§ 011. Bottom Line
Long, medium-sized core position, conviction 3/5. This is fundamentally different in kind from the NVTS probe — TXN is a quality-led, cycle-recovery, cash-generative core long whose main job in this cohort is to neutralize the two structural risks the rest of the cohort cannot hedge: Taiwan-tail concentration and AI-capex single-factor concentration. The screener's +1 long is intact, but the easy money is gone — the asymmetric trade was at $150–$180 in late 2024 (per financial.md), and at $280 the equity now trades at 14.4x EV/Sales / 34x forward P/E with the bull case substantially priced in. The trade now is "buy weakness, not multiple" — own size at this level only because it pays you a 2.0% dividend yield and hedges the cohort, then add aggressively into any sub-$240 reset that the September 13 2026 MOFCOM final determination is likely to deliver.
§ 022. Three-Sentence Thesis
TXN is at the trough-pivot of a six-year capex super-cycle that has already mathematically inflected: FY25 revenue +13%, Q1'26 +19% YoY with industrial +30% and data center +90%, capex peaking at $4.6B and cut to a $2–3B 2026 guide, FCF/share recovering from the $1.47 trough toward the $8+ management 2026 target — all confirmed in disclosed prints, not forecasts (per financial.md, customer.md, market.md). The cohort-load-bearing layer is the vertical-integration thesis converting at the system-BOM level, not the device-share level: TXN's GaN device share is only ~5–7% globally (#5–6 per Yole), but TXN was the named anchor partner at the NVIDIA GTC March 2026 800V architecture debut — capturing the surrounding 1,500+ analog SKUs per AI rack at TI's 58% gross margins on TI's own US fabs underwritten by ~$6.6B of CHIPS + §48D + state subsidy stack (per competitor.md, regulatory.md). The risk that bounds sizing and shapes the catalyst calendar is the September 13 2026 MOFCOM antidumping final determination on US analog imports — Bernstein sizes 11.4% of revenue exposed; we estimate 3–6% of FY26 revenue at meaningful adverse-outcome risk — so this is a Q1–Q3 2026 long that needs explicit defense planning through a Q4 2026 binary.
§ 033. The Asymmetry
Bull case payoff (~+30 to +55% over 24–30 months): The bull-case mechanics work cleanly: revenue compounds to $24–27B by FY28 on +12% CAGR, gross margin recovers from the 57% trough toward 62–63% as the 300mm cost-out delivers ~250bps of the promised 350bps and utilization fills the new SM1/LFAB1 capacity, capex roll-off from $4.6B → $2–3B → $2B normalizes by 2027, and FCF/share clears $13–15 by FY28 (per financial.md bull-case framing). Data center crosses the 15% revenue mark by FY27 (per CEO Ilan's "20% soon" framing) and MPWR/IFX socket-share losses on peak VRM/PMIC are immaterial because TI captures the >50% of the rack power BOM around them (per customer.md, market.md). At 25–30x P/FCF the equity goes to $360–435. Add ~5% in cumulative dividends. Probability: ~35–40%.
Bear case payoff (~-25 to -40%): The bear case is not a wipeout — it's the multiple compressing back to the 5-year average of 10–12x EV/Sales as: (a) MOFCOM delivers a punitive antidumping rate >15%, retroactive to Q4 2024 invoicing, compressing the highest-margin China analog franchise by 200–500bps for 18+ months; (b) the industrial restock stalls or rolls over on a 2H'26 / 1H'27 PMI contraction, leaving SM1 underutilized at ~30%; (c) the 300mm GM tailwind under-delivers (only 100–200bps of the promised 350bps); and/or (d) the auto franchise continues to bleed Chinese share. Revenue stalls at $19–21B by FY28, FCF/share at $7–9, multiple at 18–22x. Equity at $200–225. The dividend yield (now ~2.5–3.0% at lower price) provides a real floor — the bear case is bounded at ~40% downside, not 70–80% the way NVTS/WOLF can run. Probability: ~30–35%.
Modal case (-5 to +15%): TXN delivers cycle recovery roughly as guided but not as bulls hope: revenue $20–22B by FY27, GM 59–60%, FCF/share $7–9. MOFCOM is moderate (5–10% duty, partially offset by pricing). Equity drifts in $260–320. Dividend pays 2% per year. Probability: ~25–30%.
Probability-weighted return: ~+8 to +18% over 24 months (capital) + ~4–5% in cumulative dividends. Implied probability the market is pricing: roughly base-case execution at ~70% probability, which I think is slightly too high. Asymmetry ratio: ~1.4:1 capital, ~1.7:1 with dividend included. Verdict on asymmetry: This does NOT clear the 2:1 bar that justified medium sizing at $180. At $280 the asymmetry is adequate for a medium position only because (a) the dividend is real cash, (b) the cohort hedge value is structural, and (c) the bear case downside is bounded by quality. The asymmetric trade was 12–18 months ago. Don't kid yourself about that.
§ 044. What You Have to Believe
The long requires all five of these propositions to hold. Each is stated as a falsifiable claim a 2027 retrospective could clearly score.
- The "$8+ FCF/share for 2026" guide is materially closer to truth than the 2022 framework's "$9 FCF/share at trough" promise was. Per financial.md: the 2022 framework missed by 6x — actual trough was $1.47, not $9. Management has not formally re-issued the through-cycle framework. The $8+ 2026 guide is more credible because (a) the capex side is mechanically locked ($2–3B vs $4.6B), (b) Q1'26 revenue +19% and FCF/share TTM at $4.07 are corroborating prints, not projections, and (c) the dividend coverage math forces this number to be defended. But the asymmetry of management-credibility risk is real — investors are paying a quality multiple to a team that just missed by 6x. Falsifiable by: Q3'26 / Q4'26 prints sustaining FCF/share run-rate above $7 annualized.
- 300mm cost-out delivers at least 200bps of the promised 350bps GM tailwind by FY28. Per market.md, the 300mm-vs-200mm cost differential is real (~40% lower per-die cost, 2.5x more chips per wafer); per supply-chain.md, SM1 entered HVM December 2025 at low single-digit utilization and the depreciation curve is loading now. Q1'26 GM at 58% (up 210bps QoQ) is directionally consistent. The bull case requires utilization to fill the new fabs and the cost-out to materialize. Falsifiable by: GM stuck below 60% through FY27 despite revenue +15% YoY would invalidate the 300mm thesis.
- The vertical-integration thesis converts at the system-BOM level, not the device-share level. Per competitor.md finding #5 (refinement-log C-TXN-1): TXN GaN device share is only ~5–7% globally. Innoscience (~30%), NVTS (~17%), POWI (~17%) all lead on discrete share. The bet is that TXN captures more system-BOM dollars by feeding internal GaN into the PMIC + isolated bus converter + multiphase buck stack TI debuted at NVIDIA GTC March 2026 — the broadest-stack 800V offer. Q1'26 data center +90% YoY at $540M run-rate is supportive. Falsifiable by: a NVIDIA Vera Rubin Ultra (2027) reference design BOM that names MPWR/IFX/Innoscience but not TI as a primary system partner.
- MOFCOM final determination on September 13 2026 lands at a duty rate <15%, OR is structured in a way TI absorbs into pricing rather than losing the China franchise. Per regulatory.md and customer.md: TI is named primary import respondent; cumulative 51.77% Chinese price decline 2022–2024 cited as injury; the highest-margin commodity-analog slice (Personal Electronics + low-end industrial) is the target. Bernstein sizes 11.4% revenue exposure (~$2B at run-rate); my estimate is 3–6% of FY26 revenue at adverse-outcome risk, with the rest absorbable. Falsifiable by: MOFCOM affirming retroactive duties >20% with effect on Q4 2024+ invoicing on September 13 2026.
- Embedded-analog restock holds through 2026–27 — no PMI rollover or hard recession in 2H'26 / 1H'27. Per customer.md, market.md, macro.md: the restock is in the prints (Q1'26 industrial +30%, lead times 4–8 weeks, book-to-bill above 1.0, distributor inventory at multi-year lows). The bull case requires this cycle to sustain through 2027. TXN's macro hedging value is real across macro scenarios, but the bull-case revenue path requires the cycle. Falsifiable by: any quarter through Q4'26 printing industrial revenue growth below +10% YoY would signal cycle stall.
Honest read: Belief #4 (MOFCOM) is the binding near-term falsifier with the highest probability and largest asymmetric impact. Belief #2 (300mm cost-out delivery) is the binding medium-term falsifier — if it under-delivers, the multiple compression is permanent because the structural quality of the franchise is no longer differentiated. Belief #5 (cycle holds) is the macro precondition. #3 is the load-bearing cohort claim — it's why the screener and the cohort frame put TXN in the deep-dive at all.
§ 055. What Kills the Long
Three named scenarios with leading indicators a PM should automate around.
Kill scenario 1: MOFCOM punitive outcome on September 13, 2026
Mechanism. MOFCOM final determination affirms antidumping duties on US analog imports at >15% effective rate, with retroactive application to Q4 2024 / Q1 2025 invoicing. Chinese OEM procurement pivots to SG Micro / Silergy / 3PEAK / Bright Power / Awinic on accelerated qualification timelines (per competitor.md, customer.md, market.md). The 11.4% revenue exposure (Bernstein) translates to 3–6% of FY26 revenue at meaningful margin compression — 200–500bps GM hit on the China franchise specifically — and a structural rather than cyclical step-down because qualification cycles are 6–18 months for industrial / 18–36 months for automotive, so the share loss extends through 2028+.
