One of the challenges asset owners and institutional asset managers face is the measurement of decarbonization targets at the strategy and total plan level. While many firms set decarbonization targets to 2030 and beyond, few have detailed insight as to whether and why they're on track. Questions arise such as:
What are the drivers of portfolio financed emissions and intensity?
Are changes driven by factors controlled by the manager, such as position weights, new investments, and divestments?
Could they reflect companies lowering emission levels due to engagement?
What other factors are reflected and outside the manager's control, such as EVIC changes and data coverage?
To address those challenges and questions, the Net Zero Asset Owners Alliance (NZAOA) created an attribution model. (Understanding the Drivers of Investment Portfolio Decarbonisation) It identifies factors driving emission changes in both financed emissions and carbon intensity terms, providing insights to help managers understand whether their strategies are working.
To enable our client base to answer those questions in practice, we have incorporated the NZAOA attribution model into our suite of FactSet Carbon Diagnostic reports within Portfolio Analytics. The model inputs include portfolio market value, EVIC, and emissions, which we used to recreate the calculations from the appendix of the paper.
In this article we investigate the drivers of emissions changes over a one-year period ending June 30, 2026, using a low-carbon ETF for our analysis. Before examining real-world examples, let’s start with the calculation for financed emissions as detailed by PCAF (Partnership for Carbon Accounting Financials).
The outstanding amount is synonymous with position market value, and EVIC is the typical denominator. The following equation is the basis for much of the analysis described in the paper, and we will discuss how the levers impact results from the sample ETF.
We evaluated a Low Carbon Paris Aligned Benchmark with approximately 300 securities. Over the course of the one-year period, the fund experienced a slight decline in overall absolute emissions—from 395M (tCO₂e) to 386M—yet financed emissions increased dramatically. The divergence between absolute and financed emissions is not unusual, and the NZAOA attribution model helps explain what’s driving it.
To analyze the movements, the model decomposes changes into six attribution factors (see chart below) where the drivers are grouped into three categories:
Factors directly controlled by asset owners: weights, investments, divestments
Factors potentially influenced through engagement: company-level emissions
Factors outside the manager's control: EVIC changes, market cap changes, data coverage
The eight sections below walk through each of these drivers and what they reveal in the context of the ETF example.
1. Change in Exposure—Existing Positions Growing in Size
In the table below, the existing positions securities held throughout the entire period increased in position size. It’s logical because as the fund’s AUM grows, the manager allocates capital to both existing and new positions.
The top 10 positions contributing to higher financed emissions (seen below) increased their position size between 2x to 10x over the year. That directly impacts overall market value, and therefore the changes flow into the Change in Exposure calculation.
Because asset owners manage the portfolio and make decisions to increase/decrease position size, this driver falls into category 1 (factors directly controlled by asset managers).
2. Market Rally—Investment Value Increase
Many of the securities in the above list significantly increased in value over the reporting time period, when market values increased far more than position size factors. Similar to the change in exposure section above, the market value increase is a direct input into the Change in Exposure factor. However, unlike increasing share counts for existing positions, market value movement falls into category 3 (factors outside the manager control).
3. New Investments and Divestments—High Turnover Amplifying the Effect
There is significant portfolio turnover in the ETF we selected for this analysis. Over the one-year period, approximately one-third of the portfolio was sold off while an even larger amount of new investment entered during the same period.
The model captures that through its New Investments factor. Notably, all new investments contributed to an increase in financed emissions, regardless of carbon intensity. The effect is substantial: approximately 50% of the increase in financed emissions stemmed from new securities entering the portfolio. While that is partially offset by the Divested Investments factor, the net effect still resulted in a 20%+ increase from new securities alone.
4. Fiscal Year End Alignment and Reporting End Date Timing
The report date and fiscal year alignment impact the timing of input values such as investment market value and EVIC. Investment value (the numerator) is subject to more recent market movements whereas EVIC (the denominator) is tied to fiscal year alignment. That creates scenarios where market value can increase/decrease more than EVIC due to timing differences in fiscal year-end dates and report dates.
Using Corning as an example, (the largest financed emissions increase above) we see that exact scenario. Corning’s fiscal year ended December 2025 and is the end point for EVIC. During the one-year time period from December 2024 to December 2025, EVIC increased 1.7x. However, from June 30, 2025, to June 30, 2026, the market price moved 5x.
Timing plays into market value movements and therefore is captured in the Change in Exposure factor, while the EVIC changes are captured in the offsetting Changes in EVIC calculation. This example of timing differences flows into category 3, where exogenous factors outside manager control can influence the results.
5. High Emissions Intensity Companies—A Target for Engagement
The next driver to examine is emissions intensity, measured by Emissions/EVIC. For managers considering engagement, the metric enables direct company-to-company comparison. Typically, high-intensity companies will have Emissions/EVIC factors that are 100x to 200x the weighted average in the portfolio. Because we're working with a low carbon ETF, significant outliers are limited. Nevertheless, a handful of airline companies stand out.
Higher intensity companies ultimately have a larger impact on the Change in Exposure calculation that elevate financed emissions. Considering it is company specific, that falls into category 2, where managers can potentially influence investee companies through engagement and steer them to lower carbon policies.
