Keysight Technologies (NYSE: KEYS) — Tech Test & Measurement Deep Dive
Multi-Model Consensus Report | April 2026
5-Model Analysis: Claude Opus 4.6, Claude Sonnet 4.6, GPT-5.2, GPT-5.1, Gemini 3 Pro Framework: Universal Rules + Day1Global (Modules A/B/C/D/E/K/L/O, 6 Perspectives) Subject: Keysight Technologies, Inc. (NYSE: KEYS)
TL;DR — Executive Summary
| Metric | Value |
|---|---|
| Stock Price | ~$330 (Apr 14, 2026) |
| Market Cap | ~$57B |
| PE (TTM) | 58x |
| Forward PE | 22–28x |
| PB | ~9.7x |
| 1-Year Return | +140% |
| FCF Yield | ~2.3% (TTM $1.3B) |
| Consensus Rating | HOLD / CAUTIOUS BUY |
| 12M Conservative Target | $280 (-15%) |
| 12M Base Target | $350 (+6%) |
| 12M Bull Target | $420 (+27%) |
| Prob-Weighted 12M Return | +3% to +8% |
Summary: Keysight is a high-quality secular growth compounder with dominant market share (~34–45%) in test & measurement. The AI/5G/6G/automotive secular tailwinds are real, but at 58x trailing PE (1.5x its 5-year average of ~38x), much of the upside is already priced in. Risk/reward is balanced — not a screaming buy, not a sell. Best entry on a 15–20% pullback.
1. Company Snapshot
| Item | Detail |
|---|---|
| Full Name | Keysight Technologies, Inc. |
| Ticker | NYSE: KEYS |
| Headquarters | Santa Rosa, California |
| Founded | 2014 (spun off from Agilent Technologies; heritage traces to HP) |
| CEO | Satish Dhanasekaran (since 2022) |
| Employees | ~15,400 |
| Core Business | Electronic test & measurement solutions — hardware, software, services |
| End Markets | 5G/6G, AI data centers, automotive/EV, aerospace & defense, semiconductors |
| Market Position | #1 or #2 globally in electronic T&M |
Business Segments (FY2025: $5.38B Revenue)
| Segment | Revenue | % of Total | YoY Growth | Key Drivers |
|---|---|---|---|---|
| Communications Solutions Group (CSG) | ~$3.42B | 69% | +10% | 5G/6G infra, AI datacenter, A&D |
| Electronic Industrial Solutions Group (EISG) | ~$1.56B | 31% | +9% | Semiconductor test, automotive, IoT |
Sub-Segment Detail (CSG)
| Sub-Segment | Key Products | Growth Driver |
|---|---|---|
| Commercial Communications | 5G/6G protocol analyzers, network test | AI datacenter buildout (+33% Q1 FY26) |
| Aerospace & Defense | Signal analysis, EW test, radar | US/NATO defense spending (+18% Q1 FY26) |
2. Industry Context — Test & Measurement Market
2.1 Market Size & Growth
| Metric | Value |
|---|---|
| Global T&M Market (2026) | ~$19.8B |
| CAGR (2026–2032) | 4–6% |
| Keysight Market Share | 34% overall, up to 45% in subsegments |
| Key Growth Drivers | AI/ML infrastructure, 5G→6G, EV/ADAS, defense modernization |
2.2 Competitive Landscape
| Company | Focus | Keysight Advantage |
|---|---|---|
| Rohde & Schwarz (Private) | RF/microwave, broadcast | Keysight broader portfolio, more software-led |
| Anritsu (6754.T) | Telecom test | Keysight stronger in AI/datacenter |
| Tektronix (Fortive/FTV) | Oscilloscopes, general test | Keysight higher-end, more R&D-intensive |
