Abstract

Understanding the drivers influencing the greenhouse gas emissions associated with financial portfolios is crucial for constructing and monitoring a consistent climate investment strategy. Several frameworks have emerged in recent years to perform attribution analyses, and which aim to identify the drivers behind portfolio decarbonisation over time. This article provides a qualitative and quantitative comparative analysis of these frameworks, examining the key drivers identified and the methods used to isolate the effects of each driver. Based on this review, we formalise a generalisation approach to combine them, and recommend five models that are tailored to answer specific questions. These models should help investors to better understand the drivers behind portfolio emissions metrics changes and distinguish the exogenous drivers over which they have limited control from those where they can exert direct influence.

Key takeaways:

• Since 2022, several attribution frameworks have been developed to help investors better understand the drivers influencing change in emissions metrics – absolute emissions, emissions intensity, and emissions footprint – associated with financial portfolios over time.
• The drivers identified across these frameworks can be classified into four main categories: data coverage, portfolio reallocation, economic and financial fluctuations, and company emissions. Two methods are commonly used to attribute change in emissions metrics to these drivers: the Laspeyres indicators and the logarithmic mean Divisia index.
• A quantitative example illustrates that these frameworks are complementary, and we show that it is possible to combine them through a three-step generalisation. This flexible approach enables investors to assess the contributions of various drivers such as strategic asset class allocation, divestment, sector allocation, stock selection, instrument price volatility, emissions scopes, company activity, inflation, to any change in portfolio emissions metrics.

Introduction

Asset owners can contribute to mitigating climate change through various means, including by reducing the emissions associated with their portfolios. Regulatory and voluntary frameworks have been established to support these efforts by defining metrics for measuring the indirect emissions associated with financial portfolios (PCAF, 2022), harmonising reporting standards related to the emissions of financial entities and instruments (SFDR, 2019), and offering emissions reduction target setting protocols aligned with the Paris Agreement’s goals.

Despite these developments, investors still face challenges in understanding and controlling the evolution of portfolio emissions over time. Since 2022, several attribution frameworks have been introduced to help investors better understand the drivers influencing the emissions metrics associated with financial portfolios (Bouchet, 2023; NZAOA, 2023; Nagy, Giese and Wang, 2023; Simmons et al., 2022). NZAOA (2023) identify several purposes of attribution analysis, with the overarching goal of enabling asset owners and managers to understand the emissions drivers in a portfolio and take informed action. Depending on the type of financial management, actions may involve divestment, portfolio reallocation, corporate engagement, or dialogue with asset managers to assess and challenge their decarbonisation performance. Attribution analysis can also enhance transparency for public reporting, as frameworks like the Science Based Targets initiative (SBTi) recommend financial institutions report and attribute changes in emissions, as well as progress on target indicators, using three levels of details2. Undertaking attribution analysis is therefore increasingly recommended in frameworks for institutional investors, such as the second version of the Nez Zero Investment Framework from the Institutional Investors Group on Climate Change (IIGCC, 2024a). This article compares the main existing attribution frameworks, highlighting their similarities, differences, and explores how they can be combined to offer more flexibility to investors seeking answers to questions specific to their portfolios.

The first section compares frameworks based on the type of portfolio analysed, the emissions metric used, the identified drivers, and the attribution method. Most frameworks focus on historical analysis of absolute emissions, emissions intensity, and the emissions footprint3 of equity portfolios. The identified drivers that explain change in a metric over time can be grouped into four main categories: data coverage, portfolio reallocation, economic and financial fluctuations, and company emissions. Differences between the frameworks primarily arise from portfolio reallocation drivers, with some models emphasising the impact of investment universe variations and others focusing on sector and intrasector reallocations. Two methods are used to attribute change in emissions metrics between drivers. The first is based on Laspeyres (1871) price and quantity indicators, traditionally used to analyse price index change. The second is the logarithmic mean Divisia index (LMDI), derived from environmental economics (Ang, Zhang and Choi, 1998). While the Laspeyres method aligns with established portfolio performance decomposition techniques (Brinson and Fachler, 1985; Brinson, Hood, and Beebower, 1986), it introduces interaction terms that can be challenging to interpret and requires a hierarchical arrangement of drivers. The LMDI approach treats all drivers equally without interaction terms, though it is more complex to implement

In the second section, models inspired by these frameworks are applied to a simplified fictitious portfolio, analysing its absolute emissions, emissions intensity, and emissions footprint. The analysis shows that the frameworks offer complementary insights into the evolution of these metrics. Regarding the attribution method, the Laspeyres method seems preferable for models with fewer than three drivers, while the LMDI method is better suited for models with more drivers.

To enhance the flexibility of these models in addressing specific investor concerns, a three-step generalisation is formalised in the third section to combine previous drivers. The first step involves choosing relevant subsets of instruments (e.g., excluded instruments, instruments with missing emissions data) for which the effect on the metric must be isolated. The second step is to express the emissions metric (absolute emissions, intensity, or footprint) as a product of variables (representing drivers) at the instrument level. Finally, the third step involves selecting an attribution method and its parameters. This flexible approach is illustrated with five model variations designed to assess the contribution of strategic asset class allocation, divestment, sector allocation, stock selection, instrument price volatility, emissions scopes, company activity, inflation, to any change in portfolio emissions metrics.

Attribution analysis of changes in portfolio emissions metrics enables investors to better understand the drivers behind these changes and distinguish between exogenous drivers, over which they have limited control (e.g., instrument price volatility), and drivers where they can exert influence (e.g., divestment, sector allocation, stock selection). As a result, attribution analysis is a crucial tool for constructing and monitoring a climate investment strategy. The generalised model proposed in this article aims to make attribution analysis more flexible and easier to implement, adapting to the specific needs of investors.

Authors

Matteo Bagnara, PhD
Quant Researcher,
Scientific Portfolio ……………………………………….

Shahyar Safaee
Deputy CEO and Business Development Director,
Scientific Portfolio

Read the full Whitepaper

"*" indicates required fields

Name*

First Name Last Name

Company Email*

Phone Number

Company Name*

Organization Type*

Select your organization typeAsset ManagerBanks & Capital MarketsInstitutional InvestorWealth ManagerRegulators & Public BodiesConsultantsMediaAcademicsOthers

Job Title*

Country*

Select your countryArgentinaAustraliaAustriaAzerbaijanBelgiumBoliviaBrazilBulgariaCanadaChileChinaColombiaCosta RicaCroatiaCyprusCzech RepublicDenmarkEstoniaFinlandFranceGermanyGreeceGuernseyHong KongHungaryIcelandIndiaIrelandIsraelItalyJapanJerseyKuwaitLatviaLiechtensteinLithuaniaLuxembourgMalaysiaMexicoMonacoNetherlandsNew ZealandNigeriaNorwayPeruPhilippinesPolandPortugalQatarRomaniaSaudi ArabiaSingaporeSlovakiaSloveniaSouth AfricaSouth KoreaSpainSwedenSwitzerlandTaiwanThailandUnited Arab EmiratesUnited States of AmericaOther

City

Choices*

I acknowledge that Scientific Portfolio will process my personal information in accordance with its Privacy Policy. I can opt out from Scientific Portfolio’s marketing emails at any time by clicking an unsubscribe link.

https://scientificportfolio.com/wp-content/uploads/2025/09/Volatile-Markets-Put-‘Extreme-Risk-in-the-Spotlight.pdf

Download the PDF