Synaptics

Phase A — Understand the business

Company Overview

Synaptics is a fabless, mixed-signal semiconductor company — historically the world leader in human-interface silicon (touchpads, touchscreen controllers, display drivers, fingerprint biometrics), now repositioning as an "AI-enabled edge" supplier combining embedded compute, wireless connectivity and multimodal sensing. Founded 1986 (California; reincorporated Delaware 2002), ~1,700 employees across 15 countries (21% N. America, 69% Asia-Pacific, 10% EMEA), FY ends the last Saturday in June.

Three reported product markets (FY25 revenue $1,074.3M):

• Enterprise & Automotive — $610.1M (56.8%): PC touchpads, PC fingerprint (Natural ID), DisplayLink/DisplayPort video-interface docks/adapters, automotive touch/display/driver-sensing. The cash-cow base.
• Core IoT — $272.4M (25.4%): Veros wireless (Wi-Fi/BT/BLE/Thread/Matter/UWB/GNSS) + Astra edge-AI processors (MPU/MCU). The growth engine.
• Mobile — $191.8M (17.8%): OLED touch/display controllers for smartphones/tablets. The secular-decline legacy (>500 touch patents but commoditizing, ASPs falling).

Business model: custom silicon designed into OEM products through their contract manufacturers/distributors; the ODM/CM takes title and pays directly. Not recurring/subscription — it is design-win-driven unit sales plus a small, lumpy IP-licensing line (a ~$30M Enterprise & Auto licensing step-down hurt FY25 optics). Contracts carry limited return rights (warranty only); modest stock-rotation/price-protection for distributors. One reportable segment — the CEO (CODM) manages on consolidated net income.

Supply Chain

Upstream inputs → Synaptics → end customer, every named node:

• Foundries (wafers): Asia-based third-party fabs. The 10-KForm 10-KA company’s audited annual report to the US regulator. The most complete thing it publishes. names no foundry, but Synaptics' mixed-signal/legacy-node parts historically run at TSMC and mature-node Asian foundries; memory-component tightness in the broader electronics chain is now explicitly flagged as influencing order timing for smaller customers.
• OSAT (package & test): third-party assembly/test houses, "predominantly Asia-based," consign die/ASICs to contract manufacturers.
• Synaptics (fabless): design, IP, firmware/software. 2,516 active patents / 566 pending, expiring 2025–2045.
• Contract manufacturers / distributors → OEMs: end demand in PCs (Dell/HP/Lenovo-class), smartphones (Samsung/Chinese OEMs), smart-home/IoT OEMs, auto Tier-1s.
• Chokepoints / single-source risk: Synaptics itself discloses "dependence on a few sources of supply," long-term capacity/pricing agreements with some suppliers (a legacy of the 2021–22 shortage), and the write-down risk if end-demand falls below those take-or-pay commitments. Geographic concentration is extreme on the sell-side: FY25 billing was China $493.0M + Taiwan $307.2M + Japan $137.6M + S. Korea $71.7M — US just $7.4M. Middle-East exposure via an Israel design/subsidiary is a named geopolitical risk (Israel/Iran conflict, Strait of Hormuz).

Competitive Advantages (moats)

• Real moats: (1) Deep human-interface IP — >500 touch patents, decades of OLED touch/display-driver integration know-how; (2) design-win switching costs — parts are custom to an OEM's application with long design cycles, so incumbency is sticky within a socket; (3) breadth — one of few vendors spanning touch + display + biometrics + wireless + edge-AI processing, enabling integrated Astra+Veros+Touch platforms.
• Weak/eroding moats: touch and mobile-display controllers are commoditizing (Mobile ASPs −10.3% in FY25, −12.0% in 9M-FY26). In Core IoT wireless/processors Synaptics is a challenger, not the leader — it competes with far larger balance sheets (Qualcomm, NXP, Nordic, Silicon Labs, Espressif on connectivity; ST/NXP/Renesas on MCU/MPU).
• Bargaining power: weak on both sides — fabless dependence on a few Asian foundries/OSATs upstream, and rising customer concentration downstream (below). This asymmetry is precisely why scale-via-M&A (the onsemi deal) is the strategically rational endgame.

