Chip Stocks Lost $2 Trillion in 8 Days, Testing Investor Faith in the AI Boom

Chip stocks erased about $2 trillion in eight sessions, their longest losing streak since 2008. Yields above 5% and doubts that artificial intelligence (AI) spending will pay off weighed on the group.

The Philadelphia Semiconductor Index, a benchmark of US-listed chipmakers, bounced on Monday. However, CNBC’s Investment Committee panel split on whether that marks a buying chance or a pause in a larger unwinding.

Are Rates, Not Demand, Driving the Chip Stocks Selloff?

The index fell more than 11% last week, capping eight straight losing sessions.

Meanwhile, an inflation reading above forecasts has raised the odds of Federal Reserve rate hikes. Minutes from the Fed’s September meeting showed most officials saw another 2026 rate hike as likely appropriate.

The 10-year Treasury yield stood at 5.10% at Monday’s open, according to Schwab data.

One panelist said neoclouds, debt-funded firms that rent out AI computing power, are feeling the strain of high yields.

The panelist said yields 0.5 to 0.75 percentage points lower would ease that stress.

In contrast, another panelist welcomed hedge funds trimming chip exposure, noting that November and December are seasonally favorable for stocks.

However, a skeptic said last week’s broadening favored healthcare and energy rather than industrials such as Caterpillar.

Is the AI Spending Peak One Year Away or Four?

Panelists disagreed on when AI spending will peak, with estimates ranging from one year to four years away.

One panelist cited memory chipmaker Micron’s share buybacks and take-or-pay contracts, under which customers pay for chips regardless of usage.

Another panelist countered that such contracts would offer little protection once spending peaks, as in past cycles.

Earnings now offer the next test. Lam Research, a chipmaking equipment supplier, reports this week alongside Texas Instruments, Western Digital, Meta, Tesla and SpaceX.

One panelist said the key is returns on AI spending, not assurances that spending will continue.

Therefore, with the 10-year yield above 5%, investors judge each AI earnings report against the cost of money.

That bar applies to every AI-linked stock, from chipmakers to the Magnificent Seven, the group of largest US tech companies.