If anything I came away from today’s Monetary Policy Statement and (the bulk of it that I saw) the Governor’s press conference more convinced that I was yesterday that the OCR should have been raised by 50 basis points today.

There were a couple of elements in the minutes that were a little more encouraging than one might have feared.

There was the fact that a 50 point increase was clearly seriously considered, and debated. There was the fact that that debate was actually disclosed in the minutes (I think that is a first). There was the explicit comment not ruling out 50 point increases in the future. And there was, at last, a slow start to the process of unwinding the huge punt on the future of bond rates taken on in the LSAP intervention of 2020 and 2021.

It could have been worse. There clearly is an element of unease around the Committee table around the rise in inflation expectations, even for the longer-term horizons the Bank has often previously used to reassure itself there was no particular problem.

But…..when 2 year ahead inflation expectations of relatively more-informed observers have increased (as they have) by 29 basis points since you last met, an increase of the OCR by 25 basis points – especially as those survey respondents will already have expected such an increase and factored it into their inflation expectations – isn’t getting on top of things, even slightly, rather it is just falling a bit further behind. Whether thinking about inflation or inflation expectations, the MPC has fallen short of the Taylor principle – that at very least one should raise (lower) the OCR to the extent inflation and/or expectations rise (fall). The real OCR is now lower than it was when they descended from the mountain top in November, even as all measures of inflation (including the core inflation ones) have moved considerably higher. In the Bank’s own words, all the core measures are now above the top of the target range.

And, of course, the real OCR is now materially lower than it was two years ago, even though (core) inflation has been high and rising and – again on the Bank’s own reckoning – the labour market is unsustainably tight, the unemployment rate is too low to be sustained.

And yet the Committee made no attempt anywhere in the document to justify why real monetary conditions now would prudently, on the balance of probabilities (and they even invoked the “least regrets” language again), be so much looser than they were two years ago.

Linked to that, and perhaps my major criticism of the document itself, is that there was no sustained effort to analyse and explain why the Bank’s core inflation forecasts had been so wrong, why core inflation was now above the top of the target range, or what the MPC had learned from that experience that now gave them greater confidence that they understood the inflation process sufficiently well to keep on with the “slowly does it” approach to adjusting the OCR. In the circumstances it is a pretty inexcusable oversight – and it was a bit disappointing that no journalists asked about the issue/omission.

If one goes back to those minutes, the MPC lists a few reasons not to raise the OCR by 50 points:

• there were the LSAP sales, but that is a clearly just there to bulk out the paragraph, since the body of the document says any impact of the LSAP, including sales, is now very small, and the Governor reiterated that point even more strongly in the press conference. In any case, the bond sales don’t even start for several more months.
• there was the fact that interest rates had increased quite a bit late last year.  Which is fine, but there isn’t any sign that (say) inflation expectations have dipped.
• there was Omicron, which seems to have been a factor, even though monetary policy operates with lags that run well beyond the next 3-6 weeks of the peak Omicron wave, and
• there was this strange line: “They also noted that conditional on the outlook, the OCR is expected to peak at a higher level than assumed at the November Statement.”.   They have certainly raised the peak of the tightening cycle quite a bit, but……that would normally be an argument for getting on it with now, not just carrying on in the slow and measured way, even though you think –  as the MPC appears to –  that another 240 basis points of increases will be required.

There are more than a few puzzles in the document. For example, the peak of the OCR cycle (3.4 per cent) seems to be well above the Bank’s estimate of neutral rates – chart suggests something around 2 per cent – and both core and headline inflation eventually come down a lot. But the economic forecasts suggest this all happens by “magic”, since the output gap only goes negative in the year to June 2025, and the unemployment rate looks as though it never gets above the NAIRU. It reinforces the point that the model – the understanding of the inflation process and what has gone on in the last couple of years – is at best weak, and possibly missing in action.

At this point I should make clear that I do not have a strong view on where the OCR cycle should peak. I tend to think it is a fool’s errand given how little we know, and so I concentrate on the next few quarters. But the Bank is clearly uneasy about inflation expectations, thinks there is a lot more to do, and yet seems to want to get there very slowly, running in the process risks of things getting even further away on them. 50 basis points would have prudent, especially at a time when no one supposes there is much risk that a few months down the track there would have been cause for regret, having raised the OCR to the fearsome level of (about) -2 per cent.

Is there reason to think economic activity may not do that well this year? Indeed there is, although the eventual opening of the borders will add to (not detract from) capacity (including labour market capacity) and inflation pressures), and core inflation having once got this high – and the Bank expects it to stay this high for at least the next year – doesn’t typically come down by magic. It typically requires some policy force – a little more than is implied by real monetary conditions a lot easier than they were before Covid and this inflation surge got underway. As it is, the Bank is probably painting a rose-tinted picture all round: core inflation falls surprisingly easily, and productivity growth actually picks up a bit. Perhaps it will happen, but there is no compelling case made (and the Bank’s answers to the productivity question were particularly half-hearted.

I could go on on other matters but will end just noting three points briefly:

• since yesterday’s post, Bob Buckle and Peter Harris have been reappointed to MPC positions for further three year terms.  There was no question to the Governor about how he can possibly justify –  the more so in the current circumstances –  the blackball placed on anyone with a serious ongoing engaged analytical or research interest in monetary policy.    The final appointment is made by the Minister of Finance, but he can appoint only those nominated by the Bank’s Board and no doubts that Orr dominates the Board on such matters, 
• there was no attempt by the Bank to justify or explain away the more-than $5 billion in losses run up on the LSAP and –  more disappointingly –  no questions from the assembled media,
• there was not a single question about the appointment to the senior deputy position responsible for macroeconomics and monetary policy (and to the MPC) of someone so manifestly underqualified –  with no relevant qualifications or experience – as Karen Silk.

