I noted yesterday that the Reserve Bank had also released some papers on the proposed new LVR restrictions on Thursday.  When that release was pointed out to me yesterday, the Bank’s website suggested that the papers had been released in response to an OIA request.  The papers still appear in the obscure corner of the Bank’s website where (in a welcome development) they have started publishing some of the responses they make to OIA requests.

But when I checked again this morning, the table now says of the latest release:

Date of Response	Subject matter
25 June 2015	Loan-to-value ratio (LVR) restrictions, proactively released, jointly with the Treasury

Go through to the detailed page, and it suggests that the release is partly pro-active and partly a response to an OIA request.

This is information relating to loan-to-value ratio (LVR) restrictions that has been released proactively and in response to requests for information under the Official Information Act 1982 (the Act).

I’m a little confused.  But if there genuinely is a pro-active component to the Bank’s release then I welcome it, even if (say) they may just have released one additional paper to provide context for a few that were covered by the OIA.

It still leaves a little bit of a puzzle about the Treasury’s (entirely pro-active) choice to release.  Sometimes documents requested of one organisation cross-reference material generated in other organisations, but that does not appear to be the case here.  Perhaps the OIA request to the Reserve Bank included material the Bank held, even if it did not generate it.  If so, no doubt the Bank held copies of at least some of the Treasury papers?    But having been on the receiving end of numerous OIA extensions from the Bank (and recently one from Treasury), when documents written little more than 20 working days ago are pro-actively released, it has the feel of a genuinely deliberate timing choice.

But enough of the bureaucratic process stuff.  What I had intended to write about was the content of the Bank documents.  They released six, one of which (the 17 Feb one) I had previously seen.

Date	Released on 25 June 2015
19/5/2015	Memo to Minister on draft consultation paper on LVR policy (PDF 818KB)
30/4/2015	Memo to Minister on estimated impact of changes to LVR policy (PDF 1.36MB)
24/4/2015	Memo to Minister on potential adjustments to LVR ratio policy (PDF 2.67MB)
21/4/2015	Memo to MFC on Proposed changes to the LVR policy (PDF 350KB)
2/4/2015	Memo to MFC on revisiting the case for regional targeting of marco-prudential policy (PDF 134KB)
17/2/2015	Memo to MFC on effectiveness of a tighter investor LVR limit (PDF 99KB)

We don’t have much context for this release.  In particular, we don’t know the scope of any OIA request the Bank may have been responding to, but if these papers are intended to reflect the analysis the Bank undertook in developing the proposed control on investor lending (and the Auckland-specific nature of the new policy), they are surprisingly short and weak.  Recall that the Bank is,  implicitly, saying the banks and borrowers are so risky and irresponsible that not one single (practical) cent can safely be lent by banks to Auckland residential property investors on LVRs over 70 per cent without jeopardising the soundness of the financial system.  That is a pretty ambitious claim.  It is not supported.

I have been critical of the Bank for not making a stronger case for its proposed controls.  The one comfort I suppose that we can take from these papers is that they are not hiding anything from us.  But if this is all there is – the extent of their engagement with the law, the economics, the stress tests, the uncertainty –  it is even less surprising that The Treasury was also not convinced that a compelling case had been made for the new controls.  Unfortunately, the Reserve Bank also continues to repeat to the Minister of Finance its claims that lending to investors is generally materially riskier than other housing lending.  An attendee at the LEANZ seminar the other night told the audience that he had gone through all the references the Bank has previously invoked in support of its claim, and had found that they simply did not say what the Bank claimed they were saying.  If that assessment is correct, it is pretty concerning, and should be so to the Board and the Minister –  charged with holding the Governor to account on our behalf.

A process issue struck me in reading the papers.  In discussing the possible new policy, the 21 April paper lists a possible timeline.  It suggests that a consultation period should end in “late June” and “Release final policy position and revised conditions of registration” in “mid-July”.  Allowing perhaps 10 working days to read, analyse, and reflect on submissions, write up recommendations to the Governor, write-up a response to submissions, and finalise the new conditions of registration themselves does not suggest that the Bank had in mind a very open process of consultation.  Actual timing slipped somewhat, as the consultative document was not released until early June, but we should hope that they take a little more than 10 working days to work through all the issues likely to be raised in submissions.

As I noted yesterday, the cause of serious scrutiny of the Governor’s proposals (and of open government more generally) would be advanced if the Bank were to publish on its website, as they are received or at least on the closing date, all the submissions they receive.  They are, after all, public, official, information.

