The Point of No Return: Estimating the Irreversibility Threshold of German Chemicals
Executive Summary
German chemicals are not in a cyclical downturn that the next upswing will repair. They are close to, and on most measures beyond, a threshold below which plant closures become self-reinforcing and permanent. The 2026 oil, naphtha and LNG shock caused by the war with Iran did not create this situation: it hit an industry that had spent four years below the utilisation rate at which it earns its cost of capital, had closed a tenth of its capacity and had almost stopped investing.
The evidence is stark. Production of chemicals excluding pharmaceuticals was 21.5 percent lower in 2025 than in 2021, and a further fall of 1.5 percent is expected in 2026. Utilisation has been below the 82 percent profitability threshold for eighteen consecutive quarters and stood at 73.2 percent in mid-2026. The decline is concentrated in basic chemicals: polymer output in the first half of 2026 was 23 percent below mid-2022, petrochemicals 18 percent and inorganic basics 15 percent, against 5 percent for specialty chemicals.
We make four original contributions. First, a threshold hazard model fitted to the European closure radar places the closure threshold at 80.7 percent utilisation, with each point below it adding about half a percent of capacity to annual closures. Second, we show that closures restore the profitability of the survivors only if their output is recaptured by other domestic plants. Between 2021 and 2025, demand for German output fell 2.3 times as fast as implied capacity.
Third, a dynamic model maps an irreversibility frontier: the combinations of utilisation and demand growth from which the industry returns to the threshold within a decade. Germany lies well inside the region of no return; a return would need sustained underlying demand growth of 0.7 to 1.4 percent a year. A new quarterly Irreversibility Index rose from 24 at the end of 2022 to 87 in mid-2026.
Fourth, we translate output into trade. Germany already imported USD 12.8 billion more organic chemicals than it exported in 2025. The combined surplus in organic chemicals, plastics and other chemical products (USD 16.5 billion) vanishes after an output loss of about 8.4 percent. It turns negative in 2030 under a persistent energy premium and in 2028 under a second-round shock; the Monte Carlo median is 2029.
The temporary rebound of mid-2026, driven by hoarding and Gulf outages, is deceptive, and the 2026 to 2028 power price relief offsets at most two fifths of the gas cost gap with the United States. Not every plant must be saved, but Germany should decide deliberately which keystone nodes of its value chains to keep, because closed capacity does not come back.
Key findings.
- German chemicals have been below the profitability line for eighteen consecutive quarters. Capacity utilisation has not reached the 82 percent level the industry association treats as the profitability threshold since the end of 2021. It has drifted down by 1.8 percentage points a year since 2022, and a structural break in the second quarter of 2025 lowered its mean from 77.5 to 72.7 percent.
- The closure threshold sits at about 81 percent utilisation. A threshold hazard model fitted to the European closure radar shows that each percentage point of utilisation below 80.7 percent adds about 0.52 percent of capacity to closure announcements per year. At the second-quarter 2026 reading of 73.2 percent, the model implies closures of 3.9 percent of German capacity a year, roughly 3.4 million tonnes.
- Closures do not cure low utilisation. When a whole product line exits and imports take its market, closing 10 percent of capacity lifts the utilisation of the survivors by only about 0.6 points. Between 2021 and 2025 demand for German output fell 2.3 times faster than capacity. Restoring the threshold through closures alone would require shutting about 60 percent of capacity.
- On current demand, the point of no return has been passed. Returning to the threshold within ten years requires underlying demand growth for German output of about 0.7 percent a year, and 1.4 percent if Verbund spillovers dominate. Output fell by 5.0 percent a year from 2021 to 2026. Our Irreversibility Index stands at 87 out of 100, against 24 at the end of 2022.
- Germany is likely to become a net importer of core chemicals around 2029. Germany already ran a USD 12.8 billion deficit in organic chemicals in 2025. The combined surplus in organic chemicals, plastics and other chemical products (USD 16.5 billion) disappears after an output loss of about 8.4 percent. In 20,000 Monte Carlo simulations the median crossing year is 2029, with a 70 percent probability by 2030 and 87 percent by 2035.
- Relief is smaller than the shock. The 2026 to 2028 industrial power price relief is worth at most about EUR 1.3 billion a year to the chemical industry, against a structural gas cost gap with the United States of about EUR 3.2 billion a year and a feedstock shock of up to EUR 3 billion. In the adverse scenarios, 21 to 36 million tonnes of capacity and 80,000 to 140,000 direct and indirect jobs are at risk by 2035.
1. Introduction: From Cycle to Exit
For most of the post-war period, the German chemical industry was the textbook case of a successful European heavy industry. Integrated Verbund sites such as Ludwigshafen turned imported naphtha and natural gas into thousands of intermediate products, supplied the automotive, construction, pharmaceutical and consumer goods industries at short distance, and exported most of their output. The industry remains Germany’s third-largest manufacturing sector after automotive and mechanical engineering. In 2025 it generated sales of about EUR 220 billion, of which EUR 136 billion abroad, and employed about 478,000 people according to the association’s statistical definition.
Since 2022 that model has been under continuous stress. The loss of Russian pipeline gas drove quarterly gas prices above EUR 100 per megawatt hour, and production of chemicals excluding pharmaceuticals fell by about 10 percent in 2022 and 12 percent in 2023. When gas prices normalised, the expected recovery did not come: output edged up by about 3 percent in 2024, fell 3.3 percent in 2025, and utilisation kept drifting lower. By December 2025 the chief executive of BASF described the situation as probably the industry’s most difficult in 25 years.
Then came 2026. The war that began on 28 February, the disruption of shipping through the Strait of Hormuz and the extension of the Qatari LNG force majeure raised oil prices above USD 100 per barrel at their peak, lifted naphtha prices and pushed European gas prices up again. In the first quarter, crude oil averaged more than USD 80 per barrel, 26.5 percent more than in the previous quarter, and naphtha about EUR 532 per tonne, 17.9 percent more. The shock is broader than in 2022: it hits not only gas used as fuel and feedstock, but the oil-based feedstock that supplies about 71 percent of the industry’s material inputs.
This paper asks a simple question with a precise meaning: has the German base chemical industry crossed a point of no return? By that we mean a state in which the mechanisms that would normally restore profitability, namely the closure of the least efficient plants and the recovery of demand, no longer work, so that each round of closures leads to the next. If such a threshold exists and has been crossed, the policy debate about temporary relief misses the point. The relevant question becomes which parts of the chemical value chain Germany can and should keep, and how to manage an orderly transition to importing the rest.
The analysis relies on VCI statistics based on Destatis data, the Cefic and Roland Berger closure radar, Cefic’s chemical trends report, company disclosures and UN Comtrade data. Several series are short or revised; unobservable parameters are treated as explicit assumptions with sensitivity analysis.
2. Four Years Below the Line
2.1 Output: a fifth of the industry gone
Table 1 assembles the annual production record for chemicals excluding pharmaceuticals. We build an index with 2021 as the base from the annual changes published by the association: minus 10.3 percent in 2022, minus 12.1 percent in 2023, plus 2.9 percent in 2024 (January to November), minus 3.3 percent in 2025 and the association’s forecast of minus 1.5 percent for 2026. The resulting index stands at 78.5 in 2025, a decline of 21.5 percent from 2021. This matches the figure of about 21 percent cited in industry commentary in April 2026 and provides a useful consistency check.