Perhaps the FEC members might do a little better when the Governor next appears there?

UPDATE: I had to go and pick up a child and so missed the last few minutes of the press conference. I gather the Governor explained that the strange cover was in honour of his departing chief economist, who told us last time he was going to coach his son’s cricket team. It is a nice touch…….for someone who appears to have been forced out in the great Orr restructuring.

Any journalist at the 3pm press conference care to ask the Governor about the intended messaging in an aerial shot of the Basin. No (cricket) tests, so a ref Covid disruption? Dark shadows advancing over the ground? Or….or…..? https://t.co/jwMtRilYzi

— Michael Reddell (@MHReddell) February 23, 2022

The Reserve Bank and the MPC will tomorrow emerge from their long summer slumber to deliver a Monetary Policy Statement and OCR decision. It is quite extraordinary that in this period, of considerable volatility, uncertainty, and inflation surging above the target range (even on core measures) we’ve heard precisely nothing from any of them (individually or collectively), and had no policy actions, for three months.

But, at last, tomorrow we will hear. As I’ve said repeatedly here over the years, I’m not really in the business of trying to guess what the MPC will do. There are plenty of people with a strong interest in that, and banks have people who nurture every little contact or conversation they can secure with Reserve Bank officials, analysts, or policymakers. As the Bank says little in public, and isn’t that consistent through time in how it acts, all one can say is good luck to those people.

My interest is in what the Reserve Bank should do, given the target set for them by the government (and, of course, the wider but less immediate question of what that target should be). The target is really pretty specific, with a primary focus on keeping the persistent elements of inflation near 2 per cent.

Against the backdrop of the Remit, I think it is pretty clear what the MPC should be doing tomorrow:

• they should be raising the OCR by 50 basis points, and giving a signal that, all else equal, they expect another 50 basis point increase at the next review on 13 April,
• they should be terminating new loans under the Funding for Lending programme, the crisis programme introduced in 2020 which is still at the margin having the effect of holding interest rates lower than otherwise, and
• they should commence a programme of bond sales, designed to liquidate the LSAP bond position within 2 years (or, equivalently, have agreed a programme with The Treasury under which it would sell new bonds on market, buying back on the same day an equivalent amount of the bonds held by the Bank).

I don’t suppose either of the last two items on that list are at all likely.  I don’t even think that LSAP sales would make much material difference to monetary conditions (exactly paralleling my argument when they were buying bonds) but (a) every little helps, and (b) we should get this massive taxpayer bet on the bond market closed out ASAP, on principle.   And there is no crisis now –  the OCR is fully able to be used as desired.  Much the same goes for the FLP programme.  Closing that now might have some modest useful impact in tightening overall monetary conditions, but it would also align messages –  this isn’t the 2020 Covid crisis any more, and the focus of monetary policy now is (or should be) getting core inflation back down fairly expeditiously.

But even if they took both of those steps, the case for a 50 basis point increase tomorrow is strong. 

There are a couple of ways I find useful to look at things.

It was in last August’s MPS that the Bank really set out towards some significant OCR increases.  They didn’t act that day because they had the misfortune of a (long-scheduled) release a few hours after the Level 4 lockdown was imposed.  Back then they envisaged the OCR being 0.9 per cent for the March quarter 2022 (roughly consistent with having been 0.75 per cent moving to 1 per cent now).  But back then:

• they thought core inflation was about 2.2 per cent and would go little or no higher (headine inflation forecasts for the second half of this year and beyond were 2.2 per cent, dropping further away)
• they expected the unemployment rate for this quarter to be 3.9 per cent, and to go no lower than that.  
• their latest reads on inflation expectations weren’t particularly inconsistent with inflation near the midpoint of the inflation target range (the most recent 2 year ahead measure they had then was 2.3 per cent).

And where are we now? 

• the unemployment rate – still probably the best indicator of spare capacity (or lack of it) – had fallen to 3.2 per cent in the December quarter, and there is talk it could go lower yet.  No one seems to suppose that the unemployment rate is near the NAIRU.
• the latest headline inflation rate is 5.9 per cent, and may go higher this quarter.  Much more importantly, core measures have increased substantially –  the best of them, the sectoral core factor model measure, is now 3.2 per cent for the year to December, and
• while there is a wide range of inflation expectations measures, none of them are as low as the Bank was seeing back in August, and the best of them are pretty consistent with people assuming (and acting as if?) inflation will stay outside the target range (and well away from the midpoint) for the next couple of year.  Even with the tightening in monetary conditions observed in recent months.

It was really captured quite well in a newspaper article yesterday that reported that all economists talked to thought that the Bank was now “behind the game”.   In that sort of environment, when you’ve been repeatedly surprised by the strength of inflation, and outcomes are no longer consistent with target –  suggesting that your model of what is going on is not working that well – the appropriate response is not to edge up the OCR ever so gradually and hope.   Least regrets –  of which we heard quite a lot from the Bank over the last 18 months – calls for something a bit bolder, to get ahead of the rise in (core) inflation and assure watchers (firms, households, markets) that you really are serious, rather than reluctant.

Is there a case in Omicron for holding back (25bps rather than 50bps)?   I don’t think so.  We can see how other countries have come through the short and intense Omicron waves –  which will temporarily disrupt both supply and demand –  and there is little sign of what is required from monetary policy being revised down in the wake.

