Steno Signals

Happy Monday, and welcome to my editorial on everything macroeconomic! My old mentor, the "Macro Mufti" of Sweden, Mikael Sarwe, tweeted for the first time in a long while late last week, and it was an eye-catching chart on the US cycle, which looks nasty in his forward-looking regression.

Having worked closely with Mikael for years, I have also contributed to forming some of these models. No doubt, there are few people in the world with a more comprehensive set of cycle indicators and forward-looking models than Mikael, and I think I have a decent understanding of how these models work... but right now, they are clashing with the thesis I have laid out. Should we fear the slowdown? Let's look at the case for LOW inflation and growth over the next 6-9 months.

First of all, I think Mikael's chart will be proven wrong. And the reason is that some of the very best survey-based indicators that you could use to construct amazing leading indicators for the macroeconomic cycle broke down in 2020/2021 and especially during this second Trump era in the White House. Why is that? Well, for starters, we have seen inflation in this time period, which is a vast difference to the entire time-series history of many of the best surveys, which often run from the 80s and 90s and forward, where we in most cases had the opposite environment on a trend basis (inflation trending lower). That wreaks havoc on many correlations/relationships when that happens. We know that by now (see, for example, chart 2 on consumer sentiment versus consumption growth).

Second, we have seen an incredibly partisan response developing in surveys after Trump took office again (and also pre-election during the stand-off with Biden / Kamala). Since then, the political environment in the US has been growing increasingly hostile both domestically and internationally, between the US and its "allies" abroad. I, for example, cannot recall another observation (I have double-checked the data and no, it does not exist) where consumers are as "depressed" and fearful of the future as now, with the stock market at ATHs and employment practically within spitting distance of being "fully booked". The disconnect between reality and survey has rarely, if ever, been larger. This disconnect also holds true for most inflation surveys taken among consumers, by the way. So I am personally very careful using surveys to predict 6-12 months into the future, which was otherwise entirely plausible to do before 2020/2021 (and especially before 2024/2025), since we have seen SO many correlations breaking down, and in some cases, even reversing after the two abovementioned trends occurred. Very few macro-strategists are willing to admit to it yet (outside of myself).

I can also feel it myself... When I dare to say something just barely positive on Trump's policy mix, my inbox is FULL of angry Europeans, not least, but also Americans against Trump, while if I do the opposite (for example claiming that there is no plan in Iran, which is obvious to everyone with a pair of eyes), I get a load of accusations of suffering from TDS.

When it comes to analyzing the economy, I honestly could not care less about who is the president in the US. It doesn't matter to me. I am not in the business of being on the right side of history's "ethics". I am in the business of being right.

As a consequence of the aforementioned trends wreaking havoc on the commonly used business cycle models, I therefore decided to launch Nowcast IQ, to produce a high quality LIVE and REAL data set (no surveys, only hard numbers) to try and gain an edge in the current environment, and I think we have largely succeeded in that, as we have had a range of very strong calls right against the economic consensus often, also including the inflation trends VERY right, when it became a politically hot potato to do so (it to some extent still is).. Our evidence is still fairly strong. At least live, there is very limited reason to fear this cycle. We have a semi-recovery ongoing in growth in rate of change terms, while inflation risks have largely disappeared (if you look at the probability of rising inflation at least).

But remember that these are nowcasts, not forecasts, so I have little to say about e.g. the cycle in 4-6 months from now based on those numbers. So which forward-looking indicators do I actually trust amidst this breakdown of correlations due to inflation, and this lack of solidity in surveys due to an incredibly partisan survey response average?

For the growth cycle, my view remains that cycles never die of old age, but only of central banks killing the momentum for the layer upon layer debt structure of the global economy... And judging from our CB decision index below, we still have the peak in the business cycle ahead of us. Likely in Q4 of this year... Then we may trend a bit lower into 2027, but nothing tells me that it will be a catastrophe yet. Rather just slightly less hot than this year...

And what about meaningful leading indicators on inflation? My best guess is either to use the growth of money (which is a LONG lead), or else to look at Truflation. I know Truflation pretty well, with their pros and cons, but they have mostly been able to call out the trend directions right well in advance in US inflation, and look at the core inflation measure currently. It is off a cliff versus the actual. This is important.

It is first of all incredibly odd to see that at this stage of the business cycle. I can actually understand why e.g. Kashkari of the FOMC has said that "this is typically the time to raise rates", because you would typically see inflation ticking up a bit in response to the rising business cycle we have seen in recent quarters. So why isn't it happening?

