What risks should the state protect people from?

Later yesterday morning, before major international markets opened for the week, the US authorities announced two steps in response to the failure of SVB Bank

  • first, depositors not covered by the FDIC (amounts in excess of US$250000) would in fact be completely covered, with the costs to be covered by levies (taxes) on other US banks,
  • second, a new Fed lending facility was set up, backed by the US Treasury, under which banks could borrow at market rate against securities that for these purposes would be valued at face value not market value.   For most longer-term securities issued in the last decade, market value is currently less than face value.

Legislative changes after 2008/09 were supposed to make bailouts much harder and less likely, but at the first real test – in respect of one failed bank that was 16th largest in the US (and another a bit smaller still) – there were significant elements of a bailout anyway. In respect of a bank that mostly had large deposits (this wasn’t an entity where just a few people had a bit more than $250K on deposit), all depositors were made whole. Most of the commentary suggests that had the assets been liquidated and depositors and other creditors been paid out what was left they’d probably have got back more than 90 cents in the dollar, and the FDIC resolution procedures could readily have allowed those larger depositors access to some portion of their money upfront (this incidentally is/was similar in this respect to the way the Reserve Bank’s Open Bank Resolution model was envisaged as working). The precedent value of this action suggests that in future any depositors at even a moderate-sized US bank are likely to be made whole (“what do you mean you aren’t going to bail out depositors in a failed bank in my district when you bailed out those Silicon Valley tech companies?”)

The main focus of yesterday’s announcement seems to have been to snuff out the risk of further bank runs. Signalling to depositors that they won’t lose their money (no matter how large and otherwise sophisticated they are) is one way of doing that. Another is providing ready access to liquidity for other banks, to signal that such banks would have no problem paying any requests for accelerated withdrawals that did arise. It bears some resemblance to a classic lender of last resort (a function of central banks that few have too much problem with in principle), except that whereas Bagehot counselled that central banks should lend readily on good collateral at a high price, yesterday’s announcement really only met the first element of that test. Much of the collateral has a market value less than the valuation being used to secure these loans (that isn’t good collateral, even if the bond itself is issued by the US government), and the loans are simply at normal market prices. It is, in effect, subsidised lending by the state.

Perhaps some might be inclined to pardon less than ideal policy responses when things have to be done in a rush. And I’m sure the weekend was pretty fraught for many relevant officials and politicians. But the US is a big country with huge bureaucracies and ample time and resources to have robustly war-gamed how failures and potential failures of significant-sized institutions would and should be handled, including thinking hard about the lessons from such exercises for future incentives. If state insurance of all deposits made sense, it made sense a couple of years ago, not just today. But that wasn’t the model adopted. It is hard to believe that lending on collateral using face rather than market value, at normal market rates, would ever make a lot of sense (at least outside some deep and severe systemic crisis where wholesale securities markets had become deeply dysfunctional).

Various people point out that the moral hazard is not complete. After all, SVB’s management will have lost their jobs (but not presumably past salaries and bonuses), shareholders will have lost their money (but not past dividends), and other creditors including any bondholders will not be made whole. But the benefits of yesterday’s bailout will also flow to the management, shareholders, and other creditors of other banks with somewhat similar (albeit typically less extreme) business models. And if the quid pro quo for the heightened moral hazard is supposed to be heightened regulation and supervisory intensity, how much confidence should people really have in that given the evident failure of supervisors and regulators in this case? Perhaps exemplary bank regulation might act as an adequate counter, but in the real world of US banking/regulatory politics?

I’m not one of those opposed to all deposit insurance. Well before the current government decided to introduce deposit insurance to New Zealand I was arguing for it as a second-best response because absent a limited deposit insurance system it seemed all but certain that in a stress event for any major bank (and perhaps some less major ones) in New Zealand, all creditors would end up being bailed out, and no one would have paid the Crown anything for the insurance that was being provided. Even in conjunction with Open Bank Resolution as an option in the toolkit there is still a high risk of a full bailout of creditors of the larger banks – partly because of the pressure that will almost certainly come from the Australian government – and at the margins actions like yesterday’s from the US authorities only increase that likelihood. Perhaps in truth, our deposit insurance scheme will end up only ever being practically relevant should banks like TSB or Heartland be close to failure (in terms of relative size comparisons they are our SVB).

There are people who believe that it is practically desirable, or at least unproblematic, to provide full deposit insurance. My stance is much closer to that of Peter Conti-Brown (professor of financial regulation and author of a stimulating book on various Fed governance issues) than to the former chair of the US Council of Economic Advisers.

And I’m not uninfluenced by having observed, and been involved with, our own NZ retail deposit guarantee scheme in 2008 where once guarantees were in place money flooded towards entities (notably South Canterbury Finance) that offered slightly higher yields. It isn’t a perfect comparison, since prudential supervision of finance companies wasn’t a thing at the time, but it is a useful cautionary experience nonetheless. And my perspective on bank runs is that generally they happen too late and too rarely, rather than seeing them as typically some random or fundamentally unwarranted event. The threat of a run is an important element in market discipline.

