Chemicals · Commodity Chemicals
Mid-Cycle EBITDA for Commodity Chemicals
Why last year's EBITDA misleads at a cycle low, how Dow, Olin, Methanex and LyondellBasell frame mid-cycle earnings, and how to build it when none is filed.
Selborne Research · Commodity Chemicals coverage: 7 guides, 6 company profiles, a primer and Excel model
Educational analysis for professional use. This guide is not investment advice or a recommendation to buy or sell any security, and it is not personalised.
On this page
Last Year’s EBITDA Carries the Cycle Into Every Year
A commodity chemical producer earns a spread, its selling price less its feedstock cost, on every tonne it sells. In a downturn the spread narrows and plants run below capacity, so earnings fall on both counts. Value the business on last-twelve-months (LTM) EBITDA and you assume that bad year repeats forever.
Westlake shows how far off that gets. FY2025 EBITDA excluding identified items was $1,144M, a 10% margin against 19% the year before. GAAP EBITDA was a ($248)M loss after $1,392M of identified items, mainly write-downs and closures. Even the adjusted line bakes in downturn spreads. The reported vs adjusted EBITDA guide takes those items apart line by line.
The fix is mid-cycle EBITDA: what the business earns at long-run average spreads and normal plant loading (utilisation).
FY2025 Against Each Company’s Own Framework
Dow, LyondellBasell, Olin and Methanex each publish a yardstick for judging a downturn year; Westlake and Celanese do not.
LyondellBasell’s Value Enhancement Program is a set of plant reliability, cost and commercial improvements, measured by the recurring EBITDA they add. Chlor-alkali is salt electrolysis into chlorine and caustic soda; vinyls are the PVC chain the chlorine feeds. Methanex’s realised price is its average selling price after discounts.
| Company | FY2025 EBITDA (company basis) | Company’s own framework | What it shows |
|---|---|---|---|
| Dow | $3,256M Operating EBITDA | Trough <$6B, peak >$12B, mid-cycle marked without a value (May 2024) | Below Dow’s own trough line |
| LyondellBasell | $2,543M ex. identified items | Value Enhancement Program: $1.1B at 2017-19 margins, $1.5B by 2028 | Covers one improvement programme only |
| Olin | $651.8M adjusted | ~$2.0B in 2029 at mid-cycle; chlor-alkali and vinyls ~$1.5B (Dec 2024) | The only company-wide dollar target |
| Methanex | $808M adjusted at $361/t realised | $650M / $1,075M / $1,450M at $300 / $350 / $400/t (Mar 2026) | Earnings at three prices; no target |
| Westlake | $1,144M ex. identified items | None filed | Reported (GAAP) EBITDA was negative |
| Celanese | $1,893M operating (19.8% margin) | None filed | Management cites below-normal demand |
Methanex shows why price alone does not rebuild earnings. In FY2025 it realised more than the $350/t row of its own table, yet earned about $267M less than that row. The gap is mostly volume: the table assumes 9.3 Mt (million tonnes), and Methanex made 7.8 Mt.
Building Mid-Cycle EBITDA When No Target Is Filed
For Dow, LyondellBasell, Westlake and Celanese there is no company-wide dollar figure, so you build one: capacity times mid-cycle utilisation times mid-cycle margin per tonne, at the long-run prices listed on the spreads guide. The Excel model that comes with the primer does this for fictional producers; two show the arithmetic.
Fictional polyethylene (PE) producer on cheap US ethane feedstock:
| Input | Value |
|---|---|
| PE capacity | 5.0 Mt/yr |
| Mid-cycle utilisation | 88% |
| Margin from feedstock to finished plastic (long-run assumption) | $650/t |
| Less plant running costs and overhead | $200/t |
| Margin after running costs and overhead | $450/t |
| Mid-cycle EBITDA (5.0 Mt × 88% × $450/t) | $1,980M |
| Net debt | $5,625M (2.8x mid-cycle EBITDA) |
Fictional methanol producer:
| Input | Value |
|---|---|
| Capacity | 3.0 Mt/yr |
| Mid-cycle utilisation | 90% |
| Realised price (long-run assumption) | $380/t |
| Less all-in cash cost, gas included | $250/t |
| Margin after cash cost | $130/t |
| Mid-cycle EBITDA (3.0 Mt × 90% × $130/t) | $351M |
| Net debt | $900M (2.6x mid-cycle EBITDA) |
The $650/t margin lies between the North American PE margin of $580/t in 2025 and its 2013-22 average of $830/t (Chemical Market Analytics data in LyondellBasell’s earnings slides). The $380/t methanol price lies between Methanex’s FY2025 realised price and the $400/t it calls mid-cycle.
