Net zero in practice: Insights from equity investors (part 2)

Preparedness for transition

Daniel Gallagher, the PRI’s senior lead on climate, discussed the latest climate science and summarised current investor actions on climate risk and the strategic value of transition planning. He added that new risk-based analyses are providing insights on the system-level dimensions of climate risks as we witness record-breaking climate disruption and near-term activation of climate tipping points. Some of the key key pillars that investors can benefit from in transition planning include:

• Strategic ambition: set a baseline (the starting position) and objectives relating to the transition.
• Identify suitable metrics and specific targets i.e. to which portfolios and how many assets classes does the strategy apply?
• Implementation strategy (implement the above two points).
• Engagement strategy: identify actions to engage with peers, policy makers, etc., to deliver an enabling environment to achieve the plan.
• Governance: account for how the plan will be implemented across an organisation, including through its leadership.

Gallagher acknowledged that planning in the face of uncertainty requires forward-looking analysis and adaptive decision-making. Investment managers should consider using levers such as policy and portfolio company engagement, target setting and asset alignment.

Presentations and discussions in the subsequent four workshops focused on the following topics: using metrics, conducting company analysis, engaging with issuers and reporting progress.

Using metrics

Our Initiatives team presented an overview of the most common metrics, their key points and any associated challenges.

Rationale for using climate metrics:

• Assess (and demonstrate to clients and beneficiaries) the climate impact of investment activity.
• Measure climate performance relative to peers.
• Track progress toward climate-related goals.
• Help assess climate-related financial risk and opportunity.

Overview of four metric types:

1. Net-zero alignment

This is an assessment of portfolio companies’ alignment with net zero. The assessment is often based on a combination of data points including companies’ emissions, targets, capex, and strategic planning. The Net Zero Investment Framework from the Institutional Investors Group on Climate Change (IIGCC) is often used to help assessment.

Potential challenges

Challenges include data gaps, data choices (which decarbonisation pathways are relevant?), data ambiguity (are targets credible?) and ranking judgments (what if a company falls between tiers, or has other salient features that are not listed here?)

Net-zero alignment should be used with a consistent research methodology, either that of the investor or a research provider, to allow comparability of metrics and results over time.

2. Decarbonisation

A measure of the amount of greenhouse gases attributable to a portfolio (‘financed emissions’), typically reported in terms of carbon dioxide equivalent (CO2e) using the Partnership for Carbon Accounting Financials (PCAF) methodologies. Decarbonisation is calculated based on proportional ownership of public equity and debt for listed companies and conceptually similar methodologies for other asset classes, including private debt, sovereign debt, real estate, and mortgages.

Why use it?

Decarbonisation focuses on the core issue, is quantitative – hence popular – and is framed in the same terms as ‘net zero’.

Potential challenges

Data can be missing or estimated, and can be as much as two years old. Asset managers usually measure emissions across a portfolio using financed emissions metrics based on Enterprise Value Including Cash (EVIC) or using intensity measures. However, market fluctuations mean emissions footprints and intensity ratios can change for reasons unrelated to actual GHG emissions: this became apparent when investors, having emerged from the pandemic, thought their portfolio emissions intensity was declining rapidly before realising they were relating pandemic-era emissions to post-pandemic financial variables.

Challenges with baselines

Investors often want to track emissions intensity over time (e.g., since 2019), but it can be hard to obtain good data for a past baseline. In addition, future developments can necessitate ‘re-baselining’ or recalculating portfolio baseline year emissions, a practice that investors may want to undertake to ensure consistency and relevance of reported GHG data over time and to reliably track progress against portfolio decarbonisation objectives.

Challenges with benchmarks

Decarbonisation metrics (applied to a particular investment or portfolio) are hard to understand in isolation and make more sense when compared to a benchmark. But investors sometimes struggle to decide upon or construct an appropriate benchmark for their portfolios, and to obtain robust emissions data for it.

