On peripheral share markets, and the mess that can still lie behind them

As the wild gyrations around the sharp decline in Chinese share prices dominate the business news headlines,  the chart of the Shanghai share price index was reminding me of another chart, from our own history.

The historical data are hard to come by so I’ve had to use this chart, which covers a rather long period, from a good Reserve Bank Bulletin article about past New Zealand financial crises (mostly, the 1890s and the 1980s). In China’s case, a 150 per cent increase in share prices in a year or so, most of which has now been given back in only a few months (of course, it is perhaps worth remembering that there was an even bigger boom, to even higher peaks, in 2006/07).   In New Zealand’s case, the boom in real prices was even larger – if it took a slightly longer period of time – and the bust was complete.

NZ share prices 80s

There is lots of discussion at present – often from US-centric people – about how the Chinese share market is somewhat peripheral to the economy. And, in some senses, no doubt it is. The share market has always been relatively peripheral to the New Zealand economy too, and market capitalisation as a share of GDP has typically been low. Between farms, farmer co-ops, wholly foreign-owned companies, and state-owned companies, most economic activity simply wasn’t represented by listed companies. Take state-owned companies as an example. In 1986/87, the government still owned several banks, insurance companies, the post and telephone system, the one TV company, the railways, the airlines, the largest petroleum company, and the steel mill. There were no New Zealand listed banks, and no vehicles for direct equity exposures to our largest export industries (sheep and dairy products).  Forestry was probably the only major export industry with material listed exposure.

And yet the market went crazy, and when it burst it was the start of some very tough years for New Zealand. The whole post-liberalisation mania has not been adequately documented (unlike, say, the Nordics booms and busts of much the same time), but it was characterised by much more than just crazy share prices for the minority of people with any material direct exposure to the market. These things rarely happen in isolation. We had a massive credit boom – private sector credit growing at 30 per cent per annum – and a massive commercial property boom, skewing real resources into places that proved not really to offer economic returns. The absurdity of some of what went on during that period was captured in the suggestion of one (brief) high flyer, that New Zealand had a comparative advantage in takeovers. Few bankers had any experience in disciplined credit risk analysis in a liberalised market economy – and nor did their shareholders (whether government or Australian private). And, in any case, bonuses weren’t being paid to people who stayed clear of the boom for long. Highly leveraged investment companies were all the rage – often the principal asset was overvalued shares in other investment companies.

As one illustration of what went on, here is the commercial property picture.  A 3 percentage point of GDP fall is huge, and of course the boom-times share of GDP has never been regained, suggestive of how overblown that market had become.

non-res 80s

As they often are, central bankers were slow to recognise what was going on. The section I managed wrote a paper to the Minister of Finance a few weeks before the crash arguing (as I recall) that the strength of the share market was a pointer to the underlying strength of the economy (and hence a reason to tighten monetary policy). A week or two after the crash, my boss and I were visited by a firm of economists we were often dismissive of – they urged us to recognise that a savage commercial property and banking shake-out was about to get underway. We were polite but pretty dismissive.

Those are the sorts of episodes that leave financial crises in their wake, and which often have nasty and prolonged real economic consequences. Some of us at the Reserve Bank might have initially been a little sceptical, but the liquidity stresses soon became very apparent. Within weeks we had acquiesced in a sharp fall in short-term interest rates (we didn’t directly set them then), which eventually totalled 400-500 basis points.  The exchange rate came down as well. And it still took years for the economy to really recover.

No two situations are ever quite the same. New Zealand was trying to vanquish double-digit inflation, a situation few countries these days find themselves in. But ill-disciplined credit booms, of the sort we had in the 1980s, and of the sort the Chinese have had – on a far larger scale and with fewer market disciplines – are often enormously damaging. Often there is an equity market mania dimension to these things, but the mania is often the epiphenomenon rather than the main event; symptoms of something deeper going on. And the timing will never be quite precise – we had a mania in 1986 around a company (investment and finance) whose shareholders, in their private capacity, were backing what appeared for a time likely to be a successful tilt at the yachting America’s Cup. Of itself, it probably did little or no real economic harm, but it was symptom of the excesses, and of the undisciplined nature of the times.

So count me as a little sceptical of the notion that because not many Chinese hold shares, we don’t need to worry about what is going on there. As here, it is a matter of looking to the phenomena behind the headlines – in their case, several years of one the largest, least-disciplined domestic credit and investment booms in history.

Oh, and there are simply no parallels between the 1987 New Zealand position – which brought down several major financial institutions, including our largest bank – and the New Zealand position now. I wrote about that (lack of) parallel a few months ago.

Does anyone believe core inflation is about to rise?

The next Reserve Bank Monetary Policy Statement is scheduled for 10 September. Unless they’ve materially changed the timetable, this will be the week in which the Governor, his staff and advisers will be in long meetings, working their way through the data from the last few months and the draft economic forecasts. The forecasts are finalised to be not-inconsistent with the Governor’s preliminary OCR decision (which is usually made at this stage of the process, but can be revised very late in the piece).

One set of data was released only yesterday (although the Bank will have had it last week). That is the results of the Reserve Bank’s two expectations surveys. The household survey, of around 1000 people, asks only about CPI and house price inflation, while the other one, the Reserve Bank’s Survey of Expectations asks quite a wide range of questions about financial and macroeconomic variables. I’m not sure how many respondents they now have, but it used to be around 70. Respondents were a mix of people from business, the labour market, the financial sector, and economists (I’m now one of them). We used to treat this survey as the responses of an “informed” group, even if not many were experts in each of the areas on which they were questioned. The surveys have both been going for quite a long time now, and complement the quite different sorts of information available from business and consumer confidence survey. For some years, we ran a staff version of the Survey of Expectations, which occasionally raised interesting question when it showed up material differences between staff and respondents’ expectations.

