Investing.com — Investors are reassessing artificial intelligence investments as surging infrastructure spending could weigh on earnings growth and valuation multiples, even as demand for AI services remains robust, according to ING.

The long-term outlook for large technology companies remains positive, driven by growing adoption of AI services. Still, rising capital expenditure is expected to increase depreciation costs, slow earnings per share (EPS) growth and reduce the capacity for share buybacks, which have helped support equity valuations.

Technology stocks have experienced sharp swings this year. During the first half of 2026, Microsoft (NASDAQ:MSFT) fell 20% and Oracle (NYSE:ORCL) dropped 27%, while Alphabet (NASDAQ:GOOGL) gained 14%. Much of the volatility has been attributed to uncertainty over future returns on AI-related investments.

Current spending levels remain economically justified, with Microsoft investing about $65 billion in cloud and AI infrastructure in fiscal 2025 while generating annualised AI revenue of roughly $37 billion. Demand for AI computing continues to exceed available capacity, and most large technology companies can finance these investments through operating cash flow.

Rising capital expenditure is expected to reduce free cash flow and limit share repurchases, particularly at Microsoft, Alphabet and Meta Platforms (NASDAQ:META). Slower buybacks, combined with higher depreciation costs, could make investors less willing to pay premium earnings multiples if AI revenue takes longer to materialise.

Oracle (NYSE:ORCL) stands out as a higher-risk case. Its investment programme, including support for Project Stargate, could pressure cash flow and increase funding needs. Nvidia (NASDAQ:NVDA) may also face stronger competition over time as Microsoft, Alphabet and Amazon (NASDAQ:AMZN) develop their own AI chips to improve infrastructure efficiency.

Despite those risks, current AI spending is viewed as fundamentally sound. The key question for investors is whether future revenue growth, profit margins, and earnings expansion will ultimately justify today’s level of investment.

Related articles

Why tech investors are reevaluating AI investments

Apple sues OpenAI over alleged trade secret theft

SK Hynix CEO sees even worse memory crunch in 2027, demand above supply thru 2030

Investing.com — Morgan Stanley revealed in a note on Friday that it has upgraded Lenovo to Overweight from Equal-weight and more than doubled its price target to HK$30 from HK$14.20, arguing that AI-driven demand has fundamentally altered the memory market in a way that materially benefits the Chinese technology giant.

Analyst Howard Kao said the key insight is that “AI has fundamentally altered the memory market,” creating a structural supply constraint that differs from previous upcycles.

Unlike prior cycles when customers delayed purchases in anticipation of lower component prices, customers now “increasingly expect memory costs to remain elevated and have shown a greater willingness to absorb higher system prices,” according to the bank.

This has enabled Lenovo to “fully pass through higher component costs while preserving margins — an outcome that differs materially from previous memory cycles.”

Lenovo shares have already surged 82% over the past two months versus a 9% decline in the Hang Seng Index, driven by stronger-than-expected earnings.

Morgan Stanley noted that its latest supply chain checks and management discussions suggest this trend “could continue into at least the second half of calendar 2026.”

Beyond the PC business, which remains “a stable cash generator,” Morgan Stanley highlighted Lenovo’s Infrastructure Solutions Group as increasingly central to the investment case.

With an AI infrastructure pipeline of approximately $21 billion, the firm forecasts ISG will contribute approximately 35% of group profits by fiscal year 2029, versus near breakeven in fiscal year 2026.

Morgan Stanley’s fiscal year 2027-2029 EPS forecasts are approximately 20% above consensus, driven primarily by stronger margin assumptions.

Related articles

Morgan Stanley more than doubles Lenovo price target on AI-driven growth

Delta Air Lines earnings and guidance beat expectations, shares slide

Barclays upgrades CubeSmart, cuts Public Storage on relative valuation

Investing.com — The sustained outperformance of semiconductor stocks relative to hyperscalers may be difficult to maintain, according to JPMorgan, which said the gap in performance between AI chip makers and cloud providers “appears somewhat unsustainable in the long run.”

In a note to clients on Thursday, analyst Nikolaos Panigirtzoglou outlined two scenarios under which the performance gap could narrow.

