Buy
Market
🔥
Prediction Market

UBS views AI selloff as repricing, not unwinding; backs diversification

UBS says AI selloff is a positioning reset, not a collapse, and advises staying invested with diversification across sectors, regions, and bonds.

19/11/2026 00:1118 min read

According to UBS, the recent decline in AI-related stocks represents a repricing of positioning and valuations rather than the beginning of a collapse in AI investment. The bank advises staying invested while diversifying across sectors, regions, and fixed income, focusing on utilisation and returns instead of chip orders. UBS maintains a constructive outlook on equities over a six-to-12-month horizon.

The bank projects AI capital spending to reach $1.0 trillion in 2026, $1.4 trillion in 2027, and $1.5 trillion in 2028, with the majority coming from well-funded major platforms. New US chip restrictions could impact 5% to 10% of next year's AI demand, though the revenue effect on US chipmakers should be limited, the bank said.

Factors that could alter UBS's view include weakening compute demand, slowing AI revenue growth, poor utilisation, tighter financing for large platforms, or selling that spreads into cyclical sectors. The bank favors tech platforms with genuine AI revenue and strong balance sheets, as well as US health care and industrials, eurozone equities, and high-quality short- and medium-dated bonds.

Since the note was issued, Nvidia shares have fallen to a 16-month low despite beating forecasts, and four Federal Reserve officials have taken hawkish stances. The broader S&P 500 has declined for five consecutive sessions.

A rough week for the AI trade

The note followed a severe sell-off on Monday triggered by US restrictions on offshore data centre purchases of American chips and a cautious speech from Fed Chair Kevin Warsh about AI investment risks. Pressure has continued, with Nvidia shares dropping about 3% on Wednesday to a 16-month low, even after beating revenue expectations of approximately $116 billion and guidance of around $130 billion for the current quarter. Investors focused on the leadership gap created by Jensen Huang's resignation and a reported SEC investigation.

Four Fed officials—Neel Kashkari, Christopher Waller, Mary Daly, and Michael Barr—all sounded hawkish on Wednesday, with Waller again suggesting a possible 50 basis point rate hike. Higher borrowing costs weigh most heavily on leveraged infrastructure builders in the AI space.

Why UBS isn't calling the top

UBS contends that demand for computing capacity remains strong, citing accelerating cloud revenue and rising advance orders. The bank forecasts total AI capital spending of $1.0 trillion in 2026, $1.4 trillion in 2027, and $1.5 trillion in 2028, mostly from large, well-funded technology platforms. This should limit damage if weaker, debt-heavy infrastructure providers face difficulties. Delays, repricing, and consolidation appear more likely than a broad spending collapse.

On new chip restrictions, UBS estimates the rule could affect 5% to 10% of next year's AI demand but expects only a limited revenue hit for US chipmakers, given tight supply and the ability to sell displaced capacity elsewhere. The greater risk, UBS said, is added pressure on already elevated 2028 spending and earnings expectations. Smaller cloud and regional data centre operators and standalone Chinese AI labs look most exposed, while major hyperscalers seem relatively insulated.

What to watch in AI

The bank's message is that the bar has been raised. With cumulative AI investment reaching trillions, investors now evaluate the build-out based on revenue, free cash flow, utilisation, and sustainable returns, not just chip demand. Signs that new computing capacity is being absorbed by rising usage would support confidence. Weaker utilisation, slowing customer commitments, or difficulty converting investment into revenue could revive overcapacity fears.

UBS also outlined conditions that would change its view: weakening compute demand, slowing AI revenue growth, poor capacity use, or tighter financial conditions forcing large platforms to cut back. Breadth is another indicator. Selling concentrated in chips, data centre names, and richly valued growth stocks fits a valuation adjustment. Selling that spreads into financials, industrials, and other cyclical sectors, along with weaker economic, credit, or earnings data, would suggest something broader. In UBS's view, the evidence does not yet support that conclusion.

Where UBS sees opportunity

Rather than turning defensive, UBS recommends selectivity. It favors established technology platforms with real AI revenue, existing computing capacity, strong balance sheets, and the ability to fund investment internally. It suggests investors with outsized positions in single tech names use rallies to reduce concentration. Within the US, UBS's preferred sectors are health care and industrials.

Outside the US, the bank is positive on eurozone equities, where valuations are lower and earnings are supported by a manufacturing recovery and structural investment. It also advises investors who are significantly underweight bonds to rebuild exposure to short- and medium-dated high-quality debt, where elevated yields offer income and diversification. Broad commodities and selected alternatives round out a multi-asset core designed to keep investors in the market while mitigating further tech volatility.

Share to

Disclaimer: this article comes from third-party media and is provided for reference only. It does not constitute investment advice. Crypto and other financial products carry significant price volatility risk, so please make your own decisions carefully.

Related articles