(“Not surprising the market” isn’t a good excuse either.  When you say nothing for months, the market has little to go on re the MPC reaction function.  If you won’t talk, you have to live with surprises sometimes.  The market will cope, although might suggest that a bit more communication than once every few months would make sense.)

The second way of looking at things is to compare where are now to where we were two years ago.  Having cut the OCR during 2019 by early 2020 markets, not discouraged by the Bank, were beginning to think in terms of when a first OCR increase might be warranted (still then, it was thought) some time away.

The OCR is early 2020 was 1 per cent.  Core inflation was probably just under 2 per cent (having edged up over the previous couple of years towards the target midpoint), and the unemployment was about 4 per cent.   Where things are now we’ve already covered above –  much more inflation, more capacity pressure, and on any metric you like materially higher inflation expectations.

Any policy and/or monetary conditions?

Well, the OCR today is 0.75 per cent, still below where it was two years ago.  A 25 point increase tomorrow will only take it back, in nominal terms, to where things were in February 2020.  The nominal exchange rate is also now about where it was two years ago.

But what about the interest rates that firms and households face?   We’ve heard a lot about fixed rate mortgage rates have risen from their lows.  And that’s true: a two year mortgage rate is now about 170 basis points higher than the low (last April).

But what of comparisons to the start of 2020?

So nominal mortgage rates are perhaps 50 basis points higher than they were two years ago. But inflation is a lot higher than it was then, and so are all indications of inflation expectations. For these purposes, I’ll just use the Reserve Bank’s own survey of two-year ahead inflation expectations: 1.93 per cent in the survey done at the end of January 2020, and 3.27 per cent in the survey done a few weeks ago, an increase of 1.34 percentage points.

Here is the same chart, but expressed in real terms, adjusted for the increase in inflation expectations.

and the same chart for term deposit rates

(Real) monetary conditions have tightened a bit from the lows, but they really needed to. Compared to conditions prevailing two years ago, real mortgage rate, real deposit rates and of course the real OCR are substantially looser than they were. And whereas two years ago inflation looked to be fairly comfortably somewhere near the target midpoint, nothing remotely that optimistic describes the situation now.

Of course there are other influences – things like fiscal policy, the terms of trade, other regulatory restrictions (eg tax or the CCCFA) – but it would very very hard indeed to construct a story that suggested that real monetary conditions as we find them today (or might find them tomorrow) are likely to be consistent with the Bank delivering on its mandate, on the environment as we see it today. In fact, if core inflation becomes established at current rates (3% plus), standard economic analysis would tell us that reductions in core inflation – to the 2% the Bank is supposed to be focused on delivering – are likely to come about only by luck (ill-luck probably – adverse economic events) or by monetary conditions being tightened to above-normal levels of tightness for some time, a process that would (at very least) result in a period of sub-par economic growth, perhaps even a recession.

We’ll see tomorrow what the MPC has decided to do. But whatever the rate adjustment we should be able to expect a serious and rigorous accounting for just what has gone wrong in the last 12-18 months, what the Committee has learned, and a serious analysis of the options and risks in the monetary policy that might be required to get core inflation (and expectations of it) settling back around 2 per cent. Sadly, with this MPC it would probably be a bit of a surprise if we got even the appropriate degree of analysis, self-scrutiny, and accountability.

I was tempted to head-up this belated MPS post “Je ne regrette rien”, as that – I regret nothing (about last year’s monetary policy) – was what Orr told yesterday’s press conference as he was getting rattled towards the end. He should regret quite a bit – notably the $5.7 billion of taxpayer losses on the LSAP, and the ongoing huge risks (neither were points he was willing to engage on, whether in the MPS, in the press conference, or at FEC this morning – indeed he actively played distraction). But that isn’t really where I want to focus my thoughts on the Monetary Policy Statement.

I thought the MPC should have raised the OCR by 50 points. The MPC disagreed, and moved by only 25 points. That is their choice of course, but – once again – I was struck by just how lacking and inadequate the supporting analysis and argumentation were, on their own terms (ie relative to their own published forecasts). No informed reader – and there won’t be many other readers of their 56 pages – will have come away feeling persuaded by the insights and analysis the Bank’s big team of macroeconomists had generated. There was nothing new or insightful, at least that I could see. And much that wasn’t convincing.

As I’ve noted on various previous occasions, when there are big starting point surprises, surely we expect to hear from the MPC (a) why they think they got things wrong, and (b) what that mistake – and mistakes are inevitable in such areas – has taught them about how the economy is behaving and how, if at all, it changes their view about the road ahead. But once again, there was none of that (in fact, in the press conference again played distraction suggesting that the surprise was the lockdowns after August, whereas the September quarter unemployment and (core) inflation surprises really had nothing much to do with those lockdowns (which will, of course, have a big impact on September and December GDP). Their August projections/discussion had suggested something fairly unproblematic on core inflation, and instead they suddenly found themselves with outcomes near the very top of the range, and so on.

Their discussion of inflation expectations also lacked structure, consistency, and any sense of authority. In the minutes of the MPC meeting we were told

The Committee noted that near-term inflation expectations tend to move with actual inflation. Medium-term measures provide a better gauge of whether inflation expectations remain anchored, and these remain close to the target midpoint.

The message seems to be one of nothing to worry about at all. But even the story is misleading, at best. It is certainly true that year-ahead survey measures of inflation expectations seem to be very driven by fluctuations in headline inflation, but here is the two year ahead measure (from the Bank’s own survey) lined up against their (historically) preferred – and most stable – measure of core inflation. Inflation expectations – over almost 30 years – have typically fluctuated through materially narrower ranges than core inflation itself. The exception – potentially important exception – has been the last two years. If anything, the two year ahead expectation could be disconcertingly high already, given the extent of the rise in core inflation itself.