Table 1. Production of chemicals excluding pharmaceuticals in Germany.
| Year | Annual change, % | Index (2021 = 100) | Note |
|---|---|---|---|
| 2018 | n.a. | 97.8 | Implied: 2024 output 17% below 2018 |
| 2021 | plus 5.7 | 100.0 | Pre-crisis peak |
| 2022 | minus 10.3 | 89.7 | Gas price shock |
| 2023 | minus 12.1 | 78.8 | Demand slump |
| 2024 | plus 2.9 | 81.1 | Partial rebound |
| 2025 | minus 3.3 | 78.5 | Renewed decline |
| 2026 (forecast) | minus 1.5 | 77.3 | VCI forecast, July 2026 |
Source: VCI World Chemistry Report, VCI quarterly reports and press releases; index: author's calculations.
Two features of Figure 1 matter. The loss is large relative to any previous downturn, and the 2024 rebound faded within a year although gas prices had returned to roughly their pre-2021 range. If the 2022 shock had been purely cyclical, two years of normal gas prices should have restored much of the lost output. They restored almost none: the first sign of hysteresis.
2.2 Utilisation: eighteen quarters below the threshold
The association publishes a quarterly capacity utilisation rate based on survey data. Table 2 summarises the twenty quarterly observations we could retrieve from its reports between the second quarter of 2021 and the second quarter of 2026; the series is plotted in Figure 2. The fourth quarter of 2021 is missing from the published material we could access.
Table 2. Capacity utilisation in the German chemical industry: summary statistics.
| Statistic | Value |
|---|---|
| Highest reading (Q2 2021) | 85.9% |
| Last reading at or above 82% (Q3 2021) | 82.3% |
| Consecutive quarters below 82%, to Q2 2026 | 18 |
| Lowest reading (Q3 2025) | 70.0% |
| Latest reading (Q2 2026) | 73.2% |
| Annual average 2022 / 2023 / 2024 / 2025 | 79.5 / 77.2 / 75.7 / 73.4% |
| Linear trend, Q1 2022 to Q2 2026 | minus 1.84 pp per year (s.e. 0.32) |
| AR(1) coefficient | 0.62 (s.e. 0.19) |
| Break in mean (best split) | Q2 2025: 77.5% before, 72.7% after (F = 18.7) |
| Long-term average (VCI) | 83% |
Source: VCI quarterly reports 2021 to 2026; statistics: author's calculations.
Three results stand out. First, the series has not touched the profitability threshold for four and a half years. The association described 74.8 percent in the third quarter of 2024 as below the profitability threshold, 72.5 percent in the fourth quarter of 2025 as a level at which plants still could not run profitably, and 75.1 percent in the first quarter of 2026 as still unprofitable. Its third-quarter 2025 report put the threshold at about 82 percent. Second, the trend is downward, at almost two percentage points a year, and statistically well determined despite the short sample. Third, the best single break in the mean falls in the second quarter of 2025, after which utilisation averaged only 72.7 percent. The 2025 deterioration therefore was not noise around a stable post-crisis level. It was a further downward step.
The autoregressive coefficient of 0.62 implies quick reversion around a sample mean of about 75 percent, but that mean itself keeps falling, and a Dickey-Fuller type statistic of minus 2.0 cannot reject persistence in eighteen observations.
A caution: some reports cover chemicals and pharmaceuticals, others chemicals alone, and some figures are revised (the second quarter of 2025 from 71.7 to 72.9 percent). We use the latest vintage where a revision is explicit. No conclusion depends on these details.
2.3 Basic versus specialty: a divergence that tells the story
Aggregate figures hide a sharp difference between segments. The association reports year-on-year changes for five chemical segments in each quarterly report. Chaining the second-quarter comparisons from 2022 to 2025, and the first-half 2026 change, yields the segment indices in Table 3.
Table 3. Production by segment, chained index (Q2 2022 = 100).
| Segment | Q2 2022 | Q2 2023 | Q2 2024 | Q2 2025 | H1 2026 | Change since Q2 2022 |
|---|---|---|---|---|---|---|
| Petrochemicals and derivatives | 100.0 | 80.9 | 94.6 | 86.8 | 82.0 | minus 18.0% |
| Polymers | 100.0 | 80.2 | 85.6 | 80.3 | 77.1 | minus 22.9% |
| Inorganic basic chemicals | 100.0 | 79.5 | 90.3 | 85.3 | 84.9 | minus 15.1% |
| Fine and specialty chemicals | 100.0 | 98.2 | 99.9 | 96.9 | 95.4 | minus 4.6% |
| Consumer chemicals | 100.0 | 87.1 | 91.3 | 90.0 | 86.9 | minus 13.1% |
| Chemicals excluding pharma | 100.0 | 85.8 | 93.0 | 88.3 | 85.6 | minus 14.4% |
Source: VCI quarterly reports (year-on-year changes); chaining: author's calculations.
The base period is itself already a crisis quarter: in the second quarter of 2022, chemical output excluding pharmaceuticals was 8.2 percent below the year before. Combining both, output in the first half of 2026 was about 21 percent below the second quarter of 2021, consistent with the annual figures. The pattern across segments is unambiguous. The three basic segments (petrochemicals, polymers and inorganic basics) lost between 15 and 23 percent on top of the 2022 shock, whereas specialty chemicals lost less than 5 percent. The basic segments also show the characteristic profile of a failed recovery: a rebound in 2024 that was fully reversed in 2025 and 2026.
2.4 A shrinking energy footprint
A further indicator of structural contraction is the industry’s energy use. According to Destatis data analysed by Argus Media, the chemical industry consumed 42.79 terawatt hours of electricity in 2024, more than 9 percent of Germany’s total consumption, but almost 20 percent less than in 2014 to 2018. Only 2023 was lower in a series going back to at least 2008. Total energy use by the chemical and pharmaceutical industry was 615 petajoules in 2023, about 170 terawatt hours, after falls of about 11 percent in 2022 and 10 percent in 2023. Natural gas accounted for 40.2 percent of energy use and electricity for 25.5 percent. In addition, the industry uses about 15 million tonnes of raw materials as feedstock, of which 10.6 million tonnes are mineral oil products (mainly naphtha) and 1.8 million tonnes natural gas.
The Chemistry4Climate initiative projects that a climate-neutral industry would need 160 to 440 terawatt hours of electricity by 2045. Electricity demand is instead falling because production is falling: an industry that is shrinking cannot electrify.
3. The 2026 Shock: Feedstock, Gas and a Deceptive Rebound
3.1 What changed in 2026
The 2022 crisis was primarily a gas crisis. The 2026 shock adds a feedstock shock in oil and naphtha to higher gas prices. Press reports in May 2026 put the rise in European naphtha prices at 64 percent after the collapse of an April ceasefire, and Cefic’s president warned that the effect on feedstock costs had gone beyond the 2022 gas crisis. The association reported that the import wave in basic chemicals had eased in the first quarter, plausibly because Gulf supply was disrupted.
Table 4. The 2026 shock in quarterly indicators.
| Indicator | Q4 2025 | Q1 2026 | Q2 2026 | Source |
|---|---|---|---|---|
| Brent, quarterly average | n.a. | above USD 80/bbl (plus 26.5% q/q) | n.a. | VCI |
| Naphtha, EUR per tonne | 450.75 | about 532 (plus 17.9% q/q) | n.a. | VCI |
| Production, chemicals and pharma, q/q | plus 0.9% | minus 2.8% | plus 2.4% | VCI |
| Production, chemicals only, q/q | minus 2.4% | plus 2.0% | n.a. | VCI |
| Producer prices, q/q | 0.0% | plus 0.2% | plus 5.7% | VCI |
| Sales, EUR billion | 51.8 | 50.9 | 54.5 | VCI |
| Capacity utilisation | 72.5% | 75.1% | 73.2% | VCI |
| ifo business climate, chemicals | n.a. | n.a. | minus 17.8 (June); plus 5.5 (Sept) | ifo |
Source: VCI quarterly reports 4/2025, 1/2026 and 2/2026; ifo Institute via press reports.
