The AI trade’s second act: looking beyond the Magnificent Seven

Capital at risk. The value of investments and the income from them can fall as well as rise and are not guaranteed. Investors may not get back the amount originally invested. 

After several years of debate over whether AI’s potential was being overstated or whether we were witnessing the emergence of a genuinely transformational technology, it is now increasingly clear that large language models (LLMs) represent a meaningful leap forward. Recent advances have demonstrated capabilities that were difficult to envisage only a few years ago.

Despite the attention AI has attracted across industries, its economic benefits have been relatively concentrated on a handful of tech hyper-scalers, the so-called ‘Magnificent 7’. However, as adoption of the technology accelerates, AI’s impact on global capital markets is likely to be broader than just a bump in Nvidia’s share price.

Investment returns from AI have been focused on a small number of US technology companies

Net total return in USD: Jan ‘23 – July ‘26

The above relates to past performance. Past performance is not a reliable indicator of future returns.
Index performance returns do not reflect any management fees, transaction costs or expenses. Indices are unmanaged and one cannot invest directly in an index. Source: BlackRock and Refinitiv, 31/12/2022 – 23/07/2026. Global Equity represents MSCI All Country World Index in USD.

The widespread usage of artificial intelligence represents a positive supply-side shock to the global economy, boosting worker productivity and therefore supporting stronger economic growth. This view is increasingly reflected in both policy and market discussions. For example, BlackRock Investment Institute has argued that AI could materially boost US productivity and economic growth.

The second order and less frequently considered effect could be structurally higher interest rates. While stronger economic growth has yet to materialise in the real-world data, there are fledgling signs of change, namely increased worker productivity in high-skill and service sectors.

In our view, the consequences for portfolios are reasonably clear to see. Equity markets become more appealing, boosted by stronger economic growth. At the same time, government bonds become less appealing in a higher interest rate environment. 

Of course, AI could prove sufficiently disinflationary to support fixed income returns, which is plausible, albeit only over the long run. However, our view is that the reduction in price pressure will not be sufficient to offset the rate-raising force of higher economic growth.

While the above might sound fairly simple, at a more granular level, the impact of AI is likely to be highly heterogeneous. It will materially benefit certain countries, and to a greater extent sectors, while seriously undermining others. In navigating this complex environment, we see three areas where AI adoption is most likely to create differentiated investment opportunities.

1. Increased energy consumption

Artificial intelligence is nothing if not power hungry. Beyond the obvious concerns this presents for portfolios, like higher energy prices creating bond-harming inflation, there are ways investors can benefit from this surge in demand for electricity.

Going forward energy consumption is expected to spike

Global data centre electricity consumption (trillions of Kilowatt hrs)

Between the available and incumbent electricity generation technologies, it is unclear exactly how the spike in demand is going to be met. Thus, we would argue against a narrow focus on any one power source. Instead, we see an attractive opportunity in listed infrastructure equity. This asset class is set to benefit from the investment required across electricity generation, transport and storage, irrespective of which generation technologies ultimately meet demand.

Infrastructure has arguably been structurally under-owned by investors given that it does not cleanly fit into either the bond or equity portion of portfolios. This gives it a rare advantage as an AI trade: being attractively valued.

2. Demand for advanced hardware

The technology required to run the frontier LLMs is specialised, complex, and (currently) highly supply constrained. Investment returns will not remain concentrated just in the creators of models, but cascade down to the ‘picks and shovels’ enablers of their initial training and ongoing running.

The price of semiconductors is rising, while output growth has slowed over the past two years

Output of semiconductors and other electronic components

There is currently an acute mismatch between demand and supply of leading-edge AI computing hardware, generating excess profits for chip manufacturers. We are capturing this opportunity by running an elevated exposure to Emerging Market equities, which have a significant allocation to companies deeply embedded in the AI infrastructure supply chain. These excess returns are unlikely to persist over multiple years, but current market conditions continue to support the investment case for now.

Emerging Markets remain reasonably cheap, even despite the recent run-up in their prices. Furthermore, by expressing this view through an exposure to the whole emerging market complex, we somewhat mitigate the high levels of concentration in specific countries, such as Taiwan and Korea.

3. A persistent incumbent advantage

Training, maintaining and distributing cutting-edge AI models is a highly capital-intensive enterprise requiring significant upfront investment. Additionally, companies are increasingly using existing models to train their new generation. All of this suggests that AI leaders will quickly become entrenched and prove difficult to unseat. Moreover, national strategic priorities may reduce the likelihood of aggressive regulatory intervention in AI-related industries for both the US and China.

It is not difficult to imagine a world in which a handful of large AI players each carve out a fiefdom, competition between large players remains scarce, while acquisitions keep a lid on disruption from new entrants. Part of what makes this future so plausible is the direct parallels between it and the state of play with current tech leaders, where a small number of firms dominate cloud computing, digital advertising and smartphone hardware.

This motivates us to remain focused on large-capitalisation companies. While the benefits of the AI roll-out might trickle down to smaller companies, the existing leaders will likely be able to maintain material pricing power and thus retain a significant portion of the excess profits generated by AI.

Overall, for all investors, the challenge is shifting from identifying AI beneficiaries to identifying where economic value ultimately accrues. If history offers any lessons from previous technological shifts, it is that the largest rewards do not always materialise where the initial excitement is most intense. Thus, as the AI build-out moves from promise to implementation, we should be prepared for the opportunity set to evolve with it: broader in some areas, more concentrated in others, and increasingly shaped by the real-world constraints of power, hardware, capital and competition.

Important Information

Capital at risk. The value of investments and the income from them can fall as well as rise and are not guaranteed. Investors may not get back the amount originally invested.

This material is for distribution to Professional, Qualified Clients and Investors only.

In the UK and Non-European Economic Area (EEA) countries (excluding Switzerland): This is issued by BlackRock Investment Management (UK) Limited, authorised and regulated by the Financial Conduct Authority. Registered office: 12 Throgmorton Avenue, London, EC2N 2DL. Tel: + 44 (0)20 7743 3000. Registered in England and Wales No. 02020394. For your protection telephone calls are usually recorded. Please refer to the Financial Conduct Authority website for a list of authorised activities conducted by BlackRock.

The above relates to past performance. Past performance is not a reliable indicator of future returns.
Index performance returns do not reflect any management fees, transaction costs or expenses. Indices are unmanaged and one cannot invest directly in an index. Source: BlackRock and Refinitiv, 31/12/2025