Then there is that claim that all is fine because long-term inflation expectations haven’t changed much. The Bank asks about expectations five and ten years ahead, and outcomes are not far from 2 per cent – as you would hope, given the target, but only because respondents presumably expect the MPC to act sufficiently aggressive to keep inflation near the centre of the target range. If those five and ten year expectations started moving up sharply there really would be cause for concern, but the fact they are still near 2 per cent shouldn’t be telling MPC anything about the appropriate policy stance now.

At the press conference one offshore questioner asked the Governor about the MPC’s response to the big increase in inflation expectations, given the Risk Appetite Statement included in the MPC in which they asserted that “they had a low appetite for policies or decisions that could cause inflation expectations to become unanchored”. This was greeted with a glib and dismissive response from Orr along the lines of “we have reacted and raised the OCR”, not even engaging with the fact that (for example) two year ahead expectations are now a full per cent higher than they were in February, and yet the OCR has been increased by only 50 basis points over that time, there isn’t another review until February, and the MPC has stated that they prefer to move in 25 point bites. At best, it will be the end of March before the OCR will have been raised by 100 basis points, but even that won’t have raised real interest rates at all relative to the start of this year (let alone relative to the start of last year). Perhaps expectations will have moved even higher – outside the target range in the meantime. Perhaps not, but surely we should have expected a more thoughtful nuanced and engaged treatment of the issues and risks? Core inflation has, after all, already increased a lot (and – which we will come to – even they seem to expect it to increase further).

Similarly, we are told that the Bank’s projections have the OCR rising to above (the Bank’s estimate of) the neutral OCR. They seem to base that on this portrayal of neutral.

But this chart seems not to have taken any account at all of a jump in inflation expectations. The Governor said they mattered – and that the Bank had responded – but there is no sign they do so in this estimate of neutral. Given the Bank’s inflation forecasts it seems unlikely that medium-term inflation expectations will be dropping any time soon, and if those survey numbers are capturing something real, doesn’t that mean the OCR needs to go (quite a bit) higher than they might otherwise have thought. Now personally I’m very sceptical of the value of medium-term projections, but it is the Bank that uses them as a storytelling device and yet quite a material (and identifiable) part of the story seems to be have been left out.

And so it goes on. The best question I’ve heard about yesterday’s MPS was from a first year economics student, who wanted to know how the Bank could claim to be fulfilling its mandate when it projects that inflation next year will be 3.3 per cent. I haven’t seen any attention paid to what that number means. Recall that it is now November 2021. Nothing about the year to December 2022 has happened yet. So the Bank’s forecasts for inflation next year must be very close to a forecast of core inflation (they don’t know the inevitable one-off shocks – up or down – and they assume the exchange rate is fairly stable). Core inflation is currently about 2.7 per cent and the Bank is quite content to see it rise to 3.3 per cent – outside the target range. When a press conference questioner asked a similar question, she again got a dismissive (and obfuscatory answer). It might be one thing to take things slowly if the unemployment rate was still lingering high, but the Bank is quite open that at present the unemployment rate is below a sustainable level. So raising the OCR more and more quickly wouldn’t be kicking the economy into recession – the Governor’s claim – but would just get both dimensions of their dual mandate back towards desired levels sooner (and with less risk to those pesky inflation expectations). As it is, the Bank’s forecast for the unemployment rate in March 2022 is a tough lower than it was in the latest official release. This is an economy that – on their numbers – has been overheating, and they can’t even manage of Taylor principle scale of response, not even when the unemployment rate is – on their telling -unsustainably low. Perhaps there is a case to be made for their choice, but neither the MPC nor the Governor made it.

I could go on at some length on other matters, but just a few bullet points instead:

• it is sobering to see how pessimistic the Bank now is about productivity prospects  (0.5 to 0.6 per cent for annum across the forecast horizon).  The Bank has no particular expertise in productivity, but they just now take for granted our woeful performance
• it was curious to see a lengthy Risk Appetite Statement in the document.  Doubly curious in that more space was given to (for example) the risk of an MPC member missing a meeting than to (nothing at all) the huge financial risks decisions like the LSAP programme expose the taxpayer too,
• the Bank is all over the place on the LSAP (all while refusing to seriously address the losses, just waving the hands about “overall gains).  They seem oblivious to the international research that suggests the stock of bonds held is what makes any useful macro difference, asserting that the programme was previously making a big difference, but now makes only a small difference.  And once again they refused any serious answers about the future of the LSAP, claiming that a document is coming in February.   Similarly, they make laughable claims about why the Funding for Lending Programme needs to be kept open, Orr even suggesting to FEC that were they to close this crisis tool now (a year or more after the crisis) it might pose a future financial stability threat,
• there were pages and pages on climate change.  As far as I could tell, all they seemed to focus on was direct price effects, and even then had nothing to say other than “some relative prices will rise”.  No doubt and –  by definition –  others will fall.  Orr claimed there was going to be unusual volality in headline inflation relative to core, but offered not a shred of analysis in support of his claim (we’ve had lots of shocks and policy reforms in decades past).  And, somewhat surprisingly, they didn’t even touch on any effects on the neutral interest rates –  if there is any effect (and I suspect that any effect will be vanishingly small) it is likely to lower neutral rates a bit.
• remarkably, there was almost nothing in the document offering insights based on the experiences of other countries.  Again, with a big team of economists and access to overseas central banks you’d hope that the Bank’s thinking would be informed by the diverse experiences other central banks are observing.  But there was nothing.   