It was also noted in the workshops that: i) equity portfolios can decarbonise quickly through divestment, without real-world impact ii) stewardship can lead to real-world decarbonisation and also show up in portfolio decarbonisation metrics, but the process is much slower, and iii) some portfolios can look more carbon-intensive than average when their climate impact is arguably positive (e.g., those invested in renewables infrastructure that is currently being built).

Our speaker(s) said that decarbonisation was increasingly being tracked over multiple years (e.g., as a rolling three-year average to smooth out fluctuations) and is best disclosed to clients or beneficiaries within a narrative context, explaining the benchmark, baseline, reasons for short-term trends, and relationship to real-world impact.

3. Implied temperature rise

A measure of the projected rise in average global temperatures compared to pre-industrial levels that would result by a given year (e.g., 2050) if the economy as a whole were to resemble the investor’s portfolio.

Why use it?

Implied temperature rise can focus attention on the overall problem of climate change, raising awareness, incentivising action and enabling comparisons to benchmark. It can also be used as a constraint when setting parameters in quant investing strategies.

Potential challenges

Climate scenario modelling complexity and uncertainty is a well-known challenge. The sensitivity of temperature estimates to the choice of assumptions particularly problematic.

As a top-line metric, ITR is most meaningful when compared to a benchmark, not just a temperature goal (e.g., “while the IPCC target is 1.5 degrees, our portfolio’s ITR is a full degree lower than that of our benchmark”); however, this context should be accompanied by full disclosure of models and scenarios used.

4. Engagement

Engagement can be reported quantitatively (“X number of corporate dialogues”, “Y% of votes in support of climate-related shareholder proposals”) or qualitatively (“a comment on SEC proposal Z, in which we noted…”). To clients and beneficiaries, engagement can show investors positive activity and intent; it draws attention to relationships among financial system actors that all need to address climate change; and it focuses attention on particular issues, sectors or companies.

Challenges include difficulty in demonstrating current impact, since effects of engagement unfold over time, and quantitative measures may be misleading (e.g., more engagements are not necessarily better; some votes against climate proposals may be warranted). Disclosures can place engagement efforts in context and link to real-world management of climate-related financial opportunities and risks. 

Participant reflections on our presentation of metrics

Re ITR measures, users recognised the need for many assumptions and uncertainties arising. Some participants expressed preference for climate alignment metrics rather than ITR and noted that the latter measure is problematic regarding portfolio decision-making.

EVIC is widely used but can get distorted by volatility in the market cap and value of an issuer’s debt, share buybacks or other changes to the capital structure that appear to change the carbon footprint without any change in actual emissions. 

Carbon intensity – to WACI or not WACI?

This is a commonly used metric, is intuitive and particularly helpful when identifying exposure to carbon-intensive issuers.

But at portfolio level, while the Weighted Average Carbon Intensity (WACI) – which measures the tonnes of CO2 emitted per US$1 million of company sales – is intuitive, it is also backward-looking and may give the wrong message as it does not exclude business cycle impacts and can be affected by outliers. It is also difficult to use when comparing companies across different sectors. At portfolio level, carbon intensity should be treated with caution, recognising that different sectors can achieve different speeds of decarbonisation.

Not taking a number at face value

Analysts and portfolio managers who use certain values (e.g., emissions and sales of an issuer) to measure and compare carbon intensity need to recognise factors that may affect the accuracy and interpretation of that value. This includes errors in reported values (e.g., due to incorrect accounting of a corporate action at an issuer, misalignment of reporting periods between emissions and sales); and issues with methodologies that don’t consider differences in the accounting of emissions and revenue for issuers in specific industries or with distinct business models.

Portfolio-level aggregate carbon intensity or footprint should be used with care, and more granular analysis is desirable, as well as an understanding of the broader context. Simplistic sector rotation or changing holdings within the sector should give way to organic decarbonisation where the existing holdings make decarbonisation progress. There were also calls in the workshops for more robust frameworks to evaluate individual companies’ climate performance.

Not relying on backward-looking data

Participants also cautioned against relying on a single (typically backward-looking) metric that fails to differentiate between alignment efforts and potential of different issuers in different sectors, recognising how difficult it may be for some companies to decarbonise and how material and valuable decarbonisation of some companies may be.

