Thursday, 14 May 2015

Central bank lending and the Great Fire of London

My Forbes post criticising Senator Elizabeth Warren's attempt to restrict Fed liquidity in a crisis has caused something of a storm. It seems that people can't distinguish between providing unlimited liquidity to the financial system and providing unlimited government funding to banks. The first of these is essential. The second is not.

I do not, and never have, supported the second of these. Indeed I was one of the voices soon after the crisis that called for banks to be allowed to fail. But restricting central bank liquidity in a crisis is terribly dangerous.

Consider a fire in a closely-built city - let's call it the City of London. Because the houses are very close together - indeed many of them are interconnected - and are made of wood, fire spreads easily. Do you restrict the amount of water the fire brigade can use to put out the fire, even though there is unlimited water available from the nearby river? Do you tell the fire brigade it must only spray houses where the householder has fire insurance - the rest must be allowed to burn? Do you tell the fire brigade it must stop spraying water after thirty minutes whether or not the fire is out? No, you wouldn’t do any of these. You would regard it as madness. The priority is to stop the fire, not to decide who deserves to keep their house and who does not. That is the job of central bank liquidity. It stops the crisis.

Once the fire is out, then by all means decide who deserves to have their house rebuilt or refurbished by the government, and who does not. And redesign the city so fires can’t spread so easily, change building standards so houses are built of fire resistant materials (stone, brick, marble), tighten fire regulations, educate bakers about the dangers of leaving ovens unattended. 

Similarly, once the financial crisis is over, kill off banks whose capital has fallen too low, do radical surgery on banks that are damaged but could be restored, make extensive changes to the design and regulation of the industry to reduce both the likelihood of future crises and their damaging effects, and educate bankers on the importance of responsible risk management. We didn’t do this properly after the 2008 crisis: we did not kill off badly damaged banks, we did not restructure the rest as we should have done and we have taken far too long to tighten regulations and retrain bankers. On this, I agree with Senator Warren.

But redesigning the city and retraining the ba(n)kers doesn’t mean you can dismantle the fire brigade. Despite our best efforts, fires still happen. So do financial crises. We need to be able to deal with them. And we cannot deal with them if we hobble the central bank.

In my post, I observed that Bagehot's Dictum actually does not apply in a financial crisis. Bagehot's Dictum advises the lender-of-last-resort to "lend freely to solvent institutions, against good collateral, at a penalty rate". This is good advice for central banks in normal times. Individual banks that have good collateral and substantial equity buffers may meet liquidity shortfalls by borrowing from the central bank at penalty rates. Most central banks have this facility available at all times.

But lending of this type is emphatically not appropriate in a crisis, and Bagehot did not so regard it. This is what Bagehot had to say about central bank lending in a financial crisis:

"Lend that reserve most freely in time of apprehension". No restrictions. No "good collateral", "penalty rate" or "solvent institutions". Those are restrictions for liquidity support to individual banks in normal times, not in a crisis.

And why? Because failing to lend freely makes the crisis worse. Half-hearted emergency lending is worse than none at all. Bagehot is uncompromising: "To lend a great deal, and yet not give the public confidence that you will lend sufficiently and effectually, is the worst of all policies; but it is the policy now pursued." Stopping financial crises requires unlimited liquidity, just as ending the Great Fire of London required unlimited water. We would be foolish to turn off the taps.

The Great Fire of London was eventually put out. But the mediaeval City of London was in ruins. In the rebuilding programme, streets were widened and buildings were built of stone, not wood. There has never been another fire so devastating. But there have nonetheless been fires from time to time. We still need our fire brigade.

Related reading:

Cleaning up the mess

Image: Great Fire of London. from Wikimedia Commons


  1. A poetic analogy, but Wren's London was only possible because the fire burnt the city to the ground, despite every effort. I do not think we will see a financial St Paul's built unless it is on a tabula rasa of vested interests. This can never happen when CBs can move privately created assets on and off their balance sheet at will. The CB is not just the fire brigade with a liquidity hose but a man who can throw wood on the fire faster than it can burn. London is already burning.

  2. I am no expert on financial regulation, but I salute the Forbes sub who let through "bat shit insane."

  3. Thanks. So I haven't mistakenly read "Lombard Street". In spite of what, one and another time, endlessly, we are listening the restricted, "ortodox" version.

  4. I used to think that Bagehot didn't require a penalty rate. A while back Stephen Williamson pulled out an old Bagehot quote on me that says otherwise:

  5. I like the rather witty and appropriate analogy of liquidity with water to put out a fire. You ask three very pertinent questions:
    "Do you restrict the amount of water the fire brigade can use to put out the fire, even though there is unlimited water available from the nearby river?
    Do you tell the fire brigade it must only spray houses where the householder has fire insurance - the rest must be allowed to burn?
    Do you tell the fire brigade it must stop spraying water after thirty minutes whether or not the fire is out?"