Two more points on the stress testing issue.

I’ve mentioned a couple of times that someone who was at the Finance and Expenditure Committee hearing on the day of the Financial Stability Report had told me that the Governor deliberately refused to answer a question about the stress tests, and the implications of those results for assessments of the stability of New Zealand’s financial system.

I’m told that the transcript of that hearing is now on the public record, so here is the relevant question and answer.  The questioner is National MP Chris Bishop, who is deputy chair of the committee:

Bishop             Thanks, Governor. I’m just interested in teasing out what the specific risks to financial stability are for Auckland house prices, because the banking sector has rising capital and liquidity buffers; they exceed the minimums. The banks came through the stress testing pretty well last year. Credit growth is relatively restrained, and as a percentage of GDP, credit is below where it was in 2008-09. So given all that, what are the specific risks to financial stability—which is what we’re discussing here today—from the rising Auckland prices?

Wheeler           I think they’re very substantial. I mean, if you look at mortgage commitments, you quoted a number that credit flowing to the housing sector was low. It is on a net basis, but if you look at mortgage commitments, they’re growing at around 20 percent. House prices in Auckland are growing at around 17 percent. They’ve been growing in the rest of the country over the last year at around 2 percent. If you look at house prices to disposable income in Auckland, that ratio is 7.4 percent. But the rest of the country is 4.2 percent. If you look at rental yields in Auckland, they’re at historic lows, which suggests that there’s a lot of people basically investing for capital gain, whereas the rental yields across the country are basically where they have been for the last 10 years.
If you look at the median house price in Auckland, it’s up 60 percent since 2008. We had the highest rate of house price inflation in the OECD from 2003 to 2008. So the median house price in Auckland is now 60 percent above that. If you look at the Demographia survey that was done last year, we were 14^th out of 370 housing markets around the world, in terms of affordability. If you look at the survey that was done by ANZ Bank in terms of investor expectations, late last year, basically, investors in Auckland were forecasting that house prices would increase by 75 percent over the next 5 years. Now, our job is to try and keep inflation, on average, at around 2 percent per annum. So that’s just a phenomenal increase in house prices that are anticipated, and that would just drive house price to disposal income ratios up at a huge rate.

So there’s a whole range of reasons why there are major, I think, financial stability risks around Auckland.

The Governor raised a number of interesting issues, and possible areas of risk, but did not respond to any of Bishop’s points or questions.  Now sometimes MPs at select committees can ask questions just to be on record as having asked them, or to make partisan points.  And so there is an art in how public servants respond to such questions.  But Bishop’s questions and points don’t look as though they fit either of those categories.  They seem to be entirely reasonable questions, drawing on the Bank’s own factual material, and yet the Governor simply chose not to engage or respond.  That doesn’t seem very wise, or very accountable.

And now, back to some geeky stuff.  In discussing the stress tests this morning I mentioned the issue of what size house price fall one might reasonably assume if the stress tests were re-run today.  But as someone pointed out, neither I nor the Bank touched on the other factor that is critical in assessing the likelihood of large loan losses, and that is what happens to unemployment.

By and large, falls in house prices alone do not result in large losses for banks.  Between with-recourse lending and a general desire to avoid moving (which is costly and disruptive), owner-occupiers don’t tend to default if they can service their debts.  Banks can, typically, foreclose if the borrower has negative equity, but are unlikely to do so if the debt is being serviced.  Much the same is likely to go for lending for investment properties –  if rents are high enough to cover the debt service, banks aren’t likely to foreclose.  Foreclosing (itself expensive) crystallises a loss, which might otherwise never happen.

Similarly, high unemployment alone doesn’t typically lead to large loan losses on residential lending.  Some individuals will end up losing their houses, but if they have to sell up (or be sold up) the sale price will usually cover most or all of the debt outstanding.

What gets really nasty is the scenario in which house prices fall a long way and unemployment goes up a lot (and stays high for a while).  In that scenario, many people can’t service their debts (even if the OCR is cut) and if they have to sell, in many cases the proceeds won’t be large enough to cover the debts.

That is the scenario the Reserve Bank’s stress tests (rightly) focused on.  It is the true test of the quality of the residential mortgage loan book.

The Reserve Bank tells us that in the stress test they assumed that the unemployment rate “peaks at just over 13 per cent”.  The unemployment rate at present is 5.8 per cent, and the “natural rate” is probably around 5 per cent.  The modern low was 3.4 per cent.  So 13 per cent is a long way away.  In normal times it would take around an 8 percentage point increase in the unemployment rate to get to “just over 13 per cent”.

