AI, power demand and the carbon challenge: why investors need avoided emissions analysis
In a world of rising power demand and carbon constraints, our avoided emissions research shows that decarbonisation enablers can also be drivers of financial returns.
Authors
After decades of low to negative growth in power demand across developed markets, electricity consumption is set to surge in major economies (Figure 1), fuelled by the proliferation of generative AI and large language models which have turned data centres into some of the world’s fastest growing energy users.
This document may contain “forward-looking” information, such as forecasts or projections. Please note that any such information is not a guarantee of any future performance and there is no assurance that any forecast or projection will be realised.
AI model training, the initial driver of the next wave of data centre demand, is likely to be far more energy-intensive than traditional workloads because of the vast of amounts of data involved in training new models. Demand from AI data centres is expected to grow more than thirtyfold, from four gigawatts in 2024 to 123 gigawatts in 2035 (Figure 2).
This document may contain “forward-looking” information, such as forecasts or projections. Please note that any such information is not a guarantee of any future performance and there is no assurance that any forecast or projection will be realised.
This presents a challenge: how do we meet this energy demand without driving up emissions? Aside from commitments made by major players (such as Microsoft, Google, and Amazon) to meet this demand with carbon-free energy, in many regions renewable energy sources are also increasingly the most economically attractive source of new power generation1. Although some challenges do remain with the adoption of clean technology, cost competitiveness, policy support and concerns around energy security have driven up demand. Solar energy consumption has grown by nearly 200% and wind energy by 74% in the last five years alone (compared to an average growth of 4% across fossil fuels)2. Yet globally, around four-fifths of energy consumed still comes from fossil fuels2.
Generating stable and scalable energy supplies requires massive investment in low-carbon power, infrastructure, storage, and efficiency technologies, often delivered by companies in emissions-intensive sectors. Climate conscious investors seeking to identify risks and opportunities in these sectors can be constrained by traditional measures of climate impact. Due to a reliance on scope 1, 2 and 3 reporting, much of the capital currently deployed in climate strategies is concentrated in sectors and companies with low reported operational emissions. This can overlook companies in harder-to-abate sectors which may be critical to enabling decarbonisation elsewhere in the economy.
In 2021 we launched our Avoided Emissions Framework to address this gap. We have updated the framework with new data, methodologies, and additional sources of avoided emissions to reflect the evolving transition landscape, providing investors with a more complete view of portfolio-level climate impact. Our findings also indicate that, in theory, a diversified, low-risk global equity portfolio tilted towards exposure to avoided emissions would have outperformed a global equity benchmark by 10% over the past five years (Figure 3).
For more information on the value of avoided emissions analysis and how it can be used to identify investment opportunities, see the full report: Identifying investment opportunities using avoided emissions.
For illustrative purposes only. Past Performance is not a guide to future performance and may not be repeated.
2 Energy Institute (2025) Statistical Review of World Energy
Zapisz się na nasze Analizy
Odwiedź nasze centrum preferencji, gdzie możesz wybrać, które Analizy przygotowane przez Schroders chcesz otrzymywać.
Authors
Tematy