Will Thompson, Chief Sustainability Officer at Pacific Asset Management says that portfolio managers need to do more to ensure that sustainability objectives are not diluted by a fast-moving area of the market.
Technology stocks are common in sustainable portfolios. Their size in global ESG indices, strong earnings delivery, and exposure to structural growth themes have made them important holdings for many sustainable portfolio managers.
Yet this is starting to create a more difficult sustainability question: can portfolios continue to rely heavily on large technology companies when the data centres behind AI and cloud growth are putting increasing pressure on power systems?
For several years, there has been a reasonable argument to hold technology stocks and debt in sustainable portfolios.
On the one hand, many of these companies have performed well on indicators such as transparency of ESG reporting and human capital, whilst being businesses with low carbon footprints.
On the other, their dominance in broad equity markets means that sustainable strategies often end up with meaningful exposure to a relatively small number of names.
For sustainable minded investors, this matters because the sustainability profile of a portfolio can increasingly be shaped by a handful of companies whose operational footprint is changing quickly.
Data centre emissions are changing the discussion
The main complication is that the growth of artificial intelligence and digital infrastructure is no longer a marginal sustainability issue.
Rising demand for data centres implies materially higher electricity consumption, and recent research from the International Energy Agency has reinforced expectations that power demand from this infrastructure could increase sharply through the end of the decade.
The most recent data suggests that data centres may now represent 6% of electricity demand in the UK and US.
In practice, this means that companies often perceived as relatively asset-light may now be associated with a significant expansion in energy-intensive physical infrastructure, with consequences for both emissions and transition risk.
Many large technology firms remain committed to matching electricity use with renewable energy procurement, but the way this is measured can obscure the real-world picture.
Market-based carbon accounting approaches may allow a company to claim that consumption has been matched through contractual renewable purchases, even where the local grid serving a data centre continues to rely on fossil fuel generation.
This is because the credits they are buying may be from grids elsewhere (e.g. a data centre in Kent could be claiming 100% renewable electricity procurement even if the credit is bought from a Norwegian hydropower plan).
The challenge is compounded by the pace of infrastructure build-out. In some markets, incremental electricity demand linked to data centres appears to be encouraging new fossil fuel capacity alongside renewable deployment.
That creates a difficult tension for sustainable strategies: businesses that may be central to long-term innovation and productivity gains can simultaneously be contributing to a near-term emissions trajectory that sits uneasily with environmental objectives.
Moreover, increased electricity demand for these purposes will offset efforts to increase energy efficiency or delay electrification elsewhere.
A more constructive point, however, is this increase in power demand is leading to more deployment of clean energy solutions across the grid, creating opportunities for sustainable investors.
For example, there are ETFs available that track the producers and deployers of clean energy technologies, which have performed strongly this year, as orders for solar and wind power have increased – leading to increased earnings expectations.
We think this is a far cry from the bubbles these stocks experienced in the mid-2000s and 2020, where the cost and time to deploy these technologies made them less competitive than fossil fuel equivalents.
Implications for manager oversight and engagement
The practical implication is not that technology should automatically be treated as incompatible with sustainable investing.
Rather, it is that manager explanations now need to go deeper. Where an underlying strategy has meaningful exposure to the large technology platforms driving data centre growth, it is important for portfolio managers to understand:
- How the underlying fund manager assesses electricity sourcing.
- The credibility of renewable energy claims.
- The use of market-based versus location-based emissions data.
In other words, portfolio managers need to leverage one of their strongest tools, stewardship, to ensure we can support this transition.
This is also a useful reminder that sustainability credentials cannot be judged on portfolio holdings alone.
Two managers may hold the same company but approach the issue very differently in terms of challenge, escalation, and the weight they place on emerging emissions risks.
In other words, as sustainability propositions become more precise under current regulation, the quality of manager oversight matters more given a portfolio’s environmental credibility is increasingly linked not just to what it owns, but to how actively those holdings are being interrogated.
Conclusion
The growing prevalence of technology stocks in sustainable portfolios is therefore becoming a more complicated issue than headline product labels might suggest.
Data centre expansion, higher electricity demand, and the limitations of some renewable matching claims all point to a need for more careful analysis of what these holdings mean in real-world sustainability terms.
For the time being, this reinforces the need for sustainable portfolio managers to stay close to managers with meaningful positions in these companies, ensuring that engagement is active, claims are scrutinised properly, and sustainability objectives are not diluted by a fast-moving area of the market.
Main image: sustainable, AI, immo-wegmann-w69Z8K-HGQU-unsplash
