Friday 12 April 2013

Open Banking Resolution and Deposit Insurance

New Zealand's Open Banking Resolution system in case of bank failure makes bank failure less likely, and so makes depositor bailouts less likely, but doesn't eliminate the risk that the government might bail out depositors.

Matt Nolan has made this point more than a few times and has urged that we reinstate deposit insurance and require banks to pay for this implicit insurance; I hope instead that the RBNZ and government will find ways of making more credible that the depositors may be burned. If depositors bear risk, then they have to pay a bit more attention to where they put their money and, in doing so, encourage the banks to be more sensible. Further, the simple existence of OBR with a set sequence in case of failure should reduce the panic that pressures governments into bailouts. But if we can't make it credible, then Nolan may be right. 

Toby Fiennes, RBNZ Head of Prudential Supervision, laid it out in a speech yesterday. 
So OBR is about keeping a bank open and providing the government with real alternatives to liquidation or full taxpayer bailout – both of which may be totally unpalatable. It facilitates a rapid and orderly resolution of a bank failure. It does so without changing our basic legal framework around ranking of creditors in a wind-up or insolvency. In particular:
– It does not change the fact that depositors’ and other creditors’ funds are at risk. It is a well-established legal principle that people stand to lose money if a business that owes them money cannot meet all its obligations. Banks are the same as any other business in this regard.
– It does not change the ranking of creditors. Shareholders will be the first to lose their investment. Once shareholder funds are exhausted, subordinated creditors bear losses, followed by all other unsecured creditors on a pari passu basis, meaning that those with an equal legal claim get equal treatment. This is the same as in a liquidation.
Two features of OBR make it particularly well-suited to the principles for crisis resolution I outlined earlier. They are:
– Its flexibility. OBR deals with the immediate crisis, including payments and liquidity issues around failing banks, without closing off long-term solutions.
– It reduces moral hazard. Bank shareholders, management and investors know that in the case of bank failure the authorities have a viable option that would put their stakes at risk. The mere presence of OBR in the toolkit will impact expectations of government support.
Fiennes continues:
It is important to emphasise that OBR and deposit insurance are not in any way alternatives. OBR is also applicable in a world where we have deposit insurance as in one where we don't. Deposit insurance usually involves the establishment of an insurance fund, to which banks contribute. There are many different variations on, and within, that basic framework, but OBR can cope with all of them. For example, if there’s an insurance fund, the fund itself could stand as a creditor in the OBR. This is how the FDIC (the US deposit insurer) is treated in failed US banks.
New Zealand does not currently have any form of deposit insurance or deposit guarantee. This position was confirmed by the Minister of Finance in 2011 ( There are three reasons for this position:
– Deposit insurance is not always effective in preventing bank runs by retail depositors. UK-based Northern Rock suffered a classic retail run in 2007, despite a deposit insurance scheme being in place.
– Deposit insurance is hard to price accurately and fairly; and brings with it difficult boundary issues. Should it be just for banks – as is currently the case for OBR – or should it also include finance companies, building societies and credit unions? How would we ensure that the least risky banks do not end up subsidising the more risky?
– Deposit insurance will increase moral hazard, making the banks more susceptible to failure, which brings with it the need for more, costly regulation.He also notes that a large bank's failure could easily overwhelm a deposit insurance scheme and that the costs of deposit insurance in encouraging bad bank behaviour outweigh the benefits of avoiding depositor losses. So even in Nolan's world where the government cannot credibly commit not to bailout, the inefficiency caused by the certainty of a bailout (rather than the weighted expectation of one) can be large enough to make the scheme undesirable, even if the government does wind up bailing out the banks.

As I'm not a banking guy, here are a few things I particularly do not know much about and would materially affect the ability of OBR to reduce bailout risk:
  • The proportion of secured versus unsecured bank liabilities. If there isn't much that can be burned through before hitting depositor assets, then the haircut landing on depositors would be greater and so too would be the political pressure for a bailout. 
  • How easy it would be for banks to shift away from unsecured liabilities, knowing that risks put on depositors make bailouts more likely.
  • Bank ownership structures. The NZ banks are owned by big Oz banks that have an implicit Oz government guarantee that's actually worked into their Fitch credit ratings (as best I understand things). Suppose something bad happens and one of the Oz banks and its NZ subsidiary both are going down. Do we know what assets are really with the NZ subsidiary and which are with the Oz parent? Will there be incentive to restructure the equity profile depending on which side of the ditch is looking stronger? Can they tunnel equity back to the Oz parent in the NZ subsidiary is looking dodgy and so leave less to burn through in OBR? RBNZ runs some strong prudential regulation; I'm sure they're on top of this stuff. I just don't know anything about it.
The greater the proportion of the burden borne by bank shareholders and unsecured creditors other than depositors, the lower I'd expect to be the pressure for bailouts. If deposit holders are to take a ten percent haircut if bad things happen, I doubt that they could muster nearly as much bailout sympathy as depositors facing a fifty percent haircut. 

