By John Richardson
LET ME KICK this blog post off with a confession that is not frequently made: I don’t have the faintest clue which of the above three scenarios will come true. This is the only right answer in the much more complex chemicals and polymers world we now live in.
These three scenarios would result in vastly different global linear-low-density polyethylene (LLDPE) balances, as China is the world’s biggest net importer.
Under the ICIS base case, operating rates average just 72% in 2025-2035, leading to net imports averaging 4.5m tonnes a year—down from the 2020-2024 average of 5.9m tonnes. But imports would still remain very healthy.
Raise operating rates to an average of 81%—nine percentage points closer to the actual 2024 operating rate of 90%—and net imports fall to 2.3m tonnes.
Increase operating rates a further nine percentage points back to 90% and this happens: Net imports fall to just 0.3m tonnes on average, with China becoming a net exporter in some years.
For all three scenarios, I stuck to our base case forecast of demand growth averaging 3% per year in 2025-2035—compared with 12% in 1992-2024, most of which fell during the Chemicals Supercycle.
I could, of course, have also adjusted demand growth, given China’s macroeconomic challenges. But one step at a time. What follows is complicated enough.
The Wisdom—or Lack Thereof—of One Crowd
While I don’t have the faintest clue about the right scenario, I can suggest the key questions you need to ask. This comes from a wisdom of crowds approach—testing out ideas with my network.
We agreed that you cannot just consider plant economics when assessing China’s LLDPE operating rates in 2025-2035.
Profitability will play an important role in setting operating rates in a global market that could remain severely oversupplied. But we must also consider the following factors that may lead to higher operating rates than our base case suggests:
- Beijing is concerned about LLDPE supply security in an uncertain geopolitical world. The government instructs companies to run hard, even at a loss. Subsidies could be provided.
- In 2024, China’s total exports rose by 13% compared to 2023, contributing nearly 2 percentage points to GDP—the highest net export contribution since 2006. Last year’s growth in exports was more than double that of 2023. Trade tensions don’t prevent further export growth—either directly or via countries such as Vietnam and Mexico. LLDPE demand is boosted by strong growth in exports of finished goods, boosting the earnings of local producers. This is of course a big “if“.
- China has built a wave of world-scale, highly integrated plants that sit far to the right of global cost curves. Even though the global industry is losing money, China is in a stronger position than key import suppliers like South Korea and Singapore.
- China triples the number of polymer grades it produces. As its economy becomes more sophisticated, demand for higher-value C6 and C8 LLDPE grades rises. To support economic diversification, China produces more of these grades locally, reducing the need for imports.
Yet Another Variable: Different Outcomes for LLDPE Capacity Growth
Below is our base assumption for China’s LLDPE capacity growth (2025-2035).
The biggest capacity increases are scheduled between 2025 and 2028. This means most of the capacity due on-stream in 2025-2035 has already been built, is under construction, or has been approved.
However, LLDPE capacity could still be influenced by:
- A decline in capacity growth beyond 2028 because of a lack of new feedstock supply from local refineries. China is capping refinery capacity from 2028 onwards because of the rise of electric vehicles.
- China’s carbon reduction targets. Carbon Brief reports that a minimum 28% reduction in total GHG emissions (from 2023 levels) by 2035 is crucial for staying on track for the 2060 net-zero goal. This may restrict chemicals capacity expansion.
- China’s next Five-Year Plan (2026-2030)—will it prioritise chemicals self-sufficiency or carbon reduction?
- Older plant closures. Some steam crackers are now 20+ years old. Have they been modernised, or will they be shut down?
- Unconfirmed reports that China wants to close some coal-based chemicals capacity due to environmental and cost concerns.
Conclusion: How AI Will Transform Market Forecasting
Before 2017, AI was largely theoretical. Then came the Google paper on Transformer architecture, the Scaling Laws and a flood of funding, and AI became a reality.
Now, AI is fundamentally changing chemicals market analysis. Instead of spending hours on manual calculations, AI will do it in minutes. Data crunching will be faster, cheaper, and more accurate than traditional forecasting methods. Large language models (LLMs) will generate reports and studies in seconds, without mistakes, at near-zero cost.
Next consider this March 2024 TED Talk by Erik Guzik, an economics professor at Montana Western University. He cited a study where GPT-4 scored in the top percentile eight times in a row on the Torrance Tests of Creative Thinking.
The testers did not know they were testing AI. GPT-4 could not scrape the internet for prior test answers. The results were successfully peer-reviewed and further tests produced similar results.
Guzik concluded that AI was capable of generating “novel, unexpected, and unique ideas”.
So, does this mean the end of wisdom-of-crowds forecasting committees? Will AI completely replace humans? I think not. Trust in machines remains a major hurdle. While AI will supplement human ingenuity, it won’t fully replace it—at least, not yet.
But one thing is clear: we are in a new world.
The breakthroughs in AI over the last nine years are as significant as the inventions of electricity, steam power, and the internet. And these breakthroughs could not have come at a better time. They allow us to model today’s chemicals market complexities in ways previously unimaginable.