Tech Leads Us Services Growth As Prices Keep Soaring

I kept refreshing the calendar last week the way some people refresh a score. Manufacturing had already come in hotter than almost anyone on the desk expected, and the price side of that print was ugly enough to ruin a perfectly good afternoon. Then the services numbers landed, and the room split in two. One camp saw an economy that is still accelerating. The other saw an inflation problem that refuses to behave. Both camps, inconveniently, have a point.

September’s private-sector activity gauges for the service side of the American economy did not agree with each other. One widely watched survey ticked up to 58.8, a hair above the flash reading and well clear of August’s 56.5. That is the strongest mark in five years. The other, the purchasing-managers version that traders have leaned on for decades, eased to 54.9 from 55.4, a shade under the 55.0 consensus. Still expansion. Just not the same story.

Why Two Services Surveys Can Tell Opposite Stories

Divergence is not a bug in this data. It is the feature people forget until it bites them. One survey leans harder on larger firms and on the digital economy. The other pulls in a broader mix of smaller operators, hospitals, trucking desks, and regional banks. When they move together, the signal is clean. When they split, you are usually looking at a two-speed economy wearing a single headline.

I’ve found that the useful question is never “which print is right.” It is “who is actually growing, and who is paying for it.” On that score, September was unusually clear. Demand improved. Hiring improved. Optimism recovered. And the bill for all of that showed up in the price gauges before most households had time to notice.

A Five-Year High Against a Mild Cooling

A reading above 50 means activity is expanding. Both surveys cleared that line with room to spare. The gap between 58.8 and 54.9 is still large enough to matter for anyone building a nowcast. The stronger print implies a pace of business that, if it held, would look more like a reacceleration than a soft landing. The cooler print says the expansion is intact but no longer speeding up in the parts of the economy that employ the most people outside tech.

Perhaps the most interesting aspect is how little drama sat in the headlines and how much sat underneath them. New orders rose. Backlogs built. Expectations for the year ahead climbed to a one-year high. That combination usually shows up before official growth figures catch the turn. It also usually shows up before the inflation figures fully catch the turn in costs.

Business growth has surged to its highest in more than five years, with rising demand and better optimism pushing firms to hire at a pace not seen in over four years. Paired with a solid factory survey, the service expansion points to growth around 4 percent in the third quarter and something closer to 5 percent in September alone.

Chief business economist at a major survey firm

Take that growth sketch with a grain of salt. Survey-implied GDP is a translation, not a print from the statistical agencies. Still, a 4 percent third-quarter pace would sit well above the trend most rate-setters have been planning around. A September pulse near 5 percent would mean momentum was still building as the quarter closed, not fading into the fourth.

What the Split Actually Measures

Think of the two surveys as two microphones in the same room. One is closer to the stage. The other is near the back wall. Tech and information firms are on the stage right now. Regional service businesses are nearer the wall. You hear the same song, at very different volumes.

• The stronger survey captured a five-year high in service activity and a broad return to growth across every major sector group.
• The purchasing-managers survey stayed comfortably in expansion but cooled a tenth of a point versus expectations and half a point versus August.
• New orders and backlogs rose at faster rates on the stronger gauge, which is the part that usually leads official spending data.
• Hiring quickened to a pace not seen in more than four years, even as separate payroll figures have looked tired.
• Price measures, measured across goods and services, jumped to the fastest input-cost pace in nearly four years.

That last bullet is the one I would tape to a monitor. Growth without heat is a gift. Growth with heat is a negotiation with the central bank, and the central bank has been trying, a little awkwardly, to sound more open to easier policy.

Tech Is Not Just Participating. It Is Carrying the Print

For the first time in ten months, output rose across all five broad sectors covered by the stronger survey. Transport and storage, which had been the laggard, flipped back to growth. Consumer-facing businesses accelerated. Industrials and healthcare improved. Financial services kept a solid, unspectacular pace. And then there was information and communication, which did not merely expand. It ran.

