A study of major European sustainable investment products has found that most large Article 8 equity exchange-traded funds retain exposure to conventional weapons manufacturers, fossil-fuel businesses or extractive industries, challenging the assumptions many investors associate with the European Union’s most widely used sustainability classification.

The Justice Company examined 100 leading Article 8 UCITS equity ETFs, reviewing sector allocations and the weapons-screening provisions contained in their underlying index methodologies. The research found that 59 of the funds held companies involved in conventional military equipment, while most retained what the organisation described as meaningful exposure to fossil-fuel or extractive sectors.

The results focus attention on the distinction between a regulatory disclosure category and an ethical investment label. Under the Sustainable Finance Disclosure Regulation, Article 8 products promote environmental or social characteristics, or a combination of both, provided the companies in which they invest follow good-governance practices. The classification does not require sustainability to be the product’s sole objective, nor does it establish a single list of industries that every qualifying portfolio must exclude.

That distinction has often been obscured in fund distribution. Article 8 products are commonly described as “light green” funds and are frequently grouped within sustainable, responsible or environmental, social and governance investment ranges. Retail investors and advisers may therefore interpret the designation as evidence of broad ethical screening, even when the portfolio has been designed to satisfy a narrower climate, governance or controversy-based mandate.

The Justice Company’s most prominent finding concerned defence exposure. All 100 ETFs in the sample excluded controversial weapons, including categories such as cluster munitions, anti-personnel landmines and chemical or biological weapons. Those exclusions, however, did not necessarily extend to conventional military equipment.

As a result, companies producing artillery, missiles, armoured vehicles, combat aircraft and naval systems remained eligible for many of the indices tracked by the funds. The research identified major defence contractors including Rheinmetall, BAE Systems, Leonardo and Lockheed Martin among the companies that could remain within Article 8 portfolios. Rheinmetall appeared among the 20 largest holdings of at least one ETF marketed with an enhanced ESG approach, according to the findings.

The inclusion of such companies was not presented as an indexing error or a breach of the ETFs’ published rules. It was the expected outcome of methodologies that draw a regulatory and commercial distinction between controversial weapons and conventional defence production. A fund can therefore apply a formal weapons screen while continuing to own manufacturers whose principal products include military systems used by national armed forces.

That distinction has become increasingly important as European governments expand defence budgets and reconsider the role of military companies in sustainable portfolios. Supporters of defence investment argue that national security, democratic resilience and the protection of civilian populations can be compatible with social objectives. Other investors maintain exclusions covering conventional weapons because of concerns about conflict, civilian harm, international humanitarian law or the end use of military equipment.

Article 8 does not resolve that debate for investors. Instead, the classification permits a range of approaches, leaving the decisive rules to the fund manager and index provider. Two ETFs carrying the same SFDR classification may therefore produce materially different exposure to defence companies, depending on their revenue thresholds, business-involvement definitions, geographic coverage and treatment of corporate controversies.

The study found a similar divergence between broad sustainability expectations and actual portfolio construction in energy and extractive industries. Across the 100 ETFs, energy stocks accounted for an average 1.63% of assets, compared with a 3.78% weighting in the MSCI World Index. The 57% reduction relative to the global benchmark indicated a meaningful aggregate tilt away from the sector.

That average nevertheless concealed substantial variation. Fifteen of the funds carried energy allocations equal to or greater than the MSCI World weighting, leaving those portfolios with at least as much sector exposure as an unscreened global equity benchmark. The result suggests that an Article 8 classification alone cannot be used to infer that an ETF has eliminated, or even substantially reduced, its holdings in oil and gas companies.

Materials exposure was more pronounced. The sector, which includes mining, metals and chemicals businesses, represented an average 3.67% of the sampled portfolios, above the MSCI World weighting of 3.40%. Forty-seven of the ETFs exceeded the benchmark’s materials allocation, while the highest exposures approached 9.9%.

The Justice Company argued that some commodity and extractive businesses can remain prominent in ESG indices because they perform well on governance, reporting, disclosure and operational metrics, even when their core activities involve mining or other environmentally intensive processes. That outcome reflects the difference between selecting the highest-scoring companies within a sector and excluding the sector entirely.

Best-in-class ESG methodologies often retain companies from carbon-intensive industries while favouring those judged to have stronger governance, lower relative emissions, better safety records or more credible transition plans than their peers. Such an approach may improve a portfolio’s sustainability characteristics without producing the fossil-free exposure that some investors expect from an ESG-branded product.

Utilities represented an average 2.55% of the ETFs, broadly in line with the MSCI World weighting of 2.48%. Some European-focused products allocated considerably more to the sector because of large positions in renewable-energy and electricity-network companies, including businesses such as Iberdrola, Enel and National Grid. The study recognised that allocation as a genuine green tilt, while arguing that it did not offset continued exposure elsewhere in the portfolios.

