Abstract

This report examines how investors assess corporate biodiversity impacts in response to growing disclosure expectations under the Global Biodiversity Framework Target 15 and frameworks such as the TNFD. It compares eight widely used biodiversity impact assessment tools applied to large listed companies and evaluates their methodological foundations and outputs. The study finds substantial inconsistency in results, raising concerns about their reliability for investment decision-making.

Introduction

Biodiversity loss poses material financial risk, with over half of global GDP moderately or highly dependent on nature. Despite this, around US$7 trillion is invested annually in activities that harm biodiversity. Investors are increasingly expected to identify and manage nature-related risks, yet existing frameworks provide limited guidance on how to measure biodiversity impacts. As a result, investors rely on a growing range of proprietary tools, creating uncertainty about comparability, accuracy and decision-usefulness.

Results

From more than 212 tools listed in the TNFD catalogue, eight were analysed in depth. These tools use distinct methodological approaches, including life cycle assessment-based models, supply-chain extinction risk metrics, ecosystem footprinting and composite scoring frameworks. Documentation and transparency varied widely, particularly around data sources, modelling assumptions and links between economic activity and ecological outcomes.

When applied to S&P 500 companies, biodiversity impact rankings showed low to moderate correlation across most tools. Using aggregate impact metrics, only a few tool pairings showed strong alignment, with correlation coefficients as high as 0.87, while many others were below 0.5. Correlations declined further when impacts were normalised by revenue, indicating that methodological differences become more pronounced when company size is removed as a driver.

No single company consistently appeared among the top ten most impactful firms across all tools. Agreement across tools was limited, with some producing entirely unique lists. Sector-level results also varied: Financials were generally rated as lowest impact and Energy as highest, but within-sector variation often exceeded differences between sectors. Methodological choices, such as whether tools include upstream and downstream impacts or rely on LC-Impact or GLOBIO models, materially influenced rankings across eight of eleven GICS sectors.

Discussion

The divergence in results reflects fundamental differences in how tools source corporate data, model environmental pressures and translate these pressures into biodiversity outcomes. None of the tools assessed have been validated against independent benchmarks, making it impossible to determine which estimates are most accurate. This creates a risk that investors using different tools may reach contradictory conclusions about biodiversity risk, undermining comparability, engagement strategies and product credibility.

The findings also challenge assumptions about inherently high- or low-impact sectors. Significant overlap exists between sectors, with many companies in traditionally lower-impact sectors ranking above the median firm in higher-impact sectors. This limits the usefulness of broad sector screening without firm-level scrutiny. Inconsistent rankings also weaken the effectiveness of stewardship and advocacy efforts that rely on identifying high-impact companies.

Opportunities for improvement

The report identifies four priority areas for strengthening biodiversity impact assessment. First, greater transparency and reproducibility are required, including public documentation of assumptions, data sources and workflows. Second, improved spatially explicit data on where firms operate and what activities occur at each site is critical to link impacts to affected species and ecosystems. Third, reliance on expenditure-based input–output models can distort results and may misrepresent firms using higher-cost sustainable inputs. Regulatory incentives may be needed to improve corporate data disclosure. Fourth, existing life cycle impact models lack spatial resolution and ecological specificity, suggesting the need for bottom-up, site-level approaches that integrate species distribution and ecosystem data.

Study limitations

Access to proprietary methodologies was limited, constraining replication and explanatory power. The regression model explained only around 40% of variation in rankings, indicating that unobserved modelling choices remain influential. These limitations reinforce the need for standardised documentation and independent validation.

Next steps

In the near term, investors should exercise caution when interpreting firm-level biodiversity scores and focus on measuring direct pressures such as emissions, land use and water use. Longer term, collaboration between investors, researchers and tool providers is needed to improve data quality, transparency and methodological rigour so that biodiversity risks can be more reliably priced and managed in financial markets.