I think the reason is that the Trump administration has introduced a couple of very important inflation factors that go outright against the cycle. First, the Iran war, which has tapered off, for what it's worth, even without any solution, will lead to a very benign oil/energy inflation impulse from now until Q1/Q2 2027, given we saw an arbitrary spike in prices due to the war, and now a return to normal/equilibrium after. If we hadn't had the war, we would have likely seen a smoother upward-trending energy market. Now we got BIG TIME up during the spring, and a disinflation impact from now until the spring of next year. That wreaks havoc on the cycle.

The second thing is that the Trump tariffs policy mix has also gone from being a big inflation driver to being the opposite. We can see how, e.g., consumer goods printed very soft in the PPI, and how the soft summer of CPI prints coincided with the "returning" of some of the illegally captured tariffs from last year. If companies suddenly get a windfall from paid-back tariffs, they are unlikely to rush price hikes, especially as it would be terrible PR.

That leaves us with an odd disinflationary/deflationary impulse from tariffs, while the opposite should have been the case (it was until lately). Third, we of course have the World Cup hangover as well, which has led to a slide in prices on lodging away from home and the like, since the price increases were frontloaded into the spring, so it all means that core inflation is actually sloping lower, while the business cycle is accelerating. That is incredibly rare to see, but it is what we call goldilocks as macro strategists. You probably cannot find a better mix for markets...

But what about the price surveys screaming the opposite?

Now, I can already hear the pushback, and it is a fair one. The NFIB price plans have ticked back up and are hovering around levels that have historically been consistent with the median CPI BOTTOMING out around 3-4%. So is the bottom already in on inflation?

Here is the thing though... I just spent the first half of this editorial explaining why I no longer trust surveys in this environment, and I am not about to make an exception just because a survey happens to disagree with me (that would be a bit too convenient, right?). Again here, we can actually observe how the survey and the CPI data have decoupled since 2024/2025, which is another case study of the breakdown of the relevance of survey-based forward-looking indicators.

Companies plan hikes, and then reality gets a vote... and reality currently consists of a price-fatigued consumer, a used car market rolling over, and tariff refunds flowing back into exactly the import-heavy categories where those hikes were supposed to show up. My bet remains that the plans will follow realized prices lower, and not the other way around.

The ISM prices paid index is admittedly a bigger challenge to my case, as it has historically led the CPI by around 7 months, and it has spiked lately... Having said that, it holds a pretty terrible hit ratio, outside of looking like a decent indicator when you eyeball the chart.

But look at WHAT has been driving it: the arbitrary war-driven spike in energy input costs during the spring, plus the very real inflation in memory and components (more on that further down, as it is the single genuinely hot corner of this economy). The energy leg of that spike is already reversing as we speak, and the rest is currently being absorbed in margins rather than passed on to consumers, which you can see directly in the growing wedge between the PPI and the CPI. So yes, prices paid are up at the producer level... but the pass-through mechanism is broken for now, and with the illegally collected tariffs being handed back as we speak, it would be outright terrible PR to even attempt to pass it on.

And again, we have seen a ton of false flags in the ISM price surveys, especially when we have had a sudden energy price spike. And often, they fail to materialize. So also here, I am tempted to say that the survey closes the gap. Not the prices.

The Fed's June forecasts are about to age like milk

And that brings me to the Fed, because the June SEP is slowly turning into a genuine embarrassment. The committee penciled in a median 2026 PCE inflation of 3.6% (3.3% on core) with a Fed funds rate at 3.8%... in other words, a HIKING bias baked into their own projections. Remember that most of those forecasts were built on a base case of oil printing well above 100 due to the strait. That base case never materialized, and the projections are now hanging in the air like Wile E. Coyote a few meters past the cliff edge.

横向滑动查看完整图表

Run the numbers yourself instead. If you predict the PCE using the matching CPI categories (with PCE weights), you get a predicted index at 2.9% versus the observed 3.7%... and as I wrote about last week, the CPI ALWAYS leads the direction of the PCE.

The spread widens when inflation rises and closes when inflation falls, and it is the CPI that does the steering. So, the Fed's own preferred gauge is about to converge DOWN towards the CPI-implied level, right as the committee has to publish fresh forecasts in September.