Leading indicator (watch). (a) MOFCOM mid-investigation disclosures August/September 2026 — preliminary findings or Chinese trade-press signal-floating; (b) US-China trade-deal context heading into November 2026 mid-terms — a punitive MOFCOM outcome is more likely if US escalates first; (c) TXN preparedness commentary on Q2'26 / Q3'26 calls — "we're prepared for a range of outcomes" is the standard hedge, but specific pricing-action commentary signals worse expected outcomes; (d) Chinese trade-press (Yicai Global, JW iJiwei) coverage of domestic-share narrative.
Kill action. A duty rate >20% with retroactive effect forces immediate position cut to 1/3 size. A "deal" outcome (no duty / symbolic duty <5%) is thesis-confirming and triggers add-back. The position should be defended through the Q4 binary at 1/2 size — this is the catalyst-aware sizing constraint.
Kill scenario 2: 300mm cost-out under-delivers; SM1 utilization stalls
Mechanism. Per financial.md and supply-chain.md: SM1 entered HVM December 2025 at low single-digit utilization. The depreciation curve loads regardless of utilization. If embedded-analog cycle stalls in 2H'26 (consumer/industrial PMI rolls over, China industrial demand weakens, auto Tier-1s cut harder), SM1 utilization stays below 50% through 2027. The promised 350bps of GM tailwind from 300mm cost-out under-delivers — only 100bps materializes by FY28. Gross margin stuck at 57–59%; FCF/share stalls at $5–6. The "harvest-mode FCF inflection" narrative breaks. Multiple compresses to ADI-style 10–12x EV/Sales (~$200–230 implied).
Leading indicator (watch). (a) SM1 utilization disclosure on Q3'26 / Q4'26 calls — anything below 40% by Q4'26 is the early warning; (b) gross margin trajectory — Q2'26 / Q3'26 prints stalling below 59% with revenue growing +15%+ signals the cost-out isn't materializing; (c) capex re-creep — any 2H'26 disclosure of capex tracking above $750M/quarter (cumulative > $3B FY26) signals management has lost discipline (per financial.md warning that TI overshot the 2022 capex framework by 30%+); (d) auto / industrial restock stalling in cohort-comp prints (NXP, MCHP, ADI Q3'26 calls).
Kill action. A combination of capex re-creep + GM stuck below 59% + revenue growth decelerating below +10% YoY = exit the medium position back to small. Each individual signal is yellow-flag, not red.
Kill scenario 3: Hyperscaler ASIC integration consumes peripheral analog content
Mechanism. TI's AI-DC bull case is built on the broad analog content per server / per rack — the >50% of the BOM around the headline VRM and PMIC sockets MPWR and IFX win. If hyperscalers (NVIDIA absorbing PMIC into BlueField/Vera; AWS Trainium / Google TPU custom silicon doing the same) integrate the supervisory, sequencing, hot-swap, and isolation analog into custom silicon, TI's per-rack content compresses 30–50% by 2028. This is the AI-DC architectural disruption flagged in market.md — low probability near-term (2026–27), medium-term watch item (2028–30). It would invalidate the load-bearing claim that 1,500+ analog parts per rack is structurally durable.
Leading indicator (watch). (a) NVIDIA Vera Rubin Ultra (2027 timeline) reference architecture disclosure — does it pull more analog content into NVIDIA's own silicon? (b) Meta Catalina / Microsoft HVDC architecture roadmap commentary — they are the hyperscalers most likely to push custom power architectures; (c) any disclosed TI design-loss on a major hyperscaler reference platform; (d) MPWR's response — if MPWR starts losing socket-share to ASIC integration too, the disruption is broad-based.
Kill action. A named hyperscaler reference design replacing TI's hot-swap / supervisory / sequencing analog with custom silicon is the trigger. Reduce position to 1/3 size and re-evaluate as a yield-and-cycle name without the AI-DC growth premium.
§ 066. What Proves the Long
Three dated catalysts that, if hit, build conviction toward 4 or higher.
Catalyst 1: April 2, 2026 — ITC final determination, Infineon v. Innoscience (337-TA-1407)
Why it matters. Per regulatory.md and the cohort refinement-log.md C-NVTS-1 finding: the ALJ initial determination already favored Innoscience on two patents; Commission affirms expected ~70% probability. The cohort-relevant read-through is asymmetric: TXN's vertical-integration moat is cost-structure-based, not patent-based. If GaN patents fall, NVTS's density-mindshare moat erodes faster than TXN's vertical-integration moat. The base case (Innoscience prevails) is thesis-supporting for TXN within the GaN four-way race. A surprise patent-holds outcome would benefit Infineon more than TXN but would not invalidate the TXN thesis.
What I'm watching. Specifically: scope of any non-infringement ruling and whether it extends to product-import enjoinment. A clean Innoscience win + aggressive US-market entry from Innoscience would intensify cost-curve pressure on TI's commodity-tier 300mm bet — but TI's strategic answer is the system-BOM bet, not the discrete-FET cost competition.
Conviction shift if hit. Innoscience prevails (base case): conviction stays at 3, no upside to TXN sizing. Infineon wins ≥1 patent (surprise): conviction unchanged for TXN; mostly relevant as an upgrade signal for the broader GaN cohort.
Catalyst 2: Q3 2026 / OCP October 2026 — sustained data-center revenue print + named box-builder reference design
Why it matters. Per customer.md and market.md: data-center revenue $1.5B FY25 (9% of revenue) accelerating to +90% YoY in Q1'26. The remaining named-partner gap — Vertiv / Eaton / Schneider / Delta publicly citing TI in their 800V reference BOMs — is the next disclosure to watch. The NVIDIA GTC March 2026 anchor reference architecture is the upstream catalyst; box-builders typically follow within 6–12 months. The Q3'26 print is the cleanest signal of whether (a) DC revenue sustains 50%+ YoY growth and crosses the 12–14% mark of FY26 revenue (validating the "20% soon" CEO framing), and (b) box-builder reference-design disclosures land at OCP October 2026.
What I'm watching. (a) Q3'26 data-center revenue print — anything sustaining $600M+ quarterly and YoY growth above 50% confirms the trajectory; (b) Q3'26 / Q4'26 management commentary on data center as % of revenue — explicit reaffirmation of "20% soon" with a dated framing (e.g., "by FY27" or "exiting FY27") is the bull-case validation; (c) OCP Summit October 2026 reference-design disclosures from Vertiv / Eaton / Schneider / Delta — any BOM-level inclusion of TI 800V architecture is structurally bullish.
Conviction shift if hit. Data center sustained at +50% YoY + named box-builder BOM inclusion → conviction 4, sizing upgrade to medium-large. Just the revenue print without box-builder disclosure → conviction stays at 3 but with bull-case probability shifted upward.
Catalyst 3: September 13, 2026 — MOFCOM antidumping final determination
Why it matters. This is the single largest binary risk on the calendar (per regulatory.md, customer.md, market.md, and refinement-log C-TXN-1 finding). Bernstein sizes 11.4% revenue exposure; my estimate is 3–6% of FY26 revenue at adverse-outcome risk. The outcome has three meaningful states: (a) benign (no duty / symbolic <5% / structured deal in trade-talk context) — relief rally, multiple expansion 1–2 turns; (b) moderate (5–15% duty, absorbable through pricing) — modest negative, ~5–10% equity drawdown, thesis intact; (c) punitive (>15% duty, retroactive) — Kill scenario 1 fires, 15–25% drawdown, thesis at risk.
What I'm watching. Pre-determination signals: (a) preliminary findings disclosure (typically 60–90 days pre-FD); (b) US-China trade-deal context — does the November 2026 mid-term cycle drive escalation or de-escalation? (c) TI's pricing actions in China through Q2'26 / Q3'26 — accelerated price cuts signal management is preparing for a punitive outcome; (d) Chinese domestic-vendor trade-press positioning.
Conviction shift if hit. Benign outcome → conviction 4, sizing upgrade. Moderate → conviction stays at 3, no sizing change. Punitive → conviction 2, sizing cut to small / 1/3 of medium.
§ 077. Sizing & Expression
Initial size: medium (core, ~2.5–3.5% of NAV gross long). This is not a probe like NVTS. TXN is structural cohort hedge value plus quality-led cycle exposure plus a real 2.0% dividend yield. The core position is justified by the combination of (a) bounded bear case (~40% max downside vs NVTS's ~75%, WOLF's ~70%), (b) structural Taiwan-tail and AI-capex hedging value the cohort needs, and (c) cleanly inflecting fundamentals (Q1'26 +19% revenue, FCF/share recovering from $1.47 to $4.07 TTM, capex cut to $2–3B). Initial entry: scale-in over 4–6 weeks, full medium size by mid-July 2026 ahead of the August earnings print.
Where to add (toward medium-large ~4–5%): any of three triggers fires —
- (a) MOFCOM benign outcome on September 13 2026 (target add zone: $290–340 if it rallies on the news);
- (b) Q3'26 data-center revenue sustaining +50% YoY + a named Vertiv / Eaton / Schneider BOM-level reference-design disclosure at OCP October 2026;
- (c) price reset to sub-$240 without a fundamental fault — most likely in late summer 2026 ahead of the MOFCOM binary, when the multiple compresses on event risk rather than business deterioration. This is the highest-quality add zone because the asymmetry meaningfully improves below $240 (bear case ~$200–225 from there is 10–15% downside, bull case ~$360–400 is 50–65% upside — that's a clean 4:1 asymmetry vs the 1.4:1 you get at $280).