We noted little appetite for major reliance on backward-looking criteria (as mandated by the Paris Aligned or transition benchmarks, which often lag financial disclosures) to judge company transition preparedness or suitability for portfolio inclusion. (The EU’s Paris-aligned benchmarks require that index-level carbon intensity is reduced by half relative to the parent index and is also reduced each year by 7% relative to the previous year’s climate index intensity.)

There were strong calls for in-depth analysis of company positions, and assessment of alignment, considering sector differentiation, and, if necessary, followed by rigorous measurement and monitoring of progress and relevant engagement on an ongoing basis.

RBC Global Asset Management’s (RBC GAM) speakers noted that noted that carbon metrics can help to measure corporate exposure to transition risk and facilitate a peer comparison. However, metrics are typically sourced from third-party data providers and in certain cases cannot be relied upon for this analysis without resolving several key issues, including data errors, and methodology reviews – challenges that are common across data providers. Solutions are based on engagement with data providers and internal checks. 

How RBC Global Asset Management calculates emissions

RBC GAM contributors explained that when reporting an issuer’s carbon intensity, it brings together data from two distinct frameworks: the Greenhouse Gas (GHG) Protocol for emissions (e.g., emissions allocated based on financial control or equity share), and accounting standards for sales (e.g., impacted by the company’s role as either a principal or agent in the transition). In most cases the two standards are in alignment. Where they are not consistent it can lead to misrepresentation of calculated emissions intensity. In a case study written for the PRI, RBC GAM gives an example around revenue recognition methodologies in the case of a hotel company. Another example is presented below.

Example: Accounting for emissions at a holding company

Holding companies effectively own interests in other companies but do not themselves produce goods or services. The following example is a multi-national holding company with multiple wholly-owned, majority-owned, and/or minority-interest subsidiaries and affiliated companies. Based on this business model, it would be reasonable to expect the carbon intensity of the holding company to resemble the carbon intensity of its underlying businesses. However, the carbon intensity of this holding company, as represented by third-party providers, was nearly five times higher than its most carbon-intensive underlying business unit, (1,402 tCO₂  eq/US$M sales compared to 288 tCO₂  eq/US$M sales ).

Figure 1: Multinational holding company: previous approach

Source: RBC GAM. Data as at 30 April 2023

Based on analysis of the carbon intensity metric above, RBC GAM identified that the holding company and subsequently the data provider applied the GHG Protocol’s equity share approach to calculate the holding company’s emissions. This means the company’s absolute emissions were directly proportionate to the emissions of the various underlying businesses, based on equity stake. When the holding company owned 50% or more of the underlying unit, no issues arose and the relevant percentage of revenues were recognised. But where it was less, revenue was recognised by the holding company based on its share of the business unit’s profits and/or losses (not the business unit’s sales). This can create a misalignment of revenue and emissions.

For one particular case, RBC GAM identified that 98% of the holding company’s emissions were from entities for which there was less than 50% ownership. Effectively, the carbon intensity measure was more reflective of tons of tCO₂  eq/Co2 per US$1M of profits, not sales, resulting in a higher-than-expected metric. RBC GAM’s recommendation was to allocate emissions and revenue based on ownership percentage across all business units, regardless of whether ownership was greater than 50%.

As demonstrated in the chart below, this revised approach that has been adopted by the third-party data provider based on a methodology review. It resulted in the holding company’s carbon intensity being 96% lower than the previous approach (63 tCO₂  eq per US$M sales vs. 1,402 tCO₂  eq per US$M sales).

Figure 2: Multinational holding company: revised approach

Source: RBC GAM analysis. Data as at 14 May 2025 

Company analysis

Workshop participants agreed that company analysis is an integral part of portfolio decarbonisation processes, and helps to:

• understand the business and its commitment to decarbonisation;
• assess transition readiness;
• judge alignment with desirable pathways;
• support cash flow forecasting, valuation and investment decisions;
• inform engagement activities.