3.2 The hoarding paradox
At first sight, the second quarter of 2026 looks like a recovery. Production rose 2.4 percent on the quarter and 0.9 percent on the year, producer prices jumped 5.7 percent and sales rose 7.3 percent to EUR 54.5 billion, the first increase in more than three years. The ifo business climate for the chemical industry improved from minus 17.8 points in June to plus 5.5 in September, when expectations turned positive for the first time in a long while.
The industry does not read these numbers as a turnaround. The association spoke of geopolitical hoarding and a panic-driven temporary peak; the ifo analyst attributed the improvement to restocking and to outages in the Gulf region, adding that locational disadvantages remained and job cuts would continue.
The analytical point is that the shock is two-sided. It raises German costs but temporarily removes low-cost Gulf competitors. When that capacity returns, the cost increase will remain and the price support will disappear. The dangerous phase is therefore the normalisation of competitor supply while European costs stay elevated, which our third scenario models.
3.3 The cost arithmetic
How large is the cost shock in euros? We combine the physical input data from the association’s energy statistics with observed price changes. Table 5 shows the results.
Table 5. Annualised cost effects for the German chemical industry.
| Cost item | Physical base | Price change | Annual cost, EUR bn |
|---|---|---|---|
| Naphtha and other oil feedstock | 10.6 Mt | plus EUR 81/t (Q4 2025 to Q1 2026) | 0.86 |
| Naphtha and other oil feedstock | 10.6 Mt | plus 64% from Q4 2025 level | 3.06 |
| Natural gas, energy and feedstock | about 92 TWh | plus EUR 10/MWh | 0.92 |
| Gas cost gap with the United States | about 92 TWh | EUR 48.3 vs 14.0/MWh (H2 2025) | 3.16 |
| Power price relief 2026 to 2028 (upper bound) | 42.8 TWh | wholesale EUR 80/MWh, relief to EUR 50 floor | minus 1.28 |
Source: VCI energy statistics (2023 inputs), VCI quarterly reports, Cefic Chemical Trends Report, Argus Media; author's calculations. Gas volume: 68.7 TWh of energy use (40.2% of 615 PJ) plus 1.8 Mt of feedstock gas converted at 13.1 MWh per tonne.
Each EUR 100 per tonne on naphtha costs the industry about EUR 1.1 billion a year, and each EUR 10 per megawatt hour on gas about EUR 0.9 billion. The structural gap with the United States did not disappear after 2022: at EUR 48.3 against EUR 14.0 per megawatt hour in the second half of 2025, it is worth about EUR 3.2 billion a year, large relative to basic chemical sales of roughly EUR 90 billion.
3.4 The industrial power price: welcome but small
The federal government introduced a relief scheme for industrial electricity prices for 2026 to 2028. According to Argus Media, the relief covers 50 percent of the annual average wholesale price, with a floor of EUR 50 per megawatt hour below which the subsidised price cannot fall. Industrial power prices, as measured by the index of the industrial energy users’ association (VIK), stood at 309 in March 2026 (January 2002 = 100), against values consistently below 200 between 2009 and 2020.
Assuming all chemical electricity consumption were eligible, an upper bound, the relief is worth between EUR 0.86 billion a year at a wholesale price of EUR 70 per megawatt hour and EUR 2.1 billion at EUR 100; at EUR 80 about EUR 1.3 billion. That is about 40 percent of the gas cost gap with the United States and less than half of the spring 2026 naphtha shock. It helps chlor-alkali and industrial gases, not petrochemicals and ammonia, and it expires in 2028, far shorter than any plant’s investment horizon.
4. Closures: From Ludwigshafen to Böhlen
4.1 The European closure wave
The most systematic evidence on closures comes from the European Chemical Closures and Investments Radar, prepared by Roland Berger for Cefic and published in January 2026. It records public announcements between January 2022 and early December 2025. Table 6 summarises the annual figures, and Figure 4 compares closures with confirmed new investment.
Table 6. European chemical closures and investments, 2022 to 2025.
| Year | Capacity announced for closure (Mt) | Share of European capacity | Confirmed new capacity (Mt) | Net change (Mt) | Confirmed capex (EUR bn) |
|---|---|---|---|---|---|
| 2022 | 2.9 | 0.70% | 2.7 | minus 0.2 | 7.6 |
| 2023 | 8.7 | 2.09% | 1.7 | minus 7.0 | 2.4 |
| 2024 | 8.0 | 1.92% | 1.9 | minus 6.1 | 2.0 |
| 2025 (to early December) | 17.2 | 4.12% | 0.3 | minus 16.9 | 1.5 |
| Total | 37 | about 9% | about 7 | minus 30.2 | 13.5 |
Source: Cefic and Roland Berger, European Chemical Closures and Investments Radar 2022 to 2025 (January 2026); shares of the 417 Mt average 2017 to 2021 capacity base: author's calculations.
The radar shows a sixfold increase in annual closures between 2022 and 2025, while confirmed new capacity collapsed from 2.7 to 0.3 million tonnes. Petrochemicals account for 48 percent of closed capacity, basic inorganics for 32 percent and polymers for 15 percent; specialty chemicals for only 5 percent. Nine steam crackers are closing, a 16 percent net reduction in European cracking capacity. Energy costs were the main reason in 49 percent of announcements.
Germany accounts for 8.8 million tonnes of announced closures, a quarter of the European total, against 0.8 million tonnes of confirmed new capacity, and about 34,000 direct and indirect jobs at risk. For every tonne of new capacity announced in Germany, eleven tonnes were announced for closure.
Two other indicators confirm the picture. A Cefic survey, as reported by the European business press in May 2026, put the utilisation of European plants at the largest producers between 62 and 68 percent, well below the level of about 80 percent at which heavy chemical production is commonly considered profitable. And Europe’s share of world chemical trade has fallen from 23 percent in 2018 to 14 percent in the first quarter of 2026, while China’s has risen from 12 to 27 percent. Europe’s share of world chemical sales fell from 27 percent in 2004 to 13 percent in 2024, while China’s rose from 10 to 46 percent.
4.2 The German closure list
Table 7 lists the main German closures we could document from company and press sources. It is not exhaustive: the radar does not publish a plant-level list and many smaller closures are not individually reported.
Table 7. Selected German chemical closures and restructuring programmes.
| Company and site | Assets | Capacity or scale | Timing |
|---|---|---|---|
| BASF, Ludwigshafen | Ammonia, caprolactam, cyclohexanol/one, soda ash, TDI with DNT and TDA; less adipic acid | 10% of site asset value; about 700 jobs | Announced February 2023; savings by end-2026 |
| BASF, Ludwigshafen | Cyclododecanone, cyclopentanone, remaining adipic acid | About 180 jobs | 2025 |
| BASF, group | Cost programme, incl. EUR 1 bn at Ludwigshafen | EUR 2.3 bn target by 2026; EUR 1.7 bn reached end-2025 | 2023 to 2026 |
| BASF, Ludwigshafen | Workforce | About 2,800 jobs lost since early 2024; up to 20% core cost cut by 2029 | 2024 to 2029 |
| Dow, Böhlen | Ethylene cracker; butadiene extraction | 540 kt ethylene, 285 kt propylene, 105 kt butadiene a year | Q4 2027 |
| Dow, Schkopau | Chlor-alkali and vinyl assets | 250 kt chlorine (membrane), 740 kt EDC, about 110 kt VCM a year | Q4 2027 |
Source: ICIS (February 2023), Polyformnext/dpa (August 2024), Irish Times (February 2024), Chemical Processing (February 2026), Brussels Times (May 2026), Argus Media (July 2025).