Of all the questions in the two surveys, only one tends to get any media coverage (or even much market economist coverage). That is the question in the Survey of Expectations about what respondents expect the annual inflation rate to be in two years time. It gets focus because it is the variable that the Reserve Bank pays most attention to, and relatedly because it appears in many of the Reserve Bank’s formal models of the inflation process. The two year ahead measure gets the focus because it should largely “look through” most of the short-term fluctuations in inflation (oil prices ups and downs, tax changes, and so on, which can have a large effect on short-term inflation, but shouldn’t normally affect medium-term inflation). And, empirically, the two year measure seemed to do reasonably well (together with some measure of excess demand) in equations explaining (core) inflation; or at least it did until the last few years.

expecs and core inflation

You may recall that in the July OCR review, and the Governor’s subsequent speech, the Bank affirmed its view that although current headline CPI inflation is still very low (0.3 per cent in the year to June), it would soon be back to around the 2 per cent middle of the target range.   They talked in terms of inflation being back near the midpoint by early-mid next year.  I and others have previously made the point that, even if so, this would be just a one-off lift in the inflation rate, as prices adjusted to a lower exchange rate, and would not represent a lift in core inflation pressures.

Respondents to the Survey of Expectations seemed only partly convinced by even this element of the Bank’s story. Year-ahead inflation expectations picked up from 1.32 per cent to 1.46 per cent, and two year ahead expectations also rose slightly, up from 1.85 per cent to 1.94 per cent. Among households, year ahead expectations rose, while five year ahead expectations were unchanged.  The Reserve Bank is likely to take some comfort from these results, reinforcing the Governor’s bias (at least at the time of his speech) that the OCR didn’t need to be cut much more. But I think they would be wrong to take any comfort from them. As I dug through the numbers, I was staggered at how weak they were (and as I say that as one who is more pessimistic than the mean response on most, but not all, questions).

Let’s deal with households first. The numerical expectations that respondents reported did increase, but the survey also has a question – which, from memory, gets a better response rate –  as to whether the respondent thinks the annual inflation rate will rise, fall, or stay the same over the coming year. Households have always – every quarter since 1995 – reported a net expectation that inflation will rise over the coming year, but this quarter’s response was one of the lowest ever.

households higher

The only time a smaller proportion of people expected inflation to increase over the coming year was at the end of 2008, when the actual annual inflation rate was peaking at just over 5 per cent ($150 oil prices and all that). Quite reasonably, and in the middle of a severe recession, not many expected the inflation rate to increase from there. The Reserve Bank apparently expects inflation to rise from 0.3 per cent to something around 2 per cent by the middle of next year. That is a really large increase to publicly project – and it got a lot of coverage – but households don’t believe it. They don’t expect inflation to rise much at all. They might be wrong, of course, but they don’t seem to believe the Bank at present.

What about the more informed group, the respondents to the Survey of Expectations? Here we have the luxury of getting a bit more of sense of how respondents (collectively) are thinking, since we get answers to a whole variety of macro and financial questions (although perhaps the Bank might think of adding a house price inflation question to this survey).

Respondents are asked a number of inflation questions. They are asked for their expectations for each of the next two quarters (in this case, September and December) and for their year-ahead expectation. That means we can back out an implied expectation for the second six months (in this case the first half of next year). There isn’t much sign that respondents expect inflation to rise. They are expecting inflation of 0.67 per cent for the second six months of this year (still one of the weakest six month ahead expectations ever), and 0.78 per cent in the first six months of next year.  In other words, there is hardly any pick-up in near-term inflation even following one of the sharpest three month falls in the exchange rate we’ve seen since the  float thirty year ago. It is exactly the same extent of pick-up, from the first half of the year ahead period to the second half, as the average expected in the surveys done over the previous two years.

expcs 6 to 12 mths aheads

Like the Reserve Bank’s projections, the Survey of Expectations respondents take account of any changes in monetary policy they expect over the survey horizon.   We get a direct steer on that from the questions about the expected 90 day bank bill rate in a year’s time.   In this survey, that expectation fell sharply, down by 92 basis points  – and the only time there has been a larger fall than that was at the height on the financial crisis in 2008/09.  Respondents expect that the OCR next June will be around 2.5 per cent – probably not a lot different than the Governor had in mind in his speech. Even with these big cuts in interest rates, respondents don’t see the rebound in expected inflation.

The survey also asks respondents about their perceptions of monetary conditions (actual and expected). “Monetary conditions” isn’t defined, and respondents may be influenced by factors such as interest rates, exchange rates, share prices, and credit conditions.   Respondents think conditions at present are (quite considerably) on the easier side of neutral, but what caught my eye is that this is the first quarter in the history of the survey when respondents have both thought current conditions were loose now, and expected them to get looser still over the coming year.

expec mon cond

Survey respondents’ GDP growth expectations have also fallen   Expected growth rates of around 2.3 per cent for the year ahead and two years ahead would barely even match potential growth over that period, suggesting little prospect of an increase in core inflation. Consistent with this story about potential, unemployment rate expectations have risen notably, and are expected to stay at around 5.8 per cent for the next two years. If anything, medium-term expectations are getting more pessimistic – for the first time since 2009, the unemployment rate two years hence is not expected to be below that in one year’s time.