    To be quite honest, I am not sure that the answers to these questions will all be "No".. Why should I say this, when the obvious answer is No from the point of view of the entire society?

    Well, the first reason would be on the basis that by the time the next Great Fire of London happens, the fire service will have been privatised. We now have two semi-competing fire services for London (in reality, they are oligopolistic but everyone pretends not to notice). Each has a basic subsidy allocated by the government and concluded by contract. As private companies, they need to make a profit. Therefore, they have some secret guidance internally about how they respond to fires.

    The first is: "prioritise buildings which have taken out a special fire policy with us".
    The second is "for an uninsured building, if the fire cannot be put out in 30 minutes, abandon it and claim to have been called out to a more urgent case".

    When the Great Fire breaks out, the firemen worry that they may not be able to attend all of the buildings that are insured (even though they have a contractual oboigation to attend all buildings, paid by public subsidy). Their manager urges them not to worry too much about all fies as the whole city is one fire anyway: the important thing is to keep their paying customers happy and not let their buildings burn down. Through this policy choice, the fire service fails to deal with the fire on a logical basis, and the entire city burns down -- when at least some streets could have been saved by co-ordinated action.

    The reason I raise this future hypothetical issue is that I actually believe our politicians have got themselves into a seriously defective mode of thinking. That mode has neoliberal origins, and it consists of thinking of money as a finite commodity that needs to be conserved. It is no longer a resource allocation mechanism; and actual outcomes of policies do not matter, other than to fake the stats and talk nonsense to the electorate at election time.

    They are quite capable of destroying anything with this stupidity and delusional thinking -- including the City of London.

  6. I agree with Frances’s basic point, namely that if the system for fighting fires is to pour water on fires, then it’s silly to limit the amount of water (central bank money) , given that water (central bank money) is free.

    However, a problem arises out of all this (thus this point of mine is a bit off topic). The problem is that banks in practice are not charged a realistic rate for being rescued. That is, central bank loans to commercial banks in a crisis are not charged at a realistic rate (as I think Frances points out in her Forbes article). Thus the whole rescue operation equals a subsidy of private banks, and subsidies do not make economic sense.

    Moreover, it is unrealistic to think that governments will ever send banks a bill for being rescued. Lobbying by banks, assisted by wads of cash in brown envelopes, will ensure that doesn’t happen. So the cycle will repeat itself. Or as Geithner put it in his book “Stress Tests; Reflections on a financial crisis”: “If you reward pyromaniacs, you get fires”.

    My preferred solution is to bump up bank capital ratios so far that it’s plain unrealistic to think they’ll ever go insolvent. Martin Wolf and Anat Admati propose 30%. I wouldn’t object to 100%.

  7. >Frances
    could you please comment on the Bassel III definition of tier 1 and tier 2 capital.
    The wording in that document seems very awkward.

  8. Frances; read you post on Elisabeth Warren while traveling, so late response. Agree with your gut reaction on her proposal, but think you miss an aspect behind her thinking, i.e. that as currently structured, the big banks are taking us for a ride on their repo/OTC roller-coaster. So unless structural measures are taken to limit their recless specualtion on deposit insured funding, there is merit to her attempt to limit LLR - even in a crisis. I agree its not Bagehot, but may be part of a grander strategy to force change in the current TBTF setup. If you want further academic arguments for this view, see my article "Shadow banking - policy challenges for central banks" on Researchgate. Best regards Thorvald G Moe

  9. But lending *too* freely and without conditions is the same as transforming the central bank from a lender of last restor to a equity donor of first resort, as the Fed/Treasury (and to some extent the BoE) have been in 2007-2008, where they donated at no interest, against toxic securties, to deeply insolvent banks, lots of equity, with no strings attached.

    That our blogger's reading of Bagehot is very tendentious is for me very clear from the original text which does not need to be selectively quoted as our blogger has done, but can be read with full context here:

    So to me it is very clear that Bagehot is writing always about liquidity panics, and that the 3 rules are well motivated by him, including the two essential conditions:

    * «That these loans should only be made at a very high rate of interest. This will operate as a heavy fine on unreasonable timidity, and will prevent the greatest number of applications by persons who do not require it.» This is to ensure that the lender of last resort is indeed approached as last resort, not as a first resort.

    * «That at this rate these advances should be made on all good banking securities, and as largely as the public ask for them. .... That in a panic the bank, or banks, holding the ultimate reserve should refuse bad bills or bad securities will not make the panic really worse; the 'unsound' people are a feeble minority, and they are afraid even to look frightened for fear their unsoundness may be detected. The great majority, the majority to be protected, are the 'sound' people, the people who have good security to offer.» is meant to ensure that the insolvent don't get bailed out, because it is pointless (except for the bailed out managers).

    All this in the context of a gold based money system, and that's where the "foreign" element comes from, because exports of gold can cause a domestic liquidity crisis.

    I am very disappointed that our blogger seems to support despite protesting the opposite the "equity donor of first resort" theory of central bank intervention.