot: e.g. E=MC2. The above Cochrane banking system can be set out on about one side of a piece of A4 paper, whereas Dodd-Frank is 10,000 bits of paper and fails to solve the TBTF subsidy problem.

      So it’s game, set and match to the Cochrane system. Incidentally a very similar system was set out here:

  20. The problem in shadow banking is not just at the investor side. The money market funds invested in commercial paper. The issuers of commercial paper were seeking lower rates by borrowing short term and avoiding banks (which had reserve requirements). Some of those issuers then turned around and lent the money for longer terms. When shadow banking hit a bump, the issuers could not roll over their paper and had a liquidity crisis. GE, for example, had to borrow substantial sums from Buffet at high effective interest rates and it benefited from US government guarantees of its commercial paper and direct government purchases of its paper.

    One can be a purist and say that the government should not have propped up GE and simply allowed it to go bankrupt. Would that really have been better than government stop gap funding?

    1. In your GE example, GE is acting as a bank, isn’t it? I.e. if a “Cochrane” system were to be effectively implemented, the rules would state that ANY ENTITY, whether it describes itself as a bank or not, has to obey “Cochrane” rules. And if GE had obeyed Cochrane rules it wouldn’t have had a problem.

    2. Ralph

      GE's problem was that they had trouble rolling over commercial paper. Under Professor Cochrane's approach I suppose that amounts to a "run" on GE. There was no government guarantee of GE's commercial paper - the market participants knew they were taking on credit risk when they bought and were free to assess GE's credit worthiness.

      Professor Cochrane's rules would say that GE had to raise all of its money through equity or long term debt structured in a way that there could not be a material roll over risk.

  21. I know this is a post on financial regulation, not monetary policy (who can tell the difference nowadays?).

    What do you think of John Taylor's recent article in the FT saying that Mark Carney should adopt a “rule for all seasons” that "would allow for the possibility that bank rate would be extra low in periods following a stint at the zero lower bound, but by a measurable rules-based amount that depends in a consistent way on the degree to which the desired bank rate has fallen below zero bound."

    Is Taylor's advice impractical because it simply consists of "open-mouthed-operations"?

    Also, this rule sounds a lot like targeting the level of NGDP.

  22. There is an article by Matthew Klein published by Bloomberg which makes similar points to John Cochrane in his WSJ article. See:

    1. "What about the lending side of today's banking industry? There would be lending companies instead -- funded exclusively by equity investors, who consciously choose to put their savings at risk rather than hold them as deposits or other money-like bank liabilities."

      One thing Mr. Klein neglects to mention is the role of the primary dealers. Should equity investors be able to say no to the federal government? The primary dealer facility was set up back in the 1960's to ensure that the federal government can always find buyers for it's debt. If buying that debt puts equity investors at risk, should they be forced to buy?


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.