Momentum-based emissions forecasting projects each company's future emissions from its own recent reporting, using a Monte Carlo model. The result is a median pathway with an explicit uncertainty range, whether or not the company has published a target.
Median projected emissions, by Scope and year
Explicit uncertainty range, not a single guess
Versioned methodology tags, for traceability
Business-as-usual comparison, against a flat carry-forward baseline
PCAF data quality score, carried through from the underlying emissions
Probabilistic, forward-looking analysis, aligned with NGFS and TCFD expectations
Add TVaR to PVaR and you get CVaR: one number for total climate exposure
Run the forecast range through climate scenarios to see how much of a holding's exposure comes from its own emissions trend, and how much from the chosen scenario pathway.
Identify holdings where rising emissions or wide uncertainty bands materially affect future risk.
Test a published commitment against what observed behavior alone predicts.
We use each company's own reported emissions, weighting recent years more heavily, to estimate a direction and a typical year-to-year variability by Scope. That runs through a Monte Carlo simulation to produce a median emissions level and an explicit uncertainty band, rather than a single point forecast.
Companies without at least three years of clean reported data, or whose recent behavior is too volatile to model reliably, aren't force-fit into a forecast. Accuracy is highest for larger companies, whose emissions and reporting tend to be more stable. We'd rather flag a gap than publish a number that implies more confidence than the data supports.
A forecast tracks what a company has actually done. Its target says where it plans to go. Set the two side by side to see whether the trend supports the commitment.
Pairing a forecast against a published target produces a gap ratio: forecast emissions at the target year, divided by the target itself. Below 1.0, a company is tracking ahead of its own commitment. Above 1.0, it isn't.
Apple's Scope 1 emissions were unchanged across four straight reporting years. Projected forward on that flat trajectory, its 2030 forecast sits well above its own 2030 target, a gap ratio of 3.5. Toyota's emissions jumped 40% in a single year, and the forecast's upper bound is the widest gap in this comparison, at 6.5. Volkswagen is the exception: a sustained multi-year decline puts its forecast 31% below its own target, a gap ratio of 0.69.
Apple, Toyota, and Volkswagen: Scope 1 emissions (MtCO2e) with forecast trajectories to 2030, built from the underlying data behind the blog "When Targets and Trajectories Diverge."
Historical emissions
Forecast emissions
Forecast range (confidence band)
Published target
The gap ratio isn't a verdict. It's a distance still to close, and it's invisible without both numbers in the same view.
The shape of things to come
Our forecasting methodology, backtested against real reported outcomes across all three Scopes, showing where its accuracy holds and where it doesn't.
Request access to our detailed documents
