Growth and hiring accelerate in August, whilst selling price inflation hits three-year high
The headline S&P Global US PMI Composite Output Index rose to an eight-month high in August, edging up from 55.1 in July to 55.4, according to the ‘flash’ reading (based on about 85% of usual survey responses). Output has now grown continually for 31 months, with the latest two months seeing the strongest back-to-back expansions since the spring of 2022.
A sustained robust expansion was reported in the services economy, albeit with business activity growth dipping slightly from July’s year-to-date high. the latest two months seeing the strongest back-to-back expansions since the spring of 2022, assisted by a modest return to growth of services exports. Companies reported improved confidence from customers and new product offerings.
The ongoing service sector expansion was accompanied by a marked acceleration of output growth in the manufacturing sector, where production surged after a slowdown in July to register the steepest monthly rise since May 2022. Having come close to stalling in July, new order inflows in the goods-producing sector also picked up in August, with growth hitting the highest since February 2024 principally on the back of rising domestic demand but also helped by the largest rise in goods exports for 15 months.
While many manufacturers reported improved sales and demand, the upturn in production and order inflows was in part linked to renewed inventory building. Stocks of finished goods rose to an extent not previously recorded since data were first available in 2007, while stocks of purchased inputs showed the second-largest rise seen for over three years.
While stock building was partly fueled by expectations of rising demand, some factories also reported increased safety-stock building amid fears of supply shortages or to protect against further price rises, in turn reflecting the recent impact of import tariffs.
Tariffs were reported as the key driver of further cost increases in August. Companies across both manufacturing and service sectors collectively reported the steepest rise in input prices since May and the second-largest increase since January 2023. Rates of increase accelerated in both sectors. While the manufacturing cost rise was especially large, being the second-steepest since August 2022, the service sector increase was the second-highest since June 2023.
Average prices charged for goods and services rose at the sharpest rate since August 2022 as firms passed higher costs on to customers. Although goods price inflation cooled slightly for a second month in a row, it remained among the highest seen over the past three years. Service sector price inflation meanwhile was the sharpest since August 2022.
Employment rose for a sixth successive month, with the pace of job creation hitting the highest since January (and one of the strongest rates seen for over three years). Service providers took on staff at the fastest pace for seven months while factory job gains reached the highest since March 2022.
Companies largely took on additional staff in response to rising backlogs of work. Uncompleted orders rose for a fifth consecutive month, rising in August at a pace unsurpassed since May 2022 reflecting stronger demand and near-term capacity constraints at some companies.
Backlogs rose at an unchanged and therefore joint-steepest rate since May 2022 in the services economy, while manufacturing backlogs also rose to the greatest extent in over three years.
Having dipped in July, companies’ expectations about output in the year ahead rose to a two-month high in August, though remained below the level seen at the start of 2025 and also the survey’s long-run average.
Service sector sentiment revived partly from a drop in July but continued to run weaker than seen in May and June, and far below levels recorded at the turn of the year. This was closely linked to ongoing concerns regarding government policy. While support from policies such as tariffs helped lift manufacturing optimism in August to a level well-above the post-pandemic average, the degree of optimism in the goods-producing sector remained below January’s recent high, reflecting concerns over higher costs and the impact of geopolitical uncertainty, especially in relation to international trade and supply chains.
Chris Williamson, Chief Business Economist at S&P Global Market Intelligence:
“The data are consistent with the economy expanding at a 2.5% annualized rate, up from the average 1.3% expansion seen over the first two quarters of the year.
“Companies across both manufacturing and services are reporting stronger demand conditions, but are struggling to meet sales growth, causing backlogs of work to rise
at a pace not seen since the pandemic-related capacity
constraints recorded in early 2022. Stock building of finished goods has also risen at a survey record pace, linked in part to worries over future supply conditions. (…)“The resulting rise in selling prices for goods and services suggests that consumer price inflation will rise further above the Fed’s 2% target in the coming months. Indeed, combined with the upturn in business activity and hiring,
the rise in prices signaled by the survey puts the PMI data more into rate hiking, rather than cutting, territory according to the historical relationship between these economic indicators and FOMC policy changes.”