All in all, it was a fairly typically poor Reserve Bank performance, perhaps undershooting even my low expectations.  It was good that some questions were asked –  at the press conference and at FEC –  about the high turnover at the top of the Bank, even if Orr was allowed to get away much too easily with ludicrous claims about what a desirable place the Bank was to work, what an abundance of talent they had available etc etc.  It certainly wasn’t on display in this document.   

(On the turnover question, one almost had to feel sorry for the Chief Economist who has been restructured out:  asked by MPs about what was going on Orr rashly talked about how Yuong Ha “has chosen to go into a far more challenging role. What are you going to be doing?”  There was a noticeable, whereupon Ha lamely responded “Coaching my son’s cricket team. Taking a break”.   Orr seemed to display all the sensitivity and personnel management skill of, say, a Judith Collins.)

Or to give the book its full title, The Money Illusion: Market Monetarism, the Great Recession and the Future of Monetary Policy.

I was engrossed in the 2008/09 recession – and the associated financial crises – at the time it happened, as an official at the New Zealand Treasury, and it must have been very early in the piece that I started reading Professor Scott Sumner’s then new blog, also The Money Illusion. I’m not a regular reader now, but found much of what Sumner had to say about the conduct of monetary policy – mostly in the US – stimulating and thought-provoking, even (perhaps especially) when I didn’t end up agreeing. So I was keen buyer when his 400 page book appeared.

It is an interesting mixture of a book – partly textbook, partly personal intellectual autobiography, and partly a tract championing a different approach to policy (and history). It would be well worth reading for anyone interested in monetary policy, with a particular focus on the recession of 2008/09 and it aftermath, and possible reforms for the future. Were I an academic teaching a monetary economics class I’d encourage all my students to read it (and have commended it to my economics-student son), not as replacement for a textbook but as a practically-oriented complement to it.

I usually read books of this sort with a pen in hand. But flicking through the book again I see that the pen was hardly deployed at all in the first 60 per cent of the book, which is a really clear and useful introduction to how the monetary system works, through a slightly different lens than most will be used to. If I didn’t agree with it all – and there were a few straw men tackled – people coming to grips with the system and concepts and history will almost inevitably see things more clearly for having read it. It is an achievement in its own right.

The second part of the book is focused much more on the policies adopted, mostly by the Fed, in 2008/09, and on the way ahead (bearing in mind that the big was largely finished before Covid, although I doubt the thrust of Sumner’s arguments would have been much changed by the experience of the last 18 months). And it is there I start to differ (and thus have lots of marginal notes).

It isn’t, I think, that we differ that much on how the economy works (or even how monetary policy works). Like Sumner I’m a champion of the potency of monetary policy. It can’t make countries rich, or solve things like New Zealand’s decades-long productivity failure, but it can – and should – do all it can to keep the economy as fully-employed as (labour market regulation etc makes) possible consistent with keeping inflation in check. It really matters, given the pervasiveness of sticky wages and prices in the economy. Sumner’s previous book, on the US experience of the Great Depression, was a really nice illustration of both the potency of monetary policy, and the way that misguided regulatory interventions (in that case much of the New Deal) can mess up economic performance.

And I’d also endorse two of his specific criticisms of choices the Fed made in 2008.

The first was the failure to cut official interest rates at the FOMC meeting a couple of days after the Lehmans failure. I think everyone – including Bernanke – now recognises that it was a mistake, but it really was an almost incomprehensible one. The justification was that the FOMC saw the risks of higher inflation as balancing the risks of lower growth. Now, as I noted in my post yesterday, headline inflation in September 208 was high – oil price effects mostly – but it is still hard to see how smart people could have reached the conclusion that a cut in the Fed funds rates was more risky than sitting tight. There was, for example, nothing disconcerting about the medium-term inflation outlook revealed in the breakevens in the government bond market (and by this time the Fed had already cut the Fed funds rate by quite a lot over the previous year. This is, of course, consistent with one of Sumner’s themes: central banks really should be taking more of a lead from market-price indications (which embody more wisdom and perspectives than a few dozen economists in any central bank can, and – he hopes – with less risk of (eg) groupthink).

It was a bad call. But in isolation it can’t have mattered much. After all, the FOMC went on to cut over the next couple of months, reaching their (self-identified) floor in December 2008.

The second questionable call was the move to pay interest on excess reserves held at the Fed (“settlement cash” in New Zealand parlance). I’ve written about this move in an earlier post, reviewing a book by George Selgin. This step was taken to stop short-term interest rates falling further…..in the depths of the most serious recession in decades…..by underpinning the demand for (willingness to hold) settlement cash. Selgin argued that this move deepened and extended the US recession – an interpretation I challenged in the earlier post – but it certainly dramatically changed any relationship that had previously existed between money base measures and wider nominal variables (nominal GDP, inflation, or whatever) – one of Sumner’s points – and did nothing to assist in getting inflation and activity back on course. (Our Reserve Bank made a similar decision last March, moving to pay the OCR on all settlement cash balances and thus underpinning short-term rates, at a time when the Bank also thought it needed to do massive bond-purchasing programmes.)

And while Sumner constantly (and rightly) cautions about simplistic reasoning from price changes, one of his other points about this period – and how the Fed was too slow and unaggressive in its approach – is how surprising it was to see real interest rates trending up over much of 2008. Discount the extreme surge if you like – though Sumner will argue that changes in market prices like that (tied in with risk aversion, market illiquidity) probably in any case support monetary easing – but that real yields were higher in January 2009 than in January 2008 does not sit that comfortably with a story of an aggressively-easing Fed.