| NI (Emerson) | Automated test | Keysight more diversified end markets |
| VIAVI (VIAV) | Fiber/network test | Keysight broader technology range |
2.3 Moat Assessment
Moat Rating: WIDE (4/5 models agree)
| Moat Source | Strength | Evidence |
|---|---|---|
| Switching Costs | ★★★★★ | T&M equipment deeply embedded in R&D workflows; retraining cost is enormous |
| Intangible Assets | ★★★★☆ | 2,700+ patents; regulatory certifications; brand heritage (HP→Agilent→Keysight) |
| Network Effects | ★★☆☆☆ | Limited — software platform (PathWave) building ecosystem |
| Scale Economies | ★★★★☆ | R&D amortization over largest installed base; $1.5B+ annual R&D spend |
3. Earnings Model — Per Segment with Sensitivity
3.1 FY2026E Earnings Build-Up
| Line Item | FY2025A | FY2026E (Base) | FY2026E (Bull) | FY2026E (Bear) |
|---|---|---|---|---|
| Revenue | $5.38B | $6.50B | $7.00B | $5.90B |
| CSG Revenue | $3.42B | $4.30B | $4.70B | $3.80B |
| EISG Revenue | $1.56B | $2.20B | $2.30B | $2.10B |
| Gross Margin | 64.5% | 65.5% | 66.0% | 64.0% |
| Operating Margin (Non-GAAP) | 28.5% | 30.0% | 31.5% | 27.0% |
| Non-GAAP Net Income | $1.10B | $1.52B | $1.74B | $1.20B |
| Non-GAAP EPS | $6.38 | $9.04 | $10.44 | $7.10 |
| Shares Outstanding | 172M | 168M | 168M | 168M |
3.2 Revenue Growth Sensitivity
| Revenue Growth | Revenue | Non-GAAP EPS | Forward PE @$330 |
|---|---|---|---|
| +10% | $5.92B | $7.30 | 45.2x |
| +20% (Base) | $6.50B | $9.04 | 36.5x |
| +25% | $6.73B | $9.70 | 34.0x |
| +30% (Bull) | $7.00B | $10.44 | 31.6x |
4. Operating Leverage Analysis (Rule 6)
4.1 Fixed vs Variable Cost Structure
Keysight operates a high operating leverage model similar to SaaS:
- Fixed Costs (~55% of opex): R&D ($1.5B+), G&A, facility costs, depreciation
- Variable Costs (~45%): COGS materials, sales commissions, warranty
4.2 Breakeven & Profit Multiplier
| Metric | Value |
|---|---|
| Estimated Breakeven Revenue | ~$4.0B |
| Current Revenue | $5.38B (35% above breakeven) |
| Incremental Margin | ~50–55% (above breakeven, each $1 revenue → $0.50–0.55 EBIT) |
| Operating Leverage Multiplier | 1.8–2.2x (10% revenue growth → 18–22% EBIT growth) |
4.3 Revenue Scenario Waterfall
| Revenue | Distance from BE | EBIT Est. | EBIT Margin | vs $5.38B |
|---|---|---|---|---|
| $4.0B (BE) | 0% | ~$0 | 0% | — |
| $5.0B | +25% | $0.50B | 10% | — |
| $5.38B (FY25) | +35% | $0.93B | 17.3% | Actual |
| $6.50B (FY26E) | +63% | $1.63B | 25.0% | +75% EBIT |
| $7.00B (Bull) | +75% | $1.88B | 26.8% | +102% EBIT |
Key Insight: Keysight is well past breakeven and firmly in the profit acceleration zone. Each incremental dollar of AI/5G-driven revenue drops ~$0.50–0.55 to operating profit. This is why EPS is growing 2x faster than revenue.
5. Cycle / Valuation Framework
5.1 Is KEYS Cyclical?
Semi-cyclical: Keysight has cyclical exposure (semiconductor capex, telecom buildout) overlaid on secular growth (AI, defense, automotive electrification). It’s NOT a pure cyclical like VLCC tankers — it has structural growth that compounds through cycles.
