Late on Friday afternoon, Stuff posted an op-ed piece calling for the introduction of a (funded) deposit insurance scheme in New Zealand.  It was written by Geof Mortlock, a former colleague of mine at the Reserve Bank, who has spent most of his career on banking risk issues, including having been heavily involved in the handling of the failure, and resulting statutory management, of DFC.

As the IMF recently reported, all European countries (advanced or emerging) and all advanced economies have deposit insurance, with the exception of San Marino, Israel and New Zealand.   An increasing number of people have been calling for our politicians to rethink New Zealand’s stance in opposition to deposit insurance.   I wrote about the issue myself just a couple of months ago, in response to some new material from the Reserve Bank which continues to oppose deposit insurance.

Different people emphasise different arguments in making the case for New Zealand to adopt a deposit insurance scheme.  Geof lists four arguments in his article

• providing small depositors with certainty that they are protected from losses up to a clearly defined amount;
• providing depositors with prompt access to their protected deposits in a bank failure;
• reducing the risk of depositor runs and resultant instability in the banking system;
• reducing the political pressure on government to bail-out banks in distress – deposit insurance would actually make Open Bank Resolution more politically realistic.

Of these, I emphasise the fourth.  I’m not convinced that there is a compelling public policy interest in protecting depositors, small or otherwise (many schemes cover deposits of $250000, sometimes per depositor per bank).   There are plenty of other bad things in life that we don’t protect people from (the economic consequences of) –  job losses, fluctuating house values, road accidents, bad marriages and so on.  The ultimate state safety net is the welfare system, which provides baseline levels of income support.  Should “deposits” or “money” be different?  I’m not sure I can see good economic arguments why (although there are good reasons why in the market debt and equity instruments co-exist, and debt instruments generally require less day-to-day monitoring by the holders of those instruments).

And there is a  variety of ways of providing depositors with prompt access to funds following a bank failure.  A bailout is one of them.  OBR is another.  And deposit insurance, in and of itself, doesn’t ensure prompt access to funds; it just ensures that the insured amount is fully protected (minus any co-payment, or deductible).

I’m also not persuaded that deposit insurance reduces instability in the banking system.  International historical evidence has been that in many or most cases,  depositors can distinguish, broadly speaking, the weaker banks from the stronger banks in deciding whether to run (I would argue that the UK experience with Northern Rock is one recent observation in support of that proposition).  And anything that weakens, albeit marginally, market discipline (in this case, by reducing the incentive on deposits to monitor risk and respond accordingly) can’t be likely to contribute to greater stability in banking systems.  Deposit guarantees for South Canterbury Finance only postponed, and probably worsened, the eventual day of reckoning.

But I find the political economy arguments for deposit insurance (at least in respect of large banks) compelling.  I outlined the case more fully in my earlier post.  If we don’t want governments bailing out all the creditors of a failing bank (large and small, domestic and foreign), we need to build institutions that recognize the pressures that drive bailouts and take account of that political economy.  It is futile –  and probably costly in the long run  –  to simply pretend that those pressures don’t exist.  In its recent published material, the Reserve Bank again just ignores these arguments, but they know them.  In fact, I found a quote from Toby Fiennes, their head of banking supervision, who correctly observed a few years ago that

some form of depositor protection arrangement may make it easier for the government of the day to impose a resolution such as OBR that does not involve taxpayer support – in effect the political “noise” from depositor voters is dealt with,” said Fiennes

As I’ve noted previously, in the last thirty years:

• The BNZ was bailed out by the government
• Finance company (and bank) deposits were guaranteed by the government
• AMI was bailed out by the government

And it is not as if governments in other advanced countries have been any more willing to allow retail depositors to lose money.  Most of our major banks are Australian-owned, and Australia has relatively recently adopted a deposit insurance scheme, reinforcing the longstanding statutory preferential claim Australian depositors have over the assets of Australian banks.  In the event of the failure of an Australian-owned banking group, why should we suppose voters here will tolerate losing large proportions of their deposits when they see their counterparts in Australia –  in the same banking group –  protected?    The Australian government  –  in the lead in resolving such a failure – is unlikely to be receptive to such a stance either, and if they can’t force us to protect our depositors, there are lots of strands to the trans-Tasman relationship, and ways of exerting pressure if our government did choose to make a stand.

A deposit insurance scheme heightens the chances of being able to use OBR, and thus to impose losses on wholesale creditors, many of whom will be foreign.

But it doesn’t guarantee it.   I noticed that Geof’s article included this paragraph

Since the global financial crisis, many countries, including New Zealand, have developed policies that enable even large bank failures to be handled in ways that minimise the prospect of a taxpayer bail-out, by forcing shareholders, then creditors (including depositors), to absorb losses.

