2 Bargain Artificial Intelligence (AI) Stocks to Buy Now Down Around 25% From Their All-Time Highs

Meta Platforms and Microsoft are each down a little more than 25% from recent highs, but the article argues both look cheap relative to fundamentals and the S&P 500. Meta trades at 18.0x forward earnings despite 33% Q1 revenue growth, though concerns remain about its $135 billion capital expenditure midpoint for AI. Microsoft is valued at 21.3x expected fiscal 2027 earnings after AI revenue grew 123% year over year to a $37 billion annual run rate, reinforcing the case that it is already monetizing AI.

Analysis

The market is effectively repricing AI from a narrative premium into an execution premium. META is being punished because its AI spend is still an option on future monetization, while MSFT is being treated more like a utility where AI already shows up in current cash flow. That creates a subtle relative-value asymmetry: META has more upside if AI spend converts into measurable ad yield, but it also has a longer-duration drawdown risk if capex keeps outrunning incremental ROI.

The second-order winner is the infrastructure stack, not the app layer. If both companies keep expanding compute budgets, the near-term beneficiaries are the GPUs, networking, and data-center interconnect vendors, while the true losers are adjacent software names that depend on discretionary enterprise IT spend and get crowded out by AI budgets. For META specifically, higher capex can support ad performance in the medium term, but every extra dollar spent now raises the bar for future growth, which means the stock can remain “cheap” longer than value investors expect if estimates prove too optimistic.

The contrarian takeaway is that the selloff may be overdone on MSFT and underdone on META. MSFT’s multiple compression looks more like a normalization after years of premium pricing than a fundamental deterioration, especially since monetization is already visible. META, by contrast, is the cleaner value trap risk: if AI spend stays elevated for 2-3 quarters without a clear operating leverage inflection, the market may re-rate it as a lower-quality growth name despite strong headline revenue.

In the next 1-2 quarters, the key catalyst is not revenue growth but capex discipline and commentary around payback periods. A stabilization in META’s capex guidance or evidence of improving ad ROAS could trigger a sharp multiple reset higher; absent that, the stock likely trades on skepticism. For MSFT, the main risk is less business momentum and more that investors cap the multiple until Azure growth re-accelerates enough to justify renewed premium status.

Key Decisions for Investors

• Long MSFT vs short META, 3-6 month horizon: the market is currently rewarding monetization certainty over monetization optionality; this pair benefits if AI spend scrutiny persists and META needs another quarter of proof.
• Buy META on a staged basis only after the next earnings print if capex guidance stops rising; upside is a fast re-rating back toward market multiple, but risk is another 10-15% downside if AI ROI remains opaque.
• Add MSFT on weakness for a 6-12 month hold: current multiple compression offers better downside protection than META because AI revenue is already flowing through, making earnings revisions the more important driver than sentiment.
• Use call spreads on META rather than outright shares if initiating now: the stock can re-rate sharply on any capex moderation, but the spread limits damage if the market keeps punishing spending intensity.
• If seeking the cleanest AI infrastructure expression, consider a basket long in NVDA against META: both are leveraged to AI spend, but one monetizes the picks-and-shovels spend immediately while the other still has execution risk.