According to the AFR, the RBA has invited a “gaggle of academics and economic pointy heads” for a secret meeting on the current conduct of monetary policy.
My own head was deemed insufficiently sharp to warrant a guernsey so I thought I would blog out my thoughts on the topic from the laneway out the back of Chatham House.
The AFR lays out two key questions that the RBA wants to unpack:
“Which is the greater risk to the economy at the moment, high inflation or recession? What is appropriate policy in the current circumstances?”
“What is the best approach to forward guidance (e.g. time-based or state-based)? Is forward guidance useful when the cash rate is not at the effective lower bound?”
Storm clouds on the horizon
What is the greatest risk to the Australian economy: high inflation or a recession? I am a bit confused by this first question to be honest as high inflation is not merely a risk but a current reality in Australia. And I think even the most pessimistic analyst out there would have to concede that with an unemployment rate of 3.4% the economy is nowhere near being in a recession.
On top of that I’m not even sure the question really plays to academics comparative advantage at all. The ivory tower mostly considers monetary policy in a structural sense and is unlikely to be on top of all the latest beats on the economy - unlike the RBA’s many hardworking analysts.
That being said I think high inflation is clearly the more serious risk to the Australian economy at the current time. For example, one of the most timely reads on the economy - the NAB monthly business survey - reported that capacity utilitisation was at an all time high in July! Business conditions and confidence remain strong and all types of prices (inputs, outputs and labour) are growing strongly.
Accordingly the RBA should continue to increase the cash rate over the next few months. Given the most recent SMP still expects inflation to be at 3% by the end of 2024 I would argue that the trajectory for interest rates should be higher than the current market forecasts that the RBA uses as a forecasting baseline. I would however resist the temptation to increase the pace to 75bp as the Federal Reserve has done for five reasons:
The RBA has a higher meeting frequency than most central banks so can afford to increase interest rates at a slower rate each meeting.
Long term inflation expectations remain well anchored, so the RBA’s inflation fighting credibility is not being seriously questioned.
There will be a portion of hand-to-mouth households for whom an increase in interest rates directly lowers consumption. Taking seriously the concept of consumption habits, these (likely low income) households will be better off with a slower increase in the cash rate over a longer period of time.
Most macroeconomic models would suggest that the impact at the aggregate level of 75 vs 50bp hikes is relatively small.
We are still in a period of very high uncertainty. Inflation is high, but if the pandemic shocks continue to surprise us (perhaps by making inflation less persistent than normal) we could see much lower outcomes for inflation in 2023 then are currently forecast.
Given that the current market forecasts a cash rate peak of around 3½ per cent and this is still not enough to get below 3% by the end of 2024, I would aim for a peak interest rate of 4 point something in 2023 - and not be shy about telling Australia about this goal.
Forward to the Future
What about the future use of forward guidance? Should the RBA issue forward guidance in a time dependent manner (ie interest rates will remain at 0 per cent for 3 years) or state dependent (ie interest rates will remain at 0 per cent until inflation hits 2.5 per cent).
Time based forward guidance requires accurate forecasts of how long negative shocks will impact the economy. As the pandemic shows this can be very difficult. But even the post-GFC Federal Reserve regularly underestimated how long interest rates would be equal to zero. Estimate too short a period of forward guidance and the stimulus may be too weak, project too long a period and you may be forced to renege on your commitment or see inflation overshoot the target range.
More problematically, a time-dependent period of forward guidance could be (mis)interpreted as an indication of the severity of the shock, as opposed to a willingness to stimulate the economy.
State contingent forward guidance, especially if tied to the RBA’s mandate, requires a lower degree of certainty about the nature of the shocks to be effective. A promise to keep interest rates at 0% until inflation hits X% or the unemployment rate hits Y% should better be able to appropriately calibrate how long interest rates should stay low for.
For these reasons I think state-based forward guidance is the better of the two options. Any future forward guidance should also be combined with a program of quantitative easing. This offers the RBA a series of concrete policy steps it can use to demonstrate its commitment to re-inflating the economy. This could also be made to be state contingent with the number of bonds purchased each week increasing until the inflation rate is back in its target band.
Conventional economic theory would suggest that there is little need for forward guidance when the cash rate is well above the effective lower bound. But I think there are good reasons for continuing to provide such guidance, albeit as a probabilistic forecast rather than a firm commitment. I think there are several reasons why the RBA should implement some form of interest rate forecasting evn now that interest rates are rising again.
Providing non-ZLB forward guidance would give the RBA practice at communicating such guidance to the public so that it is better prepared for when it is truly needed at the ZLB.
It would also give the public practice at understanding the RBA’s probabilistic forecasts of a high salience economic variable that has a far greater impact on their personal budgets than CPI, GDP or the unemployment rate (which are already forecast in such a manner).
Non-ZLB forward guidance would provide private markets with better information about the likely path of future interest rates which would allow Australians to make better informed investment and borrowing decisions.
It would enable the RBA to be more rigorous and clear when communicating the future stance of monetary policy. For example instead of signalling that “the next move would be up” the RBA could give lay exactly how they see the likely future path for the cash rate unfolding.
It would provide greater clarity about the RBA’s reaction function.
It would make explicit the RBA’s best guess of the neutral interest rate.
Great post. Enjoyed it, and many others.
Wanted to ask about the logic of not going 75. I'd like to take the other side of this.
If the RBA were to hike rates 50bps this month, when the cumulative of additional information suggests a more benign picture for inflation (low wages), it proves they know current policy is clearly wrong. In that case, larger moves earlier might have been appropriate.
On the point about monthly meetings, I would have thought the only benefit of having 11 meetings versus 8 is that it allows you to move more aggressively on occasion. In 90% of years there is no benefit at all because there are few quarters which contain hikes of 100bps. On rare occasions, they let you move faster. If we have 50% more meetings, but we only use clips that are 33% smaller, there's no benefit. I suppose the counterargument would be there's a penalty on hiking too quickly, which makes 150bps per quarter a speed limit that both the fed and the RBA are subject to, so your maximum move is 150 divided by number of meetings per quarter. Is that the underlying reason?
In terms of habit formation, I understand the argument to be that one large cut to consumption hurts utility more than smaller ones that sum to the same amount. But do we really use those models because we think they reflect something about happiness? I have only seen them used because they help explain asset pricing puzzles by generating a high equity risk premium. Is there good reason to believe spreading these budget cuts over 3 months versus 2 is better? People might also prefer receiving bad news twice to thrice.