Where to trim (back to small or out):
- (a) Punitive MOFCOM outcome (Kill scenario 1) → cut to 1/3 size;
- (b) GM stuck below 59% through Q4'26 + capex re-creep (Kill scenario 2) → cut to 1/3 size;
- (c) Equity above $360 on consensus bull-case priced in (P/FCF >32x on FY28 earnings) → trim half. The equity at $360 with no further conviction upgrade is over-paying for the bull case.
The NVTS pair-trade explicitly
Per ..\NVTS\thesis.md Section 7 Option B: the recommended structure is TXN long / NVTS long sized 1:3 in TXN's favor — TXN at 1.5–2% / NVTS at 0.5–0.75% NAV. The pair logic and its limits, restated for TXN:
Why the pair structure works:
- TXN is the lower-beta, higher-conviction expression of the chip-to-grid GaN bet — vertical integration, US-fab, 95%+ internal manufacturing target by 2030, $6.6B subsidy stack, no Taiwan-foundry transition risk, proper cash flows, 2.0% dividend.
- NVTS is the higher-beta, architectural-bet overlay — single-stage 800V-to-6V, GaN-IC patent estate, asymmetric upside on Kyber design-in conversion.
- The pair captures both the "vertical integration / 300mm cost curve dominates" outcome (TXN wins) AND the "density wins on architectural first-mover lock-in" outcome (NVTS wins). They are complementary on outcome, not redundant.
Where the pair breaks: Per financial.md stress test: the pair structurally breaks at TXN $210. Below that, TXN's drag on the larger leg overcomes NVTS's optionality on the smaller. From $280, that means $70/share of TXN draw-down room (-25%) before the pair thesis breaks. This corresponds roughly to the bear case but not the worst tail. The pair holds at +/-0% to -10% in TXN's bear case ($200–225); bull case carries the pair to +30–50%.
The timing-correlation trap: Per financial.md and customer.md: NVTS's bull-case design-wins happen on the same NVIDIA Kyber/Rubin Ultra timeline (2027–28) as TXN's full-cycle FCF recovery. If both miss timing by a year, both legs of the pair under-perform together. This is the single biggest structural risk to the pair that you cannot diversify away within these two names. It's the cohort-level "calendar mismatch" risk (synthesis Section 6 contested claim #15) expressed at the trade level.
Practical recommendation: Run the pair at TXN 2.5% / NVTS 0.75% — slightly larger TXN side than the 3:1 minimum to over-weight the higher-conviction leg. The pair is structurally sound but timing-correlated; treat it as one position, not two, when sizing portfolio risk.
Holding horizon: 18–30 months structural for TXN. The trade is the cycle recovery + 300mm cost-out + AI-DC mix shift compounding through FY28. If the cycle recovery has not converted to 60%+ GM and FCF/share above $7 by Q4'27, the position should be exited regardless of where the stock is.
§ 088. Cohort Fit
TXN is the cohort's primary macro hedge AND primary Taiwan-tail hedge. This is the single most important cohort-construction finding from the C-TXN-1 expansion run (per ..\refinement-log.md finding #4). State this without burying it: the cohort's biggest unhedged risk before this analyst cycle was Taiwan-tail concentration (synthesis Open Question §1) — and adding TXN materially closes that gap.
What TXN neutralizes (per macro.md cohort-fit table):
| Cohort risk | TXN role | Magnitude |
|---|---|---|
| Taiwan-tail concentration | HEDGES structurally — ~10–15% Taiwan exposure vs 60–90% for NVDA / AVGO / NVTS through mid-2027 / TSM 92% | Largest cohort hedge value |
| AI-capex single-factor concentration | PARTIALLY HEDGES — industrial/auto diversification means three independent demand curves, not one; AI-DC is only ~9–14% of revenue | Moderate |
| Rate duration | PARTIALLY HEDGES — mature, dividend-paying, FCF-positive vs duration-heavy NVTS / NVDA-on-multiple. 2.0% dividend yield acts as buyer-of-last-resort floor below ~$200 | Moderate |
| Recession | PARTIALLY HEDGES — defended dividend (22 consecutive annual increases), through-cycle FCF positive, FY09 print -22% revenue with dividend held flat | Moderate |
What TXN does NOT hedge:
| Cohort risk | TXN role | Magnitude |
|---|---|---|
| China structural decoupling | AMPLIFIES modestly — Chengdu assembly+test (~25–30% of back-end), 22% China revenue, MOFCOM probe directly targets TI's highest-margin China analog franchise | Small drag, but binary risk concentrated in Sept 2026 |
| USD strength | NEUTRAL — EUR/USD revenue compression is the largest cohort exposure to FX, but offsetting Aizu JPY cost base + USD-cost dominance create rough wash | Neutral |
| Critical minerals (gallium for GaN) | AMPLIFIES modestly — GaN program shares the gallium supply chain with NVTS and IFX, though TI's 200mm scale makes pass-through easier | Small |
The honest cohort frame: TXN provides genuine cohort-hedge value on the four largest cohort-shared risks (Taiwan-tail, AI-capex, rate duration, recession), but it is long China structural — and that exposure concentrates in a single binary date (September 13 2026 MOFCOM). The cohort-portfolio expression is therefore: own TXN at medium size, but don't let the binary date catch the position naked. The Q4 2026 risk-management plan is non-trivial — consider a put hedge into the September FD if the implied vol cheapens, or trim 25% of the position in late August 2026 ahead of the determination.
Comparison to the WOLF short structure (cohort cautionary tale): Per ..\WOLF\thesis.md: WOLF is the cleanest structural-decline short, but the trade had to be expressed as a pair (long ON or IFX) to neutralize cohort SiC-bullish beta and re-engineer the asymmetry to ~3:1. The lesson translates: TXN long is structurally sound, but the asymmetry at $280 is 1.4:1 single-leg. Re-engineering via the NVTS pair (TXN long / NVTS long at 1:3) doesn't add hedge value — it adds asymmetry on the architectural-bet leg. The cleaner asymmetry-improvement is adding into sub-$240 weakness, not adding NVTS. Treat the NVTS pair as a distinct architectural call, not as TXN's hedge.
Comparison to NVDA / AVGO long expressions: Per ..\NVDA\thesis.md and ..\AVGO\thesis.md: NVDA and AVGO are the cohort's pure AI-capex longs, both Taiwan-tail concentrated, both at premium multiples. TXN at 14.4x EV/Sales is cheaper than MPWR (25.9x), comparable to ADI (16.9x), and structurally lower-beta than NVDA / AVGO. The cohort-construction logic is: TXN replaces a fraction of NVDA/AVGO sizing as the lower-beta, Taiwan-light, dividend-paying expression of the same chip-to-grid demand. A PM running NVDA-medium / AVGO-medium / TXN-medium gets meaningfully better risk-adjusted exposure than NVDA-large / AVGO-large alone.
§ 099. Open Questions / Known Unknowns
The things I don't know that would meaningfully shift the thesis:
- The actual pacing of the MOFCOM investigation. No public preliminary determination has been released; the September 13, 2026 final determination date is procedural, not substantive. If preliminary findings leak in July/August 2026, the equity may move ahead of the FD itself. Required research: MOFCOM precedent on similar cases (auto-related antidumping, US-export-side); base rates of US-China antidumping outcomes 2020–2026.
- 300mm GaN production qualification at customer. Per supply-chain.md: TI's 300mm GaN pilot is complete; customer samples shipping. But "shipping samples" is not "qualified at customer for HVM." Production-qualification timeline at NVIDIA / Vertiv / Schneider / Delta is undisclosed — typically 12–24 months from samples. If the qualification slips beyond 1H'28, the cost-curve advantage is delayed and the bull-case GaN system-BOM thesis weakens (per financial.md bull-case framing).
- The Silicon Labs $7.5B integration risk (closing 1H 2027). Per competitor.md and financial.md: TI's largest M&A since National Semiconductor 2011. The $450M cost-synergy plan is defensible; integration risk is non-zero. The deal is a $7.5B drain on the harvest-mode FCF deployment that bull-case shareholders are otherwise pricing into the multiple. Required research: Silicon Labs's TSMC outsource volume + how it absorbs into TI's vertical fab footprint; typical embedded-MCU integration timelines (3–5 years for full revenue synergies based on cohort precedents).
- The MPWR / IFX socket-share trajectory through Vera Rubin Ultra (2027). Per competitor.md: MPWR positioned for ~70% of NVIDIA Vera Rubin VRM share; IFX projected 60–70% Blackwell PMIC. The current data ends with NVIDIA Vera Rubin (2026 platform). If MPWR's share expands into the surrounding analog content TI currently captures, TI's per-rack content compresses and the AI-DC bull case weakens. If MPWR's share holds flat on the peak sockets while TI captures the surrounding 1,500+ SKUs as designed, the bull case validates.
- Through-cycle FCF/share framework as actually re-issued. Management has not formally re-issued the "through-cycle $9 FCF/share trough" framework after the 6x miss. Forward guidance is "$8+ for 2026" — but no explicit through-cycle anchor for FY28 / FY30. If management quietly updates the framework on the Q4'26 capital-management review, that's the cleanest forward read on bull-case credibility.