U expecs

And finally, the one question where respondents were even more pessimistic than I was. Wage inflation expectations picked up coming out of the recession back in 2009/10 (and recall there are more business-affiliated respondents in this survey – it wasn’t just household aspirations), but have been going sideways, or falling, for the last few years. The survey asks respondents about expected wage inflation one and two years ahead. In big booms one expects to see the two year ahead expectation below the one year ahead one – intense pressure on wages now, but they will abate as the market equilibrates. And in deep recessions, one expects the reverse: very weak wage pressures in the near-term but some recovery in the medium-term as the market equilibrates. You can see that pattern in the chart, both pre and post 2008.

wage inflation expecs

But what is striking is what has happened in the last year or so.   As actual wage inflation has been falling, expectations of future wage inflation have been falling, but the two year ahead expectations that have been falling faster than the one year ahead expectations. In the more than 20 years these questions have been run, medium-term wage inflation expectations have been lower only once – that in the middle of a recession in 1998, when the inflation target was 0.5 percentage points lower than it is now.   Respondents to this survey seem to have just given up on the idea of any labour market pressures in the foreseeable future.   There is barely any real wage inflation expected:  nominal wage inflation – before allowing for productivity growth – is expected to only barely above 2 per cent.

The respondents to these surveys might be quite wrong. But the sorts of people who fill in the Survey of Expectations are the sorts of people the Reserve Bank should be wanting to convince. Some of them are real decision-makers. Others might be, in Nigel Lawson’s memorable phrase, “teenage scribblers”, but they are the sort of people who pay more attention than most to this stuff. And they do not see any inflation pressures out there, at present or in prospect over the next couple of years.

Which brings us back towards monetary policy. Even after the first two cuts, the OCR is still around 50 basis points higher than it was at the end of 2013, even though inflation expectations – business and household – are probably 50 basis points lower than they were then. In other words, the Reserve Bank has substantially raised real New Zealand policy interest rates, over a period when markets tell us that real New Zealand long-term rates – the market’s view of what will be required over the next 15-20 years – have been falling. Perhaps that would have made sense if core inflation had been getting away on the Bank and it had needed to act fast to bring it under control. In fact, core inflation has just kept on being very low, surprising the Bank, and well below the target (and accountability) midpoint.

After the Governor’s speech, I posed the question of how the Bank hoped to get core inflation back to around 2 per cent, given its current very cautious approach to lowering interest rates.  Add in the lack of any confidence the Survey of Expectations respondents appear to feel, and the question is redoubled now.  It seems most unlikely that another 25 or 50 basis points of OCR cuts will be anything like enough –  cuts of that sort are already factored in to these survey responses,  And all this in a survey completed 2-3 weeks ago.  I suspect a survey done today might be a little more pessimistic still.

Grant Spencer speaking on the housing market again

Reserve Bank Deputy Governor, Grant Spencer, gave another speech on housing yesterday. I was pretty critical of his previous effort (here and here).

The latest Spencer speech has some interesting material in it.   But too much of what he is saying doesn’t seem to be based on any substantial research or analysis. A good example is around tax. Apparently the Bank now regards”tax policy as an essential part of the solution, given the historical tax-advantaged status of investor housing”. But if it is an essential part of the solution why are they content with such modest changes? And what has happened to previous Bank analysis suggesting capital gains taxes would have little sustained effect on house prices?   Grant Spencer has neither presented, nor referenced, any analysis that supports either limb of his statement.     Indeed, previous Reserve Bank work, done by someone who now works at the heart of the Bank’s regulatory interventions, showed that to the extent that the tax system favoured housing, the greatest biases (by a considerable margin) were in favour of the unleveraged owner-occupiers.    That work was done in 2008, and since then the tax situation of investor property owners has become relatively less favourable, because of the depreciation changes in 2010.

And yet there is not a mention in the speech of the awkward issue of tax on imputed rents on owner-occupied houses[1]. As I noted yesterday, the Bank seems to have adopted a disconcerting style of endorsing whatever measures the government of the day is happy with, and staying quiet on other aspects of policy that might directly affect house prices (eg first-home buyer subsidies, large scale active immigration programme, arguments about stamp duties for non-resident buyers, and the non taxation of imputed rents).   That sort of pandering has become too common in government departments, but shouldn’t be the standard adopted by an independent Reserve Bank.

There is lots in the speech I could comment on. But I want to focus mainly on one of my perennial issues, the stress tests undertaken by the Reserve Bank and APRA last year, and reported in the November Financial Stability Report. As regular readers will know, faced with a pretty severe test:

  • real GDP falling 4 per cent,
  • house prices falling 40 per cent (and 50 per cent in Auckland) and
  • the unemployment rate rising (by more than it has in any floating exchange rate post-war country) to 13 per cent.

banks came through largely unscathed. Loan loss expenses peaked at around 1.4 per cent of banks’ assets (compared to a peak of around 0.8 per cent in 2009). That seemed like the basis of a pretty sound financial system. Don’t just take my word for it: in the FSR the Reserve Bank itself observed “the results of this stress test arereassuring, as they suggest that New Zealand banks would remain resilient, even in the face of a very severe macroeconomic downturn.”

The Reserve Bank Act allows the Bank to use its regulatory powers to “promote the soundness and efficiency of the financial system”. How, I have asked, could further highly intrusive restrictions be warranted when the system was so sound, especially as such restrictions have inevitable efficiency costs? For quite a while, we got nothing in response from the Governor and his staff. The Governor simply avoided answering a direct question on the issue at Parliament’s Finance and Expenditure Committee – even though the Bank has often argued that the Governor’s appearances at FEC are a key part of the accountability framework.

The Reserve Bank finally addressed the issue in the consultative document, released on 3 June, on the proposed investor property finance restrictions. This was what they had to 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.

We don’t know what other submitters made of this argument, but in my submission I argued that this was both a weak and flawed claim.  Weak, in that there is no claim that the results would be “materially:, “significantly” or “substantially” worse.  And flawed in that higher asset prices would, all else equal, provide a larger equity buffer in the event of a subsequent fall.

