3 Unstoppable Dow Stocks Worth Buying Right Now

Nvidia expects fiscal 2028 revenue to rise 70% despite supply constraints after more than doubling fiscal Q2 2027 revenue, underscoring sustained AI-compute demand. Alphabet posted 24% Q2 revenue growth and 82% Google Cloud growth, though investors reacted negatively to elevated capex and potential negative free cash flow. American Express delivered 10% revenue growth, 14% EPS growth and 9% cardmember-spending growth, raising full-year revenue guidance while maintaining EPS guidance to fund investment.

Analysis

The actionable AI signal is not another directional NVDA call; it is that constrained accelerator availability extends the spending cycle while reallocating economics away from hyperscaler free cash flow and toward the AI infrastructure stack. Near term (days to 3 months), NVDA earnings revisions can remain positive, but the more asymmetric second-order beneficiaries are networking, memory and power/cooling suppliers that monetize each incremental cluster without carrying the same headline valuation concentration. Conversely, enterprise software names priced for rapid AI-margin expansion remain exposed if customers' AI budgets are absorbed by compute capacity rather than applications.

GOOG's multiple can re-rate only if incremental Cloud/AI revenue converts into operating leverage rather than merely validates a larger capex base. The key 1-3 month catalyst is evidence that cloud backlog, pricing, and utilization offset depreciation growth; absent that, a negative free-cash-flow narrative can persist despite strong top-line momentum. This makes GOOG a better execution-driven recovery than a pure AI beta, while MSFT and AMZN remain useful read-throughs for whether industry capex is becoming competitively compulsory.

AXP's setup is distinct: reinvestment while maintaining earnings targets implies management is prioritizing spend, acquisitions, and member-value investment over near-term margin expansion. That is constructive over 6-18 months if billed-business growth persists, but the closed-loop model has more direct downside than V or MA if affluent credit quality or travel-and-entertainment spend weakens. Consensus may be underestimating this credit-cycle asymmetry, so AXP should be owned only with explicit consumer-data discipline rather than as a generic payments long.

The article's valuation claims and management demand commentary are not independently sufficient to underwrite positions. The central falsifier for the AI complex is a deceleration in hyperscaler capex guidance or a widening gap between AI revenue and depreciation expense; for AXP, it is rising delinquency/reserve metrics alongside slower billed-business growth.

Key Decisions for Investors

• Maintain a core NVDA long only on post-results confirmation of supply-constrained demand and gross-margin resilience; take incremental exposure through a basket of AI networking, memory, and power/cooling suppliers rather than chase NVDA after a sharp pre-earnings move. Review over the next 1-3 months; exit/add no further risk if hyperscaler capex guidance rolls over or NVDA guides gross margin materially lower.
• Initiate a measured long GOOG versus an equal-weight short basket of lower-quality AI application software, not versus NVDA. The thesis is 6-12 month Cloud monetization and multiple recovery, but size modestly until management demonstrates revenue/depreciation conversion; falsify on another material capex increase without Cloud margin, backlog, or pricing improvement.
• Use AXP as a 6-18 month value/quality long only if upcoming card-spend growth and credit provisions remain contained; pair with a small V or MA hedge if the objective is to isolate AXP's reinvestment upside from broad payments beta. Reduce if delinquency trends accelerate, reserve build exceeds revenue growth, or billed-business growth slows materially.
• Do not buy standalone calls on GOOG or AXP without implied-volatility and event-date data. Set an alert around earnings: a capex-led GOOG selloff without deterioration in Cloud margins is the preferred entry window, while an AXP rally unsupported by unchanged EPS estimates should be sold into rather than chased.