6. Capital Structure—An Unlisted but Important Nuance
A smaller driver that’s not explicitly measured in the factor attribution is capital structure.
Companies with higher levels of debt, such as airlines, tend to have more stable EVIC values during market rallies. When markets rally and market cap increases, high-debt companies tend to exhibit more subdued movements in EVIC, causing the market value/EVIC ratio to accelerate and financed emissions to rise.
In holdings such as United Airlines, Lufthansa, and EasyJet, debt makes up 20% to 30% of the capital structure, muting the rise in EVIC. Combined with the fact that airlines are already carbon-intensive, their capital structures further amplify the increase in financed emissions.
While capital structure is not a distinct factor in the model, this scenario will impact the Changes in EVIC calculation where higher leverage companies will have less of an offset to Change in Exposure compared to low-leverage counterparts. As this is not a decision the manager controls, capital structure differences fall into Category 3.
7. Changes in Carbon Emissions—A Counterbalancing Effect
One factor that countered the increase in financed emissions during this period is actual company-level emissions. Company emissions declined over the period and contributed to an almost 10% reduction in financed emissions through the Changes in Carbon Emissions calculation.
This result sits in Category 2 (factors controlled through engagement), and it is perhaps the most meaningful signal in the analysis. It illustrates the value of having granular factor decomposition: absolute emissions and financed emissions are moving in opposite directions, and the model explicitly highlights magnitude and direction of those factors.
8. Changes in EVIC—An Offsetting Factor
Lastly, broad market movements impacted EVIC values across the portfolio. As markets rose over the period, EVIC values generally increased. All else equal, that reduces the emissions/EVIC ratio and puts downward pressure on financed emissions. However, as noted in the capital structure discussion above, the effect was uneven across the portfolio, particularly for higher leverage companies.
Having walked through the drivers of financed emissions, it is useful to step back and consider an alternative lens. The financed emissions model tells one story, but the carbon intensity version of the model tells another. And in some ways, it’s a more intuitive one.
Before diving into the carbon intensity attribution, it is worth highlighting a key diagnostic metric: Emissions/EVIC. That ratio can be used to target companies for engagement, as it removes the influence of manager investment decisions and enables cleaner company comparisons. A related metric, Carbon Intensity ($Million), scales financed emissions by portfolio market value. Because carbon intensity accounts for security weight, an increase in a company's market cap does not automatically increase intensity.
In our example, emissions decreased while EVIC values increased, resulting in a slight decline in both emissions/EVIC and carbon intensity. That aligns more closely with the movement we observed in absolute emissions.
When we evaluate the carbon intensity model, we are trying to explain the decline in carbon intensity over the one-year period. The model mirrors the financed emissions version, but with one important difference in how new investments and divestments are treated.
Rather than automatically contributing to or detracting from carbon intensity, the New and Divested Investments calculation in this model is measured relative to the total carbon intensity at the beginning of the period. That creates a more nuanced dynamic:
Low carbon intensity companies entering the portfolio benefit overall carbon intensity
Low carbon intensity companies leaving the portfolio through divestment actually increase overall intensity
That gives the carbon intensity model greater explanatory power over manager trading decisions. In our example, we see this exact scenario play out: low carbon intensity companies both entered and exited the portfolio. Those effects, when netted with Changes in Weight and Changes in Carbon Intensity, explain the overall decline in intensity observed over the period.
The Changes in Carbon Intensity factor itself reflects two underlying forces: a decline in absolute carbon emissions and a rise in EVIC across the portfolio. Together, they drove the intensity lower.
Using carbon intensity calculations helps isolate decarbonization from AUM inflows and growth, both of which account for much of the divergence between the two models in this example.
The goal of our analysis is to help managers track and manage decarbonization targets while simultaneously distinguishing between real decarbonization and paper decarbonization.
We observed genuine decarbonization at the company level, yet financed emissions increased due to a combination of AUM growth, a market rally, portfolio turnover, report timing, and capital-structure effects. Weighting decisions, investment choices, and factors outside the manager's control all contributed to the results in the analysis.
The NZAOA attribution model provides the granular factor decomposition needed to make sense of the dynamics. When managers understand which drivers are within their control, which can be influenced through engagement, and which are purely external, they are better equipped to evaluate investment decisions and engage directly with companies about real decarbonization.
Portfolio Analysis (PA) is FactSet's flagship analytics platform, enabling investment professionals to dynamically explore performance, attribution, risk, exposures, and characteristics across any portfolio and benchmark combination. PA is multi asset class, supporting equity, fixed income, and private markets across direct and fund investments. Its flexible report creation framework allows firms to tailor reports to virtually any analytical need or to answer any question. From Brinson attribution to NZAOA to track decarbonization progress to risk factor stress testing, Portfolio Analysis provides a robust foundation for translating raw portfolio data into actionable investment insights.
This blog post is for informational purposes only. The information contained in this blog post is not legal, tax, or investment advice. FactSet does not endorse or recommend any investments and assumes no liability for any consequence relating directly or indirectly to any action or inaction taken based on the information contained in this article.