In its response to submissions, released last Friday, the Bank went some way to acknowledging this point

It is true that rising house prices do not immediately increase the risks of losses in a stress test. Indeed any given percentage fall in house prices will leave house price levels higher in absolute terms if house prices have risen further prior to the downturn (so someone borrowing years prior to the downturn may still have substantial equity). The Reserve Bank is mindful, however, that gross housing credit originations are substantial (in the order of 30% of the outstanding stock of housing credit each year). So elevated levels of house prices tend to lead fairly quickly to higher levels of borrowing and debt to income ratios for many borrowers. Additionally, if house prices rise further relative to fundamentals they are likely to fall further in a downturn.

But again this is misleading.   The statistic that 30 per cent of the stock of housing credit is newly originated each year tells us nothing about risk. A borrower shifting his or her (otherwise unchanged) debt from one bank to another will be captured in the Reserve Bank’s gross originations data, but that shift does not change the risk in the banking system. Overall, debt is growing very sluggishly, and the Bank has still not given us a single historical example where a banking system has run into crisis when for the previous several years the stock of the debt had been growing no faster than nominal GDP.   In principle, too, a higher peak of prices might suggest a larger fall, as the Bank says, but they already allowed for a 40 per cent fall, and a 50 per cent fall in Auckland, and I’m not aware of any historical case in which house prices have fallen much more than 50 per cent.

So far, so unconvincing.   But in his speech yesterday, Grant Spencer took a new and interesting tack on stress tests.  Here is what he had to say

There is a point to clarify here around the stress testing that the Reserve Bank conducts on the banking system as part of our prudential oversight function. Sometimes commentators incorrectly interpret the aim of macro-prudential policy as preventing bank insolvency. From a macro-prudential perspective, we may wish to bolster bank balance sheets beyond the point of avoiding insolvency. Stress tests help to inform our assessment of the adequacy of capital and liquidity buffers held by the banks. However, they are only one of the tools contributing to that assessment. In a downturn, banks will typically become risk averse and start to slow credit expansion in order to reduce the risk of breaching capital and liquidity ratios. In a severe downturn, faced with a rise in impaired loans and provisions, banks may start to contract credit which can quickly exacerbate the economic downturn. In this way, a financial downturn can have severe consequences for macro-financial stability well before the solvency of banks becomes threatened.

In 2014, the four largest New Zealand banks completed a stress testing exercise that featured a 40 percent decline in house prices in conjunction with a severe recession and rising unemployment. While this test suggested that banks would maintain capital ratios above minimum requirements, banks reported that they would need to cut credit exposures by around 10 percent (the equivalent of around $30 billion) in order to restore capital buffers. Deleveraging of that nature would accentuate macroeconomic weakness, leading to greater declines in asset markets and larger loan losses for the banks. Such second round effects are not reflected in the stress test results. A key goal of macro-prudential policy is to ensure that the banking system has sufficient resilience to avoid such contractionary behaviour in a downturn.

I suspect that Grant has my comments in view here. Actually, I had long assumed that the Reserve Bank did not aim to prevent all bank insolvencies – the official line, after all, has long been “we don’t run a zero failure regime”. Prudential policy has avowedly been aimed to reduce the probability of an institution failing, but not to prevent all failures. That seems to me like the right public policy approach.

But I have also assumed that the aim of prudential measures (call them “macro-prudential” if you like, but there is no distinction in the Act) was to promote the maintenance of a sound and efficient financial system. Why? Because that is what the Reserve Bank Act, passed by Parliament, says the Bank is to use it powers to do.   It is not clear that the Bank has any statutory mandate for actions motivated by a “wish to bolster bank balance sheets beyond the point of avoiding insolvency”, and especially not when there is apparently no regard being paid at all to the efficiency of the financial system.

Anyway, Spencer goes on to argue that the real purpose of the controls – whatever the Act says – is to avoid banks contracting credit in a severe downturn.   But actually in a severe downturn – and especially one associated with a prior credit and asset price boom of the sort the Deputy Governor worries about – some contraction of credit seems like a good and desirable outcome. No doubt it would be unhelpful if the doors was closed to all loan applications, but no adjustment process ever occurs perfectly or costlessly. In fact, as the Reserve Bank noted in its discussion of stress tests in the FSR last November, “it can often be difficult to implement mitigating actions in the midst of a severe crisis”, and as a result “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”.

Spencer worries that a material deleveraging could exacerbate the extent of the fall in GDP and asset prices, and the rise in unemployment, worsening the credit losses relative to those highlighted in the stress test scenario. That is a plausible argument in principle, but it drives us back to the question of how demanding were the stress test scenarios in the first place.

I’ve already mentioned how demanding the unemployment component of the stress test was. No floating exchange rate country since World War 2 has had an increase in the unemployment rate as large as implied in the Reserve Bank tests. Countries are typically much better able to adjust to shocks if the exchange rate floats, than if they are fighting to defend a fixed exchange rate . Interest rates can be cut further and with fewer constraints, and the exchange rate itself can fall, a lot.

The Reserve Bank used a scenario in which nationwide house prices fell by 40 per cent, and Auckland prices by 50 per cent. Since 1970 I’ve found only been six episodes in which advanced country real house prices have fallen by more than 40 per cent (the stress tests rightly use nominal house prices, so a 40 per cent fall in nominal house prices is a more demanding test than a 40 per cent fall in real  prices).