Staff projections will look a LOT softer by then, and that typically impacts the thinking of at least the fence-sitters... which is exactly what Warsh needs to keep his do-nothing majority intact into 2027. Peak hawkishness is behind us, even if a couple of the semi-cornered hawks will need a few months to admit it.

The disinflation pipeline runs all the way into 2027

And the pipeline beyond the next couple of prints only gets softer from here.

On a steady-state Brent of 65-70 (and MoU 2.0 caps the tail risk), the mechanical oil impulse on the CPI turns outright NEGATIVE into 2027... arriving, with impeccable timing, roughly when the CB decision index from chart 4 says the momentum of the cycle starts to fade. The economy gets its energy tax cut exactly when it needs one, which is yet another reason to doubt the doom embedded in Mikael's regression.

The services side tells the same story, just with a lag

Job openings in services lead core services inflation (ex shelter) by roughly three quarters, and they have been grinding lower for a while now... the -23k payrolls report was another data point in exactly the same direction.

Wage pressure in services is the LAST bastion of any inflation cycle, and the leading indicator says it is fading (at least for now), not building. Add a shelter glacier that keeps melting mechanically quarter after quarter, and you simply struggle to construct a credible path back to 0.3% core prints. I have tried. It requires assumptions I am not willing to make with a straight face.

So where IS the inflation then? In the machine room, not at the checkout

Because make no mistake, there IS inflation in this economy. It is just not where the Fed measures it. DRAM spot prices have gone close to vertical, and NAND is now following, as the AI infrastructure buildout devours every wafer of memory that can be produced... and the same goes for electricity, where the data-center grid squeeze is quietly leaking into utility bills (a theme with YEARS left to run).

But this is B2B capex inflation, and it barely registers in a hedonically adjusted consumer price index, also as consumer electronics are close to irrelevant in weighted terms in the consumer basket (approx 1% of the total)... Great for anyone long the memory space, largely just noise for the inflation target... what a world we live in.

The other genuine pocket of current consumer price pressure sits with the refiners, where crack spreads still trade above the 2022 peaks, as the flow of PRODUCTS out of Hormuz remains close to non-existent.

But notice what that actually is: Lower inflation that remains stuck at the producer level. The consumer-facing energy relief is hence still ahead of us, not behind us, and a more thorough strait deal would release a whole new wave of disinflation at the CPI level (and probably break the stubbornly bullish USD consensus while it is at it, but that is a story I have covered at length elsewhere).

In layman's terms: Consumers simply haven't gotten the full benefit of the lower oil price yet, as extreme refiner margins hinder the pass-through of lower oil prices to the price at the pump. That is going to change if we get the Middle Eastern refining capacity up and running again in a more thorough second Strait deal.

You can actually see all of the above in one single chart: the wedge between the PPI and the CPI. Producer prices are running hot on war-energy, refiner margins and components, while consumer prices are grinding lower... and history is pretty clear on how that wedge resolves.

The PPI comes down to meet the CPI (the PPI leads by a month or so), and the very soft consumer goods in the latest PPI report was probably the first crack in exactly that direction. Margins normalize, the strait re-opens sooner or later, and the wedge closes DOWNWARDS.

The bottom line: VERY low inflation... and yes, also growth

So let me answer the question in my own headline. The case for VERY low inflation over the next 6-9 months is about as strong as I have seen it in years: a war-driven energy spike unwinding, a tariff regime flipped from headwind to PAYBACK, a World Cup hangover in services, a used car market rolling over, a melting shelter glacier and a services labor market with slack building... all while the only genuine and massive price pressures in the system sit in corners of the economy within infrastructure/hardware that the CPI barely measures.

And the growth part? The nowcast says semi-recovery, the CB decision index says the peak of the cycle is still ahead of us (likely in Q4), and central banks have not delivered anywhere near the kind of tightening that kills cycles... so with all due respect to my old mentor, I think the surveys feeding his regression are lying to him — exactly as they have been lying to everyone else since 2020/2021. Disinflation WITH growth is the rarest and most bullish mix in macro, and the live data says we are pretty likely going to get that over the next months.

Goldilocks is not a fairy tale this time. I stay received in the US front end, short the dollar, and long the cyclical rebound... and I will happily revisit the "Macro Mufti's" chart in six months, whichever way it goes. That is a promise.

Have a great week ahead, good luck out there, and be careful of making big conclusions from surveys drawn from the current hostile political environment. Garbage in, garbage out, as they say.