High inflation, and better-than expected recoveries, have pulled forward the expected lift-off dates for global central banks. This dynamic has been particularly notable in the case of New Zealand, where both the supply and demand effects of the post-COVID period have been clearly inflationary. While I have been hawkish on NZ, I am dovish on Australia – because Australian inflation outcomes have been so different to peers.
The divergence between Australian inflation and inflation in other advanced economies is something that the RBA picks up on in their October meeting minutes. My read on these minutes is that the RBA is trying to push back on market pricing for aggressive hikes.
Members concluded their discussion of domestic economic developments by observing that underlying inflation pressures in Australia were more moderate than in other advanced economies. This reflected a range of factors, including the relatively slow rate of wages growth in Australia. Members noted that, while it was possible that underlying inflationary pressures in Australia could build more quickly than currently envisaged, the central forecast scenario was still that domestic inflation would pick up only gradually over the medium term.RBA October 2021 Meeting minutes, published 19 Oct 2021 (my emphasis)
Market pricing for RBA hikes over the next two years is currently the most aggressive since 2010 (see chart at the bottom of this post), despite the fact that the RBA continues to miss on inflation and isn’t expected to hit their target inside the forecast horizon. Fading market pricing looks like an opportunity.
Australian Inflation: too low for too long
Inflation outcomes in Australia have been below target for some time. The RBA’s preferred measure of core inflation hasn’t met their 2.5% target since 2014. While headline is currently a bit above the RBA’s 2.5% target, the RBA’s preferred measure of core inflation has not yet re-entered the bottom of their 2% to 3% control range. Forecasts are just forecasts and they are frequently wrong — but it’s worth noting that the RBA doesn’t expect inflation to meet its 2.5% target over the projection period.
The test for a hike is inflation that’s printing near that 2.5%yoy mid-point and expected to remain there — which probably means trimmed mean inflation near 2.5%yoy and the wage-price index above 3%yoy. We are a few upgrades away from the RBA having that forecast, let alone it being achieved. As of 2020, the test became realised inflation, so trimmed mean inflation must be both achieved and forecast to be sustained.
The obvious question is why?
Part of the reason that Australian inflation has been low is that Government subsidies have held down various costs, particularly housing. Housing inflation will start to bleed into CPI in 2022, as the subsidy winds down. However, only some of the prior increase in building costs is likely to show up in CPI. A portion of the increase reflects short term supply problems and the ‘pricing in’ of the Government subsidy by builders, and so will likely drop out over time. While the HomeBuilder subsidy is a reason CPI inflation has been low to date, it’s not necessarily a reason to expect an inflation pop over the horizon. There are a few other special factors likely to depress Q3’21 CPI.
Another part of the reason inflation has been low is because it has been low for a long time. The long period of low inflation appears to have depressed inflation expectations. My analysis shows a clear break lower of trend inflation around the middle of the last decade. I associate this with the change of the monetary policy regime when Lowe took over as Governor & re-wrote the agreement to place greater emphasis on financial stability. Lower inflation expectations make it harder to get inflation back up.
What can we learn from NZ CPI?
I used to enjoy the read-through from NZ CPI to Australian CPI, but it’s become less useful. The correlation at the headline level is about 70% (since 2001), so it looks like an attractive ‘lead’ on Australian CPI. However, it should be noted that this is mostly historical: the correlation drops below 50% for the period starting 2014.
The main reason it used to work is because Australia and New Zealand are both price takers in many import markets. As a result, there was a high correlation between tradable goods prices in both countries. Again this is mostly historical, with the correlation between tradable goods inflation in the two countries falling below 50% for the period starting 2014. While the large increase in energy prices is a common shock that’s very likely to flatter the near term correlation, I’d caution that energy prices are almost certain to be stripped out of the RBA’s preferred core measures (the trimmed mean and weighted median CPI).
The correlation between non-tradable inflation in Australia and NZ is 30% for the period beginning in 2001, and declines to 24% for the period beginning 2014. I’ve never had much joy mapping between NZ non-tradable CPI and Australian non-tradable CPI. It’s mostly noise.
Of course, what really matters for monetary policy is core inflation. The RBA and the RBNZ prefer different core methods, however both aim to separate the signal from the noise. The correlation between the Australian Trimmed Mean CPI and the RBNZ’s Sectoral Factor Model of CPI is about 67% for the period starting in 2001, however this drops to -75% (yes they have an INVERSE correlation) from 2014.
The correlation between AUD and NZD core measures is more stable in differences, though it’s very weak: at 17% 2001+ and 15% for 2014+ … this means that they have a weak tendency to speed up and slow down together.
Focus on the distance to target …
The bottom line is that you can’t map from NZ inflation to Australian inflation for the purposes of predicting monetary policy (or trading the short end). It just doesn’t work. The relationship between the monetary-policy relevant CPI measures is too weak. Also, the reaction function is different: for starters, NZ targets forecasts whereas the RBA is now awaiting actual inflation’s return to ~2.5% before starting to hike.
One way to think about the value in the curve is to compare market pricing to the distance to target (the gap between CPI and the central bank target). This has been done in the chart below. For the purpose of this chart, I’ve assumed that Q3 Aussie CPI is 1%qoq & 3.25%yoy. This is a bit higher than the RBA had in their August SOMP, mostly reflecting the higher oil price since those forecasts were made.
The chart below shows that the gap between headline CPI and the RBA’s 2.5% target is the smallest of the six central banks (Fed, ECB, BoE, BoC, RBNZ and RBA). This reflects some special factors depressing Australian CPI, and the RBA’s higher CPI target (2.5%yoy v. 2% for all the other central banks). In contrast, RBA pricing (I’ve used the 2y1m v. 1m OIS spread) is a little more aggressive than USD, and a bit less than CAD and NZD.
The RBA prefers trimmed mean inflation for policy-making. Assuming that trimmed mean CPI is 0.5%qoq & 1.8%yoy (again, a bit higher than the 1.7%yoy the RBA had in their August SOMP), the RBA is the only central bank that’s going to have core CPI below their target in Q3’21 (I’ve used core PCE for USD; 30% trimmed mean for EUR; Ex Energy, Food, Alcohol & Tobacco for GBP; Trimmed Mean CPI for CAD; and the Sectoral Factor Model for NZD). Notwithstanding that the RBA is missing on inflation, and not expected to get close until 2024, there are lots of hikes priced. Market pricing for RBA hikes is more aggressive than for EUR, GBP and USD, but less aggressive than CAD and NZD. That seems wrong.
Given that the RBA has most recently said that they’d like CPI to be sustainably around the middle of their 2% to 3% target range before tightening rates, and that core inflation isn’t expected to get to 2.5% in the current forecasting envelope (out to the end of 2023) it’s hard to justify market pricing.
The best explanation I have is that the market tends to club AUD with USD, CAD and NZD rates. This can create opportunities when central banks act — as we just saw with the large move in the AUD v. NZD short end spreads.
I think the largest opportunity just now is to receive AUD short end against USD and GBP. Both central banks seem capable of delivering on current pricing. AUD pricing seems more like risk premium.
Australian inflation might well catch up later, but there are special factors that are depressing it just now (including a few in Q3’21 that make a spike unlikely). It’d take a huge upside surprise to put Australian core inflation anywhere near 3%yoy, which is what’s needed to place Core CPI a similar distance from the RBA’s target to other central banks (which is what I think would justify current market pricing).
With the RBA likely to be on hold as other central banks tighten, this should see the AUD depreciate. Ultimately a lower AUD will help to import some global inflation and may bring forward hikes — but it’s very hard to see the RBA beating current forwards. There’s presently the most priced since 2010.