Five of those six cases occurred in countries with fixed exchange rates. In only three of those cases was there a material fall in GDP, but the falls that did happen were very large: Finland’s GDP after the 1989 house price peak fell by 10 per cent, Ireland’s GDP after the 2006 house price peak fell by 9 per cent, and Spain’s GDP after its 2007 house price peak fell by more than 7 per cent[2].      Based on historical experience, to get an increase in the unemployment rate from around 5 per cent to 13 per cent, and a more than 40 per cent nationwide fall in nominal house prices, it would probably take more a more severe recession than a 4 per cent fall in GDP.    US real (altho not nominal) house prices fell by almost 40 per cent in the late 00s, but even there the unemployment rate did not rise by as much as is assumed in the Reserve Bank of New Zealand’s stress tests.  Losses on mortgage portfolios mostly arise from the interaction of falls in nominal prices and sharp rises in the unemployment rate.

What do I take from that? If the stress tests have been done robustly – and no one has raised serious substantiated doubts about that, the scenario probably already implicitly reflects any short-term deepening of the recession that might result from any active deleveraging banks might undertake. Banking crises everywhere – Finland, Spain, Ireland, Norway, Korea, Japan (and the United States) –  has seen some deleveraging, and the effects of that are reflected in the historical data.

I welcome the attempt to address the stress tests directly, and to attempt to defend current policy against the results of that work.  But it still doesn’t seem to stack up. The Reserve Bank has a statutory mandate to use prudential powers to promote the soundness and efficiency of the financial system. The Bank’s own numbers suggest the system is sound “even in a very severe macroeconomic [and asset price] downturn”, and direct controls have real and substantial efficiency costs which the Bank just does not address. And cyclical stabilisation is primarily a matter for monetary policy, a point the Deputy Governor also does not address.

This post has gone on long enough, so I’m going to largely skip having another go at Spencer’s unsubstantiated enthusiasm for high-rise apartments, pausing only to note that his continued enthusiasm for the number of apartments in Sydney – where land use restrictions are even tighter than in Auckland – , reads rather oddly given that Sydney is one of the few cities with higher house price to income ratios than Auckland.   I have no problem if people want to live in apartments – laws should allow them to be built – but I’m not sure there is any good basis for the Reserve Bank Deputy Governor to be trying to dictate people’s housing choices.  There is plenty of land in New Zealand.

And, finally, I was struck by Grant’s observation that housing issues keep the Reserve Bank “awake at night”.   He went on to note that “when something keeps you awake at night, it is good to do something about it”. Perhaps.  Sometimes, it might be good to get up, read a book for half an hour and go back to sleep. If my kids wake with a bad dream, I give them a hug and send them back to bed.   Focusing on the facts can also be a way of dealing with the night worries. Avoiding too much caffeine and rich food can help too.

Lying awake at night can induce tiredness and irritability, reducing one’s ability to analyse issues clearly. The Reserve Bank has offered lots of snippets of information, and lots of prejudices, but very little sustained analysis on the nature of the risks to the soundness and efficiency of New Zealand’s financial system. As someone put it to me yesterday, it is also a very macroeconomic speech, conveying little sense of what is going on in domestic credit markets.  I suggest the Bank should  focus again on the stress test results, and the very demanding assumptions used in those tests, keep a close eye on credit standards, and that they then look to monetary policy to do the cyclical stabilisation job. The price of financial stability is constant vigilance, but the Bank still looks as though it is jumping at (financial stability) shadows. Their own earlier hard analysis suggests a pretty robust financial system, no matter the insanity of the mix of other policies that is producing house prices as high as they now are in Auckland. That is a measure of the success of the Reserve Bank – charged by Parliament with financial system soundness and efficiency, not house price stabilisation – not the basis for a need for ever more costly and inefficient interventions.

(And in case any of this sounds complacent, I suspect I’m much more worried about the New Zealand and world economies right now than the Bank has given any hint of being.)

[1] I don’t necessarily favour such a change, but then I don’t think tax issues play a material role in explaining house price behaviour in New Zealand.

[2] In Korea nominal prices fell gradually during the real economic boom of the 1990s, in Japan nominal prices fell for 20 years, without a very severe recession, and in Norway in 1980s there was also no substantial recession. Both Norway and Korea had fixed exchange rates through most of these periods.

What rates of return do major US firms require to invest?

In my post on Saturday about the appropriate New Zealand public sector discount rate, I linked to an article from the Reserve Bank of Australia’s Bulletin which included the results of a survey in which Australian firms’ CFOs were asked what hurdle rates of return their firm used in evaluating proposed investment projects.

This was the chart of those survey results:

rba hurdle rate

Prompted by a commenter, I was curious whether there was something similar for the US.  I found this JP Morgan piece, undated but apparently from last year, which reports the hurdle rates that some major US firms have disclosed.  Here was the chart, and a little commentary, from that piece.

JPM US hurdle rates

If anything, this small sample of US firms had eve higher hurdle rates than the Australian firms.

In both cases, there are a few companies using hurdle rates as low as the 8 per cent real recommended by New Zealand’s Treasury for evaluating public sector investment and regulatory proposals, but not many.  As JPM point out, these high hurdle rates are a little surprising, and it may be that firms are missing out on some opportunities, but firms are presumably making decisions that appear rational to them, subject to the threat of takeover if they underperform.

Government finances are quite highly cyclical, and it seems to me that we would need a pretty compelling case for encouraging governments –  which face rather weak disciplines –  to use our money more freely than private businesses would do.  Things governments really need to do –  true public goods –  should easily pass high hurdle rate tests.

Birthplaces of our net PLT migrants

The chart below shows the birth countries for the net permanent and long-term (self-identified) migrants for the 14 years ending March 2002 to 2015 (SNZ has a break in the series prior to that).  SNZ don’t break out all the countries, but these are the ones they separately identify.  “Net” is emphasised by the large negative number for the New Zealand-born.