The list illustrates the mechanism formalised in Section 6. BASF’s 2023 closures removed whole chains, from ammonia to its uses, from cyclohexanone to caprolactam, and from DNT and TDA to TDI. The Böhlen cracker supplies propylene to Braskem’s polypropylene plant at Schkopau, which will be affected; Dow’s own Schkopau polyethylene units survive only because they can draw ethylene by pipeline from the Stade import infrastructure. Downstream units survive when, and only when, imported molecules can replace domestic ones.
Meanwhile BASF has invested about EUR 8.7 billion in its Zhanjiang Verbund site in China, now ramping up: the clearest signal of where integrated base chemical capacity is expected to earn its return.
4.3 The investment drought
Closures are only half of the capacity dynamics. The other half is investment. The association reported in July 2026 that fixed-asset investment in Germany was about 15 percent below its 2023 level and declining for the third year in a row, and that almost half of its members were increasing investment abroad, in the United States, Asia and other European countries. Orders were still more than 20 percent below their 2021 level at the end of 2025, and every second company reported too few orders. The number of insolvency proceedings in the chemical industry rose from an average of about 30 a year before 2023 to 48 in 2023, 44 in 2024 and 45 in 2025, according to an EY-Parthenon analysis that also counts more than 13,000 jobs lost since 2022.
The investment drought turns closures into a one-way street: with a new-to-closed capacity ratio of 0.09 in Germany over four years and 0.02 in Europe in 2025, there is no replacement mechanism.
5. Estimating the Closure Threshold
5.1 Implied capacity and the spiral coefficient
Before estimating a closure model, it is useful to ask what the production and utilisation data imply for capacity. If utilisation is output divided by capacity, implied capacity is output divided by utilisation. Table 8 performs this calculation on annual averages, with 2021 utilisation proxied by the average of the two quarters available (84.1 percent).
Table 8. Implied capacity of German chemicals excluding pharmaceuticals.
| Year | Production index (2021 = 100) | Average utilisation, % | Implied capacity (2021 = 100) |
|---|---|---|---|
| 2021 | 100.0 | 84.1 | 100.0 |
| 2022 | 89.7 | 79.5 | 94.9 |
| 2023 | 78.8 | 77.2 | 85.9 |
| 2024 | 81.1 | 75.7 | 90.2 |
| 2025 | 78.5 | 73.4 | 89.9 |
| 2026 (H1 utilisation, forecast output) | 77.3 | 74.2 | 87.6 |
Source: VCI; author's calculations.
Implied capacity fell by 10.1 percent between 2021 and 2025. If the 8.8 million tonnes of German closures correspond to this fall, the German capacity base is about 87 million tonnes, roughly a fifth of Europe’s 417 million tonnes. We use this approximation to convert percentages into tonnes.
The more revealing statistic is the ratio of the decline in output to the decline in capacity. Between 2021 and 2025, the log decline in production was 2.3 times the log decline in implied capacity. We call this the spiral coefficient. A value above one means that demand for German output falls faster than capacity is removed, so that closures fail to raise utilisation. The industry is chasing a target that recedes as it moves.
5.2 A threshold hazard model of closures
Firms close plants when expected future margins no longer cover the cash cost of keeping them open, including the maintenance investment needed to keep them safe and compliant. Utilisation is the best available proxy for those margins: chemical plants have high fixed costs and earn their margin on the last tonnes produced. We therefore model the annual closure rate h, measured as the share of capacity announced for closure, as a threshold function of average utilisation u:
h = b x max(0, u* minus u)
where u* is the utilisation threshold below which closures begin and b the sensitivity of closures to the utilisation deficit. We estimate u* by grid search between 76 and 88 percent, and b by least squares given u*, using European closure shares from the radar and German annual average utilisation for 2022 to 2025. Using German utilisation as the driver of European closures is justified by Germany’s weight (a quarter of all closures) and by the high correlation of utilisation across European producers; it is nonetheless an approximation.
Table 9. Estimates of the threshold hazard model.
| Specification | Threshold u* | Slope b (% of capacity per pp) | Fit |
|---|---|---|---|
| Baseline, contemporaneous | 80.7% | 0.52 (s.e. 0.06) | R2 = 0.89 |
| One-year lag (3 observations) | 83.1% | 0.48 | R2 = 0.57 |
| Leave out 2023 | 77.6% | 0.97 | n.a. |
| Leave out 2025 | 82.0% | 0.34 | n.a. |
| Profile band (sum of squares within factor 2 of minimum) | 79.0 to 83.1% | n.a. | n.a. |
Source: Cefic and Roland Berger radar, VCI; author's calculations.
With only four annual observations this is a calibration rather than a test. Yet the threshold of 80.7 percent was not imposed: it emerges between the 80 percent cited in European commentary and the 82 percent named by the association. The model explains 89 percent of the variance in closures, including the doubling of 2025, and three of four leave-one-out thresholds lie between 80.2 and 82.0 percent.
At the second-quarter 2026 utilisation of 73.2 percent, the model predicts closures of about 3.9 percent of capacity a year, or about 3.4 million tonnes in Germany. Applied quarter by quarter since 2022, it implies cumulative closures of 10.7 percent of capacity, against 10.1 percent announced.
A linear fit of new investment on utilisation gives only 0.09 percent of capacity per point: closures respond strongly to low utilisation, investment weakly to high utilisation. This asymmetry is the core of the irreversibility problem.
6. The Verbund Multiplier and the Point of No Return
6.1 The arithmetic of consolidation
The standard view holds that closures restore an industry’s health: the least efficient capacity exits and the survivors run fuller. This rests on an implicit assumption, that the output of closed plants is taken over by remaining domestic plants.
Consider an industry with capacity K and output Q, so that utilisation is u = Q / K. A share h of capacity closes. The closing plants run below average, at utilisation u minus g, where g is the utilisation gap of the marginal plant. Let r be the share of their output recaptured by surviving domestic plants, and s the additional output lost by other domestic plants because the closure removes their feedstock supplier or their customer (the Verbund spillover). Define net recapture as v = r minus s. After the closure, utilisation becomes:
u’ = [u minus (1 minus v) x (u minus g) x h] / (1 minus h)
Closures raise utilisation if and only if (1 minus v) x (u minus g) is smaller than u, that is, if net recapture exceeds minus g / (u minus g). With a utilisation of 73.2 percent and a gap of 5 points for the marginal plant, the critical value of net recapture is minus 0.073. If spillovers exceed recapture by more than about 7 percent of the closed plants’ output, each closure lowers the utilisation of the rest of the industry.
Even above that critical value, closures help far less than intuition suggests. Table 10 shows how much capacity would have to close to bring utilisation from 73.2 percent to the estimated threshold of 80.7 percent, without any recovery in demand.