This sequence is interesting:
June 29: We learned that the “resilient consumer” has not increased its consumption since December (actually –0.2%), the first meaningful slowdown since the pandemic and a very rare occurrence outside of recessions.
July 24: S&P Global’s flash PMI rose sharply from 52.9 in June to 54.6 powered by the services economy.
July 31: Real consumer expenditures rose only 0.1% MoM in June, 0.0% in the last 3 months and are unchanged in the last 6 months. All spending categories are weak. Real final sales to private domestic purchasers are up a slow 1.2% annualized in Q2.
August 1: The U.S. added 73,000 jobs in July. Revisions cut down the jobs growth originally reported for May and June by a combined 258,000. That left May as having added just 19,000 jobs and June just 14,000.
August 5: S&P Global’s final Services PMI jumped to a seven-month high of 55.7 in July, up from 52.9 in June. “The rate of expansion in new business picked up from June. The latest rise was solid and the fastest since January. The pace of job creation remained only modest, however.”
But everybody focused on the ISM survey: “The ISM index fell to 50.1, a reading that is consistent with the slowest possible pace of expansion. It is the third-lowest reading since the pandemic year of 2020. While the service-sector is still expanding, that expansion looks like it is girding to a halt. (…) business activity, new orders and employment are all lower. Employment fell to 46.4 from 47.2.”
August 21: S&P Global’s flash PMI rose to an eight-month high from 55.1 in July to 55.4 in August. The latest two months seeing the strongest back-to-back expansions since the spring of 2022. The last two months saw the the strongest back-to-back expansions in services since the spring of 2022. Employment rose for a sixth successive month, with the pace of job creation hitting the highest since January (and one of the strongest rates seen for over three years). Service providers took on staff at the fastest pace for seven months while factory job gains reached the highest since March 2022.”
Most large past divergences between S&P Global and the Employment ISM surveys ended up in favor of S&P Global. If this one is no exception:
- The ISM releases on September 2 (manufacturing) and 4 (services) could be surprising to many. If S&P Global is right, the US economy is actually quite strong and employment growth has strengthened in August.
- The FOMC would be wise to stay put a while longer, particularly since both surveys agree on accelerating inflation, fueled not only by tariffs but by strong underlying demand and limited supply.
Ed Yardeni: “The S&P Global flash purchasing managers’ indexes were strong in August. The M-PMI jumped significantly from 49.8 to 53.3, while the NM-PMI edged down from 55.7 to 55.4.”
This is supported by very strong real world data from corporate America in Q2: S&P 500 earnings are up 12.9% (14.8% ex-Energy), largely beating the July 1 forecast of +5.8%. Revenues are up 6.3% (7.4% ex-E) vs +3.7% expected. Corporate guidance remains solid.
Walmart’s US comps rose 4.6% in the quarter ended August 1, +3.6% in volume. The company raised its full-year sales guidance from +3-4% to +3.75-4.75%, a sign that back-to-school sales are solid.
Sustained employment growth is supported by unemployment claims not deteriorating much so far. On a YoY basis, claims are down 3.9% (4-w avg). Continued claims are up 6.1% but that’s not worsening from the past year.
Growth and demand are thus not a problem.
Inflation is more problematic: “The two business surveys show that both their prices-paid and prices-received indices remain elevated. They both recently rose at the same time as the Trump administration imposed tariffs.” (Yardeni)
Interestingly, Walmart said yesterday that it is seeing more tariff effects but is largely absorbing them as best it can. Walmart said its prices were up 1.0% YoY last quarter. The mammoth retailer is helping keep goods inflation reasonably low so far.
The problem, however, is services inflation as S&P Global put it: “Service sector price inflation meanwhile was the sharpest since August 2022.”
Pity Jay Powell.
Divisions Grow Inside Fed Ahead of Decision on September Rate Cut Cleveland Fed’s Hammack opposes cuts citing rising inflation, while Boston Fed’s Collins signals openness amid labor market concerns
(…) Boston Fed President Susan Collins shared Hammack’s inflation concerns but signaled openness to a rate cut as soon as next month.
