The street disagrees with the money in 6 places.

The street says

Goldman Sachs Research argues power is the defining constraint on AI buildout, with utility capex hitting $1.3tn by 2030 and Tyler Miller calling it 'the forever bottleneck'.

The street says

JPMorgan's Noah Wintroub describes the $90 billion Alphabet equity raise as a historic confidence signal for AI infrastructure spending, with Berkshire's participation validating the sector's long-term capital cycle.

The street says

Allen Chang at GS Greater China Technology Research says China's semiconductor industry is advancing faster than expected, with local semicap equipment makers guiding for ~40% year-on-year revenue growth over the next few years.

The street says

Morgan Stanley bankers Kalli Dircks and James Talbot say 2026 biotech M&A is the most robust in many years, with capital flowing to companies with clear clinical data in oncology, I&I and rare disease, and global healthcare M&A volumes hitting ~$4.7 trillion.

The street says

Goldman's Carly Davenport and Brian Singer see US data center power demand hitting ~108 GW by 2030, global demand up ~170% from 2025, and hyperscaler capex reaching $2.1T by 2029.

The street says

Gregory Shearer at JPMorgan says copper's medium-term outlook remains supportive, underpinned by tight mine supply and structurally growing electrification and data center capex demand.

The street says

Goldman's Rich Privorotsky says the crowded AI/semis/hyperscaler trade is actively unwinding due to rising cost of capital, power constraints, and model-layer competition, with last week marking the biggest-ever tech selling in GS's prime book.

The street says

Peter Callahan at Goldman Sachs argues the AI trade has broadened beyond semis and hyperscaler CapEx into infrastructure software, security, and data infrastructure, creating a stock picker's market rather than one thematic bet.

The street says

Morgan Stanley Research argues the next wave of AI value will move from infrastructure builders like chips and cloud to enterprise adopters embedding AI into workflows to lift productivity and margins.

The street says

Morgan Stanley Research argues investors should rotate toward AI adopters whose improving fundamentals are not yet reflected in expectations, while continuing to own select enablers.

The street says

Goldman's Jonathan Shugar sees AI picks-and-shovels and hyperscaler CapEx as a temporary air pocket with good runway, expecting ~$1.2T in hyperscaler CapEx next year and no Fed hike deterring the strongest-balance-sheet companies.

The street says

Goldman Sachs CEO David Solomon says AI deployment across enterprise, government, and applications will take 10-30 years to fully appreciate, with current compute demand exceeding supply and long-term durability not in doubt.

The street says

BofA analysts Haim Israel, Felix Tran, Savita Subramanian and Martyn Briggs argue private capital is structurally ascendant, with AUM hitting $22tn and PE outperforming the S&P 500 by ~6pp per year over ten years.

The street says

Morgan Stanley Research says the enterprise identity security market could more than double to $60 billion in two years, driven by agentic AI demand, and sees unified-platform incumbents like OKTA as best positioned.

The street says

Morgan Stanley Research and Investment Management see private markets AUM growing from 11% to 13% of global AUM by 2029, with durable demand across infrastructure, PE and secondaries, and private credit return potential more attractive today than at the start of 2026.

The street says

Goldman's Eric Sheridan argues consumer AI agents mark the emergence of a new platform layer driving action-oriented agentic commerce, with hyperscalers deploying $1.4 trillion in capex by 2027 to meet a supply-demand compute imbalance.