I'll research QCOM before writing the preview. Let me gather recent earnings, filings, and news in parallel.Let me read the most recent earnings call transcript and release, plus get more news context.Let me check the June Investor Day and recent stock performance, plus broader news context.Let me read the Investor Day transcript for the data center/long-term targets, and gather recent analyst/news context.Let me get consensus estimates and the latest context via web search.I have everything I need. Here's the earnings preview.
Qualcomm (QCOM) — Q3 FY2026 Earnings Preview
Reports: Wednesday, July 29, 2026, after the close | Fiscal quarter ended: ~June 2026
The one-sentence setup
This is shaping up to be one of the most consequential QCOM prints in years: a near-term handset trough colliding with a completely re-rated (and now sharply de-rated) data-center narrative. The quarter itself is largely pre-baked; the reaction will hinge on Q4 guidance, the China handset "bottom" call, and any incremental data-center proof points.
1. What the Street expects
| Metric (Q3 FY26) |
Consensus |
QCOM Guidance (given 4/29) |
| Revenue |
~$9.68–9.71B |
$9.2B – $10.0B |
| Non-GAAP EPS |
~$2.22–2.23 |
$2.10 – $2.30 |
| QCT revenue |
~$8.26B |
$7.9B – $8.5B |
| QTL revenue |
~$1.25B |
$1.15B – $1.35B |
- Consensus sits comfortably inside the guide, near the midpoint. Revenue would represent a ~6.5% YoY decline — a stark reversal from the ~10% growth QCOM posted in the year-ago June quarter.
- Embedded in guidance: QCT handset ~$4.9B, Automotive ~+50% YoY, IoT high-single-digit YoY growth, and opex ~$2.6B.
- The bar to "beat" is modest, but with the stock having fallen roughly a third from its late-May high, the burden of proof is on the forward outlook, not the June actuals.
2. Stock context — a violent round trip
QCOM had an extraordinary few months. Shares roughly doubled from ~$127 in early April to a peak near ~$251 on May 26, fueled by data-center excitement, then unwound almost the entire move — closing at ~$163 on July 28, essentially flat-to-down for the year. Much of the fade came after the June 24 Investor Day, suggesting the market either "sold the news" or grew skeptical of the aggressive long-term targets. Options markets are reportedly pricing a large (~9%) post-earnings move. The setup is asymmetric: expectations are low on fundamentals, but rich on the multi-year story.
3. The central near-term debate: Is the China/handset bottom real?
This is the most important swing factor for the quarter.
- Management has framed fiscal Q3 (the June quarter) as the bottom for QCT handset revenue from Chinese customers, with a return to sequential growth thereafter. The pressure is memory-driven: surging AI data-center demand for memory has pushed DRAM prices up, prompting Chinese OEMs to trim build plans and draw down channel inventory, so QCOM has been materially undershipping true end-consumer demand.
- Because QCOM sees sell-through via its licensing business, management argues it has good visibility that activations are holding up while shipments have lagged — hence the confidence in a Q3 trough.
- What to watch: Confirmation (or walk-back) of the bottom call, commentary on China build plans normalizing, and any read on memory-cost trajectory into 2027. A convincing bottom + sequential growth guide is the bull case; any hedging here would be a negative surprise.
4. The September-quarter (Q4) crosscurrents
Q4 guidance is where the action is, and it contains two offsetting forces:
- Positive: China Android shipments recovering off the bottom.
- Negative: The long-telegraphed Apple modem step-down. Management reaffirmed Apple will represent only ~20% of the phones launching this fall, with ~$2B of QCT product revenue modeled for FY27 and no product relationship beyond. The royalty stream, however, is expected to be unchanged. Historically the September quarter is a growth quarter driven by Apple's ramp — that tailwind is largely gone this year.
- Watch whether these net to sequential handset growth. Management declined to commit at Q2.
5. Margins & the new price-increase lever
- QCT EBT margin was 27% in Q2 (guided 25–27% for Q3); QTL remains a ~70%+ margin cash engine.