Collins said she saw more risks of weaker-than-expected employment trends. She flagged how higher tariffs might squeeze consumers’ purchasing power, which could weaken spending.
If data before the Fed’s September meeting hints at “the risks of worsening labor market conditions relative to those risks of elevated inflation…then it may be appropriate soon to begin dialing back” interest rates, said Collins, who has a vote on rates this year.
Hammack said she is less inclined to cut interest rates because she sees rates as being much closer than some of her colleagues do to a so-called neutral rate that neither spurs nor slows growth.
Hammack voted in favor of two rate cuts last September and November but opposed a third cut in December. “The picture is radically different today,” she said, given relative stability in the labor market and inflation moving up instead of down. Cutting rates in that environment “doesn’t seem appropriate.” (…)
Collins said she expected inflation to continue rising through the end of the year before resuming an earlier decline in 2026.
She said it was too soon to rule out larger or more persistent increases in prices. But at the same time, the Fed couldn’t “wait until we know everything that’s going” to happen, she said. “That’s going to be much too late.”
(…) Schmid pushed back on market pricing that points strongly to the Federal Open Market Committee lowering its key borrowing rate next month.
“We’re in a really good spot, and I think we really have to have very definitive data to be moving that policy rate right now,” he said during a “Squawk Box” interview that aired Thursday. “In September, we’ll get around tables and we’ll collaborate and we’ll figure it out, but yeah, I think there’s a lot to be said between now and September.” (…)
(…) “The last inflation report that came in, where you saw services inflation — which is probably not driven by the tariffs — really start shooting up,” he said. “It’s a dangerous data point, I’m hoping that that’s bit of a blip.” (…)
Germany falls back into ‘recessionary territory’ as second-quarter GDP revised down Europe’s largest economy contracted 0.3% as exports and investment drop
A Bolder Blend: The Forces Behind Stronger Productivity Growth
(…) An economy’s potential output growth can be broken into two parts: labor productivity growth and labor force growth. With stricter immigration and population aging weighing on growth in the labor supply, strong labor productivity growth is all the more necessary to keep output growth solid. When workers are more productive, firms tend to enjoy increased profitability, which can provide them flexibility to absorb higher costs, reinvest in the business or lower selling prices. Solid productivity growth has contributed to inflationary pressures easing since the pandemic.
Labor productivity measures the efficiency of production and is calculated as output per hour worked. Productivity growth is typically noisy quarter-to-quarter, and this year has been more volatile than usual due to massive swings in trade flows distorting overall output growth.
For instance, output per hour worked in the nonfarm business sector rose at a 2.4% annualized pace in the second quarter, bouncing back from a 1.8% annualized decline in the first quarter. Smoothing through the noise with a four-quarter moving average, nonfarm labor productivity was up 1.8% year-over-year in the second quarter, matching its average annualized pace since the end of 2019. At that pace, this cycle’s run-rate is noticeably higher than the prior cycle’s average of 1.5% and is closer to its historic average of 2.1%.
To frame the discussion, we decompose labor productivity growth into three broad categories: labor composition, capital input and total factor productivity.
Labor composition reflects shifts in the skills and experience of the workforce and can be thought of as a measure of labor quality. Capital input reflects investment in tangible and intangible assets, less depreciation. This category essentially tracks the tools workers have at their disposal. Meantime, total factor productivity (TFP) is the portion of productivity not accounted for by labor and capital inputs. TFP can be viewed as a proxy for process improvements and other innovations that give workers better tools and methods for production.
Labor composition has had a small effect on productivity growth over the past few decades and is not a major driver of the recent rise. Data on age and education are used to proxy worker experience and skills, and while educational attainment has risen, sustained improvements in labor composition have been restrained by cyclicity. (…)
Unlike labor composition, capital deepening has been a steadfast source of productivity growth. Capital investment has contributed an average of 1.08 percentage points to labor productivity growth since the end of 2019, roughly on par with the 1.04 percentage point contribution in the cycle before the pandemic. Although gross business fixed investment jumped following the pandemic lockdowns, elevated borrowing costs and economic uncertainty have weakened growth over the past few years, underpinning the roughly unchanged contribution from capital input in the cycle to date.