I was curious how unusual 13 per cent unemployment rates were, so I downloaded the OECD data as far back as it goes.  In most cases, that is just over 30 years (but we also know that in most OECD countries the earlier decades were decades of pretty full employment).

Here is the chart of the highest unemployment rates on record for each of the 34 OECD countries.  Only 13 of the 34 countries has had an unemployment of 12.5 per cent or above in more than 30 years.

I took a look at those countries.  First, I wanted to understand how much the unemployment rate had increased by in each of those country episodes.  If the NAIRU in one country had been 10 per cent (perhaps reflecting very restrictive labour market regulation), an increase in the unemployment rate to 13 per cent would have much different implications than if that country’s NAIRU was 6 per cent.

Of the 13 countries whose unemployment rates had peaked at over 12.5 per cent, in one case that peak was the first observation in the database (so I couldn’t tell where the unemployment rate had risen from).  Six of the other 12 had had increases in their unemployment rates of 8 percentage points or more (from the previous cyclical low to the measured all-time peak).  Thus, for example, Greece’s unemployment rate peaked at 27.8 per cent last year, but had been as low as 7.5 per cent in 2008.

But the other factor I looked at was the exchange rate regime these countries had been using when their unemployment rate rose to 13 per cent or more.  In only two of the 13 cases had the exchange rate been floating, and in neither of those cases had the unemployment rates increased by anything like 8 percentage points.

Why do floating exchange rates matter?  Simply because they act as a buffer when the economy is hit by severe shocks.  Greece, Spain, and Ireland have had very high unemployment rates (and large increases in those rates) in the last few years because they have had no independent national monetary policy, and no ability for their national nominal exchange rates to depreciate.  The US, the UK, and Iceland, on the other hand, each having had a nasty financial crisis, had nothing like the extent of those increases in unemployment.

Adjustment is a great deal harder, and more costly, without the additional flexibility the floating exchange rate provides.  But New Zealand has had a floating exchange rate for 30 years now, and when the economy has been in serious difficulties the exchange rate has fallen a long way.  No one really doubts that the same would happen again if, say, there was a serious recession here that involved the OCR being cut to, or near, zero.

I’m not suggesting that the unemployment rate could not possibly rise by 8 percentage points here.  From 1987 to 1991, the unemployment rate did rise by 7 percentage points, to around 11 per cent, even with a floating exchange rate.  But that was a pretty stringent test:

• Huge amounts of labour-shedding from public and private sector structural reform
• A serious domestic financial crisis
• And the transitional costs of both lowering inflation markedly, and closing the fiscal deficit, at the same time.

My point is simply to highlight that the Reserve Bank’s stress tests were very stringent, using an increase in the unemployment rate larger than any seen in any floating exchange rate country in at least 30 years.  It is right that stress tests are stringent (the point is to test whether the system is robust to pretty extreme shocks)  but these ones certainly were.  And yet not a single one of big banks lost money in a single year.  That might seem a bit optimistic –  it did to me when I first saw the results –  but they are the Reserve Bank’s own numbers.

And so, again, we are left wondering where is the evidence for the Governor’s latest regulatory initiative?

On 13 May, the Reserve Bank released its latest Financial Stability Report. As part of that release, the Governor announced that he intended to impose a ban on (highish LVR) lending secured on Auckland residential investment properties. The Bank indicated that a consultation document on the proposal would be released in “late May”.

The consultation document finally appeared yesterday. I have some fairly extensive comments on the contents of the document, and on the process. I will be making a submission, and will publish that here in due course.

But today I wanted to focus on just one aspect: the place of the stress tests the Bank undertook, with APRA, last year.

As a reminder, the results of the stress tests were reported in the November 2014 FSR (details on pages 9-11 here). The Reserve Bank was, apparently, then very happy with the resilience of the banks (individually and as a system – the latter rather than the former being the required statutory focus). As they noted:

The Reserve Bank’s emphasis tends to be on ensuring that banks have sufficient capital to absorb credit losses before mitigating actions are taken into account. The results of this stress test are reassuring, as they suggest that New Zealand banks would remain resilient, even in the face of a very severe macroeconomic downturn.

As I noted on 13 May, it was somewhat surprising then to find no reference to these stress test results in the latest FSR, even as the Governor was moving to impose very restrictive and intrusive new controls. I suggested that perhaps the Governor did not believe the stress tests. But if so, he owed us an explanation for why, especially in view of the fairly unconditionally positive coverage of the stress test results which he had signed off on in finalising the November report.

