Verdict: KEEP | LR 1.0 (neutral)

Worst factor profile in the selectable set — 46.5% idio variance, strongest anti-momentum loading (-0.71), deepest negative trailing alpha (-33.8%). On factor grounds alone, this is a remove candidate. But we already carry 5 anti-momentum removes. Adding a 6th concentrates into one correlated factor bet at zero IC. The market has already repriced QCOM correctly. No informational edge. Keep at benchmark weight.

Factor Decomposition

Factor regression (trailing 251d)

Factor       Beta     Var Contribution
---------------------------------------
SPY         +1.62      56.6%
XLK         +0.47      22.2%
MTUM        -0.71     -25.2%
Idio                    46.5%

alpha = -33.8% ann    sigma_idio = 26.4%    R-sq = 53.5%

Idio variance 46.5% — worst in the selectable set. More than half the variance is factor-driven: market beta (1.62) plus strongest anti-momentum loading in the basket (beta_MTUM = -0.71, worse than MAR -0.57, CMCSA -0.56, CSX -0.46, SBUX -0.45, CTAS -0.36).

The trailing alpha of -33.8% is the regression residual — stock-specific underperformance after removing SPY, XLK, and MTUM contributions. Something idiosyncratic is driving this beyond the factor story.

What's Driving the Negative Idio Alpha

Apple Modem Transition (Primary)

Revenue concentration (10-Q Note 2, Q1 FY2026): Customer (x) at 25% of revenue (≈$3.1B/quarter, ≈$12.4B annual). Almost certainly Apple — historically the largest QCT customer for MDM products.

Transition timeline:

10-Q risk factor language (filed Feb 4, 2026, p.22): "Apple utilizes its own modem (rather than our products) in certain of its smartphones and we expect that Apple will increasingly use its own modem products, rather than our products, in its future devices, which will have a significant negative impact on our QCT revenues, results of operations and cash flows."

Not hedging. "Will have a significant negative impact" is near-certain forward guidance in risk factor clothing.

Memory Shortage (Near-Term)

Q1 FY2026 earnings call (Feb 4, 2026) — CEO Amon and CFO Palkhiwala: DRAM supply redirected from consumer electronics to HBM for AI data centers. Chinese OEMs "taking cautious approach to chipset inventory." Q2 handset guided to $6.0B (vs $7.8B Q1, -23% QoQ). Management: "Whole fiscal year mobile handset determined by memory availability."

China Concentration

10-Q qualitative: "We derive a significant portion of our revenues from Chinese OEMs." Industry estimates: ≈60%+ of QCT revenue China-linked. Dual risk: tariff escalation and semiconductor self-sufficiency push (Huawei HiSilicon). QCOM doesn't disclose geographic revenue breakdown — which is itself a signal about how exposed they are.

What Market Consensus Actually Prices

This isn't "sell-side says 15% upside." This is what the capital structure tells us.

1. NTM Earnings: 8% Decline Already Priced

Fwd P/E 11.54 at $128.38 = NTM consensus EPS of $11.13. Trailing non-GAAP 4Q sum: $2.85 + $2.77 + $3.00 + $3.50 = $12.12. The market prices 8.2% NTM earnings erosion — Apple modem loss partially offset by auto/IoT growth. Not a cliff. A controlled decline.

2. Re-Rating Complete: No Longer a Growth Semi

QCOM forward P/E 11.54x vs SOX (semiconductor index) at ≈20-25x. That's a 50%+ discount to peers. The market has moved QCOM from "growth semi" to "mature cash cow with declining moat." This re-rating isn't pending — it happened. The stock isn't waiting to be repriced down. It already was.

3. Valuation Structure (Reverse-Engineered)

QTL (patents):     $6.4B rev x 77% EBITDA x 15x = ≈$74B = $69/share
QCT ex-Apple:      ≈$30B rev x 2.1x              = ≈$63B = $59/share
Total:             ≈$137B = $128/share

Market price:      $128.38

The market's math adds up. QTL provides a hard floor (≈$69/share) from patent licensing with protections running through the 2040s. QCT ex-Apple at 2.1x revenue is appropriate for a hardware semi losing its largest customer. This isn't mispriced — it's correctly priced for the transition.

4. Options: 10% Implied Earnings Move

Apr 24 (pre-earnings):   IV 38.6%
May 01 (post-earnings):  IV 48.0%

Earnings variance = (0.48^2 x 34/365) - (0.386^2 x 27/365) = 0.01044
Implied move = sqrt(0.01044) = +/-10.2%

10% implied move on a $137B company is elevated. The market sees real binary risk around April 29 despite the pre-guided soft quarter. The 37th percentile IV stat (pre-earnings) masks the 53rd percentile IV in earnings-inclusive expirations. The market has meaningful uncertainty concentrated around the event.