- The Vertiv / Eaton / Schneider / Delta box-builder named-partner gap. Per customer.md: as of May 2026, no public named TI 800V GaN reference design at the major box-builders. The NVIDIA GTC March 2026 anchor is upstream; downstream box-builder disclosure is the next confirmation. Required research: OCP Summit October 2026 disclosure tracking; Vertiv / Eaton / Schneider press cycle on 800V partner ecosystem.
- The CHIPS excess-profits clawback dormant risk. Per regulatory.md: if Sherman/Lehi vertical-integration thesis works too well and the fabs become genuinely profitable above the application projection, Commerce can recover up to 75% of the $1.61B direct grant. This is a "subsidy on conditional success" — net positive in expected value, but compresses the upside scenario by ~$500M–$1B. The clawback math is opaque; a clean disclosure of the trigger thresholds would clarify the bull-case asymmetry.
- The dividend-defense calculus if a deeper recession hits. TI defended the dividend through FY09 (-22% revenue, dividend held flat) and through the FY22–24 destock. Net debt is $9.2B (up from net cash $332M in FY22) — debt-funded the dividend through trough. A 2026 hard recession layered on top of MOFCOM punitive outcome is the tail scenario where the dividend math gets stressed. Probability low; magnitude meaningful for the equity floor.
§ 10Cross-References
- Cohort synthesis:
C:/Users/mosu9/.claude/investment-research/semiconductor-industry/synthesis.md(Section 0 cohort scope; Section 2 chip-to-grid stack at L8b/L8c; Section 3 themes #1 voltage-stack / #3 GaN four-way race / #11 embedded-analog cycle; Section 6 contested claims #2/#3/#15; Open Question §1 Taiwan-tail concentration — materially closed by TXN per refinement-log finding #4) - Refinement log:
C:/Users/mosu9/.claude/investment-research/semiconductor-industry/refinement-log.md— C-NVTS-1 (Innoscience #1 share, four-way race, TSMC GaN exit July 2027); C-TXN-1 findings #4 (TXN cohort's primary Taiwan-tail hedge) and #5 (TXN GaN device share only 5–7%, system-BOM bet not device-socket bet) — both promoted to thesis - NVTS pair-trade reference: C:/Users/mosu9/.claude/investment-research/semiconductor-industry/NVTS/thesis.md — Section 7 Option B (TXN long / NVTS long at 1:3 in TXN's favor); pair breaks at TXN $210 per financial.md stress test; structurally sound but timing-correlated
- Cohort comp positioning: C:/Users/mosu9/.claude/investment-research/semiconductor-industry/WOLF/thesis.md (structural-short cautionary tale; pair-trade structure lesson); C:/Users/mosu9/.claude/investment-research/semiconductor-industry/NVDA/thesis.md and C:/Users/mosu9/.claude/investment-research/semiconductor-industry/AVGO/thesis.md (Taiwan-tail concentrated AI-capex longs that TXN partially offsets)
- Analyst memos (all in
C:/Users/mosu9/.claude/investment-research/semiconductor-industry/TXN/): competitor.md, supply-chain.md, customer.md, financial.md, market.md, regulatory.md, macro.md
§ 11Conviction Distribution Across Analysts
| Dimension | Conviction | Direction |
|---|---|---|
| Competitor | 3 | lean long |
| Supply chain | 4 | long |
| Customer | 3 | lean long |
| Financial | 3 | lean long, quality-not-asymmetry |
| Market positioning | 4 | long |
| Regulatory | 4 | long |
| Macro | 3 | lean long |
| PM (you) | 3 | long, medium core |
The 7-analyst average is 3.43. PM conviction is 3, deliberately below the unweighted average. The reason: the two highest-conviction memos (regulatory 4, market 4) are most exposed to the MOFCOM binary that lands in Q3 2026 — regulatory's tailwind framing and market's structural-share durability framing both partially invert in a punitive MOFCOM outcome. The two 3s on competitor and customer reflect the actual socket-level competitive realities (MPWR / IFX peak-socket share losses, GaN device share only 5–7%) that bound the bull case. The financial 3 is the most important constraint — at $280 the asymmetry is 1.4:1 single-leg, not the 2:1+ that justifies higher conviction.
The PM-conviction-3 read corresponds to medium core sizing because:
- Quality and bounded downside justify medium over small (the bear case is ~40%, not 70%);
- Cohort hedge value justifies a real position (Taiwan-tail neutralization is structural);
- The cycle inflection is real and confirmed in prints (Q1'26 +19%, FCF/share recovering);
- But the asymmetry doesn't justify large — the easy money was at $150–180, and the September MOFCOM binary makes Q4 2026 a defensive-posture quarter.
This is the cohort's quality core long, not the cohort's highest-conviction trade. The screener is intact (+1 long); the deep work confirms direction with discipline on size and timing. Buy weakness, defend the binary, don't pay the multiple.
Works cited
- Texas Instruments 2024 Annual Report (10-K filed Feb 14, 2025)
- Manufacturing footprint: RFAB1+RFAB2 (Richardson TX), LFAB1+LFAB2 (Lehi UT), SM1+SM2 (Sherman TX) all 300mm
- 95% internal-mfg target by 2030 with >80% on 300mm
- 'Foundries and subcontractors used selectively to supplement internal capacity'
- + 1 more
- Texas Instruments 2025 Annual Report / Notice of 2026 Annual Meeting — Investor Relations
- >80% direct-customer revenue 2025 (vs ~33% in 2019)
- No single customer >10% of revenue
- >40% of revenue from outside top 100 customers
- + 1 more
- Texas Instruments Q3 2025 Earnings Call Transcript — The Motley Fool
- Q3 2025 industrial +25% YoY; automotive +HSD% YoY
- Customer inventories at low levels; depletion behind us
- China commentary in Q2 2025 (referenced): China +19% sequentially, +32% YoY
- TI Capital Management 2022 earlier framework
- 9 FCF/share at trough prior framework; actual trough 1.47 (6x miss)
- TI Capital Management Review Haviv Ilan CEO Feb 2026
- TI SAM >60B; 8+ FCF/share 2026 guide; capex 2-3B 2026
- TI Q3 2025 earnings call transcript (Oct 21, 2025)
- Q3 2025 capex / depreciation commentary
- End-market mix data underlying inventory-days estimate
- TXN Q1 2026 Earnings Transcript — The Motley Fool
- Q1 2026 revenue $4.83B +19% YoY; analog revenue $3.92B +22% YoY; embedded +12%
- Data center +90% YoY; industrial +30% YoY
- Gross margin 58.0%; operating margin 37%
- + 1 more
- Analog IC Market Trends Coherent Market Insights
- 101.5B 2024 TAM; 6.1% CAGR through 2033
- Analog rankings: Top 10 suppliers own 68% market share — EDN
- Top 10 analog suppliers control 68% of market — TI #1 at ~19%
- ADI #2; Infineon, STM, NXP, ON, Renesas, Microchip, MPWR rounding the top 10
- Analog Semiconductor Market Fortune Business Insights
- 87.5B 2024 TAM; 7.4% CAGR through 2034
- Analog Semiconductor Market Mordor Intelligence
- 130B 2031 projection; 6-7% CAGR
- Analog Semiconductor Market Precedence Research
- 96.4B 2025; 5.9% CAGR
- Analog Semiconductors Market GM Insights
- Conservative 3.3% CAGR 2025-2030
- ATREG — Texas Instruments Lehi, UT fab acquisition case study (Micron-to-TI Q4 2021)
- TI acquired Lehi fab from Micron for $900M in Q4 2021
- Production at LFAB1 began Q4 2022
- EDN Analog Rankings Top 10 Suppliers 68 Percent Market Share
- TXN 19% ADI 12% IFX 10-11% STM 9% NXP 8%; HHI estimation basis
- Embedded Processor Market Straits Research
- 23.4B 2024; 39.9B 2033; 6.1% CAGR
- Global Power Semiconductors AI Infrastructure Atlas Peak Research
- Power semi stack 56.9-57B 2025; AI-DC content per rack
- Global Semiconductor Market grows 26% in 2025 WSTS
- 2025 total semiconductor market actuals
- Microcontrollers Target 34B by 2030 Yole Edge AI Vision
- Yole MCU 34B by 2030; 6% CAGR; auto largest at 13B
- MPWR — The Power Behind the Brain: A Deep Dive into MPWR in the AI Era — FinancialContent
- MPWR 26.4% revenue growth 2025 to $2.8B
- MPWR sampling 800VDC solutions for Blackwell / Vera Rubin
- Positioned for ~70% of NVIDIA Vera Rubin VRM share
- + 1 more
- Power Management IC Market 69.54B by 2035 Astute Analytica
- PMIC 29.25B 2025; 10.1% CAGR; intelligent PMICs 1.50-3.00 vs 0.10 commodity
- Texas Instruments Q1 FY 2026: Data Center and Industrial Demand Lift Outlook — Futurum Group
- TI Q1 2026 data center and industrial demand commentary
- Outlook framing for FY26 by analyst coverage
- Texas Instruments vs Analog Devices comparative analysis — Artificall
- TI 19% analog share; ADI 12%
- TI vertical integration vs ADI capex-light TSMC outsource model
- Q4 2024 GM: TI 58.14%, STM 37.7%, IFX 39.2%, ADI 58.0%