(At the time New Zealand had only a single indexed bond, then with about 7 years remaining to maturity. Yields on that bond did not start falling until November 2008, even though the economy had been in recession all year.)

But there is a tension in Sumner’s book. At least early on there is a sense that he thinks the Fed could have avoided the recession together if only they’d done a better job, but the specific failings he explicitly highlights cannot credibly have been large enough in effect to have avoided the recession (and further on in the book he notes that they might only have dampened the severity of the recession, which is a much weaker – and harder to test – claim).

In many ways his starting point is that the responsibility of a central bank is (or should be) to manage nominal spending in ways that avoid big and disruptive fluctuations (which often involve recessions, and sometimes exacerbate periods of banking stress). I have no particular problem with that. I still prefer something like inflation targeting (at least in countries like New Zealand and Australia) as the operational form that responsibility takes, while Sumner now prefers nominal GDP targeting (preferably in levels form, but in growth rates still better than nothing.

His bolder claim seems to be that if there is a recession – and he makes explicit exceptions for one where, as in March 2020, governments temporarily close down economies/societies – it is the fault of the central bank. And that is a step far too far for me. He might be right that in an ideal world no one would ever unconditionally forecast a recession, since they would also forecast that the central bank would take the steps required to forestall it. And so he might be right to say that neither the housing bust nor the associated financial crisis caused the US recession – central bank failure to react in time did – but that seems to me to simply assume away the problem, in a way that there are no easy or quick substantive or technical fixes for.

Here is a chart of US nominal GDP growth (note, as we see it now, not as people first saw it at the time).

Sumner thinks the Fed should aim to keep nominal GDP on a path consistent with about 5 per cent annual growth. Clearly that did not happen in 2008/09. Nominal GDP fell by more than 3 per cent in the worst 12-months and (consistent with then policy) there was never a later overshoot to get back on that 5 per cent annual growth levels track.

Here is a similar chart for New Zealand (from the low inflation era)

Our nominal GDP path is a lot noisier than that of the US – commodity price fluctuations are a key reason why NGDP targets are not a good idea for New Zealand (or Australia) – but you can see how much nominal GDP growth fell away in the two recessions (1997/98 and 2008/09) – and actually in the double-dip recession in 2010 too. Broadly speaking that doesn’t count as a successful outcome – but then nor does the real GDP recession, the rise in the unemployment rate, and (the extent of) the sharp fall in the core inflation rate.

But here’s the thing though. Had Alan Bollard – then the sole decisionmaker – been presented with credible forecasts at the start of 2008 that nominal GDP growth was going to go negative over the coming year, I have little doubt that he would have been prepared to cut the OCR sharply (he wasn’t exactly an anti-inflation hardliner). But he wasn’t. Not just from the internal forecasters, but from the wider forecasting community or the financial markets. Inflation breakevens weren’t plummeting, long-term real interest rates weren’t falling, the exchange rate wasn’t falling much (as late as May 2008 it was still higher than it had been at the end of 2006). The share market was falling back but (a) the New Zealand share-market wasn’t very representative of the wider economy, and (b) few if any one in New Zealand has ever put much weight on local share prices as an indicator. The Bank did not cut early or hard enough (and I was one of Bollard’s advisers until August 2008 and although I was one of those more focused on global risks I wasn’t recommending deep early cuts)….but we did not have the information on which to do so. I would argue that no one did. In a sense, it was the point of that period…..for a very long time no one understood quite how bad some of the lending had been, or who was exposed, or what the macro consequences (absent monetary offset) would be.

From all I read or saw of the US at the time, and since, I don’t think anyone in the US did either. It wasn’t as if the Fed was totally blind: they had been taken by surprise in 2007, but actually starting cutting in September (you can see in the chart above, nominal GDP growth was beginning to slow).

Could the Fed or the RB have done better? Almost certainly (some identifiable Fed mistakes above), but it is inconceivable that they could have prevented the recession – not because the techniques weren’t there, but because the information and understanding wasn’t.

One of my criticisms of Sumner’s book is that he largely avoids this issue, and more or less assumes much more was achievable over 2008/09 (the issues re the recovery are different but note that in NZ and in the US markets were often keener on tightenings than either central bank – and Sumner urges paying more attnetion to market prices). An example of what I have in mind is his treatment of Australia.

We are told that

Among all the developed countries, Australia was the one with that sort of devil-may-care attitude, and it breezed through the Great Recession with only minor problems. And yet from a conventional point of view the Aussies did the least aggressive monetary stimulus. Unlike most other developed countries, they did not cut interest rates to zero.

He goes on to present a table showing that average nominal GDP growth in Australia for 2006 to 2013 was much the same as in the previous decade (unlike the US and the euro-area).

And yet….the RBA was still raising interest rates into 2008 (I recall a conversation at a conference in early 2008 at which a very senior RBA figure expressed astonishment at what the Fed thought it was doing keeping on cutting), the RBA ended up cutting by 425 basis points, there was a huge fiscal stimulus, and yet here is the Australian nominal GDP growth chart.

And yet look how much nominal GDP growth fell away in that downturn (a bit more than in the US). And it wasn’t as if there were no real consequences, with the unemployment rate rising 2 percentage points.

I’m not saying it was a bad performance…..it might even have been about as good as the authorities could have managed. But that is sort of the point. Limitations of knowledge, understanding etc…..not just in central banks, but much more broadly.

