Jim Cramer's approach to rallying AI stocks, plus updates on our breakup plays

AI and data-center-related stocks are leading the market higher, with Micron, Intel, AMD, Qualcomm, Nvidia, and Corning all benefiting; Corning is up more than 9% and has rallied 130% year to date. Honeywell rose 2.7% after filing an S-1 for the proposed IPO of Quantinuum, while Qnity Electronics gained 4.5% ahead of earnings and is up 89% year to date. The commentary is broadly constructive on data-center demand and portfolio simplification, but it is mostly stock-specific rather than market-moving.

Analysis

The tape is still rewarding the same scarce resource: near-term AI infrastructure capacity. That’s not just a semis trade; it’s a second-order capex trade that bleeds into optical components, power management, thermal solutions, and data-center adjacent industrials. The market is effectively repricing the probability that hyperscaler budgets stay elevated for multiple budget cycles, which means the losers are not the obvious AI laggards but any supplier with slower operating leverage or less direct AI exposure that is being treated like a generic industrial.

Corning’s move is the clearest tell that investors are willing to pay up for bottleneck assets rather than pure-chip beta. The risk is that these names become self-financing momentum trades: once index and factor flows chase them, near-term upside can outpace fundamental revision, but that also creates air pockets if capex commentary merely meets rather than beats. The more crowded the trade becomes, the more a single hyperscaler budget pause or digestion period can compress multiples across the group in weeks, not quarters.

Honeywell and Qnity point to a different but related theme: capital is rewarding corporate simplification and monetizable hidden assets. In both cases, the market is signaling that breakup optionality can matter more than current earnings quality, especially when the retained business has clearer strategic value than the consolidated story. That means the opportunity is less about chasing the headline pop and more about owning the residual asset after the easy de-rating has already happened.

The contrarian read is that the market may be underpricing execution risk on “clean-up” stories while overpricing perfect continuity in AI spend. Semis can keep running, but the better risk/reward may now sit in suppliers one step removed from the obvious winners, where earnings revisions lag price moves. If AI capex remains strong but less explosive, the next leg should favor infrastructure bottlenecks and restructuring beneficiaries over the highest-beta chip names.

Key Decisions for Investors

• Trim overextended direct AI beta into strength: reduce NVDA/AMD/INTC/QCOM exposure on a 1-2 week horizon and rotate toward infrastructure suppliers with lower crowding risk; use any intraday pullback in GLW to add only if volume confirms continuation.
• Long GLW vs short a basket of mature industrials over the next 1-3 months: the pair captures AI infrastructure scarcity while hedging away broad market multiple compression; target 10-15% relative outperformance if hyperscaler capex commentary stays firm.
• Hold HON into the breakup catalyst but avoid chasing the gap: use June event-driven strength to monetize a portion of the position; if the market prices in too much simplification too quickly, upside/reward becomes asymmetric to the downside on execution hiccups.
• For Q, consider a tactical long into earnings only if implied move is still below the stock’s recent realized volatility; otherwise sell downside put spreads to express bullishness with defined risk over the next 2-4 weeks.
• Avoid initiating fresh longs in DELL and MNDY until there is evidence they can convert AI/automation enthusiasm into estimate revisions; these are the types of names that often lag in a theme-led tape and are vulnerable to valuation compression.