I don’t have a point to make.  I hadn’t had a look at the birthplace data for a while, and I’m always conscious that two of my children are net PLT immigrants, so I was a little curious.

plt by birthplace
A few things surprised me a little, in the second tier of countries.  I was a little surprised at how many people had come from the United States, Japan, Germany and France –  all countries with higher per capita incomes than New Zealand.  Of course, the flows are tiny relative to the respective home populations, and some portion will be people like my kids (New Zealanders born overseas while their parents were working abroad), but it was a flow I was a little surprised by.  Perhaps relatedly, the size of the flow of Australian-born immigrants was interesting –  similar in total to the flow from the United States, despite being closer and despite Australian citizens having free entry.  I wonder what proportion of the Australian-born net flow is the children of New Zealanders who went to Australian for a few years and then came home?

On a somewhat selective account of policy measures affecting the housing market

In the somewhat party-political and evidence-light press release on housing the Reserve Bank has just released…

“Much more rapid progress in producing new housing is needed in order to get on top of this issue. Tax policy is also an important driver, and we welcome the changes announced in the 2015 Budget, including the two year bright-line test, the proposed non-resident withholding tax and the requirement for tax numbers to be provided by house purchasers.”

…it is perhaps not too surprising that, once again, there is no reference to the enhanced first-home buyer subsidies announced, as it happens, a year ago today, as the centrepiece of the Prime Minister’s launch of the National Party election campaign.   We know quite a lot about first home buyer subsidies.  In the presence of supply constraints, all they do is bid up the price of houses (house plus land).  Other countries have tried them (Australia closest to home).  They don’t raise home ownership rates or do any socially useful thing.  And, if the subsidies don’t last for ever (which they often don’t, especially in real terms –  for example, Sir Robert Muldoon had a short-lived one of these subsidies late in his term), they simply increase the risk of a nastier correction in house prices later.  Those corrections are what the Governor, and his deputy, often tell us they are worried about.

You might think it would be inappropriate for an unelected central bank to be commenting on party election promises, even when they become legislated government policy.  And I would have some considerable sympathy with that position.  But the same surely goes for other aspects of tax policy, regulatory policy, government data collection policy and so on (eg the list in the press release), all of which are contentious in some circles or other.  Oh, and was there any mention of the role of immigration policy –  the bit directly controlled by central government?

Today’s papers report on the take-up of the enhanced subsidies. Fortunately, perhaps, Auckland prices are already too high for many potential Auckland buyers to bid prices higher.  But elsewhere, in Tauranga, for example, where people worry about the spillover from the Auckland boom, those who’ve used that subsidy will have bid prices up just a little faster than otherwise.  But no mention of it from the Reserve Bank?

In getting so involved with regulatory matters, that go well beyond its areas of responsibility, the Reserve Bank is playing a dangerous game.   There is a strong, but not unarguable, case for central bank autonomy on monetary policy, but the case is much harder to sustain when the unelected Governor is weighing in, with words and actions, to decide who should and shouldn’t get credit, what general regulatory interventions make sense, and which he thinks don’t.  And when it stays quiet on really bad policy that drives house prices higher, but helped generate some useful headlines in the middle of an election campaign,

The Governor –  and his deputy –  would be well-advised to stick to their knitting.  Get inflation back to around the middle of the target range and keep it there (it would make a change), and focus on indirect instruments (capital requirements) to promote the continued soundness and efficiency of the financial system.  And leave the rest to the political debate, and decisions made by those whom we can hold to account.

Q&A on immigration

As most New Zealand readers will now know, TVNZ’s Q&A programme yesterday featured a debate around the economics of immigration. Unfortunately, the programme got prominence not for anything of substance that was said by the participants, but for a vulgar outburst by Shamubeel Eaqub. It was pretty unfortunate, but then this was the guy who only a few weeks ago was dismissing my analytical arguments as “that’s racist” , and then turned his attention to Fonterra, suggesting they were treating their investors as ‘scum’.

TVNZ got together Don Brash, Shamubeel Eaqub, and Massey University population/immigration professor, Paul Spoonley. They had asked me to go on the programme, but I don’t do work-like things on Sundays.

Don Brash articulated some of my macroeconomic arguments about the impact of our large-scale programme of targeting non-citizen migrants. I’ve argued that in an economy with modest savings, high average immigration inflows exacerbate pressures on real interest rates and the real exchange rate. We’ve had the highest real interest rates in the advanced world for several decades, and a real exchange rate that has been much higher than is consistent with our dismal productivity performance. In combination, the high real interest and exchange rates have squeezed out business investment – our levels have been low for decades – and skewed returns in ways that favour investment in the domestic or non-tradables sectors. Returns from investing heavily to tap world markets just haven’t been there, so even once we opened up our economy, business investment in the tradables sector has been subdued.   None of this is a story about the last 12 months, or any short-term window, it is about average patterns seen over the last 25 years or so.  (Unfortunately, some of the debate got bogged down in the last 12 months’ data.)

One should always try to ensure that one understands the arguments people on the other side are making, so in this post I have tried to identify the various points that Eaqub and Spoonley were making, and offer some comments on them.

Eaqub started out by simply asserting that my arguments were wrong.   He argues that volatility of net immigration is an important issue (“the short-term pressures are absolutely clear”) especially in the context of house prices. He argues that we need to “fix” housing supply responsiveness and associated infrastructure issues. He goes on to claim there is just a failure of political leadership, and that any “infrastructure” issues could be dealt with by more government borrowing, since the government faces no debt constraint.

Of course, I agree that housing supply responsiveness should be improved, and that the central and local government policies that impede these are a scandal.   But he seems to totally miss the point of my argument. Real resources that are used for building houses or “infrastructure” can’t be used for other stuff, and the real exchange rate and the real interest rate move to free up resources to meet these population-based demands. We could have more investment in, eg, Auckland roads, but all else equal that would put more pressure on scarce resources. Meeting the demands of a rapidly rising population squeezes out other stuff – in this case, business investment, especially that in our tradables sector.

