3 Smart Stocks to Buy Now

The article highlights Microsoft, Meta, and Nvidia as underappreciated buys: Microsoft trades at 19.3x forward earnings vs the S&P 500’s 21.5x with 18% revenue growth and 23% diluted EPS growth (no stock rebound yet, with a potential catalyst from late-July results). Meta is growing revenue 33% YoY (advertising-driven) but still trades at 17.5x forward earnings due to heavy AI data center spending without clear monetizable AI products, while Nvidia is seen as undervalued at 22.3x forward earnings despite Wall Street expecting 82% revenue growth this year (41% next year). Overall, the piece is constructive on AI-driven growth and argues market sentiment/valuation provides upside if fundamentals and product commercialization progress.

Analysis

The market’s real mispricing is not at the headline growth rates; it is in the durability of capex monetization. NVDA remains the cleanest way to express a multi-year AI infrastructure cycle because every incremental hyperscaler spend path still has to run through a narrow set of GPU, networking, and systems bottlenecks, which preserves pricing power and supports multiple expansion if forward supply stays tight. The second-order winners are the picks-and-shovels around deployment — networking, thermal, power, and test equipment — while the clearest losers are software names that rely on AI narrative without corresponding usage-based revenue inflection.

MSFT and META are cheaper for different reasons. MSFT is a better cash-flow story than a re-rating story: unless the next print shows acceleration in cloud monetization and AI attach, the stock can stay optically inexpensive because investors are already paying for the capex burden. META is more interesting as an advertising cash machine with a long-dated product optionality; until glasses or another consumer AI product demonstrates a repeatable revenue engine, its AI spend mostly pressures FCF conversion and limits multiple expansion.

Near term, the main risk to the bullish AI tape is not valuation; it is a digestion phase if 2H26/2027 capex commentary is delayed or customer budgets normalize. The setup would weaken if NVDA growth decelerates faster than consensus or if hyperscaler commentary shifts from build-out to efficiency, because that would compress the forward multiple before earnings catch down. Conversely, a strong July/August earnings season with raised capex guides would likely re-rate semis first, then spill into network and power infrastructure names over 1-3 months.

Key Decisions for Investors

• Go long NVDA vs. short XLK on a 1-3 month horizon; the thesis is that semis should outperform mega-cap software if AI capex commentary remains strong. Falsify if NVDA guides below expectations or if hyperscaler commentary turns defensive on 2027 spend.
• Add to NVDA on post-earnings strength rather than pre-earnings weakness; use any 3-5% pullback as entry for a 6-12 month hold. Risk/reward favors owning the bottleneck supplier into the next capex round, but trim if forward revenue growth falls below the low-40% range.
• Treat MSFT as a quality-core hold, not an immediate alpha idea; prefer selling upside calls against existing positions into July earnings if the stock re-rates toward a richer multiple without a clear Azure/AI monetization inflection.
• For META, avoid outright chasing until there is evidence of a monetizable consumer AI product. If you want exposure, consider a long-dated call spread 6-12 months out rather than stock, because the valuation can stay cheap while FCF is absorbed by data-center spend.
• Watch ANET, AVGO, and power/cooling infrastructure names as second-order beneficiaries; a confirmed 2H26/2027 capex acceleration would likely propagate there before it fully reprices the mega-caps.