Jim Cramer's top 10 things to watch in the stock market Monday

U.S. stocks are set for a higher open, with Nvidia up nearly 2.5% and Broadcom up 3.3% as chipmakers rebound after Friday's selloff. The article highlights several constructive analyst actions, including higher Oracle and Micron price targets, while Apple, Adobe, and recent IPOs Cerebras and Blackstone Digital Infrastructure Trust are also in focus. Geopolitical risk remains present after Iran and Israel traded strikes over the weekend, but investors are looking past it for now.

Analysis

The key market signal is not just a tech rebound; it is a reopening of the AI capex trade after Friday’s de-grossing. That favors the highest-quality “picks and shovels” names with scarce supply chains and visible pricing power, but it also raises the odds of a short-lived relief rally if investors are simply chasing back the same crowded factor exposure. In that setup, the better expression is not broad long Nasdaq beta, but selective longs in beneficiaries with near-term catalysts and cleaner fundamental leverage than the megacaps.

The Google-SpaceX chip capacity arrangement is strategically important because it implies hyperscalers are still willing to pay for interim compute even when internal silicon roadmaps are in place. That is bullish for NVDA in the near term and also for the optical / interconnect stack, where capacity constraints and AI cluster density keep pushing spend outward. Corning’s move looks more durable than a single deal headline because it reinforces that the bottleneck is shifting from compute availability to data movement, which should support the ecosystem names tied to fiber, networking, and power.

The more interesting underappreciated risk is that some of the biggest positive revisions are arriving after very sharp year-to-date runs, which narrows upside unless earnings explicitly re-accelerate. Micron is the clearest example: a powerful consensus upgrade can still be a sell-the-news event if guidance confirms demand strength but not enough to justify the multiple reset already embedded in the stock. Conversely, Adobe looks more vulnerable because even a beat may not change the market’s view that its growth profile is being outpaced by AI-native software peers, making it a relative short candidate versus stronger enterprise spend names like ORCL and CRM.

IPO sentiment is also bifurcating: Cerebras is being treated like a strategic AI asset, while BXDC is being priced like a slow-moving yield vehicle. That divergence suggests investors are currently paying for scarcity and growth visibility, not just AI adjacency. In cyclicals, the CMI upgrade indicates truck and power demand may be inflecting before the broader industrial tape catches up, which can matter if data center buildout drives incremental generator and engine demand over the next 12-24 months.

Key Decisions for Investors

• Add NVDA on intraday weakness only; use a 1-3 week horizon and trim into further AI-beta squeeze if the stock reclaims Friday’s breakdown level. Risk/reward favors upside continuation, but only as a tactical trade rather than a fresh core add.
• Initiate a relative-value long GLW / short ADBE pair for 1-2 months: GLW has a cleaner second-order AI infrastructure catalyst, while ADBE faces higher odds of a disappoint-beat cycle. Target 8-12% spread capture; stop if ADBE guides materially above expectations.
• Buy ORCL into earnings on a 2-4 week horizon via common or call spreads; positive AI spend read-through and deal visibility should keep revisions upward. Risk is gap risk if commentary emphasizes timing over magnitude, so size modestly.
• Short-term long AVGO vs. long NVDA is less attractive than it was pre-weekend; if expressing chip strength, prefer NVDA over AVGO for the next few days because Google’s interim capacity demand is more directly supportive of GPU intensity than of broader infrastructure monetization.
• Look at CMI as a medium-term cyclical recovery trade, especially if freight data and industrial PMIs stabilize; the setup is 3-6 months, with upside from both trucking normalization and power demand. Use a pullback entry, since the upgrade can front-run the next leg.