But I guess my wider caution is around the question that is the title of this post? What economic risks should the state be offering full protection against?

I can see a reasonable case for retail transactions balances being protected, even guaranteed. In that vein of course, people can choose to use Reserve Bank banknotes. More seriously, one reason why I have always been inclined to favour allowing the general public access to individual Reserve Bank settlement accounts (in modern parlance a CBDC) is precisely to provide such a credit risk-free option (even as, as I noted in my CBDC submission, I do not believe there would be a great deal of demand for such a product). But even then, an overnight Reserve Bank deposit account for transactions purposes might be free of credit risk and market risk, but it is hardly free of inflation risk (any more than a commercial bank deposit).

But if you or I have half a million dollars on deposit with a bank (let alone if an investment fund has $10m or $100m), there is no obvious public interest in the state guaranteeing that you will never lose the nominal value of your deposit. No doubt it would be tough to be substantially hair cut if your bank happened to fail and the assets came well short of covering 100 cents in the dollar but (a) you did have choices (under the proposed NZ system protection will be limited to $100K, so you have a reasonable option of spreading your deposit across five banks and securing full protection), and (b) there are so many other economic risks in life against which the state provides at most limited protection (all while providing a basic welfare system where entitlement in case of need is near universal, in the case of age universal).

I’ve already mentioned inflation. Out of the blue, quite in breach of their published targets, central banks in the last couple of years have delivered a quite unexpected 10 per cent boost to the price level. That is pure windfall gain if you have borrowed money in a conventional nominal loan, and pure windfall loss (not likely to be, or able to be, recouped) to those holding conventional nominal financial assets. Sensibly enough in macro terms, we don’t have price level targets, so no effort will be made to reverse these transfers, but for many the losses are real. For someone with $500000 in the bank, the real loss of purchasing power might be similar to many retail bank failure haircuts. It has surprised me a little – and is useful data after decades of low stable inflation – that more is not made of this arbitrary state set of wealth transfers.

House prices are falling at present in much of New Zealand. For many people – those of us without mortgages, and with a natural position long one (and only one) house – it doesn’t make much difference to anything. But there are plenty of people for whom it does – whether the owners of investment properties, or those who borrowed heavily at the peak of the most recent boom. If you had bought a house in Wellington 2 years ago rather than now you are perhaps 20 per cent worse off for that choice. And there is no state compensation scheme.

Share prices- and market values of Kiwisaver accounts – go up and down and no one proposes compensation (even champions of a capital gains tax are rarely keen on full offsetting of losses, which itself would still only offer partial compensation).

We have a system of accident compensation in New Zealand. I generally support it. It pays income-related compensation for loss of earnings, but only partially (80 per cent) and only up to a threshold (maximum liable income about $130000 per annum). Beyond that even for those risks, you are on your own (albeit with private options). And for many disabling conditions the state provides no specific insurance or compensation at all beyond the basic welfare system (and the health system itself of course). EQC cover is also capped.

Same goes for human capital or the fortunes of particular towns/regions. Or the real economic costs of a failed marriage. Many of these potential economic losses run far beyond the plausible scale of what an individual might have exposure to in a bank failure. And yet while we often sympathise individually, and support having in place a welfare system for basic support, we don’t as a society collectively attempt to compensate individuals for such losses. For most of such losses, no modern state – no matter how socialist in its reach – has ever really attempted to. We have debates at the margin – eg the government’s preferred social insurance scheme – but relative to many of the potential losses such instruments don’t really go very far.

So I struggle to see a strong principled case for treating larger bank depositors more generously. Yes, sometimes bank failures can appear to come from the blue, but they rarely do. Diversification is usually an option (and typically much more readily than you can, say, diversify your human capital or housing or relationship exposures), and so is private insurance. I suspect that few would really disagree as a matter of principle, and much just comes down to “its easier not to let any depositor lose their money” and associated fear of (the minority of ill-founded) bank runs, or “it is too hard to envisage our politicians even being willing to let big banks fail at all”, and living with the third-best consequences of that resigned stance. It isn’t a good place to be – especially when the beneficiaries of these resigned third-best policies will often be among the wealthier parts of society – although even as we try to change the politics, or create better options, there is no point pretending the politics are other than as they are.

UPDATE: Meant to include a mention of the NZ government’s choice – incredibly, on the advice of both the RB and The Treasury – to bail out all policyholders in AMI when that insurer failed after the Christchurch earthquakes. Not only were there no risks of contagious runs – insurance just isn’t like banking – but even if you thought there was a case for bailing out the less-wealthy policyholders, how could it possibly have been a wise, or priority, use of public money to be bailing out people with insurance on a high-end house who, at worst with a severe haircut, might have had to lower their housing sights.

I mean, in this country – as no doubt most – you can be charged by the state for serious offences and a couple of years later acquitted, or wrongfully imprisoned for multiple years and still find it a major hurdle to get serious economic compensation. (To be clear, I do favour erring on the generous side when mistakes are made when the coercive powers of the state are exercised in such ways.)