Tech companies reported by far the strongest growth. That is not a footnote. It is the plot. Software, data services, communications infrastructure, and the firms selling tools into every other industry are pulling the aggregate higher. When people say the rising tide is lifting all boats, they are half right. The tide is real. The tech boat is still the one with the outboard motor.

In my experience, markets under-weight this kind of sector split until earnings season forces the issue. A headline services number in the mid-50s can hide a world where cloud and communications firms are adding capacity while a dentist’s office and a regional hauler are mostly trying not to lose ground. September’s commentary suggested the gap narrowed. It did not close.

Breadth matters more than the tech spike alone. An expansion that lives in one sector is fragile. An expansion that finally includes logistics is harder to dismiss as a statistical quirk. The catch, and there is always a catch, is that moving goods and running data centers both burn fuel, power, and wages. Those costs do not stay inside the sector that incurred them.

Orders, Backlogs, and the Mood on the Desk

New orders rising at a faster rate is the cleanest demand signal these surveys offer. Backlogs rising alongside them means firms are not simply working through old tickets. Work is arriving faster than it is leaving. That is how you get hiring plans that survive a cautious finance meeting.

Growth expectations recovered to a one-year high. Optimism is squishy, I know. Managers talk their book. Even so, a rebound in forward views after a long stretch of caution usually lines up with real order books, not just a nicer mood on a Friday. Firms do not staff up at a four-year pace because a survey asked them to feel better.

September service snapshot, stronger survey:
  Activity: 58.8, five-year high
  Prior month: 56.5
  Hiring: fastest in more than four years
  Expectations: one-year high
  Sector breadth: all five groups expanding

Set that next to the purchasing-managers cooling and you get a useful tension. The economy is not stalling. It is also not uniform. Anyone trading a single “services beat” headline last week was trading a composite that does not exist.

The Heat Under the Growth

Here is the part that should bother anyone hoping for a smooth path to lower rates. Measured across goods and services, firms’ input costs are rising at the fastest rate in nearly four years. Some of that is fuel. Not all of it. Selling-price growth moved higher again, which is the line that matters for inflation staying stuck above a 2 percent target.

Fuel can fade. A renewed willingness to pass costs on does not fade on its own. When order books are full and customers are still showing up, managers stop absorbing the hit in the margin. They put it on the invoice. That is rational. It is also exactly the behavior that keeps consumer inflation from gliding back to target on schedule.

Concerns that the economy is running too hot will be fed by the price gauges. Input costs are rising at the fastest rate in nearly four years. Higher fuel is part of it. The worry is that selling prices have moved up again, signaling inflation that stays stubbornly above the 2 percent goal.

I keep coming back to a simple distinction. Input inflation is a cost. Output inflation is a choice, constrained by competition and by how desperate the customer is. September suggested both moved the wrong way for anyone betting on a quick disinflation story. Stronger growth and sticky prices do not sit comfortably next to dovish remarks from rate-setters or next to a softer payroll report.

Payrolls Versus the Hiring Plans Inside Firms

Last week’s employment report looked tired. These surveys say firms took on workers at a pace not seen in over four years. Both can be true for a month. Official payrolls are a count. Survey hiring is a direction. One can lag the other, and revisions have a habit of closing gaps that looked dramatic on release day.

Still, the clash is the story markets have to live with. A weak payroll print invites talk of insurance cuts. A services sector adding staff while raising prices invites talk of an economy that does not need the help. If you only trade one of those, you are going to get whipsawed the week the other one reasserts itself.

1. Official job counts cooled, which supported a gentler policy tone.
2. Service firms reported the fastest hiring appetite in more than four years.
3. Input costs hit a near four-year high across goods and services.
4. Selling prices firmed, pointing to inflation staying above target.
5. Forward expectations hit a one-year high, which argues against an imminent stall.

That sequence is awkward for a clean narrative. It is also a fair description of where the US sits: busy, a bit overheated in the price data, and uneven enough that both bulls and bears can quote a number without lying.