Segments

Revenue by product market — the cyclical scar and the IoT inflection are both visible:

The story in one line: Core IoT is nearly doubling (FY24 $177.6M → FY26E >$385M, +40% YoY) and dragging the whole company back up, while Mobile bleeds and Enterprise/Auto grinds sideways. Geography is ~99% billed to Asia (China + Greater China the single largest bloc) — the fact that makes China SAMR the key deal-approval swing factor.

Phase B — Measure performance

Earnings Result (latest print — fiscal Q3 2026, quarter ended 28 Mar 2026)

• Revenue $294.2M, +10.4% YoY — vs guide midpoint ~$290M; a beat.
• Non-GAAP EPS $1.09 vs ~$0.84 consensus — a ~30% beat; non-GAAP net income $44.1M; non-GAAP GM 53.6%.
• GAAP net loss $(8.0)M / $(0.21) basic; GAAP GM 45.3% (+190bps YoY on mix + roll-off of acquired-intangible amortization).
• Driver: Core IoT +30.8% YoY; Mobile −16% YoY; Enterprise & Auto +9%.
• Guidance (Q4 FY26, given 7 May 2026): revenue $305M ±$10M, non-GAAP EPS $1.20 ±$0.15, non-GAAP GM 53.5%, non-GAAP opex ~$105M; mix ~33% IoT / 54% Ent&Auto / 13% Mobile; full-year Core IoT >$385M (+40%+).
• Balance-sheet flags: cash + ST-investments $404.4M (Mar-26), down $48.1M from Jun-25; 9M operating cash flow $81.8M (vs $84.8M); DSO 50d (from 45d) and inventory turns ~3x (from ~4x) — receivables and inventory both loosening modestly, worth watching but not alarming.
• Market reaction is now deal-dominated — the fundamentals barely move the stock; ON's price does.

The GAAP/non-GAAP chasm is the whole accounting story. FY25: revenue +12% yet a GAAP net loss of $(47.8)M versus robust non-GAAP profit, because GAAP absorbs ~$113M/yr stock-based comp + ~$114M/yr intangible amortization + restructuring + impairment. This is a non-GAAP-dependent equity — a genuine red flag for quality-of-earnings, mitigated only by the fact that the amortization is a declining, non-cash Broadcom-deal artifact (future intangible amortization total $202.8M, stepping down after FY27).

Full-year FY26 shape: revenue ~$1,194M (+11%); non-GAAP diluted EPS ~$4.59.

Earnings Calls (sentiment trend)

No transcripts on the research shelf (transcripts=0) — sentiment reconstructed from web-sourced quarter commentary:

• Q4 FY25 → Q1 FY26 → Q2 FY26 → Q3 FY26: tone shifted from "stabilizing off the trough / cautious" to "consistent double-digit beats, Core IoT surging." Q1 FY26 Core IoT +74% YoY, Q2 +…, Q3 +31%.
• Recurring themes: "Edge AI," "Astra ramp," "design-win pipeline ahead of internal expectations," "Broadcom Wi-Fi roadmap (Wi-Fi 7/8, UWB)," and — newly — "memory-component tightness affecting smaller-customer order timing".
• What they stopped saying: the 2023–24 "inventory correction / channel digestion" language has faded. Post-25-Jun-2026 the narrative is entirely "physical AI" and the onsemi combination.

Comps

Post-deal, SYNA's "comp" is really onsemi × 1.35. But the standalone peer set (audio/mixed-signal/connectivity/IoT semis) frames the deal-break downside and the sector's consolidation wave. Multiples are where sourced, n/a otherwise — no fabricated multiples.

