On November 16 the RBA Governor gave a detailed speech with an extensive Question and Answer session to the Australian Business Economists. Westpac remains comfortable with its key views: that the first-rate hike in this cycle will be at the February Board meeting in 2023 (held since June this year despite legendary market and forecaster volatility) and that the Bond Buying Program will be wound back to $2 billion per week in February 2022 prior to the program being wound up in May.
1. Timing of the First-Rate Increase
Readers will be aware that the Bank has quite cautious forecasts for end 2022 – underlying inflation at 2.25%; wages growth at 2.5%; and the unemployment rate at 4.25%.
That compares with Westpac’s forecasts of – underlying inflation rate of 2.8%; wages growth of 2.75%; and the unemployment rate at 3.8%.
The RBA’s forecasts are consistent with the Governor’s ongoing implied preference to delay the rate hike until 2024.
But if we are right, he will have overachieved on his inflation target; will have some evidence that the trajectory for wages growth is accelerating; and will have good reason to expect that as the unemployment rate falls below 4% – (something we have not seen since the 1970’s) the response of wages growth could be somewhat non – linear. For Australia, the “flat” Phillips curve has not been tested at unemployment rates below 4%.
It is true that he raised the bar on the importance of wages growth by answering one question that asked for his policy response if underlying inflation printed 3% or higher but wages growth was lagging behind at 2–2.5%. His argument was that inflation I the target zone would not be sustainable without wages growth around 3%.
But unlike the policy prescription that rates will not be raised until inflation is sustainably within the target band and we have reached full employment a particular rate for wages growth is not part of formal policy.
We expect that ultra low unemployment complemented by an upward trajectory for wages growth will allow a rate hike even if the latest annual print for wages growth is below 3%.
For markets which are expecting a much earlier rate increase the Governor does point out the inertia in his preferred measure of wages growth (Wage Price Index). The Australian wage setting scene includes an annual minimum wage agreement; 2–3 year enterprise bargaining agreements; and a slow moving wage setting process for the public sector.
It is only the informal sector where individuals negotiate directly with employers that any immediate labour shortages appear quickly in the WPI. We saw some evidence for that in the September quarter WPI which printed on November 17 where the sector receiving the largest increase – professionals, including lawyers and accountants, registered 3.4% for the year and 1.3% for the quarter – a sector where individual agreements prevail.
This inertia in the WPI is also behind our own view that the WPI will still only be increasing by 2.75% by end 2022. Of course, the RBA will be watching for the informal component of the WPI to provide a “lead indicator” for other components. The upward trajectory in growth of WPI along with the signals from the informal sector; should be enough to allow a rate move despite the official wages growth print below that “magic” 3%. Also recall that the September quarter WPI will print after the
November Board meeting but will be the last print before the February meeting in 2023. Unlike the CPI which will be available for the December quarter the WPI measure will be a little “stale” by February, but other high frequency data should provide the Board with enough evidence to start the process despite the official WPI growth rate holding below that 3%.
2. The Quantitative Easing Program.
In the speech the Governor gave some emphasis to the revised forecasts which the Board will receive for the February Board. Recall that following the lock downs in both Sydney and Melbourne the Board deferred its decision to review its bond buying program at the November Board to the February Board.
At the November Board the forecast for underlying inflation in 2022 was lifted from 1.75% to 2.25% and 2023 was lifted from 2.25% to 2.5%. It is reasonable that if there had been a review in November that recognition in the forecasts of much faster progress on the inflation target would have justified a scaling back in the purchase program from $4 billion per week to $3 billion.
We think the 2022 forecast for underlying inflation can be further scaled up at the February Board.
That would be consistent with a reduction in the purchase program to $2 billion, incorporating a catch up from November along with further recognition about the inflation target in February.
Now that the RBA has discontinued its Yield Curve Targeting having not extended the Term Funding Facility, we are only left with one more component of unconventional policy – the bond buying program.
The program began in November 2020 with a commitment to purchase a total of $100 billion in bonds ($80 billion in AGS and $20 billion in local and semi government securities) at a pace of $5 billion per week.
A second $100 billion tranche was announced at the February 2021 Board meeting which would extend until September 2021.
In July the Board announced that it would further extend the purchase program from September at a scaled back $4 billion pace with a review at the November Board meeting.
In response to the delta related lock downs in NSW and Victoria the Board announced that it would delay the November review of the program until the February 1 Board meeting in 2022.
Following the November Board meeting where the Governor announced that the Yield Curve Target would be dropped, he confirmed that the bond buying program would be reviewed at the February Board.
He noted that the review would be based on three considerations:
1. The actions of other central banks.
2. How our bond market is functioning.
3. Most importantly, the actual and expected progress towards
our goals for inflation and unemployment. And added in the speech the importance of the revised forecasts (as discussed above).