Source: Federal Reserve Bank of San Francisco and Wells Fargo Economics
Even as the capital stock’s contribution to labor productivity growth has moved sideways over the past decade, the composition of investment has shifted. Before the pandemic, capital investment was broad-based across equipment, software and research & development. Today, business investment has pivoted away from equipment and more toward software and research & development.
The shift reflects further diffusion of automation software and broadening experimentation with artificial intelligence. Investment in structures has also picked up. Federal government incentives have underpinned private construction of manufacturing facilities that will eventually produce “high-tech” equipment and components, such as computers and semiconductors, while at the same time, investment in data centers has boomed.
After accounting for the contributions from capital input and labor composition, faster growth in total factor productivity has been the differentiating factor behind stronger labor productivity growth this cycle. Since the end of 2019, TFP has risen at a 0.7% average annualized pace, running a bit hotter than the prior cycle’s average of 0.5%. Identifying the sources of TFP is inherently difficult, as the measure itself is computed as a residual and thus is not directly observable. That said, research has pointed to several potential TFP enhancers.
One prominent theory on stronger TFP growth this cycle centers around work-from-home. Roughly 27% of paid work days were worked remotely in July, up from 7% in 2019. While the focus time afforded by working from home may help experienced employees work more efficiently, it can also make collaboration difficult. Less experienced workers cannot as easily consult with co-workers or supervisors on issues when working remotely, which can stunt their productivity growth. In short, the casual relationship between the rise in work-from-home and higher TFP growth since the pandemic remains contested.
Source: Federal Reserve Bank of San Francisco and Wells Fargo Economics
Source: U.S. Department of Labor and Wells Fargo Economics
Another theory points to the period of significant labor market churn in the wake of the pandemic. The quit rate soared to its highest on record in 2022 (3.0%), reflecting the droves of workers switching jobs in pursuit of higher wages and better skills matches. The broad reallocation of labor likely boosted TFP growth in subsequent years as employees brought fresh ideas to their new employers and employee effort improved with the more preferable job match. The recent normalization in the quit rate, however, suggests any lift to productivity from worker reallocation has likely run its course.
Another potential source is an increase in entrepreneurship since the pandemic. Applications for business formations jumped roughly 50% in the second half of 2020 from the same period in 2019 and are continuing to run well-ahead of pre-pandemic levels. The burst of new firms along with increased job-switching can help spur innovation within industries and increase dynamism across the economy.
Source: U.S. Department of Commerce and Wells Fargo Economics
Source: U.S. Department of Commerce and Wells Fargo Economics
What about today’s most gripping innovation of all—generative AI? While it is still early to see the effect on TFP growth, generative AI will likely play a more integral role in the coming years. Capital investment in software and R&D has already strengthened amid broadening adoption, which has supported the capital input contribution to labor productivity growth. As workers learn to work with and gradually integrate generative AI into their processes, the second-order effect of today’s capital deepening should be a rise in efficiency and other process improvements that manifest in TFP gains tomorrow.
Yet, there remains significant uncertainty on how long it will take AI to diffuse throughout the economy. The Census Bureau estimates just 9% of firms used AI in the production of goods or services in May, which is a far cry from widespread adoption. Although that is a low starting point, we suspect the speed of AI adoption will be faster than that of personal computers and the internet, as the adoption cost associated with AI is generally low. Many office places already have the physical hardware needed and many workers have likely experimented with AI tools outside of work.
As the tailwinds from generative AI continue to muster, the outlook for productivity growth remains positive, but momentum has somewhat dissipated this year due to other factors. The whipsawing of trade policy left many businesses scrambling to reconfigure production flows and delaying major investments amid elevated uncertainty. With more time allocated to defensive maneuvers and less time allocated to growth-enhancing activities, nonfarm labor productivity has been essentially unchanged since the end of 2024.