How This Sits Next to the Factory Rebound

Manufacturing surveys had already surprised to the upside, with the same baggage of a surge in prices paid. Services confirming expansion, even with a split between the two gauges, removes the easy excuse that factories were a one-off. Goods and services moving in the same direction is how you get a quarter that prints closer to 4 percent than to 2.

The US, on these private gauges, looks like the strongest large economy in the world right now. That is not a victory lap. It is a relative statement. Peers are struggling to generate this kind of nominal heat. Strength is good for earnings until it forces financial conditions to stay tighter than equity markets have been discounting.

Would I fade the entire growth signal because one purchasing-managers index dipped a half point? No. Would I ignore the price spike because fuel explains a slice of it? Also no. The adult read is that demand is firmer than the gloomy labor headlines implied, and the inflation problem has not been solved by wishing.

What Firms Are Actually Saying With Their Behavior

Talk is cheap. Staffing is not. When service firms add workers at a multi-year pace, they are expressing a view about the next two quarters that no press release can match. Layoffs are fast. Hiring is a commitment. The September message from the stronger survey is that managers expect the order flow to stick.

Backlogs reinforce that. A rising backlog is deferred revenue with a clock on it. It supports the idea that September’s strength was not a one-month spike in a single category. Transport returning to growth fits the same picture. You do not restart warehouse shifts for a headline.

The mood recovery to a one-year high is the softest of these signals, and still worth noting. Expectations had been bruised by rate uncertainty and by a long argument about whether the consumer was about to crack. A rebound there usually means the crack did not arrive on the timetable the bears had circled.

Consumer Services Are Not the Weak Link People Feared

A recurring worry this year has been the lower-income household. Higher prices, resumed student-loan payments in prior periods, and a cooler job market were supposed to show up first in restaurants, travel, and personal services. September’s sector notes pointed the other way. Consumer-facing businesses accelerated.

That does not mean every family is fine. It means the aggregate till is still ringing. Services are where households spend once the goods binge fades, and they are also where inflation has been hardest to kill. An acceleration there is good for GDP and awkward for the price target. Both things again.

Healthcare and industrials improving adds another layer. Those are not meme sectors. They are payroll-heavy, regulation-heavy, and slow to turn. When they join tech and consumer services in expansion, the “narrow rally in the economy” argument gets weaker, even if the stock market’s leadership stays narrow.

Financial Services Holding, Not Roaring

Sustained solid growth in financial services is the quiet line in the sector recap. Not a boom. Not a bust. Banks, insurers, and advisory businesses are still writing business. That matters for credit availability. A services expansion that coincided with a freeze in finance would be a warning. This one did not.

I would not over-read it into a credit cycle turning higher. Solid is not the same as easy. Lending standards can stay tight while fee businesses and markets-related activity keep the sector index afloat. The useful takeaway is the absence of a fresh crack, not the presence of a new boom.

Fuel, Power, and the Pass-Through Problem

Higher fuel prices explain part of the input-cost jump. They do not explain the rise in selling prices by themselves. Pass-through is a decision. In a weak demand environment, firms eat the cost and hope the competitor blinks first. In a firm demand environment, they test a higher price and watch whether the order cancels.

September’s combination, faster orders plus faster selling-price growth, says the test is being passed often enough to show up in the aggregate. That is how you get inflation that looks “stubborn” in the official data two or three months later. Survey price gauges lead. They do not always lead by the same number of weeks, which is why people get bored of them and then get surprised.

Power costs deserve a mention next to fuel, even if the survey commentary led with energy at the pump and in the tank. Data centers and communications networks do not run on goodwill. A tech-led expansion has a physical bill. If that bill is rising while demand for the output is also rising, margins can hold and consumer prices can still grind up. Markets sometimes treat those as mutually exclusive. They are not.

Rough read of the tension: strong orders + rising selling prices = growth that argues against rapid easing, even if one jobs report looked soft.

