Everything About AI
NextFin News - China's industrial profits jumped 14.4% in September from a year earlier, the fastest pace in three months, as a surge in producer prices — driven by oil, coal, and other energy costs — flowed straight to the balance sheets of upstream miners and energy producers. The acceleration from 4.2% in August, the slowest reading since November 2025, beat the 12% forecast compiled ahead of the release, but it also sharpened a divide that has defined China's 2026 recovery: the factories at the top of the supply chain are getting richer, while the factories at the bottom are paying the bill.
The gains are narrower than the headline suggests. Profits for the first nine months of 2026 rose 15.5%, actually edging down from the 15.7% pace recorded for January to August — meaning the September pop was not broad enough to lift the cumulative trend. The producer price index climbed 5.0% year on year in September, the fastest since June 2022 and well above the 4.5% economists expected, while consumer prices rose just 0.9%. For every mine and oilfield booking higher prices as revenue, there is a carmaker or furniture maker absorbing those same prices as a cost it cannot pass on to households that are barely spending.
This is not a general recovery. It is a price-driven transfer of income up the supply chain — and the question it raises is whether the transfer is durable, or whether it is a cyclical oil wave that will roll back once energy prices cool.
The Price Mechanism: Why a 5% PPI Prints as Profit for Some and a Squeeze for Others
The transmission channel is simple and mechanical. China's producer price index for industrial products rose 5.0% year on year in September, with the increase concentrated exactly where the profit data says the money went. Prices in mining and quarrying jumped 25.4%, raw materials rose 9.4%, and the processing industry gained 3.4% — while consumer goods prices fell 0.1%. Within that, the processing of petroleum, coal, and other fuels rose 8.4% year on year and a striking 21.6% month on month, the single strongest monthly impulse in the September price data.
That is not a demand boom. It is a cost shock with winners and losers determined by position in the supply chain. A coal miner or an oil producer sells into that 25.4% price increase and books it as revenue. A steel mill, an automaker, or a home-appliance factory buys into it and sees its margin thin unless it can raise prices to consumers — and with retail sales growing just 1.1% in September and consumer-goods factory-gate prices still negative, it cannot.
The official production data confirms the asymmetry. In September, value added in the mining industry grew just 1.3% year on year, while manufacturing grew 6.0%. Output volumes were not the story: rolled steel fell 3.8%, cement fell 5.3%, and motor vehicles fell 5.7%. Yet upstream profits soared. The gap between flat or falling volumes and surging profits is the signature of a price-driven recovery, not a volume-driven one. When factories are producing less but earning more, the earnings are coming from the price column, not the quantity column.
Lynn Song, chief economist for Greater China at ING, captured the policy implication of the shift:
"It marks a bit of a shift in tone," she said, adding that more policy support would be needed to hit both the near-term and long-term growth targets, including the 4.5%-5.0% GDP goal for 2026.
The tone has shifted because the price impulse is doing some of the demand work that policymakers would otherwise have to do — but only for the upstream half of the economy.
The second-order effect is where the story gets more interesting. A price-driven profit surge at the top of the chain does not stay contained. It shows up first in tax receipts and state-owned enterprise dividends, which strengthens the fiscal position of resource-heavy provinces and the central government's extraction capacity from the energy complex. It shows up second in the credit cycle: profitable miners and oil producers borrow and invest, while squeezed downstream firms draw down working capital and delay capex. The aggregate profit number can look healthy while the distribution of cash flow is quietly migrating from the competitive, labor-intensive downstream to the concentrated, capital-intensive upstream. That migration is deflationary for employment and inflationary for input costs — a combination that gives policymakers the worst of both worlds.
Who Wins, Who Loses: The Sector Split That the Headline Hides
The most detailed official sector breakdown covers January to August, and it shows the concentration with unusual clarity. Electronics — the sector most exposed to the global AI investment cycle — saw profits surge 110% year on year and accounted for roughly 62% of all industrial profit growth, according to the statistics bureau. Non-ferrous metals rose 82.9%, coal mining 51.6%, and chemical raw materials 51.0%. Oil and natural gas extraction gained 17.0%. In aggregate, mining profits climbed 35.1% to 766.15 billion yuan, compared with a 17.4% rise in manufacturing.
On the other side of the ledger: automobile profits fell 16.0%, non-metallic mineral products — the sector most tied to property — fell 46.7%, and ferrous metals fell 62.4%. The contrast between non-ferrous metals at +82.9% and ferrous metals at -62.4% is the cleanest single data point in the release: this is not a rising tide for "China commodities." It is a premium for the metals and inputs that feed AI data centers and electrical infrastructure, while the metals that feed construction continue to contract.
That split maps directly onto the two forces behind the September number. First, the AI-linked electronics chain, riding a global capital-expenditure boom in chips, servers, and data-center equipment, with high-technology manufacturing value added up 15.3% in September and 14.3% for the year to date. Second, the energy complex, lifted by higher oil prices, winter-heating coal demand, and tight supply in major coal-producing regions. Exports provided a third tailwind: export delivery value from large industrial firms rose 17.0% year on year in September, meaning overseas demand — not Chinese households — is absorbing a meaningful share of factory output.
The divergence has been building all year. Zhaopeng Xing, senior China strategist at ANZ, described the same mechanism months earlier:
"Upstream sectors and the computer industry saw sharp rises, while downstream manufacturing remained under pressure, in line with the producer price index, suggesting that price improvement was the main driver of corporate profit growth."
September did not change that diagnosis. It intensified it.
