Introduction (00:00–02:00)
Welcome to SteelOnTheNet Podcasts. I'm Dr Andrzej M Kotas, and today is the 4th of June, 2026.
Today's episode is about a problem that the global steel industry has been trying to solve for a decade — and failing. Chinese steel overcapacity. The persistent, structural surplus of steelmaking capacity that has flooded international markets, depressed prices, driven Western mills to the brink, and resisted every policy measure thrown at it.
In 2025, China exported a record 131 million tonnes of steel. To put that in perspective: that is more steel than the entire European Union produces in a year. In a single year of exports.
We are going to examine why this is happening, why it is so difficult to stop, and what it means for everyone working in or investing in the global steel industry.
The answer will take us through the collapse of China's property market, the bankruptcy of Chinese steel companies, and a political conflict that goes to the heart of how China is governed — the tension between what Beijing wants and what provincial governments will actually do.
And we are going to end with a conclusion that I think many in the industry need to hear, even if it is uncomfortable: this is not a cycle. It is a condition. And the sooner the industry plans accordingly, the better.
Segment 1: The Scale of the Problem (02:00–05:00)
Let me start with the numbers, because they are extraordinary.
China produces approximately one billion tonnes of crude steel per year. That is more than half of everything the world produces. The next largest producer is India, at around 145 million tonnes. China produces more than seven times as much as India — more than ten times as much as Japan or the United States.
Global steelmaking capacity reached nearly two and a half billion tonnes in 2025, according to data presented at the OECD Steel Committee's 99th Session in Paris in March 2026 — one of the most authoritative assessments available. Of that total, global excess capacity — the gap between what can be produced and what the world actually needs — reached 640 million tonnes. That excess alone is more than four times the annual steel output of the entire European Union.
And it is not falling. Capacity has increased for four consecutive years. The OECD projects it will continue rising through 2028.
The price consequences are exactly what you would expect. Hot-rolled coil in Asia fell to around $500–520 per tonne in 2025 — levels not seen since 2020. Rebar prices, which matter enormously for construction activity, fell to multi-year lows. European steel capacity utilisation dropped to around 65%. For context, 80% is generally considered the threshold below which pricing power collapses. Europe was running 15 points below that level.
Around one-fifth of all planned low-carbon steel investment projects worldwide were suspended during 2024 and 2025, according to the OECD, because the economics no longer made sense under these price conditions. The green steel transition — which the industry needs urgently — is being slowed by Chinese overcapacity.
This is not background noise. This is the defining market condition of the global steel industry right now.
Segment 2: Root Cause — The Property Market Collapse (05:00–08:30)
To understand how we got here, you have to understand what happened to Chinese real estate.
For decades, property development was the primary engine of Chinese steel demand. Residential construction, commercial development, and the infrastructure that accompanies urbanisation all consumed enormous quantities of steel. At its peak, the property sector accounted for between a quarter and a third of all Chinese steel consumption.
Then Beijing changed course. In 2020, the government withdrew low-interest lending from property developers as part of a deliberate policy to deflate what had become a dangerously leveraged sector. The consequences were immediate and severe.
Major developers — including Evergrande, then the world's most indebted property company with estimated liabilities of $300 billion — collapsed. Construction starts fell sharply. Around 10 million homes had been sold but were never completed. Inventories of unsold properties hit record levels.
For steelmakers, this was catastrophic. Their single largest customer had gone into structural decline. And critically — unlike the downturns of 2008 and 2015 — this time there was no rescue stimulus waiting in the wings. On those previous occasions, Beijing deployed massive infrastructure and construction spending to revive domestic demand. This time, the political commitment to that model has gone.
As one industry expert put it: in 2008, fixing the crisis meant investing more, which was good for steel. Today, fixing the crisis means investing differently — which is not.
The head of China Baowu, the world's largest steelmaker, warned publicly that the current crisis is more severe than anything the industry has experienced this century. Chinese steel demand contracted by an estimated 6.5% in 2025 alone — and analysts do not expect a structural recovery. The urbanisation wave that drove Chinese steel demand for three decades is largely complete. The property sector will not return to its former scale.
What this means in practice is that China has built an industry sized for a demand level that no longer exists — and has no prospect of returning.
Segment 3: Bankruptcies — The Human Scale of the Crisis (08:30–10:30)
The financial consequences have been severe.
By the first half of 2024, approximately three-quarters of Chinese steelmakers were reporting financial losses, according to Bloomberg Intelligence. The industry recorded combined losses through August 2024 of nearly 17 billion renminbi — around $2.4 billion — based on data from China's own National Bureau of Statistics.
The casualties are real. Dongling Group, once regarded as China's largest supplier and distributor of construction steel, and a source of considerable local pride in the city of Baoji in Shaanxi province, announced in July 2024 that it had entered bankruptcy. It was not alone. Forty-six companies across the broader steel sector — producers, trading companies, and processors — declared bankruptcy in September 2024 alone, according to China's National Enterprise Bankruptcy Information Disclosure Platform.
Bloomberg Intelligence identified several additional major producers as facing acute financial pressure, warning that further insolvencies were likely unless market conditions improved materially.
And yet — and this is the crucial point — even severe financial losses are not causing capacity to close. Mills are continuing to produce even while losing money. The reasons for this tell us everything about why the overcapacity problem is so difficult to solve.
Segment 4: The National versus Local Government Conflict (10:30–14:30)
This is where the analysis becomes genuinely interesting — and where most commentary stops short.
Beijing wants to cut capacity. The Five-Year Plan commits to strict production controls through 2030. China's National Development and Reform Commission has directed that approvals for new capacity be halted. The Ministry of Industry and Information Technology has stated that unauthorised new capacity constitutes illegal production. These are serious policy commitments from the central government.