Apply the Multiple to Mid-Cycle Earnings
On LTM earnings a stock looks expensive at a trough and cheap at a peak, because its EBITDA swings with the cycle. Mid-cycle EBITDA does not.
| Benchmark | EV/EBITDA | Notes |
|---|---|---|
| Range the model’s sensitivity grid tests | 6-9x | On mid-cycle EBITDA |
| Model’s single point | 7.5x | Illustrative, inside that range |
| Damodaran, basic chemicals (US, Jan 2026) | 8.6x | Industry aggregate, firms with positive EBITDA; trailing 12 months to Q3 2025, not mid-cycle |
| Damodaran, specialty chemicals (US, Jan 2026) | 13.4x | Same basis; not comparable with commodity producers |
At 7.5x, the fictional PE producer’s $1,980M of mid-cycle EBITDA gives an enterprise value of $14,850M. Less $5,625M of net debt, equity is $9,225M. The model’s DCF puts the same producer’s enterprise value at $11,820M. About $940M of the $3,030M gap is the lean years back to mid-cycle, which the multiple skips by assuming mid-cycle earnings arrive at once. The rest is the 7.5x paying more for mid-cycle cash flow than the model’s 9% discount rate allows.
Leverage: Divide by Mid-Cycle EBITDA
Olin alone files both a downturn-year figure and a mid-cycle target, so it shows the swing: net debt of $2,659.7M is 4.1x its FY2025 adjusted EBITDA and about 1.3x its own 2029 target. For the others, each company profile gives the ratio on the downturn year.
The model labels its fictional producers’ net debt to mid-cycle EBITDA below 2.0x conservative, 2.0-4.0x normal and above 4.0x stretched. Both fictional producers sit in the normal band.
Commodity Chemicals Sector Primer
Three hypothetical producers, in polyethylene, methanol and PVC, valued year by year as spreads recover from the trough, set beside the through-cycle multiple screen.
- 15 sections, from how a spread business earns to a year-by-year reversion DCF and leverage on mid-cycle EBITDA
- 42 pages
- an ethane-advantaged polyethylene producer, a methanol producer and a chlor-alkali and PVC producer
- 3 producer engines
- listed producers across the ethylene, vinyl, methanol and acetyls chains on filed FY2025 figures
- 6-company screen
The Excel model is the primer's cycle-normalisation valuation live across 12 sheets: three producer tabs (ethane-advantaged polyethylene, methanol, and chlor-alkali and PVC), each reverting its spread and utilisation from trough to mid-cycle through a ten-year free-cash-flow schedule with working capital and sustaining capex; a valuation summary setting each DCF beside the through-cycle multiple screen; cycle scenarios, a cost-curve position, a leverage screen on mid-cycle EBITDA and two live sensitivity grids. Change the spread, the WACC or the exit multiple and the value moves.
See what's in the Commodity Chemicals Sector Primer →£25 PDF · £59 with the Excel model · £159 for all three Chemicals industries
Frequently Asked Questions
- Why does LTM EBITDA fail for commodity chemical valuation?
- A commodity producer earns a spread, its selling price less its feedstock cost, and spreads swing through the cycle, so last-twelve-months (LTM) EBITDA carries last year's place in the cycle into every future year. Olin earned $651.8M of adjusted EBITDA in FY2025 against its own target of about $2.0B for 2029 in mid-cycle conditions.
- What mid-cycle EBITDA frameworks do companies disclose?
- Olin targets about $2.0B of adjusted EBITDA in 2029 at mid-cycle conditions (Investor Day, December 2024). Dow's May 2024 Investor Day drew a trough line below $6B and a peak line above $12B, with mid-cycle marked between them and no value given; FY2025 Operating EBITDA was $3,256M. Methanex publishes estimated Adjusted EBITDA of $650M, $1,075M and $1,450M for a full year at methanol prices of $300, $350 and $400/t (March 2026) and calls $400/t mid-cycle. LyondellBasell measures one improvement programme, $1.1B of recurring EBITDA, at 2017-19 mid-cycle margins.
- What EV/EBITDA multiple is normal for commodity chemicals?
- The Excel model that comes with the Selborne commodity chemicals primer tests multiples of 6-9x on mid-cycle EBITDA, with 7.5x as its single illustrative point. Damodaran's January 2026 US data put basic chemicals at 8.6x EBITDA and specialty chemicals at 13.4x, both industry aggregates across firms with positive EBITDA, on trailing earnings. A multiple taken on a downturn year's EBITDA misleads, because the denominator is depressed.
- How should net debt leverage be screened?
- Divide net debt by mid-cycle EBITDA. The denominator changes the answer: Olin's $2,659.7M of net debt is 4.1x its FY2025 adjusted EBITDA of $651.8M and about 1.3x its own ~$2.0B 2029 target. The Selborne model's leverage thresholds are set for mid-cycle EBITDA, so a ratio on a downturn year cannot be read against them.
Read next
Reading the Petrochemical Cycle
Spread mean-reversion, utilisation as early warning, trough language across the peer set, and why normalised prices differ from spot prints.
Ethylene and Polyethylene Spreads
Cracker margin vs integrated PE margin vs colloquial spread; FY2025 NA PE at $580/t vs $830/t history; illustrative ethane cracker margin 12¢/lb.
Reported vs Adjusted EBITDA in Chemicals
What six producers add back to reach adjusted EBITDA, why write-downs and closures dominate a bad year, and which figure answers which question.
Valuing Specialty Chemicals (Premium P/E)
Why trailing P/E fails; Damodaran EV/EBITDA +56% specialty premium; forward P/E ~18-24x band; why commodity-linked earnings weaken the case for a premium.
See it applied
These company profiles apply the concepts from this guide to real public companies.