Read: SYNA standalone at ~28x non-GAAP (GAAP loss) sat rich vs Cirrus (~15x) and cheap vs the IoT-recovery names (SLAB ~79x) — i.e. fairly valued on non-GAAP, expensive on GAAP. onsemi's ~$7B EVEnterprise valueWhat it would cost to buy the whole company: its market value plus its debt, minus the cash you would get with it. Often very different from market cap. ≈ ~5.9x FY26E sales / ~29x EBITDA, a full strategic price justified by ~$200M synergies + the "physical AI" narrative. The sector is consolidating hard (Skyworks/Qorvo; onsemi/Synaptics) — scale is the survival trade in mature mixed-signal/RF.

Stock-Price Catalysts (last ~5 years)

Pattern of >5% moves:

• FY22 boom (2021–22): IoT +87% YoY, revenue peaked $1.74B run-rate; stock ran to all-time highs ($290–300 area).
• 2022–23 bust: inventory correction, FY23 revenue −22% to $1,355.1M, FY24 trough $959.4M — stock more than halved.
• 2024: Astra edge-AI launch; Core IoT sequential recovery.
• Jan 2025: $200M+ Broadcom Wi-Fi asset acquisition (total consideration $247.9M) — strategic, dilutiveDilutionIssuing new shares, so each existing share owns a smaller slice of the same company. to GAAP via amortization.
• Feb 2025: CEO Michael Hurlston departs (to run Lattice) → interim CEO/CFO Ken Rizvi → Rahul Patel named CEO May 2025 — management-transition overhang.
• FY26: four straight beats (Q1 rev +14%, Q2 +13%, Q3 +10%; non-GAAP EPS beats each quarter) — re-rating.
• 25 Jun 2026: onsemi acquisition announced — SYNA +~13%, ON −~6%. This is now the dominant catalyst; from here SYNA tracks onsemi's tape and deal-approval headlines, not its own prints.

What the market actually reacts to for this name: (1) the Core-IoT growth rate and the semi inventory cycle turning; (2) big-customer concentration shifts; (3) management stability; and now (4) the arb — onsemi's stock and regulatory-clearance news.

Phase C — Judge people & books

Management

• CEO — Rahul Patel (since May 2025): 30+ yrs semis; ex-Qualcomm SVP/GM of the multi-billion-$ Connectivity, Broadband & Networking group; before that 13 yrs at Broadcom as SVP Wireless Connectivity. A wireless/connectivity operator — his hire is on-thesis with the Veros/Broadcom-Wi-Fi pivot, and arguably a grooming-for-sale profile (deep hyperscaler/OEM relationships; a year later he delivered an all-stock exit).
• CFO — Ken Rizvi: served as interim CEO Feb–May 2025; continuity in the seat.
• Track record & capital allocation: the prior regime (Hurlston) executed the touch→IoT pivot but rode the full cyclical whipsaw. Capital allocation has been buyback-heavy and value-destructive on timing — $128.3M repurchased in FY25 near cyclical lows plus the Term-Loan payoff swap into 0.75% converts; $82.6M repurchased 9M-FY26 (auth $150M, $67.4M remaining). ROE/ROIC are unflattering on GAAP (FY25 net loss; FY24 net income was ~an artifact of a $250.2M tax benefit, not operations).
• Skin in the game: professional managers, modest insider ownership (no insider-transactions data on shelf; typical of a mature, non-founder-run semi). Founding lineage is storied (Federico Faggin, 1986) but long gone.
• Red flags: executive-departure churn in Q3 FY25 drove large SBC-forfeiture swings; heavy reliance on non-GAAP presentation. Archetype: professional-manager, cycle-navigating — and, evidently, seller.

