Stocks gain on Iran ceasefire, plus 3 more things that drove last week's market

U.S. stocks had their best week since November, with the S&P 500 up 3.6%, the Dow up more than 3%, and the Nasdaq up 4.7% as the Iran ceasefire eased geopolitical stress. March CPI rose 0.9% month over month and 3.3% year over year, in line with estimates, while core inflation came in better than feared. AI/hardware names outperformed—Marvell +20%, Intel +23%, Corning +15.7%—while software sold off, with Salesforce down nearly 12% and the IGV software ETF off about 7%.

Analysis

The tape is signaling a sharp but fragile de-risking unwind: the market is rewarding names levered to capex, data-center spend, and cyclical beta while punishing software as if AI infrastructure winners can keep taking share indefinitely. That’s usually a late-cycle positioning tell, not a durable fundamental regime shift. The key second-order effect is that “AI spend” is becoming more concentrated in a narrower set of beneficiaries, which means the same budget dollars are now being re-rated across the stack rather than expanding the whole group.

Within that split, the relative move in hardware versus software looks overstretched in the near term. If enterprise buyers pause on software multiple expansion, the next leg lower should hit the less differentiated names first, while the strongest balance sheets with embedded AI workflows likely stabilize earlier than the broader software ETF. Conversely, infrastructure winners still need proof that orders convert into backlog and gross-margin durability; otherwise the market can easily rotate back once the next macro scare fades.

The macro backdrop matters because inflation is not yet giving the Fed room to turn decisively dovish, so this rally is more about lower geopolitical risk premia than a clean earnings re-acceleration. That means the risk to the move is a weekend headline or energy spike, not a slow macro grind. If talks fail or oil re-prices, the market will likely punish the highest-multiple cyclicals and software simultaneously, while quality megacap balance sheets remain the first refuge.

The most interesting contrarian read is that the software selloff may be too broad. If AI infrastructure spending is real, many software vendors will eventually be beneficiaries through workflow integration, security, and data orchestration; the market is currently pricing them as pure losers. That creates a window for selective accumulation in names with recurring revenue and AI attach opportunity, while staying short the most crowded AI-capex beneficiaries where expectations have run too far ahead of delivery.

Key Decisions for Investors

• Buy MRVL on a pullback over the next 3-5 sessions; use the recent strength as confirmation but size tightly, since the risk/reward is favorable only if data-center sentiment stays constructive. Target a 6-8% further upside with a 3-4% tactical stop.
• Short IGV or buy puts on the software ETF for a 2-4 week mean-reversion trade; the crowding signal is strong and multiple compression can continue another 5-7% if rates or geopolitics re-tighten.
• Pair long META / short CRM for 1-3 months; META has a clearer path to monetizing AI spend and operating leverage, while CRM faces valuation pressure if investors keep de-rating legacy software beneficiaries. Aim for relative outperformance of 8-12%.
• Consider long GLW against short ADBE on a basket basis; GLW benefits from fiber demand tied to AI buildout, while ADBE is more exposed to software multiple compression. Use a 6-10 week horizon with a 2:1 upside/downside skew.
• Keep a tactical hedge via CRWD or PANW puts into weekend geopolitical headlines; if risk sentiment reverses, these names can give back quickly because they are owned as quality software proxies. Use near-dated options to express downside convexity.