Eaqub also asserted at this point that “immigrants are good for New Zealand, as they been for centuries for New Zealand. They’ve increased our human and physical capital base”.

Don Brash challenged him as to whether the increase in the total size of the economy from the large scale immigration programme had had benefits for New Zealanders, noting that productivity growth in New Zealand had been weak for decades. He articulated a story in which the far-reaching reforms of the late 1980s and early 1990s had not reversed that underperformance, even though immigration policy had been materially liberalised at much the same time.

Eaqub responded to this by arguing that “perhaps they weren’t the right reforms”. That is a reasonable possibility to explore, but he didn’t back it up. He said that deregulating the economy had been “fantastic”, and the only policy he cited that he was uncomfortable with was the amount of student debt being “piled on” students.  Reasonable people might have a range of different views on that one, but given the surge in New Zealand tertiary participation over the last 25 years, it is difficult to know what mechanism Eaqub had in mind for how student loans policy might explain much of New Zealand’s economic underperformance.

Eaqub also repeated the argument he made in his recent book that net immigration has not accounted for much of total population growth in the last 50 years. But as I’ve pointed out previously, net migration includes the coming and goings (mostly the latter) of New Zealanders, and that a discussion about immigration policy needed to focus on the flows of non-citizens (the bit the New Zealand government controls). As I’ve explained previously, and as Don highlighted, without non-citizen immigration, our trend population would now be flat or falling.   There would be cyclical population pressures on the housing market, but no trend ones.

The interviewer drew attention to the 2006 Australian Productivity Commission study Don Brash has cited which concludes that any gains from immigration were captured by the migrants, and there was no gain to Australians. Don noted that there was no comparable study in New Zealand and that there is little or no evidence that New Zealand’s immigration programme has improved productivity or economic well-being for New Zealanders.

Eaqub rejected that proposition, advancing as evidence a recent NZIER piece. That seemed rash. The NZIER piece in question is a four page note, and its conclusion were based on a single unreported equation (the Australian Productivity Commission report was, by contrast, hundreds of page long). I went to a discussion on the NZIER paper at Treasury last year, and I think it would be fair to say that even the advocates of immigration who were at that meeting did not find it particularly persuasive. And, in any case, since the single equation looks only at total (per capita) GDP, which combines GDP accruing to New Zealanders and to immigrants, it does not shed light on the specific question at hand: if there are gains from immigration, how many (if any) of them have flowed to New Zealanders as distinct from migrants.  Notwithstanding his claims for this NZIER note, Eaqub actually went on to acknowledge that the bulk of gains from immigration flow to the migrants, before asserting that it is “absolutely clear” that the type of immigration policy we have in New Zealand is improving skills and human capital. As Don noted, actually many of the migrants (temporary and permanent) don’t seem that skilled at all.

The interviewer interjected my past references to work done at the Reserve Bank suggesting that a 1 per cent of population immigration shock had boosted house prices by perhaps 7 or 10 per cent. Eaqub came back with his assertion again that the only thing that mattered was the volatility in net migration, reiterating this claim that net migration can’t explain most of growth in population, while ignoring the distinction between immigration of non-citizens (the bit the New Zealand government controls) and other flows. (Somewhat surprisingly, later in the show Eaqub actually claimed that New Zealand had relied on immigration for its population growth throughout its history)

In the second half of the show, the interviewer introduced some of the numbers I’ve run here about the number of shelf-fillers and checkout operators who had been given work visas under the Essential Skills category.  He went on to suggest that “beggars can’t be choosers” and that perhaps the best people just wouldn’t want to come to New Zealand. Somewhat surprisingly, Eaqub rejected that suggestion (I’ve previously heard it acknowledged by leading academic advocates of immigration) asserting that we “generally” get “very highly skilled people”, [even though no more than half of the permanent residence approvals are even in the skilled/business category, which includes not just the primary applicant but their immediate family] before rushing on to the “massive shortages” in places like orchards.  Eaqub argued that the challenge is to boost human capital, and that we need “more skills” to compete in the global economy. Despite our very large inward migration programme, he made no effort to show how this has actually been happening – in, for example, an economy where the share of exports in GDP is unchanged over 30 years, and where productivity growth has been among the worst in the advanced world.   Paul Spoonley backed up Eaqub claiming that, while not defending the checkout operator approvals, “by and large we are bringing in the right people” and that we had a selection system that was the envy of the world. He has just returned from Europe, where people would “die” for our system. But here I should add that one of the advantages New Zealand has is that we have near-absolute control on who comes in – we don’t have boat people, or other illegal migrants, and only Australians have automatic rights to come here (and not many do). The question is whether there is evidence that the migrants we’ve had have added to the economic well-being of New Zealanders.

Discussion turned to the (quite large) family reunification component of our immigration programme (around a quarter of approvals). As Don Brash noted, many of these people don’t appear to be particularly highly-skilled migrants. Eaqub’s somewhat surprising response was along the lines of “there has to be a human element to it.   You can’t just allow people in and then say they can’t maintain family connections, and we need empathy in policymaking”.   But……immigration policy is supposed to set to benefit New Zealanders (that “economic lever” MBIE and ministers tell us about). No migrant is forced to come to New Zealand, and any are free to take holidays back home (or to have family holiday here). And unlike the 19th century, technology now enables people to talk to overseas family easily and cheaply as frequently as they like.

Eaqub went on to argue that there was no problem because immigrants as a whole provided a fiscal boost to New Zealand public finances.   That is by no means certain, but even if we accept that the average migrant does makes a positive fiscal contribution, that need not be true of the marginal migrants, and as economists and policymakers we should be thinking at the margin. Many migrants probably make a positive fiscal contribution. But those who come under family reunification approvals in mid-life almost certainly do not (they typically don’t pay enough taxes for long enough to cover the full NZS and health entitlements).  It was at about this point that his vulgarity occurred – the expletive perhaps being a substitute for a missing substantive argument?

