In the wake of growing geopolitical instability, the investment community ifacing a critical question: can defence be compatible with ESG principles?
Traditionally viewed as ethically incompatible with sustainability goals, defence investments, particularly in weapons manufacturing, have long been excluded from ESG and Sustainable portfolios. Yet the regulatory and political landscape is shifting, and with it, the stance of some institutional investors.
Conventional vs. Controversial Weapons: The Regulatory Distinction
At the heart of the debate lies a key distinction. Controversial weapons – including nuclear, chemical, and biological arms, landmines, and cluster munitions – are banned under international treaties and widely excluded from ESG strategies due to legal, ethical, and reputational concerns. In contrast, conventional weapons such as tanks, firearms, and aircraft are not prohibited and are often seen through a different lens, though they pose ethical and reputational risks, especially when sold to conflict zones.
The SFDR regulation in the EU requires ESG funds to disclose their exposure to controversial weapons (Principal Adverse Impact Indicator 14), but it leaves room for interpretation when it comes to conventional defence, as the related PAI on conventional weapons remains optional for investors to disclose.
Defence Exposure in ESG Funds: A Closer Look
New data from MainStreet Partners shows that 28% of ESG funds analysed are invested in conventional defence companies. This includes 18% of Article 8 funds and 1% of Article 9 funds, while the remaining 9% comprises Article 6 or non-classified funds. In the UK, the figure is even higher, with 32% of UK-based ESG funds rated by MainStreet Partners showing exposure to the sector.
These findings highlight that while outright exclusions remain common, a growing number of ESG strategies are beginning to re-engage with the defence sector – albeit cautiously.
Regulatory Ambiguity and the Role of DNSH
One of the key challenges for ESG funds navigating regulatory frameworks like the SFDR and the EU Taxonomy is the application of the “Do No Significant Harm” (DNSH) principle.
Defence activities are excluded from the EU Taxonomy, which defines what qualifies as environmentally sustainable. This exclusion is not necessarily a judgment on the sector’s environmental harm—but rather a result of its absence from official sustainability classifications. In this case since Defence is not considered as environmentally sustainable activity, the dNSH principle is not applicable.
However, the DNSH principle under the SFDR is broader and applies at the company level. This broader application leaves room for interpretation, which can lead to inconsistencies. Some investors apply DNSH narrowly, excluding only companies involved in controversial weapons, while others adopt a stricter approach, excluding firms involved in any type of controversy—even in the absence of formal breaches.As a result, Article 9 funds, which must pursue sustainability as their core objective, are generally unable to invest in defence stocks. Article 8 funds, which promote environmental or social characteristics, have more flexibility but still face constraints due to reputational risks and diverging interpretations of ESG compliance.
ESG ≠ Ethical Investing
One of the most important clarifications is the distinction between ESG integration and ethical investing. ESG is about assessing and estimatingrisks—including environmental, social, and governance risks—across all industries. It is not, by default, an ethical judgement. Exclusions are often a policy choice or a response to regulation, rather than a reflection of ESG incompatibility.
This misunderstanding has contributed to a binary view of defence: either fully excluded or fully embraced. But the reality is more complex. The sector includes not only arms manufacturers but also cybersecurity firms, logistics providers, and infrastructure companies—areas that could potentially meet ESG criteria with greater regulatory clarity.
Toward a More Nuanced Framework
Calls are growing for a more pragmatic approach. Industry leaders such as Philippe Zaouati, CEO of Mirova, have proposed solutions like “Defence Bonds”—debt instruments linked to responsible defence activities—to create pathways for aligning security-related investments with sustainability objectives. Other investors, such as Alliance IG changed their policies to be able to invest in nuclear weapons and conventional weapons for their Article 8 funds.
At the same time, policymakers in countries like the Finland, Lithuania and Poland are revisiting long-held exclusions and beginning to promote defence investments as essential to national and regional security.
This raises a key point: security is a pillar of sustainability. Without peace and stability, environmental and social objectives cannot be achieved. Yet, however justified these shifts may be in the current geopolitical context, it remains true that weapons are ultimately designed to harm—highlighting the persistent ethical tension at the heart of these evolving investment strategies.
The Role of ESG Data Providers
Amid this uncertainty, ESG data providers like MainStreet Partners play a crucial role in helping investors assess company exposure to both controversial and conventional weapons, flag controversies, and evaluate overall ESG performance. Such assessments help asset managers navigate the grey zones between exclusion and engagement, especially where regulation is ambiguous.
Conclusion: Time for Clarity, Not Exclusion
Defence cannot—and should not—be labelled sustainable under the current EU Taxonomy. But not all defence-related activities are inherently at odds with ESG principles.
The finance industry must push for regulatory clarity, not just compliance. Clearer definitions differentiated treatment of defence-related activities, and practical disclosure guidelines are essential for aligning responsible investing with real-world security challenges.
Because in today’s world, sustainable investing must not ignore the importance of safety and resilience.