I could go on, exploring some of this points and Sumner’s specific policy prescriptions in more depth. But this post has probably gone on long enough already. He favours targeting a futures contract on nominal GDP, which may be a reasonable idea (at least in the US context), but it isn’t going to change the basic problem around knowledge. In countries like New Zealand (as Sumner notes) something like an aggregate wages series would probably make more macroeconomic sense (but would have its own political problems). I’m all for using monetary policy aggressively, but there are limits to what short-term stabilisation can be hoped for, no matter the indicator, the specific target, the instruments, or the individuals.

(Rather than labour points about nominal GDP targeting, I’ll link to some remarks I made on the topic at a conference a few years back.)

Good books make you think, and think harder. The disagreements are often where the most value lies in forcing one to think harder about one’s own view. This one is worth reading and reflecting on. And I’m going to finish where the book does with a quote I endorse (even if a bit more relevant in the US than here):

In other words, the goal is a world in which policy makers don’t view fiscal stimulus or the bailout of bankrupt firms as a way of “saving jobs”, but rather as a sort of crony capitalism that favors one sector over another.

But there will still be recessions, real and nominal.

No posts here for a while as I’ve been bogged down in trying to make sense of some events – little more than one week in history – from 30 years ago, where the uncertainty as to what actually happened (a precondition for making sense of what the events mean) is greatly magnified by really poor documentation and recordkeeping by….the Reserve Bank.

I was planning to return with something a bit more longer-term (perhaps tomorrow) but wasn’t yesterday’s inflation number interesting? It seems to have taken almost everyone – notably the people who do detailed components forecasts, including the Reserve Bank – by surprise to some extent.

Almost all the media focus has been on the headline number – 2.2 per cent increase for the quarter, 4.9 per cent for the year – because (I guess) it makes good headlines. (Excluding the two quarters when the GST rate was increased) it was the largest quarterly increase since June 1987 – an unexpected rise of 3.3 per cent, at a time when the Bank thought inflation was falling away, and when the Bank’s chief economist, Grant Spencer, was interviewed about the number that night he declared himself “flabbergasted”. That one number helped prompt an overhaul of, and marked improvement in, the Bank’s short-term inflation forecasting (not previously much of a priority).

But in annual terms, it is only 13 years since we had an inflation rate about this high. It was 5.1 per cent in the year to September 2008, a rate that may be beaten when the next CPI number is released in January, since last December’s (relatively modest) 0.5 per cent quarterly increase will drop out of the annual rate. Note that in the September 2008 quarter, the Reserve Bank had already (and appropriately) started cutting the OCR.

But my main interest is in core inflation. There are all sorts of different measures, from simple ones (useful for cross-country analysis at least) like the CPI ex food and energy, through varying degrees of complexity (and occasionally even special pleading by the people constructing them). For New Zealand though, my favourite measure – and the one the Bank openly favoured for some years (it is less clear how the current Governor and MPC see things) – is the sectoral factor model measure of core inflation. It was developed a decade or so ago by one of the Bank’s researchers, and initially got little attention even inside the Bank (mostly because the Governor and his advisers on the then Official Cash Rate Advisory Group were not really advised of it). I’ve been something of a lay evangelist for this measure ever since I realised it existed, and had some small role in getting this explanation of the measure published. The gist of what is going on is this

The sectoral factor model estimates a measure of core inflation based on co-movements – the extent to which individual price series move together. It takes a sectoral approach , estimating core inflation based on two sets of prices: prices of
tradable items, which are those either imported or exposed to international competition, and prices of non-tradable items, which are those produced domestically and not facing competition from imports.

Using very disaggregated data, it is an attempt to get at the systematic elements in the annual inflation numbers, recognising that tradables and non-tradables can be influenced by different systematic influences (notably the exchange rate in the case of tradables).

But the best argument for the series has been its usefulness – in some sense it “works”, telling useful stories, not subject to much revision, about what is going on in ways that square with what is going on with other things (notably capacity pressures, but also expectations) that are thought likely to be important influences on the trends in inflation, abstracting from the noise.

And the “noise” can be considerable. Here is annual headline inflation and the annual sectoral factor measure for the period since 1993 (as far back as the sectoral factor measure has been taken).

Big deviations have not been uncommon (although less so in the last decade), and spikes in headline inflation have never (yet) foreshadowed a commensurate increase in core inflation (as,say, stickier prices caught up with more flexible prices). If you did want a prediction of where core would be 12-24 months from now, historically today’s core inflation has been a much less bad (far from perfect of course) predictor than today’s headline inflation.

And so from here on I’m focusing solely on the core inflation measure. There are a few observations worth drawing from simply this chart.

First, the range in which core inflation has moved over 28 years has been 1.1 per cent to 3.5 per cent. And although the inflation target was centred on 1 per cent until the end of 1996 and 1.5 per cent until September 2002, the low in the series wasn’t then, but in late 2014 (a time when, curiously, the then-Governor was raising the OCR).

Second, over the 28 years not much time has been spent very close to the midpoint of the respective target range. In fact, the median gap between the core inflation estimate and the target midpoint has had an absolute value of 0.7 per cent over the history of the series. As it happens, yesterday’s core inflation estimate was 2.7 per cent, 0.7 percentage points above the target midpoint.

Third, for the first 15 years of the series core inflation was almost always at or above the target midpoint, and for the decade until last year it had been consistently below.

Now it is worth pausing here to note that prior to about 2012 the Reserve Bank (a) did not have the sectoral factor measure readily available to policy advisers, and (b) was not explicitly required to focus on the target midpoint. However, neither point really diminishes the usefulness of such comparisons because (a) sectoral core inflation was simply trying to put in a single measure something the Bank had constantly thought and written about since inflation targeting began, and (b) if Alan Bollard was personally disinclined to give much weight to the target midpoint, Don Brash certainly was (and revealed evidence – see those sectoral factor numbers from 2014 – suggests that Graeme Wheeler was more focused on where he thought in principle the OCR should be heading than on the target midpoint.