5. Flow Divergence: Narrative Rotating in Real Time

Accumulated OI is bullish (from higher prices). Today's flow is 14:1 put-to-call. The market is rotating from "value opportunity" to "falling knife" — classic oversold narrative shift.

6. Analyst Distribution

2 Strong Buy, 8 Buy, 22 Hold, 2 Sell, 1 Strong Sell. 63% at Hold. Targets: $100 (Seaport, Sell) to $200, median $150. Recent actions: Bernstein downgraded to Market Perform 3/26, Seaport to Sell 3/16. Consensus is herding at "wait and see" — which tells you the market has no edge either.

Edge Assessment

Counterparty analysis: If we remove QCOM (short via filtration), we bet it underperforms QQQ. The other side: value investors at 11.54x forward P/E, dividend seekers at 2.73% yield, auto/data center bulls with multi-year thesis, 22 analysts who've already modeled the Apple transition. Informed counterparties with longer horizons. Edge = 0.

What would constitute edge: Non-public supply chain data on iPhone 18 sourcing. Specific tariff policy targeting QCOM chipsets. Evidence auto/DC pivot is failing (opposite — accelerating). Hidden balance sheet deterioration (none — fortress). We have none of these.

Is anything mispriced? The short-term oversold condition (RSI 25.8, 9% of 52-week range, 14:1 put volume) might represent a 2-4 week mean reversion opportunity — but that argues for KEEP, not REMOVE. The fundamental re-rating is correct. The options market has real uncertainty about earnings but is pricing it appropriately (10% implied move). No mispricing identified at the fundamental level.

Bull Case (Constrains Downside)

Diversification is real, not slideware:

• Automotive: $1.1B/quarter (+15% YoY), guided >35% YoY acceleration in Q2. VW, Toyota, Rivian design wins.
• Data Center: $2.3B Alphawave acquisition (Dec 2025). Revenue expected 2027. Working with 7/9 largest cloud companies.
• IoT: $1.7B/quarter (+9% YoY). Industrial expansion, robotics entry.

Financial strength: OpCF $5.0B in Q1 FY2026, returned $3.6B to shareholders ($2.6B buybacks + $949M dividends), net cash positive after $6.7B capital deployment. QTL licensing: $1.6B/quarter at 77% EBITDA margin.

Legal overhang removed: Arm v. QCOM verdict (Dec 2024) cleared Nuvia acquisition risk. Arm-based server CPUs safe.

Earnings execution: 4 consecutive beats (+1.2%, +2.0%, +4.3%, +2.9%). Q2 guided soft but market knows — memory shortage pre-guided.

Portfolio Construction

Anti-momentum concentration is the binding constraint. Current removes with anti-momentum loading:

QCOM   beta_MTUM = -0.71  (strongest — would be new)
MAR    beta_MTUM = -0.57
CMCSA  beta_MTUM = -0.56
CSX    beta_MTUM = -0.46
SBUX   beta_MTUM = -0.45
CTAS   beta_MTUM = -0.36

We carry 5 anti-momentum removes — already one correlated factor bet at zero IC. Adding QCOM makes it 6 of 11 removes (55%) driven by the same factor. This concentrates risk without diversifying the filtration signal.

Weight arithmetic: QCOM at 0.76% selectable weight. Bear case (-15% underperformance) = +11.4 bps filtration alpha. Bull case (oversold bounce, +15%) = -11.4 bps cost. At RSI 25.8, the probability-weighted EV skews against us.

Timing risk: Removing at the bottom of an oversold flow rotation — 14:1 put-to-call volume, 9% of 52-week range, earnings April 29 with beat history — is textbook negative timing skill.

What Would Change the Verdict

• FILTER if: We had 3 or fewer anti-momentum removes (room for another factor bet), OR Apple announced accelerated full-line modem transition ahead of Sep 2026, OR April 29 earnings miss AND automotive guidance disappoints (breaks diversification narrative)
• WATCH for: China tariff escalation specifically targeting QCOM chipsets, Q2 FY2026 earnings (April 29) for memory shortage resolution, iPhone 18 modem sourcing confirmation

Sources

• QCOM 10-Q filed 2026-02-04 (Q1 FY2026, quarter ended Dec 28, 2025)
• QCOM 10-K filed 2025-11-05 (FY2025)
• QCOM 8-K filed 2026-03-19 (annual meeting)
• QCOM Q1 FY2026 earnings transcript (2026-02-04)
• QCOM Q4 FY2025 earnings transcript (2025-11-05)