Other Central Banks and Market Conditions
I think the two central banks that will have the most significant impact on the Board’s deliberations will be the US Federal Reserve and the Bank of Canada.
I choose these two Banks because the FOMC is the benchmark for QE policy while the BOC is often cited as having a similar steady approach to policy, in contrast, say, with the recent more volatile policies from the RBNZ and, arguably, the Bank of England.
Following the RBA’s meeting on November 2 the FOMC decided at its meeting later in that week that it would begin reducing the monthly pace of its purchase program (currently $80 billion in Treasury bonds and $40 billion in agency mortgage-backed securities) by $10 billion for Treasury bonds and $5 billion for mortgage- backed securities.
By end December it will have reduced its monthly purchases to $60 billion in Treasuries and $30 billion in agencies.
At that pace the purchase program can be expected to cease in June. The FOMC allowed for the possibility of adjusting that pace if warranted by the economic outlook.
On October 27 the Bank of Canada announced that it was ending quantitative easing and moving into the reinvestment phase during which it will purchase bonds solely to replace maturing bonds.
Because the RBA did not begin QE until November 2020 and its purchases were in the 5–10 year maturity window it will not be confronted with a large scale maturity program until around 2025, although its much more modest purchases of April 2023 bonds under the first stage of YCT will mature in 2023.
The BOC has been ahead of the RBA in terms of its QE. It began its program in March 2020 by purchasing $5 billion in bonds per week. During 2020 it cut the purchase pace back to $4 billion and at the time of the decision to cease purchases had reduced weekly purchases to $2 billion.
As with the FOMC’s current plan the BOC’s policy has been to gradually scale back purchases. That approach, complemented by the reinvestment policy, is akin to the RBA’s condition of supporting a smoothly functioning bond market without any unintended consequences of a sharp withdrawal of central bank support for the market.
By mid February the RBA estimates that it will hold 36% of AGS outstanding and 18% of semi and local government bonds.
That share of AGS will compare with BOC of around 44% and FOMC of around 30% in central government holdings.
Note that the RBA’s run up in holdings has been much sharper than FOMC or BOC.
In early 2020 RBA held around 2% of total AGS compared to around 16% by the BOC and 18% by the FOMC.
So, RBA’s holdings of AGS in terms of the outstanding volume will be high but not excessive by February.
In the “excessive” camp we would class Japan; UK; and Euro area at near 50%.
Central banks claim to assess the impact of their QE policies in terms of the size of their balance sheets relative to the size of the economy. That is considered to be much more important than the flow of new purchases.
Deputy Governor Debelle noted (May 2021 – “Monetary Policy during Covid”) that “it is the stock of central bank purchases that matters rather than the flow…. It is the total size of the purchases that affects bond yields and financial conditions including the exchange rate rather than how many bonds the central bank is buying each week.”
In that regard the RBA’s holdings of AGS will be around 16% of GDP by February compared with nearly 20% for the BOC and around 22% for the FOMC.
So the Board should assess that its QE policies are having a comparably stimulatory impact on the economy as is the case in both US and Canada.
Actual and Expected Progress Towards Goals
When the Board meets on February 1 next year, we forecast that the key data points it will focus on will be progress on inflation (the print for the December quarter CPI will be available on January 27); employment (December employment report mid January); wages (December report only available in late February).
We expect that the Board discussion in February will be in the context of an underlying inflation print of 2.3%; an unemployment rate of 4.9%; and wages growth of 2.1%. (to the September quarter); and high frequency data indicating a sharp and sustainable lift in demand in the context of a 90% national vaccination rate amongst eligible adults.
That will be signalling an improving outlook but not sufficient improvement to justify the ending of the program.
Cutting weekly purchases from $4 billion to zero might also have some potential disruptive effects for the bond market (market functioning will be a factor in the decision).
With the Governor currently seeking to calm impatient financial markets with respect to the timing of the beginning of the tightening cycle a decision to abruptly end the bond buying program might trigger more unwanted volatility in the market.
But progress towards the goals will have been clear (especially on the inflation and growth fronts) so a tapering of purchases can be justified.
In fact, cutting the purchase program to $2 billion per week would be a sensible balance between stability and recognising the need to move away from the unconventional policies that were justified by the Covid emergency.
That $2 billion purchase program would be reviewed at the May Board meeting.
By the May meeting we expect that underlying inflation will print 2.5%; wages growth will have lifted to 2.3%; the unemployment rate will have fallen to 4.7%; and growth momentum in the first half of 2022 will be assessed at around 5%.
The FOMC will be very close to the end of its QE program (expected in June) while the BOC will have already established the precedent of a step down from $2 billion per week to zero.
Conditions will be ideal for an orderly end to a very successful unconventional policy initiative.
The scene will then be set for the first-rate increase at the February Board meeting in 2023.
