A contrarian take on the US inflation freakout

There’s a peculiar practice in financial markets: expressing shock when economic data don’t match forecasts that few had faith in the first place. It’s happening right now with US inflation.

Last week’s core CPI growth for March came in at a 0.359 per cent month-on-month, against a consensus of 0.3. A ‘whopping’ 0.059 percentage points higher. Cue frantic re-pricing of expected US Fed rate cuts for 2024.

Admittedly, it was the third consecutive above-expectation reading. So it’s entirely reasonable to think the Fed’s stance on three rate cuts for 2024 may change (Powell suggested as much on Tuesday). But it feels a bit flawed to base rate reconsiderations — pricing from 3 cuts to 1 or no cuts — on a series of missed inflation forecasts that few had confidence in anyway. Best to understand why it came in higher than expected first.

The chart below shows the contributions of various components to annual US CPI inflation. Shelter (housing) has been a driving factor, and to a lesser extent, transport. Both were behind the surprises in the month-on-month data across January, February and March. They help make up the sticky services component.

Digging further — shelter is itself driven by “Owners’ Equivalent Rent of Residences”. This is the BLS’ estimate of what owner-occupiers would pay if they rented their homes. It gets a chunky 34 per cent weight in core CPI (which excludes food and energy). Motor vehicle insurance is driving the transport services bar.

So what? Calculations of OER are dubious (it is imputed based on rents for comparable rental housing). The EU HICP and UK’s CPI measures — which the ECB and BoE target, respectively — exclude it. As Andrew Hunter, US economist at Capital Economics put it:

Much of the debate on inflation across developed markets right now is being driven by comparing apples to oranges.

(There could be another problem: circularity. Moody’s chief economist Mark Zandi argues that housing supply is being constrained by tight monetary policy, which pushes up rents and therefore OER. Also, higher insurance claim costs — which may itself be due to inflation — can push up premiums. But that’s for another time.)

Anyway, the Fed targets PCE inflation. That measure places less emphasis on OER (core PCE gives it only a 13 per cent weight). It also measures insurance inflation net of claims.

Below is annual US inflation on a CPI, HICP and PCE basis. As you can see, measures with less emphasis on OER are a lot closer to target. And there’s a headline you don’t see: US inflation is actually the same as in the eurozone (when comparing like for like)

But surely CPI still reflects the price growth US household feel? True, to an extent. But, as UBS Global Wealth Management chief economist Paul Donovan says: “Because OER is completely made up, the true the cost of living for a homeowner is more benign than headline CPI would have us believe”. (Many market players already know this, but it can still cause volatility).

Even if you’re still wedded to CPI, it should come down. OER tends to lag private indicators of rent — see the chart below. Insurance also operates on rolling contracts. Right now, auto insurers are perhaps pushing up premiums to make up for higher post-pandemic costs. Both should settle, but it will require patience.

Bottom line: the CPI is noisy right now. It makes sense to focus on PCE to get a better handle on underlying inflation pressures, which is what should matter to the Fed anyway.

As Donovan at UBS points out:

The Fed seems to be caught in a bit of a trap at the moment. It elevated the importance of CPI as a measure back in mid 2022, despite traditionally favouring PCE. That makes it difficult to be cutting rates when headline CPI is 3.5 per cent.

The dispersion of price pressures in the PCE is now being driven by a narrow portion of the index (mainly shelter, and other supply-driven inflation factors). The spread between increasing and decreasing elements of the index is now running below its historic average.

Does that mean Powell’s turnaround midweek is a mistake? To answer that, it is worth assessing PCE’s overall direction of travel.

The producer price index reflects price pressures within the supply chain, and is a decent leading indicator of the prices households end up facing. Annual PPI growth for final demand goods is now back in line with its historic range. Barring any further supply chain snags or energy shocks, PCE goods should be well behaved.

As for the labour market, Goldman Sachs has compared several measures of tightness. The average z-score — a statistical measure which relates a single data point to the mean of a group of values — across all measures is essentially back to pre-pandemic levels.

This suggests pressures on wages ought to continue easing. Indeed’s latest tracker shows annual growth in posted wages is back down to 2019 levels. That means the services component of core PCE should ease further too.

Pulling this together, the disinflation narrative appears alive and well for PCE. Pressures in good, services and shelter (food and energy are less of a problem now) are abating, or already have.

But to be complete, there may be a political incentive for cuts too, notes State Street Global Advisors, which is now considered a contrarian for still backing rate cuts. The Fed may ideally want to make cuts before the US election campaigning really kicks off, given the optics.

The asset manager also cites cracks in America’s growth story despite aggregate resilience on the surface — which Alphaville and Free Lunch have also emphasised recently — such as rising debt delinquencies, rapidly declining SME hiring expectations, and the fact that the mean perceived probability of losing one’s job in the next 12 months is now above pre-pandemic levels according to the NY Fed Survey. ..

In summary:

— The market’s recent re-pricing is partly based on data realities, but also impatience, an overemphasis on headline CPI, and panic.

-The disinflation narrative remains alive and well. Current stickiness is driven by idiosyncratic factors, including measurement differences.

-The Fed’s emphasis on data dependence has tied its hands. It is finding it hard to ignore the run of three above expectation CPI prints, despite the detail, forward dynamics, and PCE all looking benign.

-Given how restrictive real rates are, pockets of economic weakness and where underlying inflation is — excluding all the noise — the Fed probably still needs to cut rates.

Still, given Powell’s midweek comments the Fed looks more likely to cut later, and potentially by less this year.