Once Again, Trimmed Mean Gets the Casting Vote

Surprisingly strong labour market data means the RBA’s February decision will (again) hinge on next week’s inflation data. Chance of a hike shifts closer to a 50:50 call but could end up being an awkward ‘one-and-done’.

• RBA does not really run policy by just reacting to the latest trimmed mean inflation print, but sometimes it looks that way. The inflation data (due next week) will once again have the casting vote in the RBA policy meeting in February.
• Stronger labour market and consumer spending data are feeding into a narrative that the Australian economy is running too hot. We think the true situation is more nuanced, with supply capacity expanding faster than the RBA and some others believe. Even so, these stronger data offset the downward revision to our expectations for Q4 inflation. An upside surprise next week relative to these downwardly revised inflation expectations will tip the balance over to a February hike.
• While the near-term starting point for domestic demand and the labour market is stronger than previously understood, a near-term hike will dampen the outlook further out. It will also pose some awkward questions about why policy needs to be almost as tight to squeeze out the last half a point of inflation as it was to reduce from 8% to 3%, and why it needs to be so much higher than expected in peer economies.

In a recent interview, RBA Deputy Governor Andrew Hauser complained that some people seemed to think the RBA runs policy as a knee-jerk response to the latest inflation data. That is not really how they do it, of course. The RBA looks at every piece of data, including some things the private sector does not get to see. Yet for the past year or so, quarterly CPI prints, especially the trimmed mean, have been unusually consequential for monetary policy decisions.

If it is any consolation to Deputy Governor Hauser, we are as annoyed about this as he seems to be. Things should never be this line-ball, but it is partly an outworking of the RBA’s decision-making process. When the other data are finely balanced, the data point that comes out last gets the casting vote. And because the forecast rounds ahead of RBA meetings are timed to take place just after that CPI release, it appears as if everything turns on it.

In addition, the RBA’s current assessment of the balance of supply and demand in the economy – and the weight this has in its decision-making – makes the inflation print even more consequential. If inflation surprises on the upside, its thinking goes, it must therefore be that the balance of supply and demand was tighter than previously believed. For any given GDP outcome, this therefore leads the RBA to conclude that supply capacity is weaker than previously thought, and this conclusion from the inflation print is carried forward into the RBA’s view of the broader outlook. While more recently there have been some attempts to triangulate this view of supply against business survey measures, by design recent inflation surprises heavily influence policy outcomes.

With this week’s release of surprisingly strong labour market data for December, we are – once again – in the position of the cash rates view hinging on the CPI print next week. We have previously flagged the possibility that another uncomfortably high inflation result will induce a near-term hike in the cash rate at the February meeting. That possibility shifted closer to a 50:50 bet with the labour market data. It could tip over the line if trimmed mean inflation prints higher than expected.

At the same time, we note that the first two months of the quarter point to some downside risk to the inflation print. We have incorporated some of this into our nowcast for inflation, with the quarterly trimmed mean measure now expected to print at 0.7%qtr (3.1%yr). An outcome at this level or lower should be enough to stay the RBA’s hand, at least for now. The rhetoric would remain hawkish, but such a result would support the RBA’s earlier assessment that some of the surprise September quarter inflation result reflected temporary factors.

If, however, the December quarter inflation result surprises us on the upside next week, the RBA will likely see little alternative to raising the cash rate in February. Such a response would buy into a narrative that the Australian economy is running too hot with interest rates at current levels. The RBA might even take the view that policy has been stimulatory with the cash rate at current levels and that the previous cuts were in hindsight a mistake. We would not draw that conclusion, but we can see how recent consumer spending and labour market data would lend themselves to it. And although the near-term demand picture is stronger, a February hike would induce a weaker outlook further out.

If the RBA does end up hiking in February, it would be unlikely to follow that up with a back-to-back hike in March. Even with a more hawkish view of where ‘neutral’ policy settings are, a 50bp increase would take the cash rate almost all the way back to the peak in rates that everybody at the time thought was very restrictive. It would seem odd if almost the same policy setting was needed to get underlying inflation from the low 3s to 2½% as was needed to get inflation from 8% to 3%, but that is the implication of a two-hike forecast. Instead, a ‘one-and-done’ (or at least ‘one then wait and see’) path might be the outcome. Typically, people rule out a single 25bp move and expect multiple moves, given how little difference a 25bp move makes to inflation forecasts. Nowadays, though, the RBA is in the business of fine-tuning policy to hit that 2½% target midpoint. It might flag the possibility of further hikes in the post-meeting communication but not end up following through.

Also odd is the implication that – if market pricing and bond yields are any guide – Australia apparently needs higher rates than all its peers for years to shake out that last half a point or so of disinflation. Some observers will note that Australia did not hike as far as many of its peers and has seen less of a slowing in the labour market (which was the intent!), but it seems like a big difference now to offset what were not huge differences at the peak.

Part of the issue seems to be the belief that Australia cannot grow faster than about 2% before stoking inflation, much slower than in the past, and yet is actually growing faster than this. We have previously discussed how sensitive these judgements are to underlying assumptions. The thing is, the only way to judge how fast supply capacity is growing in real time is to see how inflation performs given actual growth in output. Once again, trimmed mean inflation gets the casting vote.