Table 10. Capacity that must close to restore 80.7% utilisation from 73.2%, with no demand recovery.
| Net recapture v | Gap 3 pp | Gap 5 pp | Gap 8 pp | Utilisation gain per 10% of capacity closed (gap 5 pp) |
|---|---|---|---|---|
| 1.0 (full recapture, idle capacity) | 9.3% | 9.3% | 9.3% | plus 8.1 pp |
| 0.5 (consolidation within firms) | 16.4% | 16.1% | 15.6% | plus 4.3 pp |
| 0.3 | 23.8% | 22.8% | 21.4% | plus 2.8 pp |
| 0.0 (exit; imports take the market) | 71.4% | 60.0% | 48.4% | plus 0.6 pp |
| minus 0.2 (Verbund spillover) | not attainable | not attainable | not attainable | minus 1.0 pp |
Source: author's calculations.
In the world of consolidation, a firm moves output from a closed line to a sister line and closing 16 percent of capacity restores the threshold. In the world of exit, customers turn to imports: closing 10 percent of capacity raises utilisation by little more than half a point, and restoring profitability would need shutting about 60 percent of the industry. In the world of Verbund spillover, closures lower the survivors’ utilisation.
The German closures in Table 7 belong overwhelmingly to the second and third worlds: they end products (TDI, caprolactam, ammonia chains at Ludwigshafen) and upstream keystone assets (Böhlen), consistent with energy cost rather than overcapacity being the dominant motive.
6.2 The irreversibility frontier
The arithmetic of consolidation is static. To capture the dynamics, we combine it with the closure hazard in a simple annual model. Capacity falls at the hazard rate estimated in Section 5. Demand for German output grows at an underlying rate that reflects market growth and competitiveness, and is reduced each year by the output lost through closures that is not recaptured. Output is the lesser of demand and 88 percent of capacity. A small amount of investment (0.25 percent of capacity a year, roughly the German rate of 2022 to 2025) occurs only when utilisation is above the threshold. All equations and parameters are listed in Annex A.
For each combination of starting utilisation and demand growth, we ask whether utilisation returns to within one point of the threshold within ten years. The boundary is the irreversibility frontier (Figure 6).
The results are striking in their simplicity. With underlying demand falling by more than a quarter of a percent a year, no starting point below the threshold returns to it within a decade, whatever the recapture. The industry then never stabilises: it shrinks continuously, because each round of closures removes output along with capacity and demand keeps falling. With positive demand growth, a return is possible from starting points above the frontier, and the frontier is lower the higher net recapture is. Table 11 reports, for Germany’s current utilisation of 73.2 percent, the minimum sustained demand growth required to return to the threshold within ten years, and the share of capacity that would be lost over ten years at zero growth.
Table 11. Growth needed to escape from 73.2% utilisation, and capacity lost at zero growth.
| Net recapture v | Minimum demand growth for return within 10 years, % a year | Capacity lost in 10 years at zero growth |
|---|---|---|
| 0.5 | minus 0.10 | 15.4% |
| 0.3 | plus 0.15 | n.a. |
| 0.1 | plus 0.50 | n.a. |
| 0.0 | plus 0.70 | 29.6% |
| minus 0.1 | plus 0.90 | n.a. |
| minus 0.3 | plus 1.40 | 44.3% |
Source: author's calculations; threshold 80.7%, hazard slope 0.52, closing-plant gap 5 pp.
The required growth rates must be compared with experience. Production of chemicals excluding pharmaceuticals fell by an average of 5.0 percent a year between 2021 and 2026, and by 0.7 percent a year between 2023 and 2026, a period in which gas prices were close to normal for most of the time. The association’s forecasts for 2026 and the ifo assessment of persistent locational disadvantages do not suggest a reversal. Unless underlying demand for German output turns positive and stays positive by at least half a percent to one and a half percent a year, the industry is on the wrong side of the frontier. In that precise sense, it has passed the point of no return.
A return would require a durable improvement in German cost competitiveness, above all in energy and feedstock, and a regulatory environment that makes investment rational again: a structural project of several years, not a cyclical event.
6.3 An Irreversibility Index
To track the approach to and crossing of the threshold in real time, we propose a simple composite indicator. The Irreversibility Index averages three components, each scaled from 0 to 1 and the total multiplied by 100:
- the depth of the utilisation deficit, measured as the distance below 82 percent, scaled by 12 points;
- its duration, measured as the number of consecutive quarters below 82 percent, scaled by 20 quarters;
- the investment gap, measured as one minus the ratio of confirmed new capacity to announced closures in Europe in the same year.
The index is zero when utilisation is at or above the threshold and investment matches closures. It reaches 100 when utilisation is twelve points below the threshold for five years and nothing is being built. We propose, as a matter of judgement rather than estimation, to read values below 33 as a cyclical downturn, values between 33 and 66 as a zone of risk and values above 66 as the irreversible zone, consistent with the frontier analysis above.
Table 12. The Irreversibility Index for German chemicals.
| Quarter | Utilisation, % | Deficit, pp | Quarters below 82% | Cumulative modelled closures, % of capacity | Irreversibility Index |
|---|---|---|---|---|---|
| Q4 2022 | 76.5 | 5.5 | 4 | 0.7 | 24.2 |
| Q4 2023 | 77.2 | 4.8 | 8 | 2.5 | 53.5 |
| Q4 2024 | 74.7 | 7.3 | 12 | 5.2 | 65.7 |
| Q2 2025 | 72.9 | 9.1 | 14 | 6.5 | 81.4 |
| Q3 2025 | 70.0 | 12.0 | 15 | 7.9 | 91.1 |
| Q4 2025 | 72.5 | 9.5 | 16 | 9.0 | 85.8 |
| Q2 2026 | 73.2 | 8.8 | 18 | 10.7 | 87.2 |
Source: VCI, Cefic and Roland Berger radar; author's calculations.
The index entered the zone of risk in early 2023, when the gas price normalisation failed to bring output back, and the irreversible zone in the second quarter of 2025, when utilisation broke to a lower level and investment collapsed. The 2026 shock hit an industry already beyond the threshold.
6.4 Why the threshold is a point of no return
Three mechanisms make the threshold a point of no return. First, the Verbund: integrated units share feedstock, by-products and site costs, so the loss of a cracker or chlor-alkali plant forces the rest to import inputs or carry higher fixed costs, turning net recapture negative.
Second, the option value of waiting declines with time: the longer low utilisation lasts, the more firms treat it as permanent, which is why closures doubled in 2025, the fourth year of low utilisation. Third, capabilities are lost: skilled teams disperse, customers requalify foreign suppliers and restarting requires a greenfield-type decision at German costs.
7. Scenarios to 2035: Capacity, Output and the Trade Balance
7.1 Three scenarios
We use the dynamic model of Section 6 to project German chemical capacity and output from 2026 to 2035 under three scenarios. All start from the 2026 position: production at 77.3 (2021 = 100), utilisation at 74.2 percent (the first-half average) and implied capacity at 87.6. All use the estimated threshold of 80.7 percent and hazard slope of 0.52, a closing-plant gap of 5 points and the 2022 to 2025 investment rate when utilisation is above the threshold. They differ in the path of underlying demand growth, which summarises market growth and the effect of energy and feedstock costs on competitiveness, and in net recapture.