In the positive scenario, hyperscalers, AI model providers and end-users begin to see improvements in monetization, revenues and earnings, allowing them to “catch up, capturing a bigger share of the overall AI value-added pie.”

In the negative scenario, semiconductor outperformance comes at the direct expense of customers such as hyperscalers and AI model providers, which could “start to depress capex intentions” and “eventually act as a headwind to demand for the semiconductor companies’ products.”

JPMorgan said its house view favors the more positive outcome. However, the bank noted that the consensus among bottom-up analysts points to “a sharp deceleration in hyperscalers’ capex growth from next year onwards,” which it said, “taken at face value would tilt towards the negative scenario.”

Semiconductor stocks, including AI chip and memory makers, have outperformed hyperscalers almost continuously since last September, a trend JPMorgan now views as a potential source of vulnerability for the sector.

On broader macro and crypto markets, JPMorgan said U.S. money creation is on track to increase from $1.6 trillion in 2025 to $1.8 trillion in 2026, and flagged that MicroStrategy has “introduced avoidable two-way risk into crypto markets inducing more uncertainty and volatility.”

Related articles

AI chip winners may face tougher road ahead as hyperscalers catch up: JPM

Goldman’s Garrett sees investors dialing back tech exposure into H2

Kepler Cheuvreux starts GL Events at “buy,” sees 17% upside

Investing.com — Nomura said the AI infrastructure investment cycle has likely not yet peaked despite a recent pullback in semiconductor shares, arguing that supply constraints and hyperscaler spending trends point to further upside ahead.

Analyst Aaron Jeng told investors in a note on Tuesday that the Philadelphia Semiconductor Index (SOX) has surged 85% since the firm’s last cycle update in March and 211% since Nomura revisited the AI theme in May 2025, before recently pulling back.

Jeng said “a pullback is healthy following such a surge over such a short period,” pointing to risks including component supply mismatches, hyperscalers’ 2027 free cash flow pressures, and macro risks tied to a rising yield trend.

Even so, Nomura said it does not believe the cycle has peaked, citing hyperscalers’ spending that “may need to show upside further into 2027” despite insufficient free cash flow, driven in part by surging memory costs.

The firm’s proprietary global data center build tracking also points to further upside from its March estimates.

On the supply side, Nomura said the two-year timeline for greenfield data center construction beginning in late 2025 suggests insufficient capacity heading into 2027, with the supply bottleneck likely shifting from larger players such as TSMC to smaller component makers.

Jeng added that price hikes and ongoing earnings upward revisions “would still be the biggest catalysts,” adding that Nomura “would still be buyers into weakness.”

Related articles

Where are we in this AI infrastructure investment cycle?

SanDisk stock target lifted at Bernstein to $3000 on new memory LTAs

Chip stocks hit record 19.7% of S&P 500, quadrupling since 2020

Investing.com — A surge in artificial intelligence-related capital spending is unlikely to undermine share buybacks across the broader U.S. equity market, according to a Deutsche Bank strategy note, which said record corporate earnings continue to support shareholder returns despite a sharp rise in investment spending.

Capital expenditure across the S&P 500 has climbed from an annualized pace of roughly $1 trillion to about $1.5 trillion over the past two years.

Around two-thirds of that increase has come from five hyperscalers: Amazon (NASDAQ:AMZN), Microsoft (NASDAQ:MSFT), Alphabet (NASDAQ:GOOGL), Meta Platforms (NASDAQ:META) and Oracle (NYSE:ORCL).

Those companies have sharply reduced share buybacks as they channel more cash into AI infrastructure and data centers while also raising fresh capital. Alphabet’s recently announced secondary issuance alone exceeds total S&P 500 secondary issuance recorded during the first quarter.

The spending, though, is also driving strong earnings growth among companies supplying AI infrastructure. Semiconductor makers, technology hardware firms, engineering companies, utilities, independent power producers, and data center REITs have all begun increasing their own share repurchases.

Beyond those sectors, the remainder of the S&P 500 continues to generate the bulk of corporate earnings and buybacks. Quarterly net repurchases by this group have increased by nearly 30% over the past year, with further growth expected alongside earnings.

