Annual CPI inflation fell to 1% in the September quarter, which puts it at the bottom of the 1-3% inflation target range; noting Governor Wheeler has stated he is targeting the mid-point (i.e. 2%).
Are structural factors at work that mean inflation will remain low even if the Reserve Bank (RB) allows economic growth to run above the rate it estimates to be consistent with low inflation (i.e. the RB's 2.75% estimate of the "sustainable" growth rate)?
The factors include:
(1) low global inflation or possibly deflation (i.e. falling global prices);
(2) downward pressure on local retail prices because of increased competition from online shopping; and
(3) developments that mean labour cost inflation will be lower than normal.
This Raving investigates these three issues and concludes that we haven't yet, at least, entered a new inflation paradigm.
The implication is that if Governor Wheeler allows economic growth to run above the current estimate of the sustainable rate for much longer an inflation problem will develop in time.
If the governor waits for signs an inflation problem has developed before hiking again he risks repeating a watered-down version of Governor Alan Bollard's boom-bust experiment that ultimately resulted in 13 OCR hikes (i.e. timely hikes reduce the ultimate need for hikes).
Have we entered a new inflation paradigm?
The 1% annual inflation rate for the year ended September was well below what the RB predicted (left chart), with inflation in both the domestic/non-tradable component of the CPI and the tradable component coming in below the RB's forecasts (right chart).
This appears to have played a significant part in the more dovish stance adopted by Governor Wheeler last week.
It raises the question of whether local and/or international factors have resulted in a new inflation paradigm that mean GDP growth can run above the RB's 2.75% estimate of the sustainable rate without it fuelling an inflation problem.
The counter view is that a range of one-offs resulted in a temporary dip in inflation, while it could be a bit of both.
One possibility is that low global inflation or even deflation is at work (i.e. a range of developments, including the impact of technology, resulting in downward rather than upward pressure on the global prices of goods and services).
If this was the case NZ import price inflation should be running lower than would be expected just on the basis of what changes in the exchange rate suggest should be occurring.
The left chart below shows that annual import price inflation has been behaving much as would be expected given the appreciation in the NZD TWI (the right hand TWI scale is inverted).
Falling oil prices should contribute to lower import price inflation, but the fall in the NZD TWI is starting to point to the risk that annual import price inflation will turn positive again.
See below for a survey that points to retailers planning on putting up prices, which is most likely in response to the fall in the NZD. This suggests there isn't yet at least a global deflation at work resulting in lower NZ inflation than would be otherwise expected.
The next possibility is slippage between import and export price inflation on the one hand and inflation in the tradable component of the CPI on the other hand.
The tradable component of the CPI measures the local prices of internationally traded goods and services (i.e. exports and imports).
The right chart shows that tradable inflation is running reasonably closely in line with what would be expected based on annual inflation in import and export prices combined. We used a 70% weight for import prices because imports are largely consumed and a 30% weight for the export price index because exported goods make up a smaller share of local consumption.
The fall in tradable inflation in the September quarter fits with what is likely to have occurred with import and export prices in the September quarter.
The next possibility is that online retailing and especially local consumers purchasing items from overseas websites is putting local retailers under pressure to cut prices to compete (i.e. the technological revolution is resulting in at least a transitory reduction in local retail price inflation).
Anecdotes we have heard suggest this is happening to various degrees, but it is somewhat hard to measure. We have used the ANZ monthly business survey to shed some light on this issue.
We conclude that this issue could be of some significance (i.e. online retailing depressing prices), although this should be largely a transitory effect that ends when local retailers have fully adjusted to the challenge posed by increased online purchasing.
If this was a major issue we would expect retailers to report poorer profit expectations than firms in general (adjacent chart).
Retailers are a bit less positive about profits than all firms, but no more than has been the case at times in the past. If retailers were under major pressure to cut prices in response to increased competition from online shopping we would expect retailers' price intentions to be weaker than all firm intentions, which isn't the case.
The left chart below shows periods when retailers' price intentions have deviated significantly from all firm price intentions.
The right chart below compares the gap between retailers' and all firm price intentions (left hand scale) and the annual % change in the NZD TWI (right hand scale, inverted).
The exchange rate has played a major part in driving the divergences in retailers' and all firm price intentions. There have been two periods since early-2013 when the gap between retailers' and all industry intentions was negative and not justified by changes in the exchange rate. So there is some evidence suggesting retailers have had to cut/curtail pricing potentially in response to increased competition from online retailing.
But most recently there are signs retailers are responding to the fall in the exchange rate in a pretty normal manner (i.e. the adjustment phase may be over, at least for now).
If this is the case it should be a warning sign for Governor Wheeler.
The next question is whether structural changes in the labour market mean the unemployment rate can run at lower levels than in the past without fuelling an inflation problem, which would mean economic growth could run above the RB's 2.75% estimate of the sustainable rate without causing an inflation problem.
The left chart below shows a high negative correlation between the unemployment rate and the productivity-adjusted measure of labour cost inflation the RB focuses on most. The peak correlation is with the unemployment rate advanced or leading by three quarters. The chart also includes the RB's September predictions for both.
The productivity-adjusted measure of labour cost inflation measures the pure inflationary component of labour cost increases (i.e. increases not justified by increased productivity). Increases in productivity-adjusted labour costs mean labours costs per unit of production increases, which firms will respond to by putting up prices and the end result will be a self-defeating wage-price spiral. This is distinct from "good" labour cost increases that are justified by improvements in productivity (i.e. increased output per hour worked) that don't result in a wage-price spiral developing.
Recently labour cost inflation hasn't increased in response to the initial fall in the unemployment rate (adjacent chart). The unemployment line has been advanced or shifted to the right by three quarters to allow for how long it takes on average for changes in the unemployment rate to be reflected in the productivity-adjusted measure of labour cost inflation.
The fall in the unemployment rate over the last three quarters points to potential upside in productivity-adjusted labour cost inflation over the next three quarters.
But there is nothing new in this. The same thing has occurred twice in the past, with labour cost inflation not responding to the initial falls in the unemployment rate (see the green boxed areas in the chart above).
In the previous two cases labour cost inflation did start to increase after the unemployment rate had fallen below "threshold" levels.
The unemployment rate has this year fallen to levels that in the past have started to fuel labour cost inflation.
Consequently, I believe it is too early for the governor to conclude the lack of upside in labour cost inflation so far means we have entered a new inflation paradigm in the labour market.
In the September Monetary Policy Briefing report we discussed in detail whether a structural increase in the participation rate was underway that would result in lower labour cost inflation for any given level of economic growth.
We concluded that the structural increase in the participation rate was probably over (i.e. the percentage of the working age population working or seeking work).
The big story has been the increase in the female participation rate, but the third order polynomial trend line in the left chart below suggests the trend increase in the female participation may have ended for now at least. In the case of the male participation rate we showed in the Monetary Policy Briefing reports that it has largely responded to cycles in the labour market with no clear increase in recent years. There is even the suggestion of a trend fall by the third order polynomial trend line (right chart below).
If the structural increase in the participation rate is over it means future increases in employment will largely result in a lower unemployment rate that with around the normal lag will drive up productivity-adjusted labour cost inflation.
The question of whether we have entered a new inflation paradigm is still open to debate.
But the analysis above doesn’t support Governor Wheeler embarking on a go-for-growth experiment in the hope there won't be inflationary consequences.
-----------------------------------------------------------------------------------
*Rodney Dickens is the managing director and chief research officer of Strategic Risk Analysis Limited.