"A Spectre Is Haunting Markets": There Is Good Chance We Are Now In The Policy Error Zone

By Benjamin Picton of Rabobank

A spectre is haunting markets – the spectre of hawkishness. This could prove to be a watershed week for traders who have been paring-back optimistic bets on monetary easing since the implied path of policy rates hit euphoric levels in mid-January (see "Is The Fed's Next Move A Rate Hike").

The perennial canary in the monetary coal mine, the Reserve Bank of New Zealand, will be holding a policy meeting on Wednesday. Markets are pricing a touch over a one-in-four probability of a 25 basis point hike at that meeting, with a greater than 60% probability that a hike will be delivered by the end of May. Ouch.

Obviously, this would not suit the prevailing narrative of impending rate cuts the world over. How much attention will traders in other markets be paying to what the antipodeans are doing with their monetary settings? Will a hike, or even a hawkish hold, prompt further repricing in the term structure of rates for major markets around the world? Could NZ again be a lead indicator for other central banks (recall that the RBNZ was among the first to hike, back in October 2021)? And could this ultimately threaten an equity rally that continues to expand the frontiers of mathematics?

Our house view is for no rate hike this week, but household inflation expectations figures released late last week are enough to make us nervous. That data showed two-year expectations rising from 3% to 3.2%, while five year expectations rose to 3% after having been anchored at the midpoint of the RBNZ’s target range (2%) for four consecutive quarters. That’s a problem, because Keynesian central banker types believe that when it comes to inflation, expectations become reality.

The household survey is an interesting contrast to the views of business and professional forecasters. Figures released back on the 13th of February showed further declines in one and two year inflation expectations for this group, with two-year expectations within the target range at 2.5%.

The views of households probably matter most in the eyes of the central bank. Taken together with stubborn non-tradable inflation, a resilient unemployment rate and recent hawkishness from RBNZ speakers, it’s fair to say that this week’s meeting is live. The arguments against a hike hinge on the fact that New Zealand unexpectedly recorded negative growth in Q3 (the most recent period that we have data for), headline inflation has been falling faster than the RBNZ’s projections, Q4 retail sales reported last week were grim (-1.9% m/m) and the fiscal impulse is turning more contractionary as a new government seeks to reign in deficits.

There is also the fact that the unemployment rate tends to a lagging indicator. By the time it moves substantially higher the horse has generally bolted and the economy is hurtling towards recession.

Therein lays a conundrum for central bankers: to actually achieve the soft landing that market pricing says is all but assured, rates will probably have to be cut pre-emptively.  But central bank speakers are still telling us “not so fast” on policy easing, even as the TIPS market tells us that real rates are at levels that we last saw in 2007 (immediately before a deflationary crash), and transmission lags mean that the full effect of any cuts won’t be felt for 18-24 months. Central bankers are going to try to catch a falling knife 18 months in advance.

Nevertheless, Janet Yellen declared victory back in December when she said that economists are “eating their words” on predictions that the unemployment rate would have to rise dramatically for inflation to return to target. I’ve filed this for posterity alongside sub-prime is “likely to be contained” and “stock prices have reached what looks like a permanently high plateau”. Our own US Strategist, Philip Marey, continues to expect a recession in the United States later this year.

So, there is a good chance that we are now in the policy error zone. We’re comforted by model-driven forecasts that show GDP growth converging to trend, and inflation at-target over time (our own forecasts mostly suggest this), but the world of economic modelling is hostage to assumptions of Gaussian distributions, when in reality the distribution of possible outcomes often exhibits kurtosis or skew and can be hijacked by the introduction of new variables that were not incorporated into the model. That means that Keynesian DSGE models always predict a return to target, and that no problem is so great that it cannot be ‘transitoried’ away.

Consequently, one-in-a-hundred-year events have a nasty habit of occurring more frequently than they should, and seemingly small cogs in the larger economic machine can be outsized contributors to gains or losses. In an economic environment where monetary aggregates have experienced historically rapid expansion followed by historically rapid contraction in the space of just a few short years, there is plenty of scope for even the most conscientious forecasts to be mugged by the occurrence of events that were previously impossible to predict, or at least impossible to assign a probability to.

Markets continue to price for a soft-landings that history tells us almost never happen, and the economics profession, cognisant of risks in both directions, likewise adopts this as the base-case scenario. This follows the logic that if I put my head in the freezer and my feet in the oven, the room temperature is pleasant on average. Since the distribution of outcomes for important variables doesn’t always follow our assumption of a random walk, we should be mindful that forecasts are almost certain to be flawed, biased by what is known at the time, and hostage to the quality of our assumptions.

Getting those assumptions just a little bit wrong can easily result in a random walk spoiled