Second Order Smoothing
The RBA should keep hiking to keep their policy options open in 2023
One of the core tenets of central banking is the concept of interest rate smoothing. While nominal interest rates are adjusted in response to changes in inflation or the unemployment rate, they are usually done so in a gradual manner. Interest rates slowly rise and fall over several months instead of jumping instantly in response to sudden changes in the data.
That is why despite inflation rapidly soaring across the developed world, central banks have responded by increasing interest rates at a smooth and steady pace over the past year. Most central banks rolled out 3-4 percent points of interest rate increases steadily over 2022 instead of doing it all at once when the scale of the inflationary surge first became apparent (although some economists have argued for a more forceful approach!).
In the most recent minutes the RBA has gone one step further and introduced the concept of smoothing the rate at which the interest rate is changing over time:
“In considering the size of the increase, members also discussed the value of the Board acting in a consistent manner. Having moved by 25 basis points in the previous month, they considered whether the flow of information since then warranted a 50 basis point move at the November meeting. The Board agreed that acting consistently would support confidence in the monetary policy framework among financial market participants and the community more broadly.”
What does this actually mean?
Consider two hypothetical paths for increasing the cash rate over a four meeting period:
50bp, 25bp, 50bp, 25bp.
50bp, 50bp, 25bp, 25bp.
Under both scenarios the cash rate increases by the same amount over the four months (1.5%), so from an interest rate smoothing point of view they are essentially the same. But option 1 involves yo-yo-ing back and forth between 50 and 25bp hikes - effectively accelerating and decelerating the path for the cash rate. In option 2 the rise in interest rates is somewhat more orderly. The minutes indicate that the RBA Board thinks that option 2 is preferable - all else equal.
Instead of keeping the first derivative of the nominal cash rate low (traditional interest rate smoothing) they are placing some weight on keeping the second derivative of the cash rate low as well - i.e. second order smoothing!
I think this approach is sensible. While the RBA will always need to respond to economic data, it should ideally implement monetary policy in a way that is predictable and consistent. Option 2 is clearly better from that point of view.
Back to the Future
A policy of interest rate smoothing (either first or second order) affects your monetary policy decisions in two ways. The first is that you have to look back and ask if the decision you are making today is consistent with the previous path for the cash rate - or if it would be a large break from it. For example, if the cash rate was set at 3% last month, it would require a large and widespread change in recent data to justify changing to the cash rate 4% as that would be a very unsmooth adjustment!
But you also need to look to the future and ask yourself how your decision today will constrain your options tomorrow. Take for example the RBA decision to hike by only 25bp in October 2022. Coming after a series of 50bp hikes this sent a clear signal that the pace of interest rate increases was going to be slower going forward. It thus made it much harder for them to consider a 50bp increase in November - even if they might have wanted to. They acknowledge this explicitly in the minutes.
They certainly could have hiked by 50bp if new data had radically changed their view of the economy (for example if the CPI print was wildly higher than expected) - but the bar for going back to 50bp increases was higher once that initial reduction had been made. If the economy was only marginally stronger than expected you might want to increase rates by 50bp, but find it hard to justify reversing course only four weeks after announcing that a single 25bp hike was all that was required.
Macroeconomic data comes out relatively infrequently. It would be a brave central bank who would re-pivot their policy stance after only four weeks given that it may just be a noisy data print or two.
“Having moved by 25 basis points in the previous month, they considered whether the flow of information since then warranted a 50 basis point move at the November meeting.”
This same logic applies to the RBA going forward. Once the RBA stops hiking interest rates (or ends with a small 15bp rise to “round off” the cash rate target) that will be a strong signal to the markets that the RBA is going to stay put for at least a couple meetings. They could change their mind at the next meeting, but they would find it awkward to do absent a big change in the data. If the economy turns out to be slightly stronger they might want to increase the interest rate, but not enough to backtrack on their previous pause.
However by keeping the rate of increases unchanged (ie hiking by another 25 bp in December) the RBA will preserve their option space. That will enable them to keep slowly ratcheting up rates if demand continues to look hot. It would also be consistent with a full pause in February if the global outlook worsens. Either future outcome would be consistent with the current plan of smoothly renormalising the cash rate to a higher level.
In 2022 a rapid rise in interest rates was needed, and it makes sense to increase them at a constant and predictable speed. But the desire to smooth the speed of interest rate increases means that the decisions made today constrain the RBA’s option set tomorrow. They should bear this in mind when they meet to make the final call of the year.