Forensic Red Flags

Income statement / balance sheet / cash flow, every figure labeled:

• GAAP–non-GAAP gap (biggest flag): ~$115M/yr SBC + ~$95–114M/yr intangible amortization convert non-GAAP profits into GAAP losses (FY25 net loss $(47.8)M; 9M-FY26 net loss $(43.4)M). Non-GAAP is directionally fair (amortization is a real non-cash Broadcom artifact that steps down after FY27), but investors are underwriting an adjusted number.
• Tax-benefit-flattered net income: FY24 GAAP net income $125.6M was essentially a $250.2M income-tax benefit (check-the-box election for the Israel sub, TCJA transition-tax reversal), not operating earnings. Quality-of-earnings caveat.
• Working capital: DSO 45→50d and inventory turns 4→3x over the year — receivables/inventory outrunning revenue modestly; inventory $139.5M (Jun-25), a named critical audit matter. Watch, not alarm.
• Goodwill/intangibles: goodwill $872.3M (Jun-25) on a ~$5B-cap company; qualitative impairment tests passed FY23–25, but a broken deal + a cyclical downturn would pressure the Broadcom-related carrying values.
• Customer concentration (rising): Q3 FY26 Customer A = 18% of revenue (unnamed); 9M-FY26 three customers >10% (A 17%, B 10%, C 11%). A single-customer loss is a material risk.
• Leverage: total debt $850M (2031 converts $450M @0.75% + 2029 notes $400M @4.00%) vs cash+STI $404.4M → ~$446M net debt. Manageable; the converts strike ~$99.69 (capped-call to $150.48).

Regulatory findings (required):

• SEC: No Litigation Releases and no AAERs naming Synaptics, 2021–2026.
• 10-K Item 3 / 10-Q legal: ordinary-course claims only; management believes none material, individually or in aggregate.
• Non-SEC / web: no material FTC/DOJ/FDA/CFPB enforcement surfaced. The forward regulatory event is not an enforcement action but the merger's antitrust + foreign-investment review — notably China SAMR (management calls it "standard," but Synaptics' ~40%+ China billing makes it the swing factor).
• Verdict: No material regulatory or legal findings — verified via SEC EDGAR EFTS (LR, AAER), web search, and 10-Q legal-proceedings note as of 2026-07-10. Clean KPMG audit, effective ICFR.

Phase D — Project & stress-test

Forward Projection

The projection that matters is now the arb, not EPS. Standalone EPS is provided for the deal-break scenario; no our model forecast logged (unattended watchlist run; base case is deal-contingent, not a clean EPS bet).

Standalone non-GAAP diluted EPS (if the deal broke and SYNA re-listed on its own trajectory):

• FY26E ~$4.59.
• FY27E — base ~$5.00; bull ~$5.75 (Astra/edge-AI ramps to production H2-CY26, IoT +35%, GM 54.5%); bear ~$3.90 (IoT decelerates, mobile accelerates its decline, GM 52%).
• FY28E — base ~$5.60.

The real "projection" — deal value: SYNA fair value ≈ 1.350 × onsemi price. At onsemi $97.87 → implied SYNA ≈ $132.1. Quoted SYNA ~$130–137 ⇒ a thin/near-zero-to-slightly-negative spread for a ~12-month deal — i.e. the market prices high completion odds and is really just long onsemi. Deal-break floor ≈ standalone ~$115–125 (pre-deal level, ~25x non-GAAP) plus the $320M reverse-termination-fee cushion (~$8.14/sh).

Bull vs Bear

Bull case. The onsemi deal completes ~mid-2027 and SYNA holders end up owning onsemi at 1.35:1 into a semiconductor + "physical-AI" upcycle; onsemi re-rates as auto/industrial recover and the combined ~$7.8B-revenue entity delivers the ~$200M synergies. Standalone-if-broken, Synaptics is a genuine Core-IoT + edge-AI inflection (Astra SL2600 into production H2-CY26, Veros Wi-Fi 7/8 roadmap from Broadcom, IoT +40%) with declining amortization drag lifting GAAP toward non-GAAP.