Table 13. Scenario assumptions.
| Scenario | Narrative | Underlying demand growth | Net recapture v |
|---|---|---|---|
| S1 Normalisation | Gulf supply and LNG normalise in 2027; reforms improve competitiveness | plus 1.0% a year from 2027 | 0.3 |
| S2 Persistent premium | Costs stay elevated through 2028; no structural reform | minus 1.0% (2027), minus 0.5% (2028), 0 thereafter | 0.1 |
| S3 Second round | Gulf exports return cheaply while European costs stay high; Verbund cascades | minus 3.0% (2027), minus 2.0% (2028), minus 0.5% thereafter | minus 0.1 |
Source: author's assumptions.
7.2 Results for capacity and output
Table 14. Projected capacity, production (2021 = 100) and utilisation of German chemicals excluding pharmaceuticals.
| Scenario | Capacity 2030 | Capacity 2035 | Production 2030 | Production 2035 | Utilisation 2030 | Utilisation 2035 | Capacity lost 2026 to 2035 |
|---|---|---|---|---|---|---|---|
| S1 Normalisation | 80.2 | 80.3 | 75.9 | 79.5 | 79.7% | 83.3% | 8.4% |
| S2 Persistent premium | 76.1 | 66.3 | 67.6 | 60.2 | 74.7% | 76.4% | 24.4% |
| S3 Second round | 72.0 | 51.7 | 59.3 | 41.2 | 69.3% | 67.0% | 41.0% |
Source: author's calculations.
Even in the benign scenario, Germany loses a further 8.4 percent of capacity, about 7 million tonnes, before utilisation returns to the threshold around 2032, and output in 2035 is still a fifth below 2021. In the persistent premium scenario, closures run at 3 to 4 percent of capacity a year and output falls to 60 percent of its 2021 level. In the second-round scenario the industry spirals: by 2035 capacity and output are about 40 and 47 percent below 2026.
In the persistent premium scenario, capacity lost by 2035 ranges from 15.5 percent (hazard slope 0.35, net recapture 0.3) to 40.3 percent (slope 0.7, recapture minus 0.1); in all nine combinations it exceeds 15 percent.
7.3 When does Germany become a net importer?
The trade balance is where the thesis of this paper becomes concrete. Using UN Comtrade data for 2025, Table 15 shows Germany’s exports and imports in the four main chemical groups of the harmonised system.
Table 15. German trade in chemical product groups, 2025, and critical output losses.
| Product group | Exports 2025, USD bn | Imports 2025, USD bn | Balance, USD bn | Output loss that eliminates the surplus |
|---|---|---|---|---|
| Organic chemicals (HS 29) | 27.5 | 40.3 | minus 12.8 | already in deficit |
| Plastics (HS 39) | 68.3 | 51.0 | plus 17.3 | 17.0% |
| Miscellaneous chemical products (HS 38) | 35.3 | 23.3 | plus 12.0 | 22.7% |
| Core chemicals (HS 29, 38 and 39) | 131.1 | 114.6 | plus 16.5 | 8.4% |
| Pharmaceutical products (HS 30) | 134.1 | 85.4 | plus 48.7 | 24.3% |
Source: UN Comtrade via Trading Economics; critical losses: author's calculations with export share 0.62 and downstream absorption 0.2.
The first finding requires no model: Germany is already a net importer of organic chemicals, by USD 12.8 billion, partly reflecting imports of pharmaceutical intermediates and vitamins (China supplied USD 3.8 billion), partly the closures of the basic organic chain.
To project the balance, we use an accounting model: lost output disappears from exports in proportion to the export share (0.62, the 2025 foreign sales share), and the domestically sold part is replaced by imports except for a downstream absorption share (0.2). The critical loss is the surplus divided by this product. For other chemical products, closer to specialty, losses are scaled by the specialty-to-basic closure ratio of the radar (0.38).
Table 16. Projected trade balances and crossing years.
| Scenario | Core chemicals: balance 2030, USD bn | Core chemicals: balance 2035, USD bn | Core chemicals turn negative | Plastics turn negative | Other chemical products turn negative |
|---|---|---|---|---|---|
| S1 Normalisation | plus 13.6 | plus 21.1 | not by 2035 | not by 2035 | not by 2035 |
| S2 Persistent premium | minus 3.9 | minus 19.6 | 2030 | 2033 | not by 2035 |
| S3 Second round | minus 21.5 | minus 59.7 | 2028 | 2029 | not by 2035 |
Source: author's calculations at 2025 prices and exchange rates.
Under the persistent premium scenario, the combined balance of core chemicals turns negative in 2030 and plastics in 2033. Under the second-round scenario, the core balance turns negative in 2028 and plastics in 2029. Only in the normalisation scenario do the surpluses survive. Other chemical products remain in surplus in all three scenarios, which reflects the greater resilience of specialty chemicals. The crossing years are robust to the trade parameters: across export shares from 0.55 to 0.70 and downstream absorption from 0.1 to 0.4, the core chemical balance turns negative in 2029 or 2030 under S2 and in 2028 under S3, and never under S1.
These projections are at 2025 prices and ignore relative price and foreign demand changes; they indicate orders of magnitude and timing, not forecasts of a trade statistic.
7.4 A Monte Carlo view
The scenarios fix parameters that are uncertain. To give a probabilistic view, we ran 20,000 simulations in which the key parameters are drawn independently: the threshold uniformly between 80 and 82 percent, the hazard slope from a normal distribution centred on the estimate with its standard error, the closing-plant gap between 2 and 8 points, net recapture between minus 0.3 and 0.5, short-run demand growth for 2027 between minus 3 and plus 1 percent, long-run growth between minus 1.5 and plus 1.0 percent, the export share between 0.55 and 0.70 and downstream absorption between 0.1 and 0.4.
Table 17. Monte Carlo results, 20,000 draws.
| Indicator | P10 | Median | P90 | Probability |
|---|---|---|---|---|
| Capacity lost 2026 to 2035 | 15.5% | 26.7% | 41.5% | n.a. |
| Production lost 2026 to 2035 | 8.3% | 25.7% | 48.3% | n.a. |
| Utilisation below 80% in 2035 | n.a. | n.a. | n.a. | 85% |
| Core chemicals balance negative by 2030 | n.a. | 2029 | n.a. | 70% |
| Core chemicals balance negative by 2035 | n.a. | n.a. | n.a. | 87% |
| Plastics balance negative by 2030 | n.a. | 2032 | n.a. | 27% |
| Plastics balance negative by 2035 | n.a. | n.a. | n.a. | 71% |
Source: author's calculations.
In about 70 percent of simulations the core chemical balance turns negative between 2028 and 2030; in 13 percent it stays positive to 2035, under positive demand growth and high recapture. The median simulation loses a quarter of capacity, and in 85 percent utilisation is still below 80 percent in 2035. Germany becoming a net importer of basic chemicals is the central case, not a tail risk.
8. What Is at Stake: Output, Value Added and Jobs
8.1 Direct effects
To translate capacity losses into economic magnitudes, we use three conversion factors. First, the German capacity base of about 87 million tonnes implied by Section 5. Second, the job content of closures in the radar: about 34,000 direct and indirect jobs for 8.8 million tonnes of German closures, or about 3,900 jobs per million tonnes, of which about 700 direct (the radar uses an indirect multiplier of 4.5). Third, the turnover of basic chemicals, which we approximate at about EUR 90 billion, or 41 percent of the industry’s EUR 220 billion sales, using the production weights of the three basic segments in the association’s reports as a proxy for their value shares. For value added we assume a range of 20 to 30 percent of turnover. These last two factors are assumptions, flagged as such.