I was further puzzled when someone who had attended the Finance and Expenditure Committee hearing on the FSR told me that when the Governor was questioned about the stress test results and their implications for conclusions about the soundness of the financial system, he had simply avoided answering that element of the question.

However, the Reserve Bank then referred to the stress test results in responding to questions to the Herald’s personal finance columnist, Mary Holm. I covered those comments earlier.

On 16 May, there was this extract:

What does the RB think about the possibility of a property plunge. “Whether property prices could drop by half from today’s values is purely speculative,” she says. “Nevertheless, a 50 per cent drop matches some of the more severely affected economies in the global financial crisis such as Ireland.”
So they’re not ruling it out. But would such a drop cause banks to “collapse”? “The short answer is no, we do not believe so,” she says.
“The Reserve Bank conducts regular bank stress tests in collaboration with the Australian Prudential Regulation Authority. The most recent one was last year, and the results of it are featured in the November 2014 Financial Stability Report, pages 9 to 11, on our website.
“This stress-test exercise featured two imagined adverse economic scenarios over five years, one of which involved a sharp slowdown in economic growth in China, which triggered a severe double-dip recession in New Zealand. Among the impacts were house prices declining by 40 per cent nationally, with a more pronounced fall in Auckland – similar to your reader’s worst case scenario.”
So how would our banks fare?
“The Reserve Bank was generally satisfied with how the banks managed their way through the impacts of these scenarios, and we are comfortable that the New Zealand financial system is currently sound and stable, and capable of withstanding a major adverse event.”
Note that present continuous tense in the final sentence: we are comfortable “that the New Zealand financial system is…capable of withstanding a major adverse event”.

That was reassuring, but it did appear inconsistent with proposals for heavy-handed new controls on the other.

And then the following Saturday, we got some more comment from another Bank spokesperson.

“We repeat our comments from last week that the Reserve Bank was generally satisfied with how banks managed their way through the impacts of two adverse economic scenarios in the 2014 bank stress tests, which included a scenario similar to what your reader describes.
“We are comfortable that the New Zealand financial system is capable of withstanding a major adverse event, such as a collapse by up to 50 per cent of the Auckland housing market.”
These words, given to the Herald after the publication of the FSR and published less than two weeks ago, stated quite explicitly that the Bank is “comfortable that the New Zealand financial system is capable of withstanding a major adverse event, such as a collapse by up to 50 per cent of the Auckland housing market”.

And there I half-expected the matter to rest. Since policy development around LVRs seemed to be on a (rather distant) parallel track to stress-testing analysis, I half-expected the consultation document to avoid any mention of the stress-testing results at all, the Bank having reaffirmed only 10 days or so ago the resilience of the system.

But in yesterday’s consultation document, they do address briefly the stress-testing issue. Here is what they say:

The Reserve Bank, in conjunction with the Australian Prudential Regulation Authority, ran stress tests of the New Zealand banking system during 2014. These stress tests featured a significant housing market downturn, concentrated in the Auckland region, as well as a generalised economic downturn. While banks reported generally robust results in these tests, capital ratios fell to within 1 percent of minimum requirements for the system as a whole. Since the scenarios for this test were finalised in early 2014, Auckland house prices have increased by a further 18 percent. Further, the share of lending going to Auckland is increasing, and a greater share of this lending is going to investors. The Reserve Bank’s assessment is that stress test results would be worse if the exercise was repeated now.

First, they note that “capital ratios” in the stress test fell to within 1 per cent [percentage point] of minimum requirements for the system as a whole. But here is how those results were described in the November FSR.

Common equity Tier 1 (CET1) capital ratios declined by around 3 percentage points to a trough of just under 8 percent in each scenario, but remained well above the regulatory minimum of 4.5 percent (figure A3). Banks are also required to maintain a 2.5 percent conservation buffer above all minimum regulatory capital requirements, or else face restrictions on dividends. On average the banking system fell within this buffer ratio in both scenarios, due to total capital ratios falling close to minimum requirements (figure A4). Average buffer ratios reached a low of 1 percent in both scenarios. As a result, some banks would have been faced with restrictions on their ability to issue dividends. The intention of the buffer ratio is to provide a layer of capital that can readily absorb losses during a period of severe stress without undermining the ongoing viability of the bank. Given the severity of the scenarios, capital falling within buffer ratios was an expected outcome.

