Microsoft and Nvidia finally became cheap enough for this value investor – and he plans to buy any dips

Value investor Chris Grisanti said he bought Microsoft and Nvidia in Q1 2026 after their early-year pullbacks, arguing both now fit his valuation criteria and could benefit from continued AI spending. Microsoft is down more than 12% year to date and trades at 19.6x next year's earnings, while Nvidia is up more than 13% YTD, with Grisanti citing 18x forward earnings and potential upside to 30x historical norms. He said he will buy dips unless valuations become too rich or the AI capex thesis breaks.

Analysis

The important signal is not that these names are "cheap" on absolute terms, but that a respected value allocator is effectively underwriting the AI capex cycle at a lower entry multiple. That matters because it can pull incremental capital from the large pool of fundamental managers who have been underweight mega-cap tech on valuation discipline, creating a self-reinforcing demand base on dips rather than a one-off sentiment trade.

For MSFT, the market is implicitly debating whether AI monetization is a 2026 story or a 2028 story. If revenue acceleration remains lagged for another couple of quarters, the stock can stay range-bound even with strong fundamentals because investors will continue to discount elevated infrastructure spend; the setup is therefore more about second-half rerating than near-term earnings beat potential. A decisive move back above the 200-day and a clean print showing Azure/AI uptake would likely force systematic and discretionary re-entry, while another guide-down on margin trajectory would keep value buyers patient rather than aggressive.

For NVDA, the bigger second-order effect is that every dip is now being underwritten not just by momentum funds but by long-only valuation screens that have historically avoided the name. That broadens the buyer base, but it also makes the stock more sensitive to any sign that hyperscaler capex growth is decelerating, because the "cheap relative to history" argument only works as long as the earnings runway keeps extending. The asymmetry is favorable over months, but in the next few weeks the stock remains vulnerable to crowded positioning and any guidance that fails to re-accelerate consensus.

The contrarian miss is that a re-rating back to historical averages may be harder for MSFT than NVDA because software monetization is still proving out, while chip demand is already visible in shipment trends and backlog. In other words, MSFT may be the better longer-duration compounder, but NVDA is the cleaner near-term trade if the AI spend thesis holds. The main risk to both is not valuation compression alone; it is a narrative break in capex durability, which would hit semis first and then spill over into enterprise software spend expectations within one reporting cycle.

Key Decisions for Investors

• Long MSFT on a 3-6 month horizon only if it reclaims the 200-day moving average; target a 15-20% rerating move, with downside capped by using call spreads rather than stock.
• Buy NVDA on post-earnings dips or any 5-8% pullback tied to capex anxiety; maintain a tighter risk stop because the stock will likely react faster to any guide miss or hyperscaler commentary shift.
• Pair trade: long NVDA / short a weaker AI-exposed software basket over 1-2 quarters to express the view that hardware demand is still visible while software monetization remains more uncertain.
• If already long MSFT, sell upside calls against the position into earnings to monetize elevated event premium while retaining exposure to a rerating if AI adoption accelerates.
• Reduce exposure in both names if management commentary suggests AI capex growth is normalizing faster than expected; that is the catalyst most likely to unwind the value-buyer thesis within 1 reporting cycle.