There are a couple more relevant observations. First, core inflation now (2.7 per cent) is about the same as it was (2.6 per cent) in the last year or so of Don Brash’s term (2001/02), and back then the target midpoint was 1.5 per cent, not the 2 per cent the Bank is now charged with. And, second, core inflation is still well below the 3.4/3.5 per cent seen in late 2006 and throughout 2007.

But perhaps the change in the inflation rate has been unusually sharp.

I put this chart on Twitter yesterday before the Bank published the sectoral core numbers.

On the series now published, the sectoral core inflation rate rose by 0.3 percentage points in the latest quarter (so large but not exceptional). However, this sort of model is prone to end-point revision issues – new data leads the model to, in effect, res-estimate which recent prices moves were systematic and which were not. The previous estimate for sectoral core inflation for the year to March 2021 was 2.2 per cent. But that has now been revised up to 2.4 per cent. I don’t have (but the Bank should really publish) a database of historical real-time estimates, but a change from a previous estimate of 2.2 per cent in the year to March to one of 2.7 per cent for the year to June is likely to have been large by any standards.

What about changes from year to year? Again, I don’t have a real-time database, but here is how the annual rate of core inflation has changed from that a year earlier.

What we’ve seen so far – on current estimates which are subject to revision – is not exceptional. The rise in the rate of core inflation over the last year has been less than we saw around the turn of the century, and the magnitude of change is less than than the fall seen over 2009. But it isn’t a small change either.

When (last quarter, per the Bank’s published estimates) annual core inflation was estimated at 2.2 per cent, I was prepared to say (and did) that that rate of core inflation was unambiguously a good thing, given the target the government had set. After a decade of core inflation below the target midpoint, it was good to finally see an outcome on the other side, which would help to underpin medium-term expectations near the goal set for the Bank. That was doubly so because 2.2 per cent inflation went hand in hand with an unemployment rate right back down to pre-Covid levels (4 per cent) and probably pretty close to the NAIRU (itself a rate the Bank can’t meaningfully do anything about). I’d not have been uncomfortable with a core inflation rate going a bit higher still – not as a desired outcome, but not something to be too bothered about for a short period (as the MPC raised the OCR, which works with a lag). 2.7 per cent is somewhat less comfortable.

But quite a lot might have depended on where the unemployment rate (or other measures of excess capacity) was going. There have been two previous troughs in the unemployment rate. The first was in the mid 1990s, when the NAIRU appeared to be around 6.2-6.5 per cent. Core inflation reached its cyclical peak then at much the same time unemployment dropped into that range, and showed no signs of going higher. The second was just prior to the 2008/09 recession, when the unemployment rate was in the 3.4-3.9 per cent range. Core inflation had risen as the unemployment rate fell, but core inflation was not going higher in 2007, nor was it forecast to into 2008. In both cases, the Reserve Bank had been raising interest rates (or allowing them to rise) and things stayed more or less contained (before core inflation fell away in the two following recessions).

One of the great unknowns now is how things might have unfolded here without the Delta outbreak and the ongoing restrictions and lockdowns. We will get the HLFS numbers for the September quarter early next month, and the unemployment rate there is unlikely to have been much affected yet by the lockdowns etc. Most likely, the unemployment rate will be lower than 4 per cent, but how much?

But the outbreaks and restrictions did happen, and so even if the unemployment rate for the September quarter was in fact 3.6 or 3.7 per cent, it probably isn’t safe to assume anything of the sort as a December quarter starting point. Yes, most likely economic activity will eventually rebound when controls are finally lifted but (a) there isn’t the fresh policy impetus there was last year, and (b) for those who believe in house prices stories, the worst of this particular house price boom has most likely passed. It isn’t implausible that the unemployment rate for December and March could be back at or above 4 per cent.

What does it all mean for policy? No doubt the MPC is feeling vindicated in having raised the OCR at the last review, even amid the-then extreme Covid uncertainty, and even though the MPC is likely to have been very much taken by surprise by yesterday’s core inflation number. Absent Covid there was a strong case for a robust tightening of monetary conditions – reversing the LSAP bond purchases, ending the funding for lending programme, and getting on with OCR increases – and that case would have been considerably strengthened by yesterday’s outcome.

Perhaps fortunately, the MPC does not need to make another OCR decision until late next month, and that review will come with a full MPS which will allow them space to provide some careful and considered analysis of their own. We might hope that by late next month, something close to normality has returned or is on the brink of returning. There are no guarantees, but if that is the situation, the MPC should be starting to sell off the bonds, and ending the FfL programme (most likely they will do neither), and should probably still be considering seriously a 50 basis points OCR increase (albeit with one eye on the emerging China slowdown). We were told they had considered the option in August. There isn’t a need for panic or headless-chookery about the Bank having lost the monetary policy plot. But a fairly robust response does seem likely to be warranted next month, especially as the MPC has (most unwisely) scheduled decision dates in a way that gives them a long summer holiday with no OCR review at all in December and January.

Finally, I have been highlighting for a long time how the market-based indications of inflation expectations (from the indexed bond market) had consistently undershot the target midpoint for some years. Yesterday’s data seems to have prompted a move to (or above, depending on maturity) 2 per cent for the first time in a long time. That isn’t concerning – rather the contrary – but it will be worth keeping an eye on how those spreads – the breakevens – develop over the period ahead.