Combined allocations to energy, materials and utilities averaged 7.85% across the Article 8 universe studied. Separate reporting on the research indicated that 54 of the 100 funds had at least 5% combined exposure to energy and materials, the threshold used in that analysis to define a meaningful allocation.

Paris-Aligned Benchmark ETFs demonstrated another form of divergence. Those products follow methodologies designed to cut carbon intensity by at least 50% relative to their investable universes and pursue an annual decarbonisation trajectory. Their rules can remove fossil-fuel extraction companies and reduce energy exposure to zero or near zero.

Paris alignment, however, does not automatically create a broad human-rights or conventional-weapons screen. A climate-focused ETF may exclude large oil producers while continuing to hold defence contractors if those companies satisfy its carbon and controversy requirements. The resulting portfolio can be consistent with its stated climate methodology while conflicting with an investor’s separate ethical objection to weapons manufacturing.

The finding illustrates why broad ESG classifications cannot substitute for examining the investment objective. Climate transition, carbon reduction, good governance, social standards and values-based exclusions are related but distinct strategies. A fund optimised for one of those goals may offer limited coverage of the others.

The Justice Company also questioned the reliance of many indices on United Nations Global Compact controversy assessments. Such screens typically respond after data providers determine that a company has been involved in a sufficiently severe breach or controversy. They do not necessarily exclude every company operating in a conflict-affected region or every business whose products may be used in active hostilities.

For index-based ETFs, this reactive structure is particularly consequential. Passive funds generally follow predetermined eligibility and rebalancing rules rather than making discretionary judgments about each holding. Unless an index methodology includes a specific activity-based exclusion, revenue threshold or controversy trigger, the ETF may continue holding a company despite investor objections that fall outside the index’s formal criteria.

The research does not establish that Article 8 ETFs are misclassified or that their managers failed to follow SFDR requirements. Article 8 is not a guarantee that a portfolio is fossil-free, defence-free or compliant with every investor’s interpretation of responsible investment. The regulation requires managers to disclose the environmental or social characteristics promoted by a product and explain how those characteristics are met.

The analysis instead highlights the risk that investors treat Article 8 as a uniform product label. European authorities have previously acknowledged that SFDR has functioned as a de facto classification system, even though it was originally designed as a transparency framework. The European Commission has proposed replacing or supplementing the current structure with clearer product categories and more comparable consumer-facing disclosures.

For ETF issuers, the study creates a communication challenge. Providers must describe sustainability methodologies concisely enough for retail distribution while also disclosing the detailed rules governing exclusions, carbon targets, sector selection and controversy monitoring. Marketing that emphasises an Article 8 classification without explaining its limitations may encourage expectations that the underlying methodology was never designed to meet.

Index providers face similar scrutiny because their definitions ultimately determine which companies enter or leave many passive portfolios. Conventional-weapons exposure may depend on whether an index excludes all defence revenue, only controversial weapons, or companies exceeding a specified percentage of revenue from military contracts. Fossil-fuel exposure can vary according to whether the screen covers reserves, extraction, power generation, services, transportation or only the most carbon-intensive activities.

Investors evaluating Article 8 ETFs can therefore obtain more useful information from index documents than from the classification alone. Relevant questions include whether exclusions are absolute or revenue-based, whether they apply to corporate parents and subsidiaries, how frequently holdings are reviewed, which controversy provider is used and whether a company can remain eligible while undergoing a transition.

Portfolio holdings should also be assessed alongside the methodology. ETF exposures change as securities are added, removed or reweighted, and a point-in-time study may not capture subsequent index reviews. Holdings data can nevertheless show how abstract screening rules translate into actual allocations and whether the portfolio’s largest positions are consistent with an investor’s stated preferences.

The findings also underline the trade-offs involved in stricter exclusions. Removing defence, fossil fuels, mining and other contested industries can increase tracking differences relative to broad market benchmarks, alter country and sector weights, and concentrate portfolios in technology, financial or consumer companies. Those consequences may be acceptable to values-driven investors, but they should be understood as part of the investment strategy rather than treated as a costless overlay.

Article 8 ETFs continue to provide measurable sustainability tilts, including lower average energy exposure and, in some cases, substantial reductions in portfolio carbon intensity. The study’s central conclusion is not that those characteristics are meaningless. It is that they remain narrower than the comprehensive ethical screen many investors may associate with ESG terminology.

For fund selectors and advisers, the practical implication is that regulatory classification should be the beginning of due diligence rather than the conclusion. Matching an ETF to a client’s sustainability preferences requires identifying which issues matter most, distinguishing between climate and social objectives, and verifying that the index methodology addresses those priorities directly.

The research is likely to intensify demands for more precise fund naming, clearer exclusion disclosures and standardised reporting of controversial activities. Until such information becomes easier to compare, investors seeking to avoid weapons manufacturers, fossil-fuel producers or extractive businesses will need to review the underlying index rules and current holdings rather than relying on the Article 8 designation by itself.