The Policy Argument These Numbers Complicate

Rate-setters spent part of the prior week sounding more open to a gentler stance. Weaker payrolls gave them cover. These surveys take some of that cover away. An economy tracking something like 4 percent in the quarter, with September possibly faster, and with input costs at a multi-year high, is not an economy crying out for urgency on cuts.

That does not mean cuts are off the table. Policy works with a lag, and labor-market cooling can be real even when a monthly survey of managers looks busy. It does mean the dovish interpretation has to work harder. Stronger growth and soaring prices are a direct counter to the idea that the inflation fight is finished and the only remaining risk is over-tightening.

I’ve sat through enough cycles to distrust any single week of Fed commentary. Officials talk. Data accumulates. When the data and the talk diverge, the data usually wins the next meeting’s statement, even if it loses the press conference’s tone. September’s services batch belongs in the “data” column, and it leans hot.

Markets That Want a Simple Story

Equity markets prefer a clean setup. Soft landing, gentle cuts, earnings that grow because nominal GDP is fine and rates drift down. September’s services picture supports the nominal-GDP half and challenges the gentle-cuts half. Tech leadership inside the real economy also rhymes with tech leadership inside the index, which is either confirmation or concentration risk, depending on the hour you ask.

Bond markets have the harder job. A five-year high in service activity plus a near four-year high in input costs is not a rallying cry for duration, unless you believe the purchasing-managers cooling is the true signal and the other survey is noise. I do not think it is noise. I think it is a different sample hearing a louder tech cycle.

Credit spreads can live with this for a while. Defaults care more about cash flow than about the policy path, and cash flow looks supported if orders and backlogs are rising. The risk is a rates backup that tightens conditions faster than revenue can offset. That is a 2022 lesson people quote and then forget when the index is green.

A Practical Way to Read the Next Round

One month is a mood. Two months is a pattern. The next services batch will tell you whether September was a tech-led burst or the start of a broader fourth-quarter acceleration. Watch three things, not the headline alone.

• Whether information and communication stays the outlier or the rest of the sectors catch up.
• Whether selling prices keep rising after the fuel impulse fades or cools with it.
• Whether hiring plans inside the surveys converge with official payrolls, and in which direction.

If breadth holds and prices cool, the soft-landing crowd gets its chart back. If breadth holds and prices do not cool, you are looking at a hotter nominal economy than the rate path implies. If breadth fades and only tech remains, the concentration story moves from the stock market into the real economy, which is a less comfortable place for it.

There is a fourth watch item I would add from the factory side. Prices paid in manufacturing already jumped. If goods inflation and services inflation firm together, the “services are the sticky part, goods will save us” framework takes another dent. That framework has been doing a lot of work in forecasts.

Where the Growth Math Can Mislead

A survey-based sketch of 4 percent quarterly growth and 5 percent in September is a translation of diffusion indexes into a GDP metaphor. Diffusion indexes measure breadth of change, not the size of change. A lot of firms saying “up a little” can print a high index without producing a boom in dollars. A few huge tech firms saying “up a lot” can do the opposite.

So treat the 4 and 5 percent figures as a directional shout, not a forecast you should paste into a model. The shout is still loud. Activity is up. It is up across sectors for the first time in ten months on the stronger gauge. It is up enough to hire. Those are harder to wave away than a point estimate of GDP.

The purchasing-managers 54.9 is the soberer microphone. Expansion, slightly slower, slightly under consensus. If your process only trades surprises versus expectations, that print was a small miss. If your process trades the level, it was another month of growth. Both processes are valid. They should not pretend to be the same trade.

Small Firms, Large Firms, and the Sample Gap

Part of the survey split is mechanical. Larger, more global, more digital firms weigh differently across the two panels. Smaller domestic operators feel rates, wages, and local demand with less of a tech cushion. When the gap is this wide, I assume the median business is closer to the cooler print and the mean business is closer to the hotter one, because a handful of scaled platforms can lift an average without lifting a typical shop.

