Key facts
- Investors are focusing on long-term AI winners as Big Tech's infrastructure spending proves robust.
- Hyperscalers like Amazon and Microsoft are seen as major beneficiaries of the AI paradigm shift.
- Neocloud providers, which rent computing power, have capitalized on scarce AI capacity.
- Analysts suggest a long-hyperscalers, short-neoclouds trade.
- The field of AI winners is expected to narrow as the market matures.
Investors are increasingly focusing on identifying the long-term winners in the artificial intelligence space, shifting their attention from the immediate infrastructure build-out by Big Tech companies. Recent earnings reports from Microsoft and Amazon have reassured markets about the sustained demand for AI-related infrastructure, despite ongoing capacity constraints.
Asset managers are maintaining significant positions in semiconductor stocks while also increasing exposure to hyperscalers, such as Amazon and Microsoft, who are seen as major beneficiaries of the AI paradigm shift. Brian Barbetta, co-head of the technology platform at Wellington Management, noted that hyperscalers are core holdings and their positioning has been increased recently.
While shares of the largest AI capital expenditure spenders have lagged behind the broader semiconductor index and some neocloud providers like CoreWeave and Nebius, analysts believe hyperscalers will eventually see their investments translate into faster profit and cash flow growth. Richard Clode, portfolio manager at Janus Henderson's Bankers Investment Trust, stated that today's capital expenditure is tomorrow's sales, with an expectation of accelerated growth by late 2027.
Investors are encouraged to view AI as an expanding ecosystem rather than a dichotomy between chipmakers and cloud providers. John Lamb, equity investment director at Capital Group, emphasized the importance of having both in a portfolio, noting the typical 12-to-18-month lag for data centers to become revenue-generating. Companies that control computing capacity and offer efficient AI deployment layers are expected to gain a competitive edge.
Despite compressed valuations this year, hyperscalers like Microsoft and Meta are considered to have more lasting advantages than neocloud providers due to their scale and customer relationships. However, neocloud providers may face vulnerability if new computing capacity eases pricing pressures. Noah Weisberger, Chief U.S. Equity Strategist at BCA Research, recommends a long-hyperscalers, short-neoclouds trade.
Challenges remain, with estimates suggesting AI monetization needs a substantial increase to justify current spending. Experts anticipate that competition will eventually lead to fewer dominant AI winners, with companies possessing broad technology portfolios, deep customer relationships, and control over their infrastructure likely to emerge ahead of more specialized rivals.