- The Analog Giant's Rebirth: A Comprehensive Research Feature on TXN — FinancialContent
- TXN narrative framing entering 2026: capex-cycle to harvest-mode transition
- Analog $14B FY25 revenue; embedded $2.7B
- TI & Silicon Labs: a strategic move reshaping the embedded & wireless landscape — Yole Group
- Wireless 'front end' of connected systems: protocols, certification, software stacks
- Embedded MCU + wireless platform consolidation context
- Trendforce — TI to Receive USD 1.6 Billion Funding for Building Three 300mm Fabs (Aug 2024)
- Independent CHIPS Act terms corroboration
- Sherman SM1+SM2 + Lehi LFAB2 site allocation
- TXN's Market Share Relative to Competitors, Q1 2025 — CSIMarket
- Analog segment competitive share data Q1 2025
- WSTS Semiconductor Market Forecast Spring 2025
- Analog +7.5% YoY for 2025; total semi +11.2%
- Yole Group Data Center Semiconductor Trends 2025
- AI reshaping compute and analog/power market segments
- Yole Group — Power GaN 2025 (industry report; market share)
- GaN device 2024 share (Yole/BoA): Innoscience ~30%, NVTS ~17%, POWI ~17%, EPC ~12-15%, Infineon ~10%
- TI GaN device share estimated ~5-7% (#5-#6)
- China Analog IC Probe Benefit Chinese Suppliers Electropages
- Chinese beneficiaries: Silergy SGMicro Southchip Joulwatt Novosense
- China launches anti-dumping investigation into analog IC chips from US — Global Times
- Investigation context as retaliation for US Entity List expansion
- Jiangsu Semiconductor Industry Association as filing party
- China Probe on US Analog Chips Could Unlock USD350 Million Market for Local Firms — Yicai Global
- $350M Chinese domestic-market capture potential from MOFCOM probe outcomes
- Investigation 1-year duration framing
- China tariff investigation analog chip Bernstein Yahoo Finance
- Bernstein: TXN 11.4% revenue exposure China antidumping probe; ADI 7.8%; ON 10.2%
- China's Latest Analog IC Probe To Benefit Chinese Suppliers — Electropages
- Chinese domestic-share beneficiaries: SG Micro, 3PEAK, Silergy, Southchip, Joulwatt, Novosense
- Chinese analog chip vendors brace for impact as Texas Instruments slashes prices — JW iJiwei
- TI broad price reductions in China analog 2024-2025
- Chinese competitors named: SG Micro, Bright Power, Awinic — power management primary target
- Data center boom continues to buoy Texas Instruments — Manufacturing Dive (Q1 2026)
- Q1 2026 data-center growth and Silicon Labs deal context
- Industrial / automotive / data-center mix at 75% of revenue
- MPWR Falls Amid Risk to Nvidia Allocation, Edgewater Research Reports — Yahoo Finance
- Infineon projected 60-70% share of NVIDIA Blackwell power management
- Renesas projected 'meaningful' share gains in NVIDIA digital power for Blackwell/Hopper
- Competitive structure for AI-DC PMIC sockets
- NVIDIA Developer Blog — Building the 800 VDC Ecosystem for Efficient, Scalable AI Factories
- NVIDIA 14-vendor 800V silicon-partner list including TI alongside NVTS, Infineon, EPC, MPS
- Participation, not exclusivity
- Semiconductor Today — TI adds 200mm GaN power semiconductor production in Japan (Oct 2024)
- Independent corroboration of Aizu 200mm GaN qualification
- 4× internal GaN capacity claim
- Texas Instruments breaks ground on new 300-mm semiconductor wafer fabrication plant in Utah (LFAB2; Nov 2, 2023)
- $11B LFAB2 investment
- First production target 2026
- 100% renewable; ~800 additional jobs
- Texas Instruments opens its second assembly and test factory in Melaka, Malaysia (Nov 2025)
- Confirmation of seven A/T sites globally
- Melaka second fab opens Nov 2025
- Texas Instruments plans to invest more than $60 billion to manufacture billions of foundational semiconductors in the U.S. (Jun 18, 2025)
- $60B+ headline US investment plan
- Sherman + Lehi + Richardson sites anchor the reshoring footprint
- Texas Instruments Q1 2026 earnings beat on AI data center demand — Yahoo Finance
- Data center segment +90% YoY in Q1 2026
- AI data center demand a primary growth driver
- Texas Instruments Q1 2026 Earnings Yahoo Finance
- Q1 2026 4.83B +19% YoY; analog +22%; data center +90%; industrial +30%
- Texas Instruments sees data center revenue surge 50% — Digitimes (Sep 2025)
- TI data-center revenue trajectory 2025; >$1B FY25 forecast
- Texas Instruments signs preliminary agreement to receive up to $1.6 billion in CHIPS Act proposed funding (Aug 16, 2024)
- $1.6B CHIPS direct grant for SM1, SM2, LFAB2
- Estimated $6-8B Investment Tax Credit (Section 48D) on qualified manufacturing investments
- TI announces award agreement for CHIPS and Science Act funding — TI.com (Dec 20, 2024)
- $1.6B CHIPS Act direct funding award
- Three 300mm fabs: Sherman SM1+SM2 (Texas), Lehi LFAB2 (Utah)
- Investments through 2029 underwritten
- TI begins production at Sherman, TX 300mm fab (SM1) — TI.com
- Sherman SM1 in production December 2025 — 'tens of millions of chips per day' at full ramp
- 65nm-130nm analog/embedded process technologies
- TI expands internal manufacturing for GaN, quadrupling capacity — TI.com (Oct 24, 2024)
- 200mm GaN production at Aizu Japan + Dallas Texas — 4x capacity expansion
- 300mm GaN pilot complete; processes transferable to 300mm
- Target >95% internal GaN manufacturing by 2030
- TI Plans Broad Price Hike on 3,300+ Parts — E-Z-Key (Jun 2025)
- TI raised prices on 3,300+ parts in 2024 — selected legacy products
- TI signals 70% data center growth as industrial, automotive, and data center reach 75% of 2025 revenue — Seeking Alpha
- FY25 data-center revenue $1.5B +64% YoY = 9% of total revenue
- Industrial / automotive / data center = 75% of FY25 revenue
- Haviv Ilan: data center could reach 20% of total sales 'soon'
- TI signals slower semiconductor market recovery — Manufacturing Dive (Q3 2025)
- Industrial / automotive / data center = 75% of FY25 revenue
- Q3 2025 semiconductor recovery framing
- TI Slashes 2026 CapEx Outlook, Targets $8+ FCF/share — Yahoo Finance (Capital Management Review, Feb 2026)
- 2026 capex guide $2-3B vs $4.6B in 2025 — six-year elevated cycle ending
- FCF/share doubled to $3.20 in 2025; targeting $8+/share for 2026
- Path to >95% internally sourced wafers (>80% on 300mm) by 2030
- + 1 more
- TI to acquire Silicon Labs — TI.com (Feb 4, 2026)
- $7.5B EV all-cash; $231/share; ~30% premium
- Closing H1 2027
- $450M cost-synergy target within three years post-close
- + 1 more
- TI to acquire TSMC customer for $7.5B — Manufacturing Dive (Silicon Labs deal context)
- Silicon Labs is a TSMC outsource customer — TI absorbs into vertical fab footprint
- Embedded processing competitive context
- TI Unveils 800VDC Power Architecture for AI Data Centers at NVIDIA GTC 2026 — EverythingPE
- Complete 800V power architecture details at NVIDIA GTC 2026
- TI unveils complete 800 VDC power architecture for AI data centers with NVIDIA — TI.com (Mar 16, 2026)
- TI 800V architecture: 800V hot-swap, 800V-to-6V isolated bus converter at 97.6% peak efficiency / >2000W/in³, 6V-to-<1V multiphase buck
- Complete BOM offer at NVIDIA GTC 2026: 30kW 800V AC/DC PSU, 800V capacitor bank with EDLC supercaps at 40W/in³
- Two-stage architecture (vs Navitas single-stage), positioned as TI's vertical-integration flagship product
- Tom's Hardware — New Texas Instruments fab will pump out tens of millions of chips per day (Dec 2025)
- SM1 first-production date Dec 17 2025
- 3.5-year build cycle from May 2022 ground-break
- TrendForce — Innoscience Scores Major Patent Win Against Infineon as ITC Rules No Infringement (December 5, 2025)
- ITC §337-TA-1407 ALJ initial determination December 2025 — two patents not infringed
- Final determination scheduled April 2, 2026
- Cross-cohort GaN IP enforceability read-through; relatively favorable to TI vertical-integration moat
- TXN Q1 2026 Earnings Highlights GuruFocus
- Q2 2026 guidance above seasonal; EPS 1.77-2.05
- AEC (Automotive Electronics Council) — AEC-Q100, AEC-Q101 qualification standards
- AEC-Q100 IC qualification baseline; AEC-Q101 discrete-device qualification
- TI's broadest AEC-Q-qualified analog/embedded portfolio in industry — design-in moat
- AI Diffusion IFR (90 FR 4544, January 13, 2025)
- Country-tier framework for advanced AI accelerators
- TI products outside scope; cohort regulatory differential vs NVDA/AVGO
- Analog Recovery 300mm Moat Simply Wall St
- 300mm ~40% lower per-die cost vs 200mm; 2.5x chips per wafer
