Investing.com’s stocks of the week

Intel surged 21.9% intraday after quarterly results beat consensus, with Q1 revenue of $13.6 billion versus $12.41 billion expected and $12.7 billion a year ago, supported by growth in data center and AI. Roth/MKM upgraded INTC to Buy from Neutral, citing improved execution under CEO Lip Bu Tan and better participation in AI infrastructure growth. The article also notes strong TXN results but software-sector weakness in NOW, IBM, and cautious META commentary.

Analysis

The key read-through is not “earnings beats help semis,” but that the market is rewarding companies that can credibly prove exposure to AI capex without relying on perfect end-demand visibility. INTC and TXN are being treated as the safer expression of an industrialized AI buildout: more compute, more power management, more sockets, less narrative dependence on software monetization. That makes them a relative beneficiary of the current rotation away from higher-multiple software, especially if investors continue to de-rate businesses where AI is still a future promise rather than a present revenue line.

NOW and IBM are the canaries for a much broader second-order shift: AI is compressing the forgiveness window for execution misses and slowing growth. The market is no longer paying for “good enough” in enterprise software; it wants either re-acceleration or clear operating leverage, and any sign that compute costs, deal scrutiny, or regional noise can dent near-term metrics gets punished disproportionately. This raises the risk that software multiple compression persists for several weeks, even if macro stays constructive, because the issue is not just demand but confidence in pricing power and time-to-monetization.

META sits in an awkward middle ground: efficiency actions can support margins, but they also telegraph that the company sees enough cost pressure to pre-empt disappointment. The market is likely underestimating how much capex intensity can mute earnings leverage over the next two quarters even if revenue stays healthy. In contrast, the semis with clearer AI infrastructure linkage may continue to absorb incremental flows as investors seek earnings visibility with a hard asset angle.

The contrarian setup is that the selloff in software may be overdone on a 1-week horizon but still underdone on a 1-2 month horizon if guidance commentary continues to emphasize caution. The winners from this tape are not just the obvious chip names; they are also the suppliers of power, test, and packaging into the AI buildout, while the losers are software vendors that cannot demonstrate immediate attach rates or cost discipline. This argues for staying selectively constructive on infrastructure semis while fading software strength until next guidance season resets expectations.

Key Decisions for Investors

• Go long INTC versus short NOW for 2-6 weeks: favor the name with improving execution and tangible AI infrastructure exposure over the one facing higher scrutiny on monetization and costs; target a 10-15% relative spread, stop if NOW stabilizes on accelerating bookings commentary.
• Add TXN on pullbacks over the next 1-3 weeks as a second-order AI infrastructure beneficiary; risk/reward favors a slower, steadier rerating if the market keeps rewarding analog/power content tied to data center expansion.
• Initiate a software underweight basket versus semis: short NOW and IBM against a long basket of AI infrastructure names; thesis is continued multiple compression in software if guidance remains cautious, with 8-12% downside in the shorts if the next macro print weakens risk appetite.
• Buy META downside protection into the next earnings window: 1-2 month put spread to express concern that higher compute intensity will cap margin upside even if revenue remains resilient; structure for a defined-risk payoff if the market focuses on cost rather than growth.
• If semis gap higher again, trim into strength rather than chase: the trade is becoming consensus, so keep position sizes tactical and reassess if the next set of results fails to broaden participation beyond a few AI-linked leaders.