Table 18. Capacity, jobs, turnover and value added at risk by 2035.
| Scenario | Capacity lost by 2035, Mt | Direct jobs | Direct and indirect jobs | Basic chemical turnover lost, EUR bn a year | Value added lost, EUR bn a year |
|---|---|---|---|---|---|
| S1 Normalisation | 7.3 | 5,100 | 28,300 | output gain of 2.5 | gain of 0.5 to 0.8 |
| S2 Persistent premium | 21.2 | 14,900 | 82,100 | 19.9 | 4.0 to 6.0 |
| S3 Second round | 35.7 | 25,100 | 137,900 | 42.1 | 8.4 to 12.6 |
Source: author's calculations using the Cefic and Roland Berger job ratios, VCI sales and production weights; value added share assumed at 20 to 30%.
Even in the normalisation scenario, about 7 million tonnes close, with some 28,000 direct and indirect jobs. In the adverse scenarios, 80,000 to 140,000 jobs and EUR 4 to 13 billion of value added a year are at stake, against chemical and pharmaceutical employment of 471,500 and a car industry that lost 42,300 jobs in a single year.
8.2 Downstream: pharmaceuticals, plastics and automobiles
The radar’s indirect jobs do not capture the main downstream risk: customer industries losing competitiveness or supply security when production nearby goes.
Pharmaceuticals are most exposed in supply security: the USD 12.8 billion organic chemicals deficit already reflects heavy reliance on imported intermediates, much of it from China, and pharmaceutical output fell 1.5 percent in the first half of 2026.
Plastics processors and the car industry are exposed in cost and proximity: polymer output has fallen by more than a fifth since mid-2022, and falling car production and falling chemical production reinforce each other.
In the model, these effects enter through net recapture: shrinking downstream demand lowers recapture and accelerates the spiral.
9. Policy Implications
The analysis has five implications for policy.
First, stop treating the problem as cyclical. Relief that expires within two or three years does not change closure decisions for assets with lifetimes of decades. Policy should be designed on a ten-year horizon.
Second, identify and prioritise keystone assets. Closing assets that feed other domestic units turns net recapture negative and makes contraction self-reinforcing. Germany should map its value chains, identify the crackers, chlor-alkali, ammonia and key intermediate units whose loss would cascade, and concentrate any support there rather than on everything.
Third, attack the energy and feedstock gap structurally. The gas cost gap with the United States (about EUR 3.2 billion a year) is more than twice the maximum value of the power relief. Faster grid connections (now five to ten years), lower network charges, predictable compensation for indirect carbon costs, long-term low-carbon power contracts and LNG diversification away from chokepoints do more per euro than subsidies; as an economist quoted by Argus Media put it, the government cannot subsidise permanently.
Fourth, accept imports of commodity molecules, but from reliable partners. Ethane-based polymers from the United States or basic chemicals from European neighbours fit a pro-market, pro-Western strategy; growing dependence on China for intermediates of pharmaceutical and defence relevance does not. Trade defence and the carbon border adjustment should be used where evidence supports it, within WTO rules.
Fifth, monitor the threshold. The Irreversibility Index and the dashboard in Annex C can be updated quarterly. If utilisation returns above 80 percent and the investment-to-closure ratio exceeds one half, this diagnosis was too pessimistic; if utilisation stays in the low seventies after Gulf supply normalises, the second-round scenario should become the base case.
10. Conclusion
The German chemical industry has spent four and a half years below the utilisation rate at which it earns its cost of capital, with output a fifth below 2021 and a quarter of Europe’s closures.
This paper has argued that the industry has crossed a threshold. The estimated closure threshold lies at about 81 percent utilisation; the industry has been below it since 2022. Because German closures are mostly product exits rather than consolidation, and because the integration of Verbund sites transmits closures from one unit to the next, closures have not restored the utilisation of the survivors. Demand for German output has fallen more than twice as fast as capacity. On current demand trends, no starting point below the threshold leads back to it within a decade. The Irreversibility Index moved into its highest zone in 2025, before the 2026 shock, and stands at 87.
The 2026 shock raised costs by several billion euros a year, and its temporary benefit will vanish when Gulf supply returns. Germany already imports more organic chemicals than it exports; the broader core chemical balance is likely to turn negative around 2029.
None of this implies that decline is inevitable or that it should be resisted at any cost. A smaller, more specialised, more electrified chemical industry, importing some commodity molecules from reliable partners, is a perfectly viable outcome for a high-wage, energy-importing economy. The danger is not contraction as such. It is an unmanaged contraction in which the loss of keystone assets drags down the specialty chemicals, pharmaceuticals and manufacturing customers that Germany does want to keep. The point of no return for commodity chemicals may have been passed. The point of no return for the value chains that depend on them has not, and that is where policy should now concentrate.
Annex A. Methodology
Production index. Annual changes in production of chemicals excluding pharmaceuticals from the VCI World Chemistry Report (2021 to 2024) and VCI quarterly reports (2025), chained with 2021 = 100. The 2026 value applies the VCI forecast of minus 1.5 percent. The 2018 value is implied by the VCI statement that chemical output in 2024 was 17 percent below 2018.
Utilisation. Quarterly capacity utilisation from VCI quarterly reports, latest vintage where an explicit revision is reported. Annual averages are simple means of available quarters. Trend: OLS on a time trend, Q1 2022 to Q2 2026 (18 observations). AR(1): OLS of utilisation on its lag. Break: single break in mean chosen by minimum sum of squared residuals with at least four observations per segment; the F statistic compares the break model with a constant mean and is indicative only, since the break date is estimated.
Closure hazard. h = b x max(0, u* minus u), with h the annual European closure announcements as a share of the 417 Mt capacity base, and u the German annual average utilisation, 2022 to 2025. u* by grid search (step 0.05 points) between 76 and 88 percent; b by least squares through the origin given u*. Standard error of b conditional on u*.
Consolidation arithmetic. u’ = [u minus (1 minus v)(u minus g) h] / (1 minus h), where v is net recapture (recapture minus spillover) and g the utilisation gap of the closing plants. The capacity share h that restores u* is (u* minus u) / (u* minus (1 minus v)(u minus g)).
Dynamic model. Annual. h(t) = b x max(0, u* minus u(t)); K(t+1) = K(t) x (1 minus h(t)) plus 0.0025 x K(t) if u(t) is above u*; D(t+1) = D(t) x (1 + g(t+1)) minus (1 minus v) x (u(t) minus 5) x h(t) x K(t); Q = min(D, 0.88 K); u = Q / K. Irreversibility frontier: lowest starting utilisation from which u returns to u* minus 1 within ten years, for constant g, by grid search.
Irreversibility Index. 100 x (D/12 + N/20 + (1 minus I/C)) / 3, with D the deficit below 82 percent (capped at 12), N the consecutive quarters below 82 percent (capped at 20) and I/C the European ratio of confirmed new capacity to closures in the year (2025 value used for 2026; 1 before 2022).
Trade model. For product group j with exports X and imports M, output value V = X / phi, with phi = 0.62 (2025 foreign sales of EUR 136 bn divided by total sales of EUR 220 bn). A fractional output loss L reduces the balance by V x L x (phi + (1 minus phi)(1 minus delta)), with downstream absorption delta = 0.2. Critical loss L* = (X minus M) / (V (phi + (1 minus phi)(1 minus delta))). For HS 38 the output loss is scaled by 0.38, the ratio of specialty to basic capacity closures in the radar; the core chemical aggregate uses the corresponding weighted average. Losses are measured from 2026 while balances are 2025 values, which slightly delays the projected crossings.