I am less optimistic than Geof here.  Countries have been moving in the right direction, of trying to establish resolution mechanisms that would enable bank failures to occur without taxpayer bailouts, and in which large and wholesale creditors would face direct losses.  But none of these mechanisms has really been tested yet.  I’m yet to be convinced that the authorities in Britain or the US would be any more ready to let one of their major banks fail, with creditors bearing losses, than they were in 2008.

I’m reminded of a story Alan Bollard once told us about his time as Secretary to the Treasury. Faced with the prospect of Air New Zealand failing, the Prime Minister of the time asked if Treasury could guarantee that if Air New Zealand failed the koru would be still be flying the following week.  Unable to offer any such assurance, the government decided on a bailout.  Faced with the prospect of the failure of one of our larger banks, the Prime Minister might reasonably ask the Reserve Bank and Treasury whether they could assure him that, if he went ahead and allowed OBR to be imposed, other New Zealand banks and borrowers would still be able to tap the international markets the next week.  At best, officials could surely only offer an equivocal answer.  Bailouts remain likely for any major institution (especially as, in our case, resolution of any major bank involves two governments).

I hope I am too pessimistic in respect of wholesale creditors.  And we shouldn’t simply give up because there is a risk that governments might blanch and bail out the entire institution.  But the best chance of governments being willing to impose losses on larger creditors in the event of failure, is to recognize that the pressures to bailout retail depositors will be overwhelming, and to establish institutions that internalize the cost of that (overwhelmingly probable) choice.  A moderately well-run deposit insurance scheme does that, by imposing a levy on banks for the insurance offered to their depositors.

As Geof notes, he has changed his stance on deposit insurance.   Looking around the web, I stumbled on  “Deposit insurance: Should New Zealand adopt it and what role does it play in a bank failure” a 2005 paper, by Geof and one of his colleagues (now a senior manager at TSB) on deposit insurance, which has been released under the OIA.   It is a useful summary of some of the counter-arguments.

One of the issues it covers is the question  of whether, instead of adopting deposit insurance, we could achieve much the same outcome by using the de minimis provisions in the OBR scheme.  Under those provisions (built into the prepositioned software) deposits up to a certain designated amount can be fully protected, and not subject to the haircut.

As I’ve noted previously, this provision might be useful if it was only a few hundred dollars –  effectively, say, protecting the modest bank balance of a very low income earner or superannuitant, who needed each dollar of a week’s income to survive.  It might be tidier to have all these small balances protected than to have all these people turning to food banks. It might also keep down the ongoing administrative costs of the statutory management, by keeping many very small depositors out of the net   But the de minimis provisions are not a serious substitute for deposit insurance, on the sort of scale that it is typically offered at.  Any preference for very small depositors comes at the expense of the rest of the creditors.  That might be tolerable for small balances in a large institutions with lots of funding streams.    It is much less so in a bank that is largely retail funded, and quickly becomes impossible in such banks once the level of protection rises above basic weekly subsistence levels.  And, of course, no one knows what the de minimis level is, so the risk (facing other creditors) cannot be properly priced.  By contrast, a deposit insurance scheme can be set, at priced, at pre-specified credible levels.

If we were to establish a deposit insurance scheme in New Zealand, there are many operational details to work through.  One, of course, is the pricing regime.   In his article, Geof notes that

‘the cost is small –  no more than a small fraction of a percentage point per annum on each dollar of bank deposit”

I’m less convinced that that is the correct answer.  There is a market price for insuring against the risk of bank failure, and associated losses on debt instrument.  That is what credit default swaps are for.  Historically, in the decade or so prior to the crisis, premia on Australian bank CDSs were very low.  We used them in setting the price for the deposit guarantee scheme in 2008, and from memory they had averaged under 10 basis points.  That isn’t so any longer,  and for the last few years the average premium has been more like 100 basis points (fluctuating with global risk sentiment) –  nicely illustrated here. Bank supervisors would, no doubt, tell us that these premia far overstate the risk of loss –  and I would probably agree with them (and certainly did in 2008, when we used historical pricing) –  but it is the market price of insurance.  Is there a good reason why government deposit insurance funds should charge less?

It is time to adopt a deposit insurance scheme in New Zealand –  not, in my view, because people necessarily should be insulated against losses, but because governments will do so anyway.  In the face of such overwhelming pressures (and track record here and abroad) we are best to build institutions that help limit and manage that risk, and which charge people for the protection that governments are offering them, while making it clearer and more credible that others –  outside that net –  will be expected to bear losses in the event of a bank failure.