# AI dip-buying faces caution as data center costs and funding strains mount

By Riza Dagoc

Canonical URL: https://www.tokenpost.com/news/investing/39729
Published: 2026-10-24T06:40:34.000Z
Updated: 2026-10-24T06:40:34.000Z
Section: Investing

> Vermillion Research’s CIO cited rising data center costs and financing strains as reasons for caution on AI dip-buying, while highlighting stronger technical patterns in software stocks.

Falling AI stocks may offer limited grounds for bargain hunting as data center construction costs and financing pressures weigh on the sector. Within technology, software shares show relatively stronger technical patterns than semiconductor stocks.

In an October 23 interview with David Lin, David Nikoski, chief investment officer of Vermillion Research, said money was rotating out of technology into other sectors, leaving AI stocks under pressure.

Nikoski identified October 2 as the turning point in technology’s relative strength. Almost every other sector had subsequently outperformed technology, he said, while selling in major technology companies had weakened chart patterns across AI stocks. He cited widening credit spreads at CoreWeave as another sign of investor concern.

He distinguished a possible rebound in CoreWeave from its appeal as a long-term holding, saying a bounce would represent a short-term trade. His technical concerns included a descending triangle on the longer-term chart and a break below its lower boundary.

Weakness among data center builders provided another warning sign. Nikoski cited engineering and construction stocks that had fallen 50% to 60% from recent highs, suggesting potential difficulties with financing, approvals or access to electricity. He viewed those declines as grounds to question the pace of AI infrastructure expansion.

Rising costs were also squeezing construction profitability. Citing Sterling’s earnings release, he said higher pay for electricians and plumbers, alongside specialty steel costs, was sharply compressing margins. Opposition to data center locations and constraints on electricity supply added further obstacles: completed buildings could remain idle until power became available.

Nikoski interpreted sideways semiconductor trading as a potential topping pattern. Although an oversold rebound remained possible, he pointed to high-yield credit spreads of 351 basis points at the previous close as evidence of pressure on stocks sensitive to market swings. By comparison, he said HubSpot, Monday.com and Adobe displayed more compelling technical patterns than SMH or the SOX index.

A cooling investment cycle would not mean AI technology itself was disappearing, he added. Drawing parallels with railroads, telecommunications and the internet, Nikoski distinguished the persistence of a technology from the booms and busts associated with financing its expansion.