Overall, S&P 500 net buybacks reached a record $270 billion during the first quarter, with gross buybacks climbing to about $300 billion. Current buyback levels remain broadly in line with historical relationships to both earnings and market capitalization.

Attention has also turned to the recent jump in IPOs and secondary offerings. Equity issuance has accelerated sharply during the second quarter, with several additional large listings expected in the coming months.

Even though the higher supply of shares could reduce net buybacks, previous issuance waves have typically coincided with strong investor demand and positive equity market returns.

Rising inflows into U.S. equities, elevated household cash balances, and continued earnings growth are expected to provide support despite the increase in new share issuance.

Related articles

Capex boom threatens to crowd out buybacks, key equity demand driver

Hong Kong share sales hit five-year high as AI boom fuels fundraising

Why BMW’s transformation story hinges on NEUE KLASSE execution

Investing.com — Bernstein said Google Cloud has emerged as the leading AI cloud platform after a strong first quarter, with accelerating cloud growth and rising profitability helping it pull ahead of rivals in incremental cloud revenue despite continued heavy spending on AI infrastructure.

Google Cloud posted 63% year-over-year revenue growth in the quarter, outpacing Amazon Web Services, Microsoft Azure, and Oracle Cloud. The business also generated the largest increase in incremental cloud revenue, supported by strong demand for Gemini models, AI services, and enterprise cloud workloads.

Operating margins expanded to about 33%, with improving profitability helping Google gain ground as enterprises increased spending on AI infrastructure and cloud services.

Microsoft remained one of the strongest long-term AI beneficiaries, with Azure growing about 40% and annualized AI revenue surpassing $37 billion. Demand for Copilot, GitHub, and Azure AI services continued to drive growth, although capacity constraints remain a challenge.

Amazon Web Services continued to post solid momentum, with cloud revenue rising 28% and backlog expanding to $364 billion. Triple-digit AI demand, Trainium chip adoption, and enterprise workloads continued to support growth despite supply constraints.

Oracle remained the fastest-growing major cloud provider, with Oracle Cloud Infrastructure revenue climbing 93% year over year. Remaining performance obligations rose to $638 billion as AI infrastructure demand accelerated.

Across the sector, AI infrastructure investment continues to rise sharply. Combined capital spending by the largest hyperscale providers is projected to exceed $620 billion this year as companies race to expand data center capacity and meet surging AI demand.

The report said capacity constraints have shifted from GPU availability to powered data center capacity, making deployment speed and infrastructure availability increasingly important competitive advantages in the AI race.

Related articles

How did the hyperscalers do in 1Q26?

UBS sees security and safety market at $1.19tn by 2029, led by cybersecurity

SpaceX closes $25 billion bond offering across five tranches

Investing.com — Morgan Stanley upgraded Qualcomm to Equal-weight from Underweight and increased its price target to $231 from $146 in a note to clients on Thursday after the chipmaker’s investor day revealed a $5 billion data center revenue forecast for fiscal 2027 that the bank said “clearly pushes them into the AI beneficiaries category.”

In a note by analyst Joseph Moore, Morgan Stanley acknowledged it had been wrong to remain skeptical of Qualcomm’s AI ambitions.

“We have been wrong to be skeptical,” Moore wrote, adding that management’s $5 billion revenue forecast for next year is “at least 2x higher than expected.”

The bank was more cautious on Qualcomm’s longer-term target of $15 billion in data center revenue by fiscal 2029, describing it as “more aspirational.”

Morgan Stanley also flagged execution risk around the accelerator opportunity and server CPUs, noting that a mid-2028 entry into the CPU market “may not meet with the reception that it would have today” given rapidly expanding supply from competitors including NVIDIA, AMD, Intel and cloud providers’ custom silicon.

Moore stopped short of an Overweight rating, citing better value elsewhere. “We simply see better value from the dominant players vs more speculative new entrants, and we remain mindful of smartphone headwinds,” he wrote.

Despite lingering questions on the product roadmap, Morgan Stanley said Qualcomm’s credible management team and year-to-date underperformance relative to other AI winners made the upgrade difficult to resist, particularly in “the current narrative driven market.”

https://ca.finance.yahoo.com/news/morgan-stanley-upgrades-qualcomm-5b-133208098.html