Bear case. (1) Deal-completion / regulatory — China SAMR (or a foreign-investment/antitrust snag) delays or blocks; the spread offers almost no compensation for that tail. (2) You're now long onsemi, not Synaptics — ON's own cyclical exposure (auto/industrial softness, its −19% drawdown into the deal) is the real driver, and ON fell ~6% on announcement (the market questioned the price). (3) Standalone fragility if broken — GAAP-lossmaking, Customer A at 18%, Mobile in secular decline, ~$446M net debt, buybacks likely frozen under the merger agreement.

Pre-mortem (18 months out, thesis broke): most likely — China SAMR withheld or conditioned clearance through 2027, the deal is terminated, onsemi pays the $320M reverse fee, and SYNA re-rates down to standalone in a softer semi tape (~$100–115) even as fundamentals were fine. Second most likely — the deal closes but onsemi de-rates, so SYNA holders "won" the arb yet lost money on the ON leg.

Contrarian view the market is under-weighting: everyone is watching the spread; few are pricing that Synaptics' standalone business is the best it's been since FY22 right as it's being sold — meaning a deal-break is not a pure disaster, it re-exposes an improving asset with a $320M cash cushion. The optionality is more symmetric than the tight spread implies.

Devil's Advocate (short-seller)

The clean short is not Synaptics — a fixed-ratio all-stock target with a $320M reverse fee is a poor outright short. The vulnerability is the onsemi leg: if you think ON is overpaying (~$7B EV / ~29x EBITDA for a cyclical mixed-signal recovery) and its auto/industrial end-markets stay soft, then owning SYNA = synthetically long an overvalued acquirer at 1.35× leverage. On the target's own merits the bear points are real: customer concentration rising to 18%, Mobile in structural decline (ASP −12%), GAAP losses masked by aggressive non-GAAP + a one-time-tax-flattered FY24, ~40%+ China revenue creating both SAMR-approval risk and export-control exposure, and fabless dependence on a few Asian foundries/OSATs. If standalone Core-IoT growth disappoints 20–30%, the deal-break floor is well below today's price. The single scenario that permanently impairs value: SAMR blocks the deal into a semiconductor downturn — but the reverse fee blunts even that.

Management Questions (ordered by information value)

1. China SAMR — what conditions/remedies, if any, do you anticipate, and what is your realistic probability and timeline for clearance given the combined China revenue exposure?
2. Under what specific circumstances does the $320M reverse-termination fee trigger, and what is the drop-dead ("End") date before either party can walk?
3. Is the 1.350 exchange ratio truly collar-free — no adjustment for onsemi share-price moves before close? What protects SYNA holders if ON de-rates 30%?
4. What are the interim operating covenants — can Synaptics still repurchase stock, take on debt, or pursue design wins/pricing freely between signing and close?
5. If the deal terminates, what is the standalone capital-allocation plan — resume buybacks, or preserve cash against the cycle?
6. Which end-customer is "Customer A" (18% of Q3-FY26 revenue), and how concentrated is the design-win pipeline behind it?
7. What is the realistic FY27–28 Core IoT / Astra edge-AI revenue ramp, and how much is contracted vs. pipeline?
8. When does GAAP profitability arrive as Broadcom-related intangible amortization steps down after FY27?
9. How much of the Broadcom Wi-Fi roadmap (Wi-Fi 7/8, UWB, GNSS) is delivered vs. still owed through Sept 2027, and what's the revenue attach?
10. What is the true gross-margin trajectory as Mobile shrinks and higher-margin IoT/licensing grows — is 54%+ non-GAAP GM structural?
11. How exposed are you to memory-component tightness and tariffs in the next two quarters' order patterns?
12. What are the retention terms for key Synaptics engineering talent (esp. ex-Broadcom wireless hires) through close?
13. How should investors think about the $450M 2031 converts (strike ~$99.69) in an all-stock acquisition — assumed, converted, or settled?
14. What synergy dis-synergies (customer overlap/conflict) exist between onsemi and Synaptics in automotive and IoT?
15. If a superior proposal emerged, does the board view $235M as a low enough break fee to entertain it?