Several frameworks and tools are available to help conduct company analysis, including backward-looking reported carbon emissions metrics, the EU Taxonomy and the Science Based Targets Initiative (SBTi), as well as more forward-looking Climate VaR and ITR. Many investors supplement these with proprietary net-zero alignment frameworks. 

Neuberger Berman’s proprietary indicator

Company engagement

Back in the workshop series, an investment manager explained their experience with multi-year engagement and cited an example of one company with which the investment team held 60 engagements over more than six years. The individuals they engaged with included the CEO, CFO the company’s General Counsel and the Chief Sustainability Officer.

The topics they engaged on included targets around net zero, renewable energy usage, talent retention, scope 3 emissions, executive team quality, environmental sustainability product releases, emission reporting, and inclusion of sustainability metrics in remuneration.

The investment manager outlined engagement results along the way as well as occasions where they voted against board members. The results included:

• Net zero: inaugural emissions reporting and commitment to reduce emissions by 50% by 2030.
• SBTi target verification.
• Meaningful progress with board diversity, senior management succession planning and executive compensation.
• Influence on company product design to reduce environmental footprint of customer products .

The key conclusion was that it was possible to achieve real-world decarbonisation, but success requires change to be embraced and driven from the top down. Ultimately, engagement objectives need to go beyond requiring timely and credible disclosure from the issuer, but also impact decarbonisation in a real way.

The investment manager said that investors must recognise bumps along the way and that they stick to the business case, ask questions, understand the context, offer solutions and remain flexible about prioritisation. 

Attributing and reporting (portfolio decarbonisation) progress

Emissions attribution involves breaking down emission calculations into different drivers and components. In December 2023, the Net Zero Asset Owner Alliance (NZAOA) published guidance on attribution. In addition, the IIGCC’s NZIF implementation guidance recommends that investors seek to understand which factors are driving the changes in portfolio emissions by conducting decarbonisation attribution analysis.

Workshop participants noted several positive reasons for conducting attribution analysis, including a desire to be transparent and report findings internally and externally in a way that is not misleading, to understand what is causing the footprint change as well as what is causing the change in portfolio emissions, and answer potential questions regarding the split between portfolio decarbonisation and real-world decarbonisation.

Workshop contributors explained that their attribution analysis typically splits drivers of emissions into several categories:

i) company changes – emissions change due to change in the operations of the underlying assets.

ii) portfolio composition – emissions change due to divestment, new investment, and inter- or intra-sectoral weight change.

iii) financial metrics – emissions change due to movements in non-emissions-related metrics, such as EVIC and revenue.

iv) data – emissions change due to changes in data, including coverage, quality, and methodology.

Some of the above factors are under the investment manager’s control, while others are subject to market movements or are under the control of the issuers.

The following example shows how Varma conducts attribution analysis in its listed equity portfolio and captures factors such as portfolio changes (additions, disposals), inflows and outflows, stock price changes, changes in issuer capital structure, real-world decarbonisation and changes in terms of emissions scope.

The diagram below shows that:

• the greatest reduction comes from sold or reduced positions
• reduction in emissions related to existing positions is somewhat offset by an increase in the ownership share of those issuers
• data quality, possibly a major driver in reported positions across alternative asset classes, plays a very limited role as a driver in the listed equity context
• the listed equity portfolio has increased in value in 2024, yet financed emissions have decreased by 33%.

Figure 3: Varma’s portfolio decarbonisation attribution analysis (2024)

To gain further understanding of how and which portfolio changes impacted carbon footprint, the team analyses sector and geographical weight changes (through trading or by values changing) and the impact of investment decisions within sectors or regions, noting the influence of greater or diversification or concentration.

Varma said analysts can narrow the focus on new positions, which would provide a picture of where the bulk of the emission increases come from. Equally, shining the light on deleted positions would highlight the emission reductions from divesting from certain sectors. For further context, the investment team at Varma looks at multi-year attribution across the different drivers.

Varma also pointed out that investing in climate-positive companies that manufacture sustainable products e.g. renewable energy infrastructure may still result in higher emissions than, say, investments in the technology sector. Measuring progress towards climate-positive investments should be pursued in conjunction with other climate-related targets.