Should economic growth continue to moderate, the trend in productivity is unlikely to pick up meaningfully this year. Slower sales pose a particularly acute threat to small businesses who often operate on thin profit margins and are at greater risk of closure, which would unwind some of the lift from entrepreneurship since the pandemic. Other firms may hunker down rather than ramp up capital investment or launch new products and processes this year.
As economic activity improves in 2026 with less restrictive monetary policy and more stimulative fiscal policy, we expect labor productivity growth to edge higher again. Stronger growth is poised to bolster capital investment, which stands to benefit from deregulation as well; a less restrictive regulatory environment could enable businesses to focus resources on their primary areas of production. Improvement in the labor market could also rekindle job-switching.
The degree of improvement in the medium term remains uncertain with the countervailing forces of lower immigration and higher trade barriers. While a cut-off in low-skill immigration could force domestic businesses to invest more heavily in capital, the current environment could deter high-skill immigration as well, which would curtail knowledge spillover and innovation.
Meantime, higher tariffs reduce foreign competition and weigh on the incentive for domestic firms to improve efficiency. That said, the manufacturing sector has a higher output per hour worked than most of the service sector, so faster growth in this industry than the broader economy could bolster overall productivity growth.
The Congressional Budget Office projects nonfarm labor productivity growth to settle at an annual rate of 1.5% by 2030, essentially in-line with its pre-pandemic average. We are optimistic that productivity will run stronger than that— likely closer to, if not a bit higher than, its historic trend of 2.1%.
Capital deepening has already gathered momentum this cycle and is likely to continue to intensify amid the diffusion of generative AI and re-shoring of “high-tech” production. We have less conviction on TFP, as the timing and degree of efficiency gains from AI remain uncertain. We are somewhat cautious in the near term given pandemic-related tailwinds have faded and policy cross-currents could dampen the jolt from AI. Even so, productivity growth a little over 2% would counteract slowing growth in the labor force and still keep potential output growth stronger than it was in the cycle preceding the pandemic.
Time will tell if this “capital deepening” will prove profitable enough for the mega spenders. Read below:
DeepSeek unveils V3.1 model with agent focus and China chips
The 685-billion-parameter model introduces what DeepSeek calls a “hybrid inference structure,” allowing users to toggle between rapid “non-thinking” responses for basic queries and slower “thinking” mode for complex reasoning tasks through a “DeepThink” button on the platform. According to the company’s WeChat announcement, this represents “our first step toward the agent era”.
DeepSeek-V3.1 delivers significant performance improvements over its predecessors, scoring 71.6% on the prestigious Aider coding benchmark and achieving comparable results to proprietary models like Anthropic’s Claude Opus while costing substantially less. The model processes tasks at approximately $1.01 each, compared to nearly $70 for equivalent workloads from competing systems.
The release coincides with DeepSeek’s preparation for next-generation Chinese-made AI chips, utilizing FP8 (8-bit floating point) processing formats optimized for domestic hardware. This technical specification suggests the company is positioning itself for China’s emerging semiconductor ecosystem as Beijing pushes to reduce dependence on Western chip manufacturers.
The timing appears deliberate, coming as the Trump administration recently permitted Nvidia to resume limited AI chip sales to China through its H20 model, though Beijing’s response has been notably tepid. Chinese officials have advised domestic firms against relying on the American chips, reflecting broader ambitions for technological self-sufficiency.
Cambricon Technologies led a rally of Chinese chipmakers on Friday after artificial intelligence start-up DeepSeek unveiled an updated model that would be compatible with domestically made semiconductors. (…)
“If DeepSeek can use China-made chips, then the rest of the semiconductor universe will fly,” said Wee Khoon Chong, a senior strategist at BNY. “The potential demand for Chinese chips is going to be huge.” Huawei is widely seen as the main competitor in China to Nvidia, with its Ascend AI chip series being broadly adopted by state-owned enterprises and telecoms companies. (…)
“The fact that China can tell people not to buy H20 must mean they know there’s an alternative. Otherwise, they would just be shooting themselves in the foot,” said BNY’s Chong. (…)
