Major asset managers are shifting their AI investment strategy from semiconductor chipmakers to hyperscalers like Amazon and Microsoft. As capital expenditure anxiety fades, investors are betting that major cloud providers will capture long-term profit growth and generate billions in operating cash flow through 2028.
The recent earnings season has fundamentally altered how Wall Street evaluates the artificial intelligence boom. Rather than fixating on whether Big Tech’s massive spending spree will ever yield a return, major investors are pivoting toward the companies best positioned to deliver sustained profits over the long haul, according to market reporting.
Cloud Growth and the Shift Toward Hyperscalers
Cloud growth is accelerating, and persistent capacity constraints continue to define the market. While a sector rout in July cast doubt on semiconductor valuations amid rising Chinese competition and heavy spending concerns, the world’s largest asset managers are aggressively increasing their exposure to hyperscalers—the massive cloud service providers capable of rapidly expanding AI infrastructure.
“The hyperscalers are being recognised in this moment as companies that are likely to be very large beneficiaries of this AI paradigm shift,” said Brian Barbetta, co-head of the technology platform at Wellington Management, which manages about $1.3 trillion in assets. “They remain core holdings in our portfolios, and we’ve in fact increased our positioning in many of these companies recently.”
Brian Barbetta, co-head of the technology platform at Wellington Management
Wellington Management manages approximately $1.3 trillion in assets, making its portfolio adjustments a bellwether for institutional sentiment. Although shares in the four biggest capital expenditure spenders have lagged behind the broader Philadelphia Semiconductor Index and skyrocketing neocloud providers, major fund managers view the temporary lag as an entry point.
Cash Flow Projections and the Case for Amazon and Microsoft
Optimism surrounding hyperscalers is grounded in concrete cash flow expectations. A Reuters analysis estimates that hyperscalers will generate about $340 billion more in annual operating cash flow in 2027 than in 2025, even as capital expenditures rise by roughly $534 billion during the same timeframe.
“By later next year into 2028, we think you’re going to start seeing these companies growing profits and cash flow faster than the incremental capex growth.”
Richard Clode, portfolio manager at Janus Henderson’s Bankers Investment Trust
Clode noted that Amazon represents one of his fund’s largest overweight positions, summarizing the bullish institutional outlook with a concise mantra: Today’s capex is tomorrow’s sales
.
Balancing Semiconductor Chipmakers With Cloud Infrastructure
Despite the rotation into cloud providers, institutional investors emphasize that artificial intelligence should be treated as an expanding ecosystem rather than a zero-sum game between hardware and software.
“It’s not about whether chips are better investments than hyperscalers. It’s about having both in your portfolio.”
John Lamb, equity investment director at Capital Group
Capital Group manages about $3.6 trillion in assets. Lamb pointed out that enterprise data centers typically take 12 to 18 months to transition from active construction to revenue production, noting that latest quarterly earnings mark the beginning of this vital inflection point.
Valuations and the Vulnerability of Neocloud Competitors
Valuations across major hyperscalers have compressed throughout the year, keeping them below post-pandemic peaks. Microsoft currently trades at a forward earnings multiple of about 24.6 times, while Meta trades at the lowest end of the group at 17.6 times.
Meanwhile, neocloud providers such as CoreWeave—which has surged around 50%—and Nebius, up over 200%, have capitalized on elevated spot pricing for scarce computing power. However, analysts warn that this business model carries hidden risks.
BCA Research Chief U.S. Equity Strategist Noah Weisenberger cautioned that neocloud providers could face severe vulnerability if fresh computing capacity comes online and market pricing normalizes, pointing to their heavier reliance on debt and high spot rates. Because hyperscalers have shifted toward a more capital-intensive model, their valuations may remain restrained despite robust earnings, prompting some analysts to recommend a long-hyperscalers, short-neoclouds trading strategy.
Monetization Pressures Ahead of 2028
Even as investor angst over heavy spending begins to dissipate, the road ahead features steep performance hurdles. Swiss wealth manager LGF+ZEST CIO Alberto Conca estimates that artificial intelligence monetization will require a fivefold to thirteenfold increase to fully justify current corporate spending plans.
As the market matures, industry experts expect competition to winnow the field. Companies that successfully control both underlying computing capacity and the software layers required to help customers deploy AI efficiently across multiple models will capture the most enduring competitive advantages.