Cyclical Wave or Structural Shift? The Answer Is Both — and That Matters
The central analytical question is whether this profit surge is cyclical and mean-reverting, or structural and durable. The answer is that two different forces are at work, and they must be separated rather than blended. Getting this distinction wrong flips the conclusion, so the evidence floor for each leg matters.
The oil-and-coal leg is cyclical, and the historical record is clear about why. China's industrial profits have spiked on energy-price impulses at least three times in the past fifteen years and given back the gains each time: the 2011 commodity supercycle peak, the 2016-2017 supply-side capacity-cutting campaign, and the 2021 coal and energy price spike. Each episode produced a sharp, price-led profit pop in mining and raw materials, followed by a reversion as energy prices normalized and downstream demand failed to keep pace. The 25.4% mining price increase is not a permanent new level for the price structure; it is a swing factor that can and does reverse. When the energy leg rolls over, the September 14.4% print will look like a high-water mark rather than a new baseline. The ingredients of this particular wave — winter-heating demand, temporary tight supply in major coal-producing areas, and expectations of rising international coal prices — are by their nature transitory.
The AI-electronics leg is structural. It rests on a regime change in global technology capital expenditure — data centers, AI accelerators, and the electrical infrastructure that supports them — that is multi-year in nature and not dependent on China's domestic property cycle. Electronics contributing 62% of profit growth is not a price spike; it is a demand shift that will not self-correct on its own. This is the part of the profit story that can persist even after the oil wave fades.
So the correct read is narrower and more specific than either extreme: own the pricing power, not the cycle. The headline 14.4% is a cyclical number wearing structural clothing. Strip out the energy leg and the electronics leg, and the middle of the industrial distribution — the firms most exposed to Chinese household demand and the property chain — is still contracting.
The Counter-Thesis: This Is a Real Recovery, Not Just a Price Illusion
The strongest case against this reading is that the profit recovery is broader than the sector data implies, and that the price impulse is itself the sign of a genuine demand turnaround. The argument runs as follows: producer prices are rising because demand is finally returning; Beijing's "anti-involution" campaign — the push to curb destructive price wars where firms cut prices below sustainable levels to win share — is restoring pricing discipline across industries; and once prices start moving, profit margins follow, lifting investment and employment through the rest of the economy. Under this view, September is the inflection point, not the peak.
That case has real support. Industrial profits posted their first annual increase since 2021 earlier this year, and the year-to-date 15.5% growth is the strongest cumulative pace in years. High-technology manufacturing is growing at double-digit rates on real output, not just prices. And export strength at 17.0% shows that Chinese factories remain globally competitive even as domestic demand lags. A sustained exit from producer-price deflation is, in any large economy, a genuinely constructive signal — it means firms can finally pass costs through, which is the precondition for a capex cycle.
But the counter-thesis breaks on the volume data and the cumulative trend. If this were a genuine demand-led recovery, output would be accelerating alongside profits — yet rolled steel, cement, and vehicles all contracted in September, and the nine-month cumulative profit growth of 15.5% is lower than the 15.7% pace for January to August. A real recovery broadens the trend; this one narrowed it in the very month the headline surged. The price improvement is real, but it is a transfer within the industrial sector, not a creation of new demand. And the anti-involution campaign, while real, is a policy preference, not a demand shock — it can lift margins for survivors, but it cannot manufacture the household spending needed to absorb higher prices.
The falsifying signal is specific: if the next two monthly PPI prints hold at or above 5.0% year on year while industrial output accelerates above 6.5% and the cumulative profit growth rate re-accelerates above 16%, the broad-recovery thesis is confirmed and the narrow-transfer read is wrong. Until then, the default assumption is a price-driven, sector-concentrated profit wave.
What Comes Next: Beneficiaries, the Exposed, and the Signals to Watch
The mechanism translates into a clear asymmetry across sectors and time horizons.
Short term (sentiment and liquidity): the momentum favors upstream energy, coal, non-ferrous metals, and the AI electronics chain. These sectors have both the price tailwind and the earnings momentum, and the September print will reinforce capital flows into them. The exposed are downstream manufacturers — automakers, home-appliance producers, and property-linked materials makers — whose input costs have risen while their selling prices remain trapped by weak household demand. Property investment, still down 20.6% for the year to date, is the anchor dragging on that downstream half.
Medium term (fundamentals): the base case is a gradual normalization. Oil and coal prices cool from their September peaks, the 25.4% mining price impulse fades, and headline profit growth settles back toward the mid-to-high single digits — still positive, but far below 14.4%. The upside case is that AI-driven electronics demand accelerates further and offsets the energy fade, keeping aggregate profit growth above 12%. The downside case is that energy prices fall faster than expected while domestic demand stays soft, compressing margins across the whole chain and pulling cumulative profit growth below 10%.
Long term (structural): the profit structure of Chinese industry is being rewritten around two poles — the AI technology chain and the energy-and-materials complex — while the old property-and-construction pole continues to shrink. Ferrous metals down 62.4% alongside non-ferrous metals up 82.9% is not a cycle; it is a reallocation of the economy's profit pool that will outlast any single commodity-price move.
The signals to watch are concrete. First, the October and November PPI prints — sustained readings above 5.0% would confirm the price impulse is durable rather than a one-month energy spike. Second, the sector profit breakdown for September and October, which will show whether electronics' 62% contribution is holding or rolling over. Third, property investment and retail sales — if either re-accelerates, the downstream squeeze begins to ease and the recovery broadens.
The bottom line: China's industrial profits surged in September, but the surge is a story about who captures the price increase, not about how much new demand exists. The oil wave will fade; the AI chain may not. The market that prices this as a broad cyclical recovery is pricing the wrong recovery.
Explore more exclusive insights at nextfin.ai.
