Qualcomm Is the Most Underrated AI Chipmaker to Buy

Qualcomm’s June 24 data-center/AI update raised its fiscal 2029 guidance target to $40B (nearly 2x the prior target) and aims for $15B+ in AI infrastructure revenue by fiscal 2029. Management projects fiscal 2026 Q3 revenue of $9.2B–$10.0B (midpoint $9.6B), implying a 7.4% YoY decline, but pairs the transition with continued capital returns: $5.4B of buybacks in 1H FY2026 plus a new $20B authorization. At a 21.2x P/E with a P/E-to-Growth below 1, the article argues the stock is undervalued versus peers while investors look for fundamental inflection beginning in 2027+.

Analysis

QCOM is less a "new AI leader" than a plausible second-order beneficiary of hyperscaler diversification. If even one or two design wins scale, the market will start valuing the data-center option value rather than the handset franchise, and that can expand the multiple faster than the revenue base does. The real losers are the incumbent AI profit pools: NVDA and, to a lesser extent, AVGO face a narrative risk that incremental custom-silicon spend broadens to a second source, even if QCOM never becomes a primary accelerator supplier.

The near-term setup is mostly sentiment, not fundamentals. Over the next 1-3 months, the stock only rerates if management converts the story into independently verifiable milestones: named customers, backlog, tape-out cadence, or guidance that shows the AI line is pulling forward the 2027-2029 slope. If that evidence stays absent, the move likely fades because the current cash flow still depends on a mature phone market; buybacks can cushion downside, but they also risk masking weak organic growth.

Consensus appears to be underpricing execution risk and overpricing timing confidence. The market may be extrapolating a platform shift before the commercial proof exists, which makes this a better relative-value or staged-entry idea than an outright momentum chase. The thesis is falsified if QCOM fails to show data-center contribution in the next two earnings cycles, or if NVDA/AVGO maintain share and pricing power while QCOM stays a story stock; structurally, the upside is 6-18 months out, not days.

Key Decisions for Investors

• Initiate a starter long in QCOM on pullbacks over the next 1-2 weeks; add only if the next earnings call produces customer names or backlog evidence. Risk/reward is favorable for a 6-18 month rerating, but the near-term could chop if the AI story remains unverified.
• Pair trade: long QCOM / short NVDA for a 1-3 month relative-value expression. The bet is that QCOM's valuation has more room to expand on modest AI proof, while NVDA's premium is more vulnerable to any perception of custom-silicon share leakage.
• For higher-conviction investors, use a QCOM bull call spread with 12-18 month expiry instead of common stock. This isolates upside to the 2027+ inflection while defining risk if hyperscaler conversion slips.
• Avoid outright shorting NVDA or AVGO on this headline alone; wait for evidence that QCOM's wins are displacing, not merely supplementing, incumbent spend. The first falsifier is a lack of named design-win follow-through by the next two quarters.