Workshop participants were reminded by the Varma and other speakers that to achieve real-world change, institutional investors should focus on engaging with existing investments to help them achieve climate goals as opposed to overly focusing on changes in portfolio holdings. Regarding engagement, participants noted that it is much easier to encourage companies and managers to provide good quality emissions reporting compared to achieving change in how companies and managers operate. Successful engagement that results in new commitments and targets for the investee does not necessarily provide immediate results for portfolio financed emissions but has significantly more impact on real-world emissions. 

Robeco’s evolution in carbon performance measurement

Re-baselining is the process of calculating what a baseline carbon footprint would have been at a start date, accounting for changes in financial valuation metrics, asset mix, or emissions scopes. In 2023, we published a case study   on re-baselining methodologies, particularly focusing on Enterprise Value Including Cash (EVIC) adjustments required for EU Paris Aligned Benchmarks   and advised by PCAF. The guidance advises on using EVIC inflation adjustments to correct for artificial footprint changes caused by company valuations, using the formula: Adjusted (Re-Baselined) Footprint = Original Footprint / EVIC Adjustment Factor. This approach aims to isolate genuine decarbonisation from valuation-driven changes.

However, after five years of applying EVIC adjustments in our own decarbonisation performance assessment, we have identified flaws in the standard methodology. The first issue was the weighting used in adjusting for EVIC: our analysis of the top 20 MSCI World holdings revealed that EVIC growth predominantly occurred in low-footprint companies such as technology stocks like Apple and Microsoft, while high-footprint companies showed minimal EVIC changes. This created a paradox where the adjustment factor corrected for portfolio footprint changes that hadn’t occurred, leading to misleading results — in one example, a -12% raw decarbonisation became a +20% adjusted increase. The second problem involved estimation correlation, where estimated emissions data, often revenue-based, can correlate with EVIC changes. When EVIC increases alongside estimated emissions, the footprint may remain stable, reducing the need for adjustment and leading to flawed EVIC inflation correction.

In response to these challenges, we developed an enhanced approach, weighting EVIC changes by footprint contribution rather than market weights: ADJ = Σ(Footprint Contribution × EVIC Return). This reduced the adjustment factor from 1.4 to 1.1 in our test case, yielding more intuitive results (-4% vs +20% adjusted decarbonisation). However, even this improvement couldn’t resolve fundamental issues with estimated versus reported emissions data, particularly during periods of high revenue correlation with EVIC (notably in 2022). The complexity of distinguishing genuine decarbonisation from data artifacts remained problematic.

Recognising these limitations, we propose to focus on the use of attribution analysis rather than continuing to refine EVIC adjustments. This approach decomposes footprint changes into constituent factors including active management effects from portfolio construction decisions, weight drifts from market-driven allocation changes, company emissions from underlying business changes, and EVIC effects from valuation-driven changes. This methodology provides transparent insights into decarbonisation drivers without requiring complex adjustments to historical data. Figure 1 illustrates this approach.

Figure 4: Attribution analysis   of the footprint change over time (Dec 2019-Dec 2024) for an illustrative portfolio

Instead of adjusting footprint figures, we now propose to adjust pathway targets once based on the cumulative EVIC effect observed. If EVIC effects contribute 3.4% market tailwind, we adjust our 7% Paris-aligned target to 10.4%, interpreting this as: “We need to achieve 7% decarbonisation from active decisions plus 3.4% from favourable market conditions.” This approach maintains nominal footprint reporting while providing clear accountability for different decarbonisation sources. An example is shown in Figure 2.

Figure 5: Decarbonisation pathway adjusted for EVIC inflation

Our experience demonstrates the value of proactive methodology evaluation in carbon performance measurement. While EVIC adjustments address legitimate concerns, their practical implementation faces technical challenges with estimated data, correlation effects, and choice of weighting. By transparently breaking down and distilling carbon performance drivers through attribution analysis, we gain actionable insights while maintaining robust accountability for our net-zero commitments.