Climate risk lives in the same DCF, the same comparables, the same credit model as any other risk. It does not require a separate framework. What it requires is that the assumptions in those models reflect climate-related changes in operating conditions, regulation, and capital costs. For some companies this changes the valuation modestly; for others it changes it by an order of magnitude.
The four DCF adjustments
- Revenue trajectory: market share trends in declining categories (combustion engines, thermal coal, virgin polymers) and growing categories (battery storage, green steel, plant-based protein).
- Operating margin: carbon-tax pass-through, energy efficiency capex, regulatory compliance costs, asset retirement obligations.
- Capex schedule: transition capex (refurbishments, new technology, decarbonised processes) — often front-loaded in 2025-2035.
- Discount rate / terminal multiple: higher beta for unhedged transition risk, lower terminal multiple for businesses where the long-run unit economics are uncertain.
Where the value lives
For most carbon-intensive sectors, the biggest valuation impact comes from the terminal multiple, not from near-term cash flows. A coal-fired utility with 20 years of operating life left has substantial 2025-2035 EBITDA — but its terminal value is approaching zero. Standard DCF practice (Gordon growth to perpetuity) systematically overvalues such businesses. Climate-adjusted modelling caps or zeroes the terminal value.
NGFS scenarios
The Network for Greening the Financial System (NGFS) — a coalition of central banks and supervisors — publishes a standard scenario set used for climate stress testing. As of 2026, the main scenarios are:
- Net Zero 2050: orderly transition; warming limited to 1.5°C; high transition cost in the 2020s but lower physical cost long-term.
- Below 2°C: orderly transition; warming limited to ~1.7°C; slightly later policy action; comparable transition cost.
- Delayed Transition: policy action delayed to 2030; then sharp, disorderly response; high transition AND physical cost.
- Current Policies: only existing policies implemented; warming reaches ~3°C+ by 2100; very high physical cost.
- Fragmented World: regions decouple; uneven policy; weak coordination; both transition and physical risks elevated.
Pick at least two scenarios
Best practice for stress testing is to run a 1.5°C-aligned orderly scenario (e.g. Net Zero 2050) and at least one disorderly scenario (Delayed Transition or Current Policies). The spread between them is the climate-risk range. A single scenario is no scenario.
When climate risk is small
Not every business needs a climate-adjusted DCF. A software-as-a-service company with no physical infrastructure, low energy intensity, and customers across diverse sectors has limited near-term climate exposure. The exception is its data-centre energy use, which can be material at scale but typically affects margin by basis points, not percentage points. The discipline is to ask the materiality question first; doing climate analysis on immaterial issues wastes time and credibility.
When climate risk dominates
Long-lived fossil-fuel infrastructure (coal plants, oil refineries, gas transmission), carbon-intensive heavy industry (cement, steel, aluminium without abatement plans), and physically exposed assets (coastal real estate, drought-vulnerable agriculture) sit at the other end. For these, climate-adjusted valuation isn't a sidebar — it's the load-bearing assumption set. Recent analyses of oil major valuations show 20-40% TV differences between climate-adjusted and conventional models.
Exercise
Take a DCF you've built or seen recently. List the assumptions: revenue growth, margin trajectory, capex / sales, terminal growth, WACC. For each, ask: how would this change under (a) Net Zero 2050 and (b) Current Policies? Calculate the valuation under both. The range you produce is the climate-risk uncertainty range for that asset.