Steno Signals

Happy Monday, and welcome back to my editorial on everything macroeconomic!

Let us start with the curve, because it is the chart that hangs over everything else this autumn. The 2s10s Treasury curve has flattened to a few tenths of a percent, and it has done so despite a week of data that should have steepened it: core PCE at less than 0.25% with a downward revision to July, and payrolls at 29k against 84k expected with wages at 3.0%. Yields dipped on both prints and then climbed straight back. Front-end yields are now back above the level that prevailed before the NFP, which kinda makes sense once you accept what Warsh and the rest of the committee have said in plain terms: the growth and labor side of the mandate is not what they are looking at for the time being. For good reasons, from their chair: Headline inflation is above 3%, and it is their job.

So those parameters, growth and labor, will keep surprising on the low side, and for now it will not matter to the Fed. I also stick to my view that core inflation will surprise on the low side, and that of course WILL matter, eventually. But eventually is doing a lot of work in that sentence. Which brings me to the question of the week: once a central bank is hiking into an energy shock, can anything actually convince it to stop? Is this a mid-cycle adjustment, the 1995 or 2019 kind, or is it the late-cycle killer, the 2000 or 2007 kind, where the curve inverts, and the clock starts? Today I will address three questions:

What are the three best arguments that this is a late-cycle killer?

What are the three best arguments that this is a mid-cycle adjustment?

And whichever it is, where exactly are we on the inversion clock, and what does that mean for risk assets?

Follow along below.

Theme 1: Three arguments in favor of the late-cycle view

First, the Fed turned hawkish right when inflation rolled over, and it has not turned back. Our natural-language model of Fed communication has climbed to its most hawkish reading of the cycle at the exact moment the one-year inflation swap peaked and started to fall.

That is the classic late-cycle error: a central bank that tightens into decelerating inflation because the level is still too high, and keeps tightening because its reaction function is anchored to the level rather than the direction. Warsh has made that anchor explicit. If the committee keeps hiking on a 3% headline while the monthly core runs at 0.1-0.2%, it will over-tighten, and the curve is telling you it suspects exactly that.

Second, the gasoil crack. Rising refining margins in diesel and gasoil have historically been a worrying harbinger for CPI, because a wide crack means the refiner has pricing power, and a refiner with pricing power passes cost increases straight through to the pump. The gasoil crack is back at 2022 levels. If that pass-through runs for another quarter, headline CPI does not come down the way the base effects promise, the Fed gets its justification to keep going, and the curve inverts for the standard reason: policy too tight for too long against an exogenous price shock. This is the strongest of the three late-cycle arguments, because it is the one that keeps the Fed hawkish regardless of what core does.

Third, the lag. Our central-bank decision index, which simply adds up hikes and cuts across the major central banks, leads US activity by roughly ten months. It has rolled over hard this autumn as the ECB, the BoE, the RBNZ and the Fed moved within weeks of each other. Taken at face value, that points to a softer ISM around the middle of next year, which is also where the inverted-curve clock would put a slowdown. The late-cycle reading is that the damage from a synchronized global hiking round is already done; we just cannot see it yet.

In other words, within a quarter or two from here we need to see the average, or median, central bank returning to softer policies for this not to turn into a central-bank-driven slowdown at the very least. And while I can see that pivot coming, it requires lower oil product prices at the pump. Iran and Ukraine hold the keys to the central bank cycle, and therefore to the business cycle, in the West.

Theme 2: Three arguments in favor of the mid-cycle view

First, where we actually are in the cycle. The Pictet composite of US cycle indicators, which sums the standard business-cycle gauges and ranks them against their own history since 1970, sits at the 62nd percentile. That is mid-cycle.

Not late, where the readings sit in the 80s and 90s before every recession on the chart, and not early. The closest analogue is March 1997: a Fed tightening into a strong economy with a benign core, an inverted curve a year later, and three more years of expansion before anything broke. Late-cycle killers do not start from the 62nd percentile. They start from the 90th.

Second, the necessity impulse. Our basket of input costs for food, energy, and chemical producers is the best leading indicator we have for the curve, and it has already rolled: the impulse has halved from its peak. In 2022 and 2008, the same basket was up 100% or more at the point the curve inverted. Today it is up around 25% and falling.

A curve flattener driven by a cost shock lasts as long as the cost shock, and this one is fading. Which is also why I keep saying the hiking cycle will be short: the thing the central banks are hiking against is already behind its peak. The Iran war is past peak escalation, Qatari gas is flowing again, the IRGC is out of money, and the sine wave is turning.

Third, the PCE-CPI spread, which is the hardest argument for the bears to answer. Core PCE sits around 0.6 points ABOVE core CPI, close to a record, when it normally sits below because of the shelter weights. Historically, the spread has only opened this way once, in 1983, and it opened for the same reason it is opening now: inflation was collapsing, and PCE, with its stickier imputed components, lagged the CPI on the way down.

The spread is a symptom of core trending DOWN. As long as Warsh is anchored to the PCE, his life is difficult, but the CPI is the directional guide, and the PCE will follow it lower. A central bank hiking on the headline into a core that is already rolling over is the definition of a mid-cycle adjustment: painful, short, and reversed within a year.

Theme 3: Where are we on the inversion clock, and what does it mean for risk?

Here is the bottom line… I am not sure which of the two it is.

I typically start to worry when I get the macro right, because the bond and FX markets then have a habit of not moving the way the macro says they should, and that has been the case for a few weeks now: soft data, firm yields, a dollar that will not fall. That is uncomfortable, and I will not pretend otherwise.

But everything considered, three things we DO know. We have NOT inverted yet; the 2s10s is still positive by a few tenths. The Fed is early in its process, one hike in and one more signaled, not six in and exhausted. And we are still in an environment where things go up for another while no matter what the Fed does, because that is what the history of inversions says.

Look at the chart: equities do not sell off on the inversion. Rebased to the first close below zero, the MSCI World across eight episodes since 1978 and the S&P across five since 1998 both RISE into the inversion and then go sideways for the following year to eighteen months. The average path is higher at month plus-twelve than at month zero. The sell-off, when it comes, comes with the recession, and the recession comes ten to eighteen months after the inversion, which has not happened yet. So it is way too early to worry about stuff like that still.

So where exactly are we? With the 2s10s at a few tenths and flattening at the pace of the past month, the first close below zero is three to six months away, which puts us at roughly minus three to minus six on the x-axis of that chart. Historically, that stretch delivers another 4-6% of upside in the index before the inversion, and then a sideways market rather than a bear market. That is the window. It is the window in which Korea says semis double, in which positioning is sour enough that any de-escalation wrong-foots the crowd, and in which the Fed hikes once more into a core that is rolling over. Mid-cycle or late-cycle, the next two quarters look the same from here: up, then sideways. The difference shows up in 2027, and we will deal with 2027 in 2027.

The bottom line

Nothing convinces a central bank to stop once it has started, except the data it has chosen to look at, and the Fed has chosen the headline. So the hikes continue for a while, the curve flattens towards zero, and the late-cycle arguments (a hawkish Fed into falling inflation, the gasoil crack, the ten-month lag) are real. But the mid-cycle arguments are stronger: a 62nd-percentile cycle composite, a necessity impulse that has already halved, and a PCE-CPI spread that only opens when core is collapsing. Either way, we have not inverted; equities historically rise into the inversion and go sideways after it, and the next two quarters are a window to be long the things that deliver earnings. Trade the window.

Have a great week ahead, and good luck out there!