- AVGO/customer.md — comparative concentration / contract structure frame
- Hyperscaler concentration comparative frame: AVGO has named >10% customers (Apple ~20%, Google ~10-12%); TXN has none — anti-AVGO concentration profile
- Contract structure comparative frame: AVGO has formal RPO + multi-year ASIC tape-outs + ELA-locked software; TXN has socket-level switching cost as moat substitute, no formal RPO
- Demand quality comparative frame: AVGO has clean pull-through with some Tomahawk pre-buy; TXN has majority direct pull-through with bounded ~20% distribution channel-fill exposure
- Barclays via X (Jukan) — power semi BOM $140k per AI rack 1MW; 14-vendor 800V partner list
- Power semi content per rack scaling 10x to $140k for 1MW racks
- GaN ~30% / SiC ~10-15% of BOM; 14 vendors named in NVIDIA 800V program (TI included)
- BestAnchorStocks — TI 2025 Capital Management Update analysis
- Independent analysis of TI capital plan; FCF/share trajectory; capex cycle ~70% complete
- BIS Advanced Computing IFR (87 FR 62186, October 7, 2022)
- Original advanced-computing TPP / performance-density thresholds
- TI's analog and embedded products fall outside scope (foundational baseline)
- BIS December 2, 2024 HBM rule (89 FR 96790)
- HBM density thresholds for FDPR coverage
- TI not affected — no HBM product line
- BIS Export Controls on Semiconductor Manufacturing Items update (88 FR 73424, October 17, 2023)
- Tightened TPP, removed performance-density safe harbor
- FDPR extension; H20 origination
- Reaffirms TI products outside advanced-computing scope
- China Initiates Antidumping Duty Investigation into Analog Chips from the US — US ITA / Trade.gov
- US government acknowledgment of MOFCOM investigation
- Investigation timeline and procedural framing
- CHIPS and Science Act of 2022 (P.L. 117-167)
- Statutory basis for §9902 direct grant program and §48D Advanced Manufacturing ITC
- 10-year PRC advanced-node guardrail codification
- Cohort companies.json TXN entry (v2, 2026-05-04)
- TXN cohortRationale: vertical-integration anchor in three-way GaN race; spans chip→board through analog/embedded power management content
- TXN catalysts: 95%+ internal manufacturing target by 2030; 200mm production / 300mm pilot complete; Morroni EV-to-rack supply-chain crossover; embedded-analog destock-to-restock cycle absorbing under-utilized capacity
- TXN risks: capex-heavy vertical integration carries cyclical risk; vertical-integration bet may underperform Infineon scale or Navitas density
- Cohort companion — AVGO market memo (TAM triangulation methodology)
- Cohort TAM frame methodology — triangulate independent vs company-disclosed sizing
- Cohort companion — NVTS market memo (bifurcated GaN frame)
- Discrete vs IC tier bifurcation in GaN; Yole DC GaN $380M 2030 anchor; cycle calendar 2027-2030
- Cohort
corpus.md— Note 1 'The AI Power Crisis — Part 2' (TXN/Morroni references)- Direct quote (line 100): 'As TI's Jeffrey Morroni put it, the technology and semiconductor infrastructure that safely supports an 800V EV looks very similar to what an 800V rack infrastructure needs.'
- Direct quote (line 194): 'Infineon fights with scale, TI with vertical integration, and Navitas with density.'
- Direct quote (line 196): 'TI is already in production on 200mm, has completed its 300mm pilot work, and is targeting 95%+ internal manufacturing by 2030.'
- + 1 more
- Cohort
refinement-log.md— C-NVTS-1 cross-ticker brief for TXN- Cross-ticker learning #1 for TXN: 'Address TI's GaN vertical strategy head-on. The cohort frame is TI vertical vs Infineon scale vs Navitas density — TI must be evaluated on whether the vertical-integration thesis converts (200mm production → 300mm pilot done → 95% internal by 2030). The March 2026 800V-to-6V product (97.6% efficiency, >2000W/in³) is an anchor.'
- Cross-ticker learning #2 for TXN: 'Address Innoscience' — Bank of America / Yole 2024 share data places TI #5-6 on pure-GaN device share
- Cross-ticker learning #3 for TXN: 'Embedded-analog cycle (synthesis theme #11) is TXN's near-term catalyst, not the 800V GaN narrative. Don't conflate.'
- + 1 more
- Cohort sibling — AVGO/macro.md (macro lens)
- Equity-duration baseline for cohort comparison (AVGO higher duration than TXN, both materially lower than NVTS)
- Cohort-typical fabless + TSMC FX profile (USD revenue / TWD cost) — contrast with TXN US-fab vertical model
- ASIC pricing-power baseline transferable to TXN's analog ASP-stickiness analysis
- Cohort sibling — NVDA/macro.md (macro lens)
- Hyperscaler operating-cash-flow funding model — basis for TXN AI-DC rate-insulation argument
- AI-capex super-cycle modal expectation (~$600B / ~50 GW 2026 hyperscaler capex)
- Taiwan-tail revenue-magnitude framing for cohort-relative comparison
- Cohort sibling — NVTS financial.md (pair-trade counterweight)
- NVTS at 82.7x EV/Sales vs TXN at 14.4x — opposite valuation poles in same cohort
- NVTS bear case -75%, bull case +150% — pair break-point at TXN ~$210
- TXN as cohort anchor with bounded downside vs NVTS architectural optionality
- Cohort sibling — NVTS/macro.md (carry-forward C-NVTS-1 brief)
- C-NVTS-1 finding: NVTS amplifies (does not hedge) cohort Taiwan-tail and AI-capex concentration
- TXN identified as candidate primary cohort macro hedge — direct carry-forward instruction
- Cohort Taiwan-tail probability framing inherited (2–4%/yr kinetic, 5–8%/yr blockade)
- + 1 more
- Cohort sibling — WOLF/macro.md (macro lens)
- Auto-loan rate channel for EV end-market modeling — mechanism applied to TXN auto segment
- CHIPS / IRA cost-side tailwind framework — extended to TXN's $1.6B + 25% ITC analysis
- US-domestic fab footprint as Taiwan-tail hedge — extended to TXN US-fab analysis
- + 1 more
- Cohort Synthesis (semiconductor-industry/synthesis.md) — macro lens
- Section 0 cohort scope decision and TXN deep-dive promotion under chip-to-grid framing
- Section 3 theme #1 (voltage-stack redesign at every layer simultaneously)
- Section 3 theme #3 (GaN/SiC competitive structure: 'Infineon scale, TI vertical, Navitas density')
- + 5 more
- Cohort synthesis — chip-to-grid value chain, GaN three-way race, supply-chain-chose-800V, embedded-analog cycle, Taiwan-tail Open Question
- TXN at L8b/L8c spans chip→board
- Section 3.3 GaN three-way race: TI vertical / NVTS density / Infineon scale
- Section 7 Open Question §1: Taiwan-tail concentration as cohort's biggest unhedged risk
- Cohort synthesis.md (semiconductor-industry, 2026-05-04 refresh) — TXN-relevant sections
- Section 2 value-chain map: TXN positioned at L8b (high-density GaN conversion) + L8c (board-level VRM/BCM/eFuse analog power management) — only deep-dive name spanning both layers
- Section 2 L8c row explicitly names MPWR at 'last-mm VRM' — establishes MPWR as named share-gainer at AI-server VRM
- Section 3.3 GaN three-way race: 'Infineon scale, TI vertical, Navitas density'
- + 2 more
- Department of Commerce — Biden-Harris Administration Announces Preliminary Terms with Texas Instruments (Aug 2024)
- Government-side disclosure of TI CHIPS Act terms
- Mature-node and current-generation chip capacity expansion framing
- Department of Commerce — Preliminary Memorandum of Terms with Texas Instruments ($1.61B CHIPS Act direct funding, August 16, 2024)
- $1.61B CHIPS Act §9902 direct funding award; final binding agreement signed December 20, 2024
- Sherman SM1/SM2 (Texas) + Lehi LFAB2 (Utah) project scope
- Capacity covenants, buyback restriction (5-year), excess-profits clawback up to 75%, 10-year PRC advanced-node guardrail
- Edge AI Vision — Microcontrollers $34B by 2030, 6% CAGR
- MCU TAM $34B by 2030 at 6% CAGR; Auto MCU $13B by 2030
- Electropages — China's Analog IC Probe to Benefit Chinese Suppliers
- Mechanical share-transfer mechanism described; consequences for Western analog IDMs
- EU CSRD / ESRS reporting framework
- Corporate Sustainability Reporting Directive applicability to TI EU subsidiaries
- Wave-1 thresholds — Freising, Munich, Reading, Nice operations
- EU Regulation 2021/821 (Dual-Use Regulation, recast)
- EU dual-use export controls
- TI products outside scope — minimal incremental burden
- Findchips — Gap between Chinese and overseas signal chain manufacturers
- China local players concentrated in mid-to-low-end; SG Micro $513M LTM (Sep 2025)
- Futurum Group — Analog Devices Q1 FY 2026: Broad-Based Recovery