Monte Carlo. 20,000 draws, fixed seed. u* uniform (80, 82); b normal (0.52, 0.06), floored at 0.15; g uniform (2, 8); v uniform (minus 0.3, 0.5); 2027 demand growth uniform (minus 3, 1); long-run growth uniform (minus 1.5, 1.0), with 2028 the average of the two; phi uniform (0.55, 0.70); delta uniform (0.1, 0.4).
Limitations. The closure model rests on four annual observations and uses German utilisation to explain European closures; utilisation coverage varies between reports; announced closures may not be executed on time; the trade model ignores price effects and re-exports; some Cefic figures come from press reports. These affect the precision of dates and magnitudes, not the direction of the results.
Annex B. Data Tables
Table B1. Data used in this paper.
| Indicator | Value | Date or period | Source |
|---|---|---|---|
| Production, chemicals excl. pharma | minus 21.5% vs 2021 | 2025 | VCI; author’s index |
| Production forecast | minus 1.5% | 2026 | VCI, July 2026 |
| Capacity utilisation | 75.1% (Q1), 73.2% (Q2) | 2026 | VCI |
| Profitability threshold | about 82% | 2025 | VCI quarterly report 3/2025 |
| Sales | EUR 220 bn (EUR 136 bn abroad) | 2025 | VCI |
| Employment | 471,500 | H1 2026 | VCI |
| Electricity use | 42.79 TWh, about minus 20% vs 2014 to 2018 | 2024 | Destatis via Argus Media |
| Energy use | 615 PJ; gas 40.2%, power 25.5% | 2023 | VCI energy statistics |
| Feedstock | 15 Mt, of which 10.6 Mt oil products, 1.8 Mt gas | 2023 | VCI energy statistics |
| Naphtha | EUR 450.75/t (Q4 2025); about EUR 532/t (Q1 2026) | 2025 to 2026 | VCI |
| Industrial gas price, EU vs US | EUR 48.3 vs 14.0 per MWh | H2 2025 | Cefic |
| Power price relief | 50% of wholesale average, floor EUR 50/MWh | 2026 to 2028 | Argus Media |
| European closures | 37 Mt (9% of capacity), Germany 8.8 Mt | 2022 to 2025 | Cefic and Roland Berger |
| Utilisation at major European producers | 62 to 68% | May 2026 | Cefic survey via press |
| Europe’s share of world chemical trade | 23% (2018), 14% (Q1 2026) | n.a. | Cefic via press |
| Fixed-asset investment | about 15% below 2023 | 2026 | VCI |
Annex C. A Quarterly Dashboard for the Point of No Return
Table C1. Dashboard for the irreversibility threshold. Thresholds: author's proposal.
| Indicator | Publication | Signal of recovery | Signal of irreversibility | Current reading |
|---|---|---|---|---|
| Capacity utilisation, chemicals | Quarterly, VCI | Above 80% for two quarters | Below 75% for two quarters | 73.2% (Q2 2026): irreversibility |
| Irreversibility Index (this paper) | Quarterly | Below 50 | Above 66 | 87.2: irreversibility |
| Basic segment production, y/y | Quarterly, VCI | Positive for four quarters | Negative after Gulf normalisation | Petrochemicals minus 5.5%, polymers minus 4.0% (H1 2026): irreversibility |
| Investment to closure ratio, Europe | Annual, Cefic radar | Above 0.5 | Below 0.2 | 0.02 (2025): irreversibility |
| Core chemical trade balance (HS 29, 38, 39) | Monthly, Destatis / Comtrade | Rising | Below USD 10 bn | USD 16.5 bn (2025): watch |
| EU to US industrial gas price ratio | Semi-annual, Cefic | Below 2 | Above 3 | 3.4 (H2 2025): irreversibility |
Five of the six indicators point to irreversibility, and the sixth, the trade balance, is the one that responds last. Global Strategy Files will update this dashboard as new quarterly data become available.
References
Argus Media (2025). Dow to close German cracker, other assets by 4Q 2027. 7 July 2025.
Argus Media (2026). German chemical industry power demand could resume fall. 30 April 2026.
BASF (2026). Results for 2025 and outlook 2026, as reported by Chemical Processing, 27 February 2026; cost programme as reported by Brussels Times and dpa, 20 and 21 May 2026.
Börsen-Zeitung (2026). Chemie- und Pharmabranche spürt leichten Aufwind. 3 September 2026.
Cefic and Roland Berger (2026). European Chemical Closures and Investments Radar 2022 to 2025. Brussels, January 2026.
Cefic (2026). Chemical Trends Report, April 2026 edition.
CHEManager (2026). Europe’s chemical sector under stress. 5 February 2026; Chemiegeschäft bleibt weiter unter Druck, 3 September 2026.
chemie.de (2025). Das Jahr 2024 endete für die chemisch-pharmazeutische Industrie mit einem empfindlichen Rückschlag; (2026) Geschäftsklima in der Chemie deutlich verbessert, 3 July 2026.
Der Aktionär / dpa-AFX (2025). Chemie funkt SOS: Anlagen historisch niedrig ausgelastet. 10 December 2025.
Destatis (2026). Produktion im Juli 2026 (press release 316) and Produktion im August 2026, 7 October 2026.
Deutsches Ärzteblatt (2026). Chemisch-pharmazeutische Industrie: Prognose für 2026 gesenkt. 16 July 2026.
European Business Magazine (2026). How the Iran war just broke European chemical industry. 15 May 2026.
European Rubber Journal (2026). German chemical output falls under constant stress. 1 June 2026.
EY-Parthenon (2026). Analysis of the German chemical industry, as reported by chemeurope.com, 14 April 2026.
ICIS (2023). BASF to close several major assets at Ludwigshafen on high production costs. 24 February 2023.
Indian Chemical News (2026). German chemical industry in the red as war risks and weak economy deepen pressure, 16 March 2026; German chemical industry warns of deeper slump as costs surge, 31 May 2026.
intellinews (2026). Europe’s petrochemical collapse: plants close as investment heads for China. 14 August 2026.
Irish Times (2024). BASF extends cost cuts by EUR 1bn as demand for chemicals falls. 23 February 2024.
Polyformnext / dpa (2024). BASF closes further production plants at Ludwigshafen. 30 August 2024.
SCI (2026). Chemical plant closures surge in Europe, investment drops. 28 January 2026.
Trading Economics (2026). Germany exports and imports by category, 2025 (UN Comtrade data).
VCI (2021 to 2026). Quartalsberichte zur Lage der chemisch-pharmazeutischen Industrie: 3/2021, 1/2022, 2/2022, 4/2022, 2/2023, 4/2023, 2/2024, 3/2024, 1/2025, 2/2025, 3/2025, 4/2025, 1/2026; press release on Quarterly Report 2/2026, 3 September 2026.
VCI (2024). Stocktaking 2024 and World Chemistry Report, November 2024 edition.
VCI (2025). Annual press conference, 10 December 2025.
VCI (2026). Key business figures, first half of 2026; Energiestatistik.
wallstreet-online / dpa-AFX (2026). Ifo: Geschäftsklima in Chemiebranche hellt sich deutlich auf. 2 October 2026.
Driss EFFINA, PhD
Independent Researcher, Global Strategy Files
The opening pages of our four country reports: United States, United Kingdom, Japan and China. A free PDF, sent to your inbox.
You will also receive occasional analysis from Global Strategy Files. Unsubscribe in one click, at any time.