Fiennes concludes:
But there’s always the remote possibility that a bank does get into trouble, at some point. If that happens there are no simple solutions. It will be messy, people will lose money and how it is dealt with will depend on circumstances at the time.
OBR is a tool that gives government an additional option to taxpayer bailout or liquidation. It is not the only option that will be available on the day. Its mere existence provides important incentives for bank shareholders and management to minimise the risk of failure.
New Zealand does not currently have deposit insurance, for reasons that are more to do with moral hazard and the sheer difficulties of defining boundaries and pricing than consumer protection. We believe it is better to keep the risk of failure very low, including through a strong regulatory framework, than to build structures that can distort incentives and behaviour.
If, however, deposit insurance were to be introduced, it could easily be accommodated within our toolkit of OBR and other crisis measures. It is not a case of choosing between one or the other – they have different objectives and can work alongside one another if need be.
We have to distinguish between best-case deposit insurance and achievable deposit insurance. How well you think the government can price the bank's underlying riskiness and how well the political system can ensure fair actuarial rates rather than implicit subsidisation of riskier banks will matter in your evaluation of deposit insurance's desirability. But note too that the world in which the government is really good at assessing bank's risks is also the world in which prudential regulation likely works well. I lean towards staying out of deposit insurance, but I'm not hugely confident in my point estimate here - as I noted earlier, I'm not a banking guy and really don't have enough information to be very confident.

Update: I failed in copying the second bit of Fiennes's speech the first go-round. My computer was trying very hard to crash while I was posting.


  1. Secured liabilities can be very useful for cheaply reducing bank risk. For example, covered bonds shift unmanageable/undiversifiable duration risk from banks to foreign pension/insurance firms, while NZ banks retain the relatively more manageable credit risk. Secured lending also frees up the interbank repo market quite dramatically. These are both probably risk-reducing ex ante for depositors.

    Although I'm sure there are examples as you suggest where secured loans are basically stripping the good assets from the bank, leaving the weaker assets for the depositors/unsecured creditors.

    There are regulations prohibiting NZ banks from lending to their Australian parents. Now, one way to get around that is to build up a net debt position with the parent via short term borrowing and repos, which could be discontinued at the first sign of trouble. The Core Funding Ratio prevents this behaviour to some extent.

  2. Eric advises us on Banks
    “If depositors bear risk, then they have to pay a bit more
    attention to where they put their money and, in doing so, encourage the banks
    to be more sensible.”
    Well. thanks Eric that’s what I thought myself and I said so to the ASB,
    I said to them you have to be more sensible, and well they just laughed at me; … , so I went across to the ANZ and they said you don’t have the credentials to set up bank
    account we can’t be sure of your address,:.,, and so I said to the woman, why don’t you come over and sleep in my room for a week or so and see if you can remember where I live, And then Wetpac, I said can you guarantee my deposit, and they looked slyly at each
    other and then said ‘ thanks so much for coming to visit us ‘

    So I thought I would get to the bottom of all this .,, and there it was
    From Matt Nolan, a Pigoulasn theory go ther for didillusionment
    7.2 Choosing the Pigouvian tax rate Comparing (18) with condition (7) shows that
    implementing the planner’s choice of c1 requires³1 +ηθ´(1−q)μ∗α= (1−q)μ∗α+qμ∗β.14
    Kocherlakota (2010) also advocates using taxes to offset the ncentive
    distortions associated with bailouts and suggests a market-based mechanism for
    determining the appropriate tax rate.Suppose the policy maker setsη =θqμ∗β(1−q)μ∗α≡

    So I rang up Winston, said what you think Pal. and he is laughing and he is laughing,

    I rang up Winston and
    I said what do you think, and he is laughing and laughing