Setup

Cirrus Logic makes mixed-signal ICs — audio codecs, camera controllers, haptics — for Apple, which is 91% of revenue. The stock is up 71% in a year, driven by a tax rate reset the Street has been catching up to. The FY2026 10-K clarifies what's actually driving that reset, adds a new product category, and confirms the thesis is real — but the risk-reward is tight at current levels.

What the Filing Says

The ETR reset, correctly attributed. FY2026 effective tax rate: 16.6%, down from 25.5%. The 10-K rate reconciliation is explicit: UK Patent Box incentive = −8.5% of rate (≈$42M/year). GILTI/Subpart F net of FTC = +0.4% (was +14.1% in FY2025 before OBBBA). The UK Patent Box is the dominant structural driver — it is permanent UK tax policy for IP-intensive companies with significant UK R&D operations, unrelated to US legislation. OBBBA eliminated the GILTI amplifier but is not the primary mechanism. The distinction matters: UK law is not politically reversible in the way OBBBA is. On current pre-tax income (≈$497M), the Patent Box alone accounts for ≈$42M in annual tax savings that existed before OBBBA and persist regardless of US political outcomes.

FaceID smart power IC — new product category. Management: "a recent announcement from our largest customer highlights our collaboration on a solution to support their FaceID implementation in future products. We are designing our first product in this area, a smart power IC, which represents an exciting new application space for Cirrus Logic." This is incremental to existing audio/haptics/camera content. Apple's American Manufacturing Program press release (March 26, 2026) independently names CRUS for "advanced ICs to power Face ID systems" at GlobalFoundries Malta, NY — a Tier 2 source corroborating the 10-K. Apple names suppliers in press releases only when the relationship is binding. CRUS stock rose 7.2% on that announcement. Revenue contribution: FY2027-2028 (18-month design cycle).

The business. Revenue $1,997M (+5.3%). Gross margin 52.8% (+30bps, three consecutive years expanding). Operating cash flow $650.6M (+46% YoY). Total liquidity $1,154M. $274M remaining on buyback authorization. HPMS segment (camera, haptics, power ICs) growing at 10.3% vs Audio at 2.0% — mix shift is real. GlobalFoundries Capacity Reservation Agreement expires CY2026, eliminating ≈$180M in fixed purchase commitments and restoring full sourcing flexibility.

The tail. IRS proposed $168.3M additional tax + $63.7M penalties (≈$232M before interest) for FY2017-2019 transfer pricing, in Appeals since July 2024. No accrual. Survivable with $1.15B liquidity but a one-period EPS crater (≈55% of one year's net income) if resolved adversely.

What the Market Thinks

At $171, the market implies ≈$9.50 FY2027 EPS at an 18x multiple — meaning 9-10% revenue growth and ≈16-17% ETR already priced. Four consecutive earnings beats (+37.9%, +19.1%, +21.9%, +10.7%) have educated the Street. This is not an undiscovered stock.

Probability-weighted EV:

At $171: EV = −1.2%. At $160: EV = +5.6%. At $155: EV = +8.9%.

Why the Gap Exists

Two specific misattributions, not a general mispricing:

1. Tax mechanism. The market has priced a lower ETR but attributes it broadly to OBBBA (political risk, reversible). The rate reconciliation shows the UK Patent Box is the dominant driver. A reader who misattributes the mechanism applies a political-reversal discount to earnings that are actually protected by UK law. This creates a 1-2pp ETR pricing gap between our view and the market's, worth ≈$3-5/share.

2. FaceID IC not in consensus estimates. The AMP press release was March 26, public for two months — the $10/share market move on announcement day already priced the announcement. But ASP × volume in FY2027 is not yet in sell-side models. If HPMS grows >15% YoY in Q2-Q3 FY2027 (vs 10.3% baseline), FaceID contribution is visible — $0.30-0.40 incremental EPS, ≈$5-7/share.

What's fully priced: The ETR level itself, HPMS mix expansion at the current pace, buyback velocity.

Risks

1. Apple concentration (structural, ongoing). 87% → 89% → 91% annually. Every new design win deepens it. A single Apple supply chain decision — insourcing, renegotiation, content shift — is existential. This is not a tail; it is the permanent architecture of the business. No diversification path visible at scale.

2. FY2027 ETR uncertainty. 10-K notes "modifications to existing international tax provisions will take effect in fiscal year 2027." Direction unquantified. UK Patent Box floor (≈17.5% ETR minimum) limits downside, but OBBBA international provisions could move ETR 0-2pp in either direction. Resolves August 4.

3. IRS tail. $232M+ proposed assessment in Appeals. 30% adverse probability. One-period EPS impact ≈$4.50/share if triggered, without existential risk (liquidity covers it).

4. Under-display FaceID timeline. iPhone 18 Pro launch September 2026, but full under-display implementation may slip to 2027 per supply chain reports. Smart power IC ships regardless of FaceID display architecture, but volume ramp timeline uncertainty persists.

5. %Idio variance ≈50%. Below 75% target. Apple supply chain is a systematic factor for CRUS — 91% revenue concentration means iPhone cycle beta is real, not diversifiable. True idio alpha (ETR mechanism + FaceID) is 1-4%. Sizing should reflect that most of the expected return is beta exposure.

Catalysts

August 4, 2026 — Q1 FY2027 earnings. Full-year ETR guidance. If management guides 16-18%, confirms structural reset and triggers consensus EPS upgrades. This is the near-term thesis resolution event. Probability of 16-18% guide: 60%.

September 2026 — iPhone 18 Pro launch. First vehicle for FaceID smart power IC. Volume confirmation begins.

October 2026 — TechInsights iPhone 18 Pro teardown. Independent confirmation of CRUS content in FaceID subsystem.

November 2026 — Q2 FY2027 earnings. First quarter with FaceID IC revenue potentially visible in HPMS growth rate.

December 2026 — GlobalFoundries agreement expires. Working capital flexibility restored, $180M in fixed purchase obligations cleared.

What Would Change Our Mind

Bearish:

• Aug 4 ETR guide >19%: UK Patent Box threatened or OBBBA international provisions materially adverse. Exit immediately — the structural ETR thesis breaks.
• HPMS growth <10% in Q2-Q3 FY2027 despite iPhone 18 Pro launch: FaceID IC either delayed, smaller-ASP than expected, or lost to a competitor.
• Apple publicly discloses CRUS content reduction or internal PMIC expansion into FaceID subsystem.
• IRS Appeals adverse ruling crystallizes.

Bullish (thesis strengthens):

• TechInsights teardown confirms CRUS FaceID power IC content in iPhone 18 Pro.
• Q2 FY2027 HPMS >15% YoY — FaceID ramp visible in the numbers.
• Stock at $155-162 with thesis intact.

Evidence