- ADI Q1 2026 industrial +38% YoY; comms +63% YoY; cycle confirmation independent of TI
- Futurum — Texas Instruments Q4 FY 2025 Earnings Highlight Industrial, Auto, DC
- Q4'25 segment performance — analog +14%, embedded +8%, DC +64%; FY25 industrial +12%
- IC Insights / SMM — Texas Instruments Analog IC 19% in 2021
- TI peak analog share data point ~19% in 2021 — reference for share trajectory 2021-2025
- IEC 61508 — Functional safety standard
- Industrial functional-safety certification framework
- TI MCU and isolated-driver family compliance
- IEC 62443 — Industrial communication networks security
- Industrial cybersecurity certification levels
- TI Sitara industrial Linux and connectivity portfolio compliance
- Industrial Analog Semis TXN vs ADI TechInvestments
- ADI fab-light vs TXN IDM comparison; industrial analog moat framing
- Industry analyst consolidated TXN segment / customer / channel models (Bernstein, Citi, Morgan Stanley, BofA Global Research)
- AI-server analog content per server estimated ~$50-150 (8-GPU baseline) + $200-400 per AI rack at platform level (industry analyst aggregation; not TI-disclosed at SKU level)
- Auto Tier-1 named set within Automotive segment (~70-80% of segment revenue): Bosch, Continental, Denso, ZF, Aptiv, Magna, Forvia, Valeo, Mando, Hyundai Mobis, plus auto OEM direct relationships
- Estimated top 5 customer concentration ~15-18% of revenue; top 10 ~25-30% (consistent with TI's >40%-outside-top-100 disclosure)
- + 1 more
- INTC regulatory analysis — §48D ITC scale and CHIPS covenants comparator
- §48D ITC scaling methodology
- CHIPS buyback / dividend / clawback covenant interpretation
- Investing for Beginners 101 — Publicly Traded Analog Semiconductor Spring 2024
- Top-5 analog share TTM: TI/IFX 19.5%, ST 19.1%, NXP 15.4%; HHI computation source
- Investing.com — TI Q4 2025 earnings call transcript
- Q4 2025 segment commentary; Q1 2026 guidance $4.3-4.7B; ASP discipline preserved through trough
- IRS — Advanced Manufacturing Investment Credit (Section 48D guidance)
- ITC mechanics: 25% (now 35%) of qualified property
- Placed-in-service after Dec 31 2022 to qualify
- ISO 26262 — Road vehicles functional safety
- Functional safety certification framework (ASIL grades)
- TI Jacinto TDA4, AWR/IWR radar, Hercules safety MCU portfolio compliance
- Macro regime baseline (general-knowledge synthesis, 2026-05-04)
- US 10y ~4.0–4.5% / Fed funds 3.75–4.25% / 5y real ~1.5–2.0%
- DXY mid-100s; EUR/USD trading 1.05–1.12 range
- Auto-loan rate environment 2023–2026 suppressing US/EU EV unit demand by ~5–10% per 100bp
- + 3 more
- MOFCOM Spokesperson on Anti-Dumping Investigation into Analog IC Chip Imports from US (Sep 13, 2025)
- MOFCOM antidumping investigation initiated September 13, 2025
- TI and ADI named as primary import respondents
- Cumulative 51.77% Chinese market price decline 2022-2024 cited as injury
- + 1 more
- Mordor Intelligence — Analog IC Market
- Analog IC TAM $83.8B in 2025; mid-single-digit CAGR 2025-2030
- NIST CHIPS Program Office
- CHIPS Act §9902 direct grant program structure
- Milestone audit framework for disbursement tranches
- Standard guardrails (buyback / dividend / PRC expansion / clawback)
- NVDA regulatory analysis — BIS export-control scope comparator
- BIS advanced-computing rule scope demonstration
- Cohort regulatory differential — TI relative insulation in analog/embedded category
- NVIDIA Developer Blog — 'NVIDIA 800 V HVDC Architecture Will Power the Next Generation of AI Factories' (May 2025)
- Infineon named lead partner in NVIDIA 800V HVDC ecosystem (May 2025); Navitas named ecosystem partner; TI not named at that time
- Establishes the named-partner gap that the March 16 2026 NVIDIA GTC complete-architecture announcement subsequently closed for TI at the architecture-anchor level
- Cross-checked through May 2026: no public Vertiv-TI / Eaton-TI / Schneider-TI / Delta-TI 800V GaN reference-design joint announcement found at the box-builder level
- NVTS regulatory analysis — subsidy asymmetry comparator
- NVTS receives zero CHIPS direct grant or DOE LPO; TI receives $1.61B + ~$4.5B §48D
- Subsidy-disadvantage asymmetry directly cited per refinement-log C-NVTS-1
- NVTS supply-chain analysis — TSMC GaN exit by July 2027, foundry transition framing
- TSMC ends GaN foundry by end-July 2027
- NVTS triple-foundry transition (PSMC + GF Burlington + X-Fab) — the 'two halves of the same story' brief
- TXN vertical integration as the structural hedge to NVTS foundry-loss
- NVTS/customer.md — Reference-Design Partners cross-reference
- Cross-reference establishing the 800V GaN named-partner gap as of late April 2026: NVTS customer analyst found no public Vertiv / Eaton / Schneider / Delta reference design naming Navitas at the GaN partner level
- Same gap applied to TXN at the box-builder level pre-March 2026; TI's March 16 2026 NVIDIA GTC complete-architecture announcement closes the gap at the architecture-anchor level (NVIDIA itself); box-builder-level disclosure remains pending
- C-NVTS-1 cross-ticker brief: 'reference-design partner naming is the critical disclosure for the cohort'
- Peer valuation — Analog Devices (ADI) statistics
- ADI: $194B mcap, 16.9x EV/Sales, 36.3x EV/EBITDA, 33.1x forward P/E, 2.4% FCF yield, 1.1% dividend yield
- Peer valuation — Microchip (MCHP) statistics
- MCHP: $51B mcap, 12.8x EV/Sales, 57.1x EV/EBITDA, 38.6x forward P/E, 1.6% FCF yield, 1.9% dividend yield, 6.3% operating margin (cyclical bottom)
- Peer valuation — Monolithic Power Systems (MPWR) statistics
- MPWR: $78B mcap, 25.9x EV/Sales, 89.4x EV/EBITDA, 62.5x forward P/E, 0.5% dividend yield — closest AI-DC narrative comp
- Peer valuation — NXP Semiconductors (NXPI) statistics
- NXPI: $74B mcap, 6.5x EV/Sales, 17.2x EV/EBITDA, 18.8x forward P/E, 3.6% FCF yield, 1.4% dividend yield — relative-value standout in comp set
- Peer valuation — ON Semiconductor (ON) statistics
- ON: $41B mcap, 6.9x EV/Sales, 23.7x EV/EBITDA, 35.1x forward P/E, 3.5% FCF yield, no dividend
- Power Semis in the AI Data Center TechInvestments
- Infineon 12k-15k per 130kW rack; Onsemi 50k-100k per MW next-gen
- Refinement log — C-NVTS-1 carry-forward
- Subsidy asymmetry citation requirement (NVTS zero CHIPS, TI $1.61B)
- April 2, 2026 ITC Infineon-v-Innoscience FD as cross-cohort GaN IP catalyst
- SEC Final Rule — Climate-Related Disclosures for Investors (March 2024)
- Reg S-K Subpart 1500 climate disclosure requirements
- Status: stayed pending Eighth Circuit consolidated litigation
- AI-DC cross-exposure mechanism (Scope 2/3 customer disclosure pulling through to energy-efficient power-semi design-in)
- South China Morning Post — Chinese analogue chipmakers join price hikes
- China local analog (SG Micro etc.) pricing dynamics; substitution narrative validated
- Stanford Securities Class Action Clearinghouse
- TXN securities class action docket review (FY2022/FY2023 inventory disclosure cases)
- Strait Research — Embedded Processor Market
- Embedded processor TAM $24.8B (2025) → $39.9B (2033) at 6.1% CAGR
- Successful Daily — TI Q1 settles analog inventory question
- Q1 2026 inventory days normalized into long-run band; destocking ended Q1; cycle resumed above prior peak
- Texas Comptroller of Public Accounts — Chapter 313 / Texas Jobs and Security Act (HB 5) value-limitation agreements
- Sherman site state/local incentive package — ~$340M PV
- Texas Enterprise Fund + property-tax abatement structure
- Texas Instruments $60B U.S. fab investment plan announcement
- $60B+ planned investment across 7 U.S. fabs / 3 mega-sites (Texas + Utah)
- Sherman site alone up to $40B
- Analog and embedded processing chip production focus
- Texas Instruments 2025 Capital Management Review (Feb 2025)
- Industrial+auto ~70% of revenue (vs ~40% prior cycle); 95%+ internal manufacturing target by 2030
- TI SAM frame >$60B post-2030 (company definition incl. embedded)
- Texas Instruments begins production at Sherman 300mm fab
- Sherman SM1 in production (2025); SM2 shell complete; Lehi LFAB1 ramping at 45-65nm; 28nm qualification underway
- Texas Instruments Capital Management Review (Haviv Ilan, CEO)
- Free cash flow per share is the primary metric for measuring shareholder value creation
- Dividend grows every year regardless of cycle (22-year streak)
- Buybacks opportunistic and suspended during capex peak
- + 3 more
- Texas Instruments CHIPS Act funding award (December 2024)
