Key takeaways
- AI remains the dominant market driver: Continued investment in artificial intelligence (AI) infrastructure is supporting economic growth, corporate earnings and investor risk appetite, with demand broadening beyond chatbots towards more advanced 'agentic' AI applications.
- A higher-inflation world may be here to stay: Fidelity's investment experts believe central banks are increasingly comfortable with inflation closer to 3% rather than 2%, creating a backdrop that favours selective risk-taking and real assets such as commodities.
- Selectivity is becoming more important: While opportunities remain across equities, commodities and parts of fixed income, not all AI-related spending will translate into returns, making company fundamentals and valuation discipline critical for investors.
The remainder of 2026 is likely to be defined by continued AI capital expenditure (capex) spend supporting global markets and keeping inflation higher.
Salman Ahmed, Fidelity’s head of macro and strategic asset allocation, explains that “three is the new two” – that is, most central banks are no longer aiming to achieve their proclaimed 2 per cent inflation target, but are happy settling closer to 3 per cent as the capex cycle pushes on.
As a result, inflation is likely to prove the buzzword for the next six months (alongside AI) – a shift from the start of the year when the focus was on growth. But jobs data has improved since then and the market is now expecting rate hikes more than it expects cuts. The consumer broadly remains strong, and that is generating cash flow for the hyperscalers that are providing most of the AI capex.
This environment means Fidelity’s investment teams are happy to take on risk.
AI growing up
The AI capex cycle itself has also evolved. Niamh Brodie-Machura, CIO for equities, explains ’agentic AI’ has become the dominant narrative: “We’ve gone from the LLM chatbot-style single query that all of us have become used to, to actual process management.” This has broadened out demand for AI and is putting ever-greater strain on supply capacity.
For markets to remain this buoyant, demand for AI needs to stay at these levels. Earnings momentum is helping, and Brodie-Machura is not overly worried about that momentum coming to a halt imminently, but it is clearly a risk. She also addresses two other risks that investors should continue to bear in mind.
First is that capital allocation now is far broader than the eventual pool of winners. So some of that capital will have been misallocated.
Second is how companies monetise the trend over the long term as ethical questions start to press on the market. A growing backlash among young people against how AI is deployed, for instance, is throwing some of those concerns into the spotlight.
Brodie-Machura also stresses the need to distinguish between those companies that have earnings based around solving real-world problems, and the frothier parts of the market where investors are buying expected growth for the decades to come. Many companies are reducing their returns to investors (through share buybacks or dividends) and preferring instead to invest in even more capex given investors continue to reward higher expected returns.
Again, not all of this capital spend will generate those expected returns - being discerning about the companies you back is all the more important in this new environment.
The cross-asset view
One way to mitigate against the risks highlighted above is to broaden your equity allocation. Fidelity’s research teams see potential in financials companies, for which the macro environment is supportive, and in healthcare, where valuations are encouraging and AI should boost innovation.
George Efstathopoulos, a portfolio manager and co-head of multi-asset research, explains that he is positive on emerging markets, especially those parts of Asia benefitting most from the AI capex cycle. He also looks favourably on Japanese equities, particularly mid-caps, owing to renewed wage inflation in the country, which is supporting domestic consumption.
The other obvious place to look remains commodities, for two reasons. The first is that certain metals, like uranium and aluminium, form the literal building blocks for structural themes like AI capex and electrification (datacentre construction and grid regeneration, for instance, are commodity-intensive).
The other reason is that real assets – gold in particular – have historically protected portfolios through inflationary episodes. Gold has struggled recently but this most likely represents an expected consolidation after a strong rally. Many of the underlying drivers supporting the precious metal remain in place.
The fixed income space looks harder to navigate. Yield curves have flattened following the shift in expectations from rate cuts to potential hikes. This leaves long-term credit looking vulnerable. Lei Zhu, head of Asian fixed income, is more positive on short-duration bonds, and in particular the three-to-five-year range of the curve has been a strong performer.
Zhu has also seen opportunities in those sectors benefitting from energy supply disruption, such as utilities. Many companies in this sector have re-rated, which means investors have not only enjoyed positive carry but also tightening spreads.
She also sees idiosyncratic support for Asian investment grade. Spreads have tightened as regional investors have brought surplus US dollars closer to home, putting them into higher-quality local paper. She believes this trade has further to run, with spreads still attractive.
Observations from the road
Fidelity’s research team recently returned from meeting companies at the forefront of the AI charge in Silicon Valley. Niamh Brodie-Machura described some of their findings.
First, analysts see demand for AI broadening out, largely off the back of the shift to agentic models. Second, they anticipate continued bottlenecks in memory and CPU (Central Processing Unit) through 2026 and 2027. Third, they observe companies talking increasingly about “valuemaxxing” – that is to say, companies are thinking more rigorously now about the value they derive from AI and the costs involved, rather than investing blindly into AI’s promise.
“The cost is not an afterthought at this point,” she says, “because it is becoming much more material for corporate management teams.”