That distinction shows up in markets as well. Index-level earnings can look excellent while a median listed company looks ordinary. September’s real-economy data rhymed with that market structure. Tech carried. Others improved. Nobody, on the stronger survey, was contracting. Improvement is not the same as prosperity, and it is worth keeping the words separate.

What “Running Too Hot” Actually Means Here

Too hot is not a moral judgment. It means nominal spending and pricing power are strong enough to keep inflation from settling at target without a tighter stance than markets currently enjoy imagining. The phrase showed up in the survey commentary for a reason. Price gauges were the evidence, not a vibe.

A too-hot reading can coexist with pockets of weakness. Payrolls can cool while service invoices heat up. Manufacturing can rebound on a narrow base of orders while services broaden. The US economy is large enough to host several climates at once. Policy has to pick a national setting anyway, which is why these months feel argumentative.

If I had to park a view, it would be this. The growth signal is real, tech-heavy, and broader than it was a month ago. The price signal is also real, only partly about fuel, and hostile to a rush toward easier money. The jobs data is the swing vote, and one soft print is not a verdict.

Implications for Anyone Allocating Capital

None of this is a trade recommendation. It is a map of tensions. Growth exposure, especially anything tied to information services and to businesses that sell into a still-spending consumer, has fundamental wind at its back. Duration exposure has a headwind if selling prices keep firming. The mix inside a portfolio matters more than the slogan on the slide.

Quality still earns its keep when the debate is hot growth versus hot prices. Firms that can hold margin without leaning on a single rate cut are less hostage to the next survey. Firms that needed a rapid easing cycle to make the math work just got a more complicated September.

Internationally, a US that is “by far the strongest” on these gauges keeps pulling capital toward dollar assets when the alternative is a weaker nominal pulse abroad. That support is not unconditional. It fades if the price side forces rates so high that US growth itself cracks. We are not at that crack. We are at the argument about whether one is coming.

The Consumer Is Still the Swing Factor

Services are the consumer’s economy. Goods booms come and go with inventories and with rates on big-ticket items. Haircuts, software subscriptions, medical visits, freight on the packages people still order, and the data pipes under all of it are the steady spend. An acceleration in consumer-facing businesses says that steady spend got less steady and more upward in September.

Can that last if prices keep climbing? For a while, yes, especially if employment income is holding even as headline payroll growth cools. It cannot last indefinitely if real wages stall. The survey cannot see that limit. It can only tell you the limit was not reached last month.

That is the suspense worth sitting with. Demand has not rolled over. Costs have reaccelerated. The next few inflation prints and the next payroll revision will decide which of those facts markets respect more. Until then, anyone claiming the services sector “confirmed” a single story is editing out the half of the data that inconveniences them.

A Note on Optimism and Its Half-Life

Expectations at a one-year high feel good in a write-up and age badly if October disappoints. Managers have been wrong before, particularly when a single strong month flatters the backlog. The hiring response is the check on that. You can talk up the outlook in a survey. You hesitate before you post a job and mean it.

The four-year hiring pace is therefore the line I trust more than the optimism line. It can still reverse. Plans get cancelled. But as a snapshot of late September, it says the people closest to the order book were not behaving like a recession was six weeks away.

Putting the Two Prints on the Same Page

Call the stronger survey the expansionist read and the purchasing-managers survey the cautious read. Expansionist: 58.8, five-year high, all sectors up, tech in front, hiring and prices both hot, growth metaphor near 4 percent for the quarter. Cautious: 54.9, still growing, a small step down, no drama, no collapse. The overlap is expansion. The argument is about speed and about heat.

Most of the market commentary last week picked one microphone. A more honest note admits both were on. Tech is leading a genuine upswing in measured service activity. Prices are soaring enough to complicate the easier-policy story that weaker payrolls invited. That is the whole piece, and it does not need a forced winner.

If you remember one tension from September’s service sector, make it this. The boats are rising. The tech boat is rising fastest. And the water itself is getting more expensive. Growth like that can run for a while. It rarely runs quietly, and it almost never runs cheap.