- New wildcard: Bloomberg reported (July 24) that Qualcomm is telling customers to expect double-digit price increases on chips — a response to rising input/memory costs. This is a potential margin/ASP offset to volume weakness. Look for management to address whether pricing can cushion gross margins as memory inflation flows through.
6. Diversification: Auto & IoT still the near-term proof points
- Automotive remains the standout: Q2 was a record $1.33B (+38% YoY); Q3 is guided to accelerate to ~50% YoY. QCOM crossed a $5B annualized run-rate and expects to exit FY26 above $6B, backed by a $65B design-win pipeline and the 5th-gen Snapdragon Digital Chassis (its largest-ever content step-up, with ADAS ramping).
- IoT returned to growth (+9% in Q2), guided high-single-digits in Q3, led by industrial and consumer.
- Combined Auto + IoT grew ~20% YoY in Q2 — the tangible, already-happening piece of the diversification thesis while handsets bottom.
7. The data-center story — now the stock's swing factor
The June 24 Investor Day radically reset the long-term framing, and this will be the most-scrutinized topic on the call:
- New long-term targets: FY29 non-handset revenue raised to $40B (from $22B set 18 months earlier) — a ~40% four-year revenue CAGR off FY25.
- Data center specifically: $5B revenue targeted in FY27 (with ≥$1B from two global hyperscaler customers), scaling to $15B in FY29, and a goal of >5% share of a ~$1T TAM in 5–7 years.
- Product roadmap: custom silicon shipments begin FQ1'27 (this December quarter); AI accelerators (built on proprietary "HBC" high-bandwidth-compute memory architecture) ramp 2H FY27; data-center CPUs (the "C1000" fleet) ramp 2H FY28.
- Marquee validation: on-stage endorsements from Microsoft (Azure HBC deployment), Meta (multi-generation CPU supply deal), plus the acquisition of Modular (AI software stack) and a Hugging Face partnership.
- Skeptic's lens: these are bold targets from a data-center new entrant, custom-silicon gross margin was flagged as slightly below corporate average (though operating-margin accretive), and the stock's post-Investor-Day fade signals the market wants execution, not slides. Any concrete revenue quantification for the December custom-silicon ramp would be a meaningful catalyst.
8. Capital returns & balance sheet
- QCOM is returning cash aggressively: $3.7B to holders in Q2 ($2.8B buybacks + ~$0.9B dividends), $5.4B of buybacks in 1H FY26, and a new $20B authorization. Expect continued heavy repurchase support into a depressed stock.
- Note: a one-time $5.7B non-cash GAAP tax benefit in Q2 inflated GAAP EPS ($6.88) — it's excluded from non-GAAP; don't be confused by prior-quarter optics. Inventory has been building (working-capital watch item).
What to watch / how to frame the reaction
Bullish triggers: (1) clear confirmation the China handset bottom is in with a sequential-growth Q4 guide; (2) evidence pricing increases are protecting margins; (3) any hard data-center revenue detail on the December custom-silicon ramp; (4) continued auto acceleration toward the $6B exit rate.
Bearish triggers: (1) softer-than-guided Q4 as Apple step-down outweighs China recovery; (2) any hedging on the "bottom" call or on memory-cost pressure bleeding into FY27; (3) margin compression from input costs; (4) vague/pushed-out data-center milestones that further erode confidence in the $5B FY27 target.
Bottom line: The June quarter is likely a controlled, in-line trough. The share-price outcome depends on whether management can (a) credibly ring the bell on the handset bottom and (b) begin converting the ambitious data-center narrative into dated, quantified revenue — against a stock that has already given back nearly all of its 2026 AI-driven rally.
Note: figures above are drawn from QCOM's Q2 FY26 release/call (4/29/26), the June 24 Investor Day, recent press reports, and sell-side consensus compiled ahead of the print; actual reported results and updated Street numbers may differ.