The problem is enforcement.
China's steel industry is not a monolith controlled from Beijing. It is distributed across dozens of provinces, each with its own government, its own fiscal pressures, and its own political priorities. A steel mill that Beijing classifies as surplus capacity looks very different from the perspective of a provincial governor.
- It is a source of value-added tax revenue.
- It is a provider of employment for thousands of workers, with many more thousands employed in suppliers and services around it.
- It is a supplier of steel to local construction and manufacturing.
- And its continued operation reflects well on local economic performance metrics that matter enormously to local officials' career prospects.
Closing that mill means losing tax revenue, managing large-scale redundancies, missing GDP growth targets, and explaining the shortfall to higher levels of government. That is politically costly in ways that simply not enforcing a national directive is not.
The OECD Steel Committee's data reveals precisely how this plays out in practice. In 2024, the median Chinese steel firm received fifteen times more government subsidies relative to its asset size than a median firm elsewhere in the world — up from ten times more in previous years. China's overall steel subsidy rate nearly doubled between 2019 and 2024. And in 2025 alone, fifty-nine new provincial and municipal subsidy programmes to support the domestic steel industry were introduced. Not cut. Added.
So while Beijing announces capacity reductions, provincial governments are quietly paying mills to stay open. The centre and the periphery are pulling in opposite directions.
There is also a powerful economic logic keeping mills running regardless of subsidies. Stopping and restarting a blast furnace is extremely expensive — the refractory lining degrades, and the process of banking down and relighting can cost many millions and take months. Once a blast furnace is lit, the economics strongly favour keeping it running even at a loss, because the cost of shutdown may exceed the cost of continued loss-making operation. Mills also carry a rational expectation — grounded in the history of Chinese industrial policy — that if they can survive long enough, government support will eventually materialise. It has before.
Beijing issued new capacity-cut pledges at the 2025 National People's Congress. Independent analysts noted that the outcomes were mixed. The OECD Steel Committee, meeting in March 2026, stated plainly that existing policy tools are insufficient to address the scale of the problem.
Segment 5: Exporting the Problem (14:30–16:30)
The result of all this is an export surge that has reshaped global steel markets.
Facing no viable domestic outlet, mills have priced aggressively to move volume. As one economist observed, Chinese steel firms near bankruptcy are dumping inventory abroad simply because there is nothing else to do with it — it is the only way to raise cash while waiting for a recovery.
The methods have become increasingly sophisticated as importing countries have responded. As tariffs on finished products multiplied, Chinese producers shifted to exporting semi-finished goods such as billet, which faced fewer restrictions. Exports of semi-finished products surged dramatically. Chinese steelmakers have also invested in production facilities offshore — in Indonesia, Vietnam, Malaysia, Serbia, and Saudi Arabia — allowing steel manufactured under Chinese ownership to enter markets as non-Chinese product, effectively circumventing the trade measures designed to restrict it.
The OECD Steel Committee's March 2026 statement specifically highlighted this circumvention problem, noting that exporters are resorting to a widening range of techniques: slightly modifying products, investing in overseas plants to change the declared origin of the steel, and exporting steel embedded in downstream manufactured products not subject to trade measures. China's exports in 2025 surpassed the combined exports from the rest of Asia for the first time in recent history.
In 2025, seventy-five anti-dumping and countervailing duty investigations were initiated against Chinese steel globally. Vietnam, Chile, the European Union — country after country has moved to protect its domestic industry. But the OECD's own assessment is that these measures, while delivering some early results, are being systematically undermined by circumvention activity.
Conclusion: A Structural Condition, Not a Cycle (16:30–18:00)
So let me bring this together — and give you the conclusion that the data demands.
The conventional wisdom in the steel industry has long been that overcapacity is cyclical. Prices fall, weak producers lose money, capacity closes, prices recover. The market self-corrects. That is how it worked in 2016. That is how analysts assumed it would work again.
The Chinese case has disproved that assumption.
Price pressure does not close mills when the state pays the losses. Anti-dumping measures do not reduce exports when producers invest offshore to circumvent them. Capacity pledges from Beijing do not reduce capacity when provincial governments introduce fifty-nine new subsidy programmes in a single year to keep mills running.
The OECD's own data — the most authoritative global assessment available, gathered from 42 governments in March 2026 — projects global excess capacity continuing to rise through 2028. Not stabilising. Rising. The green transition is making it worse: electric arc furnace capacity is being added to replace blast furnaces, but the blast furnaces are not being retired. Both are running simultaneously.
The overcapacity is not a market failure waiting to self-correct. It is a policy outcome — the product of subsidies, of the national-provincial governance conflict, and of an industrial model that has not yet found a politically acceptable way to shrink. This episode has examined three threads: the overcapacity itself, the bankruptcies it is causing, and the promises — from Beijing and from international bodies — that have so far proved insufficient to address it.
For steel executives, the message is this: Chinese export pressure is a permanent feature of the market environment, not a temporary headwind. Strategy built around its eventual disappearance is not a strategy.
For banks and governments assessing steel investments: the question is not whether Chinese overcapacity will persist — it will — but whether the international trade response will build a coherent and circumvention-proof framework before more capacity outside China becomes permanently unviable. As of mid-2026, that framework does not yet exist.
The OECD set itself a target of June 2026 to develop that framework. Whether it will prove adequate remains to be seen.
This has been a SteelOnTheNet Podcast. I'm Dr Andrzej M Kotas. For full show notes and related resources, visit steelonthenet.com/insights/podcasts. Thank you for listening.
ORCID iD