In other words, the Bank seemed pretty comfortable. As they should have been. A banking system that can withstand a very severe asset market correction and adverse macroeconomic shock with, at worst, “some banks would have been faced with restrictions on their ability to issue dividends”, while all were always above minimum required ratios (themselves calculated using risk weights that are demanding by international standards) is an extremely strong banking system. Plenty of banks abroad raised additional capital during 2008/09 without ever coming close to failure, but not one of the big New Zealand banks ever needed to raise any new capital in these stress test scenarios. But that it is what one would expect when capital buffers are large, and credit to GDP ratios have been going nowhere for seven or eight years.

As the Bank also notes, it is not even that the loan losses in the scenario were large enough to cut into the dollar level of capital banks held: the deterioration in capital ratios arises only because the risk weights on bank loan books rise in the course of the severe downturn. Not a single bank had less capital at the end of the severe stress scenario than at the beginning.

CET1 (tier one, common equity) is the focus of the Bank’s capital framework.  Here is the chart from the stress test results.

The Bank also rightly notes that the scenarios for the stress tests were finalised in early 2014 and things have changed since then. Of course, they have not changed materially in the two weeks since the Bank publically reaffirmed the resilience of the system, but let’s put that detail to one side for the moment.

Unfortunately, neither the Bank nor I can easily tell what this set of facts means for the results if the stress tests were to be run today. Auckland house prices have certainly increased a lot in the last year, the share of lending going to Auckland has increased, and a greater share of this (Auckland) lending is going to investors.  But other things have changed too – among other things, nominal incomes are higher than they were then, and interest rates look to be lower for longer than the earlier scenario envisaged. Those owing the large accumulated stock of debt (a stock that continues to worry the Bank) have had more time, and more income, to strengthen their own ability to handle adverse shocks.

Perhaps the much higher level of Auckland house prices now suggests that any future stress test scenario should use an even larger fall than the 50 per cent used last year. But 50 per cent is about as large a fall in house prices as has been seen, on any sustained basis, anywhere. If a 50 per cent fall is still a reasonable scenario from the new higher level (as I’d argue it is, given that no one has a good basis for knowing the “equilibrium” level of prices in the presence of ongoing regulatory constraints and policy-fuelled population growth), then all else equal there would be fewer loan losses for banks in an updated test not more.

It is certainly true that a higher share of residential lending is now taking place in Auckland (although I suspect the share of the stock can’t have changed much in one year). In the stress test scenario that would, mechanically, mean a higher level of losses (since the scenario assumed a larger fall in house prices in Auckland than elsewhere). And of the lending in Auckland a little more has been going to investors. But note that final point carefully – as Figure 3 in their consultation document illustrates, the proportion of house sales being made to “multiple property owners” (the proxy for investors) is now no higher than the average in the series since 2008.

The investor property share is higher than previously in Auckland but (a) the difference from the rest of New Zealand is small, and (b) the greater role of investors is partly due to the earlier LVR restriction, which will have forced some first home buyers out of the market, to be replaced by investors. Moreover, in the consultation document the Bank indicates that the earlier LVR restriction has improved the overall “resilience” of the financial system. Even if one believed that lending to investors was riskier than lending to owner-occupiers, all other characteristics of the loan held equal (and the Bank has still not yet persuasively made that case), the overall implications of any changes in portfolio structures over the last year look likely to be small.

Stress tests run today would certainly produce different results to stress tests run a year ago.  But housing loan losses have hardly ever been at the heart of a banking crisis, the stock of debt is rising only slowly, and the 2014 results were so strong that it is difficult to believe that the Bank’s analysts are seriously wanting us to believe that stress tests run today would suggest that the financial system was now imperilled. Indeed, I noted the careful way their claim was worded – they suggest that the results today would be “worse”, but not “materially” or “substantially” worse. Given how strong the 2014 results were – as the Bank itself told us – a slight deterioration, in a business so fraught with uncertainty, should not really be a matter of particular concern.  Recall that in last year’s tests –  an exercise to which the Bank and APRA devoted a lot of resource –  not a single bank had a single year of losses.    It might sound too good to be true, but it is the Bank’s own work, and “no losses” leaves rather large room for them to be wrong without the soundness of the system being in jeopardy.

Unfortunately, the Bank seems to be all over the place on this issue. It is difficult not to feel some sympathy for the staff who are required to dream up rationalisations, and explain away past robust results, to provide some support for the Governor’s strong pre-determined views.  But if they really do believe that stress tests run today would result in a materially greater threat to the financial system then (a) they should probably have steps in train already to raise required levels of bank capital, and (b) it might have been helpful if they made the case in the Financial Stability Report.