- Up to $1.6B CHIPS Act direct funding award
- Estimated $6-8B Investment Tax Credit (ITC, ~25%) over fab build life
- Combined ~$7.6-9.6B subsidy offset against $60B+ headline capex commitment
- Texas Instruments Citizenship Report FY2024
- Scope 1 ~1.6 Mt CO2e; Scope 2 ~3.4 Mt CO2e
- Water reclamation targets; Sherman site environmental commitments (>80% recycle target)
- 95th-percentile peer disclosure
- Texas Instruments February 2026 Capital Management Review coverage (Yahoo Finance)
- 2026 capex slashed from $4.6B (FY25) to $2.0-3.0B guide
- FY2025 actual FCF/share: $3.23
- FY2026 target: $8+ FCF/share
- + 2 more
- Texas Instruments FY2025 Annual Report and Notice of 2026 Annual Meeting
- FY2025 revenue $17.7B (+13% YoY)
- FY2025 free cash flow $2.6-2.9B / 14.7% of revenue (+96% YoY)
- FY2025 capital expenditures $4.6B
- + 3 more
- Texas Instruments — TPS1685 48V hot-swap eFuse press release (Mar 2025)
- TPS1685 industry-first 48V integrated hot-swap eFuse with power-path protection
- LMG3650R series GaN power stages: 650V, >98% efficiency, >100W/in3
- The AI Power Crisis — Part 2 (cohort vault note, ingested 2026-05-03) — macro/Morroni anchor
- Direct corpus quote: 'TI is already in production on 200mm, has completed its 300mm pilot work, and is targeting 95%+ internal manufacturing by 2030'
- Morroni quote: 'the technology and semiconductor infrastructure that safely supports an 800V EV looks very similar to what an 800V rack infrastructure needs'
- Competitive frame: 'Infineon fights with scale, TI with vertical integration, and Navitas with density'
- + 1 more
- The Register — AI gobbling up power and management chips for servers (Apr 2026)
- Power-management chip demand inflection driven by AI servers
- TI / Vertiv 5.5kW server PSU reference design
- Vertiv PowerDirect Rack DC powered by TI GaN technology delivers up to 132kW per rack
- TI Capital Management investor materials (capex roadmap, depreciation, internal-mfg framing)
- Capex roadmap; 95% internal mfg target framing
- Effective subsidy stack treatment
- Depreciation guidance ($1.8-2.0B for 2025)
- TI Capital Management Review presentation — Haviv Ilan / Rafael Lizardi (Feb 2026)
- Capital management framework: maximize long-term FCF/share growth
- RFAB2, LFAB1, Sherman fab roadmap
- Manufacturing-leverage thesis underwriting Silicon Labs synergies
- TI gallium nitride (GaN) — product/technology page
- 300mm GaN status: 'delivering customer samples' — sampling, not production-qualified for HVM
- Important caveat: don't conflate sampling with production qualification
- TI Lehi, Utah: 300mm wafer fabs (manufacturing site disclosure)
- LFAB1 acquired from Micron Q4 2021; production Q4 2022
- LFAB2 broke ground Nov 2 2023; first production target 2026
- Starting nodes 65nm/45nm analog/embedded; expandable beyond
- TI Q4 2025 financial results press release (Jan 27, 2026)
- Q4'25 revenue $4.42B, EPS $1.27
- FY25 capex $4.6B; 2026 guide $2-3B
- FCF $2.9B (+96% YoY) including ~$670M of CHIPS / ITC cash benefits in 2025
- TI reports Q4 2025 and 2025 financial results
- Q4 2025 analog $3.6B +14% YoY; embedded $662M +8%; data center $1.5B +64% YoY (~9% of total)
- FY25 capex $4.6B; FY26 guide $2-3B (capex tapering)
- TI Sherman, Texas: 300mm wafer fabs (manufacturing site disclosure)
- SM1 in production Dec 17 2025 (3.5-year build)
- SM2 shell complete; cleanroom + tools begin 2026
- SM3-SM4 sequencing through end of decade
- + 2 more
- TI worldwide manufacturing overview (company-page disclosure)
- Seven captive A/T sites globally: Aguascalientes (MX), Chengdu (CN), Kuala Lumpur (MY), Melaka×2 (MY), Baguio (PH), Pampanga/Clark (PH), Aizu (JP), Miho (JP)
- Wafer fab list: Sherman (SM1-SM4), Lehi (LFAB1-2), Richardson (RFAB1-2), Dallas (legacy 200mm + GaN), Aizu, Miho
- Cheonan, Korea NOT listed as a TI manufacturing site (corrects prompt assumption)
- TIKR Blog — Texas Instruments 2025 free cash flow doubled analysis
- FY2025 FCF/share $3.23 (up 97% YoY from $1.63 in FY2024)
- Capex $4.6B in 2025 → $2-3B guide for 2026 (end of 6-year elevated investment cycle)
- Depreciation $1.9B in 2025 → $2.2-2.4B in 2026 (~$400M step-up)
- + 1 more
- Tom's Hardware — Semiconductor giga cycle as AI rewrites compute
- AI-driven structural cycle context; giga cycle framing for forward semi outlook
- Tom's Hardware — TI Sherman fab background, $60B program
- $60B 6-year capex program; tens of millions of chips per day capacity at Sherman
- Treasury / IRS Final Rule — §48D Advanced Manufacturing Investment Credit (26 CFR Part 1)
- 25% refundable Advanced Manufacturing ITC on US semiconductor capex placed in service from January 1, 2023
- Final rule issued October 23, 2024
- Recapture mechanics on PRC-guardrail violations
- + 1 more
- TSM regulatory analysis — Section 232 / CHIPS PMT covenant comparator
- Section 232 framing methodology
- CHIPS Arizona PMT covenants comparable to TI Sherman/Lehi
- TSM supply-chain analysis — Taiwan-tail framework, tier-2 chokepoint analysis
- Cohort baseline for Taiwan exposure (~92% TSMC vs <5% TXN)
- Tier-2 chokepoints (ZEISS, Inpria, Aixtron) with TXN's differential exposure (no EUV → no ZEISS/Inpria; Aixtron is the only shared chokepoint)
- TXN income statement, cash flow, balance sheet, statistics — StockAnalysis.com
- FY21-FY25 revenue / margin / earnings time series (rev: $18.3B / $20.0B / $17.5B / $15.6B / $17.7B)
- FY21-FY25 OCF / capex / FCF time series (FCF: $6.3B / $5.9B / $1.3B / $1.5B / $2.6B)
- FY22-FY25 balance sheet: cash drained $9.1B → $4.9B; total debt $8.7B → $14.0B; flipped to net debt $9.2B
- + 3 more
- TXN Q3 2025 earnings call transcript (October 21, 2025)
- Q3'25 revenue +14% YoY / +7% sequential
- Q3'25 industrial +25%, auto +upper-single-digits, comms +45%, enterprise +35% all YoY
- Q3'25 Analog +16% YoY, Embedded Processing +9% YoY
- + 2 more
- TXN Q4 2025 earnings call transcript (January 27, 2026)
- FY2025 segment revenue: Analog +14% YoY, Embedded Processing +8% YoY
- FY2025 end markets: Industrial $5.8B (+12%), Auto $5.8B (+6%), Personal Electronics $3.7B (+7%), Data Center $1.5B (+64%), Comms ~$500M (+20%)
- FY2025 CHIPS Act cash benefit received: $670M
- + 2 more
- US Customs and Border Protection — CSMS #67400472 Section 232 import-duty guidance
- Section 232 implementation mechanics and HTSUS application
- USTR — Section 301 China Tariff Actions (Four-Year Review and modifications)
- Section 301 List 4A baseline rates
- 50% rate increase on HTSUS 8542-class semiconductor subheadings effective January 1, 2025
- Bidirectional tariff exposure framing for TI Chengdu assembly and US-origin shipments to China
- Utah Governor's Office of Economic Opportunity — EDTIF program
- Lehi LFAB2 Economic Development Tax Increment Financing
- ~$150-200M PV state/local incentives
- Uyghur Forced Labor Prevention Act (P.L. 117-78)
- UFLPA enforcement framework; Xinjiang rebuttable presumption
- Indirect gallium upstream sourcing exposure for GaN supply chain
- Vault corpus — Morroni quote and three-way GaN competitive frame
- Jeffrey Morroni (TI): 'the technology and semiconductor infrastructure that safely supports an 800V EV looks very similar to what an 800V rack infrastructure needs'
- Competitive frame: 'Infineon fights with scale, TI with vertical integration, and Navitas with density' (AI Power Crisis Part 2)
- TI 200mm in production, 300mm pilot complete, 95%+ internal manufacturing target by 2030
- White & Case — President Trump orders narrowly targeted 25% Section 232 tariff on certain advanced semiconductor articles (January 2026)
- Section 232 narrow-scope analysis
- Derivative-product expansion contemplated; relative-cost asymmetry vs EU-origin parts
- White House Proclamation — Section 232 narrow tariff on advanced computing semiconductors, January 14, 2026
- 25% Section 232 narrow scope on H200/MI325X-class advanced computing semiconductors
- TI's analog and embedded products explicitly out of scope
- Explicit contemplation of derivative-product scope expansion
- WSTS Fall 2025 Forecast
- $975.5B 2026 global semis; analog moderate recovery 2026 vs Logic/Memory steepness
- WSTS Spring 2025 Forecast
- Analog +7.5% growth in 2025; global semis $772.2B 2025
- Yicai Global — China Probe on US Analog Chips $350M unlocked pool
- Sept 2025 Chinese MOFCOM anti-dumping probe; $350M+ near-term unlocked Chinese pool (Citi)