Paul Spoonley offered a slightly more sophisticated defence, arguing that our family reunification policies help provide migrants with a reason to stay in New Zealand (the centre of the family is now here), and that the cheap childcare the grandparents could provide was an economic gain for New Zealand.   His attachment story is a plausible argument, although whether it stands up to empirical scrutiny is another question. Perhaps instead primary migrants stay just long enough to get families into New Zealand and secure health and welfare rights? Having “bought the insurance” and secured the family, there is no compelling need for the migrant to stay. But even if Spoonley is correct, it simply puts more of a burden of proof on the advocates of immigration – if we need to bring in family members and parents, and provided associated welfare support, to get migrants to stay, we should be very sure that the primary migrants themselves are substantially benefiting New Zealanders by being here.   There is little real evidence of that so far.

The interviewer also raised the question of our ability to attract high net worth investor migrants. Don Brash noted that it wasn’t matter of how many we attracted, or how much money they might bring, but of whether any inflow (of financial or human capital) boosted productivity for New Zealanders. For that, he asserted, the evidence was skimpy. Eaqub reiterated his claim that in fact the evidence is yes, that gains accrue over long periods of time, and that we have relied on immigration for population and productivity growth across all our history. In his view, it made no sense to arbitrarily look at the previous 25 years or so. This made little sense to me. First, as Don Brash pointed out our productivity growth (relative to other countries) has been atrocious for decades. And second, we actually had a major change of immigration policy 25 or so years ago, and a conventional approach to policy evaluation would suggest looking at what had happened in the economy in the subsequent years.

Eaqub was asked by the interviewer what level of immigration he would regard as too high. He argued that we need to focus on medium-term trends, not annual fluctuations. This seemed inconsistent with his story that the volatility of net migration was the only problem, but perhaps it wasn’t. He seemed to favour bringing in any number of “high quality and skilled people”, and that the only problem is other “broken markets” (presumably housing supply and immigration.

In concluding the discussion, neither Don Brash nor Shamubeel Eaqub seemed keen on the recent change to immigration policy that will give more points to those willing to settle somewhere other than Auckland. As I noted recently this will, more or less as a matter of algebra, lower the average quality of the “skilled” migrants we do get.

Spoonley didn’t weigh in much on the bigger economic issues. He thinks we choose migrants well – even though seriously high-skilled migrants make up only a moderate share of our overall immigration programme. But as Spoonley noted in his 2012 book “understanding exactly how immigrants contribute to economic outcomes is still somewhat fraught in the New Zealand context”.

Eaqub made most of the economic arguments in defence of New Zealand’s immigration programme, but in the end I was a little surprised at how thin his case seemed to be. Despite his comments:

  • We simply have nothing comparable in New Zealand to the Australian Productivity Commissions studies on the economic impact of immigration.
  • Housing market problems certainly could be fixed without touching immigration settings, but there is no doubt that policy-controlled immigration (the inflow of non-citizens) is now the only source of trend population pressure on the housing market.  And annual fluctuations simply aren’t that controllable.
  • A surprisingly large proportion of New Zealand permanent residence approvals are not coming in under the skilled or business stream, despite the official rhetoric about a skills-focused immigration programme. Even among those who are, many are dependents, not those who have independently demonstrated value to New Zealand, and a disconcertingly large share of the “skilled” migrants don’t seem that skilled at all
  • Despite the repeated mantra about the need for more skills and more human capital, there is no evidence I’ve seen that New Zealand is now particularly short of human capital, or that large scale migration can easily resolve the gaps there might be.  We now have high levels of tertiary participation, and rather low estimated economic returns to tertiary education, suggesting that availability of “skills” isn’t the pressing constraint on lifting New Zealand’s medium-term productivity performance.
  • We need to be thinking about immigration (like all policies) at the margin, not as an average. I’ve no doubt that there are some migrants who make a huge contribution to New Zealand, and to New Zealanders.  But I think there are some pretty good arguments that the 45000th permanent residence approval is unlikely to be generating a net gain at all.  If there really are material gains to be had from immigration, we could presumably capture most of them with, say, an annual target of 20000 permanent residence approvals, especially if we wind-back further the family categories and focus on genuinely high-skilled migrants.

There may be other good explanations for the stylised facts around New Zealand:

  • Persistently high real interest rates
  • A persistently high average real exchange rate
  • Persistently weak productivity performance, despite much-lauded liberalisation in the 1980s and early 1990s
  • Persistently weak business investment,
  • A persistently modest export sector
  • A persistently high level of net international debt, despite the strong government balance sheet.

But against that backdrop, the very large scale non-citizen programme that New Zealand has run over the last 25 years – a really big policy experiment – looks a plausible candidate for what might have gone wrong. At best, the advocates of the large experiment are still struggling to show that there have been any real benefits for New Zealanders. At worst, if my story is right, much of the expansion in total GDP that we have had in the last 25 years might have been at the expense of the sort of productivity gains that would benefit all New Zealanders.   Since 1990, not many advanced countries have had more growth in total real GDP than New Zealand, and not many have had less growth in real GDP per hour worked.

There is apparently a great deal of faith among New Zealand elites that our large-scale immigration programme is “a good thing”. As I’ve noted previously, in the biblical book of Hebrews, there is a verse that reads “Now faith is the assurance of things hoped for, the evidence of things not seen”.       The evidence of the economic benefits to New Zealanders from one of the largest controlled immigration programmes anywhere still remains largely a thing not seen,