Compare Financed-Emissions Metrics Without Mixing Their Questions

Financed-emissions analysis connects financial positions with the emissions of borrowers or investees. The result can support disclosure, portfolio monitoring, target setting, engagement, or research. One number rarely serves all these uses. Absolute attributed emissions, carbon footprint, intensity, data coverage, and data quality answer different questions.

A comparison must keep the numerator, denominator, attribution factor, boundary, and decision use visible.

Start with attributed absolute emissions

Absolute financed emissions estimate the share of an entity’s emissions attributed to a financial position. The attribution factor depends on the asset class and available financial data. This measure can show the scale associated with a portfolio, but it also changes when holdings, enterprise values, debt, and exchange rates change.

Do not interpret a decline as operational decarbonization until the analysis separates company emissions change from portfolio trades, valuation movement, methodology change, and data replacement.

Define the emissions boundary

State whether the numerator includes Scope 1, Scope 2, and which Scope 3 categories. Record the accounting approach for purchased electricity and the reporting year. A comparison is weak when one company includes material value-chain emissions and another does not.

The GHG Protocol provides official standards and guidance for corporate and value-chain greenhouse-gas accounting. A financed-emissions study must still define how company inventories are selected, adjusted, estimated, and attributed.

Treat denominators as questions

Portfolio carbon footprint can divide attributed emissions by the value invested. Weighted average carbon intensity can weight company intensity by portfolio share, often using revenue as the company denominator. Other measures use production, floor area, energy, or another activity.

Each denominator creates a different interpretation. Revenue intensity can change with prices, currency, acquisitions, and business mix. Investment-value denominators change with market values and portfolio flows. Physical intensity can be comparable within a sector but not across unrelated activities.

Write the question beside the metric. “How large is the attributed inventory?” differs from “How emissions-intensive are the companies relative to revenue?” and from “How efficiently does a sector produce a physical output?”

Show coverage and estimation

Report the share of portfolio value and attributed emissions covered by reported data, estimated data, and missing data. A high coverage percentage by value can still omit a small set of emissions-intensive positions.

Record estimation method, source year, sector proxy, and any scaling. Do not replace missing data with zero. Test the result with reported-only data and with reasonable alternative estimates.

Reconcile change over time

Build a bridge from the prior result. Separate changes caused by company emissions, portfolio purchases and sales, valuation, foreign exchange, corporate actions, boundary changes, and data updates. This makes progress claims testable.

Use a consistent base when possible, but do not preserve old poor data only to maintain a smooth series. Restate the history when practical and disclose the break when it is not.

Match the metric to the action

For disclosure, completeness and methodological consistency can be central. For engagement, company trajectory and decision leverage can matter more than portfolio-weighted intensity. For allocation research, the metric may need to connect to valuation, risk, or a stated mandate. None of these uses turns emissions into a prediction of returns.

Present several measures only when each has a job. A compact table should show definition, unit, coverage, boundary, main sensitivity, and intended use. The goal is not to find the one true carbon number. It is to prevent a metric from answering a question it was not designed to answer.