5.2 Historical Valuation Bands
| Period | PE Range | Revenue Cycle | Driver |
|---|---|---|---|
| 2019 | 22–30x | Growing (+10%) | 5G rollout begins |
| 2020 | 28–35x | Flat ($4.2B) | COVID dip + recovery |
| 2021 | 30–42x | Strong (+17%) | 5G + semiconductor boom |
| 2022 | 20–35x | Growing (+10%) | Multiple compression, rate hikes |
| 2023 | 22–30x | Flat ($5.5B) | Destocking cycle |
| 2024 | 25–35x | Declining (-9%) | End-market correction |
| 2025 | 30–45x | Recovery (+8%) | AI datacenter demand emerges |
| 2026 YTD | 45–60x | Strong (+20–30%) | AI super-cycle pricing |
5.3 PE Compression Path (Forward-Looking)
| Phase | Timeline | Forward PE | Narrative |
|---|---|---|---|
| Phase 1: Re-Rating (2024–2025) | H2 2024–H1 2025 | 25x → 35x | Revenue trough recognized; AI theme emerges |
| Phase 2: Momentum (2025–2026) | H2 2025–H1 2026 | 35x → 55x | AI capex super-cycle; beats & raises |
| Phase 3: Maturation (2026–2027) | H2 2026–H1 2027 | 55x → 35–40x | Growth decelerates; PE compresses on high base |
| Phase 4: Normalization (2027+) | H2 2027+ | 35x → 25–30x | Steady-state growth; market reprices to historical |
Current Position: Late Phase 2 / Early Phase 3. The easy money has been made. Forward PE of 36x on FY26E EPS is not unreasonable but leaves little margin of safety.
6. Multi-Model Consensus — Target Prices (Rule 7)
6.1 Model-by-Model Assessment
| Model | Rating | 12M Target | Key Reasoning |
|---|---|---|---|
| Claude Opus 4.6 | HOLD | $340 (+3%) | Quality compounder fairly valued; wait for pullback |
| Claude Sonnet 4.6 | BUY (cautious) | $365 (+11%) | AI datacenter thesis underappreciated; FY26 beats |
| GPT-5.2 | HOLD | $320 (-3%) | Trailing PE too rich; forward PE fair but no margin of safety |
| GPT-5.1 | BUY | $380 (+15%) | Operating leverage inflection + defense tailwinds |
| Gemini 3 Pro | HOLD | $330 (0%) | Consensus target = current price; upside priced in |
6.2 Consensus Target Price Matrix
| Scenario | 12M Target | FY26E EPS | Implied PE | Upside/Downside |
|---|---|---|---|---|
| Conservative | $280 | $7.10 (bear) | 39x | -15% |
| Base | $350 | $9.04 | 38.7x | +6% |
| Bull | $420 | $10.44 | 40x | +27% |
6.3 24-Month Targets
| Scenario | 24M Target | FY27E EPS | Implied PE | Upside/Downside |
|---|---|---|---|---|
| Conservative | $310 | $9.50 | 32.6x | -6% |
| Base | $400 | $11.50 | 34.8x | +21% |
| Bull | $500 | $13.00 | 38.5x | +52% |
7. Day1Global Framework Application (Rule 12)
Module A: Revenue Quality ★★★★☆
| Factor | Assessment |
|---|---|
| Revenue Diversification | Good — CSG (69%) + EISG (31%); no single customer >10% |
| Recurring Revenue | Growing — software/services now ~35% of revenue (SaaS transition underway) |
| Geographic Mix | Americas 38%, Asia-Pacific 35%, Europe 27% — well diversified |
| Revenue Visibility | Moderate — mix of project-based + recurring; order backlog provides 1–2Q visibility |
| End-Market Diversity | 5G/6G, AI/datacenter, A&D, auto, semi — no single market >30% |
Module B: Profitability / Operating Leverage ★★★★★
| Metric | FY2025 | FY2024 | Trend |
|---|---|---|---|
| Gross Margin | 64.5% | 63.2% | Improving |
| Non-GAAP Op Margin | 28.5% | 26.1% | Expanding |
| R&D as % of Revenue | ~28% | ~30% | Leveraging |
| Incremental Margin | ~50–55% | — | High operating leverage |
| ROIC | ~18% | ~14% | Inflecting higher |
Module C: Cash Flow — “THE BIG ONE” ★★★★★
| Metric | FY2025 | FY2024 | Assessment |
|---|---|---|---|
| Operating Cash Flow | $1.45B | $1.10B | Strong |
| Free Cash Flow | $1.30B | $0.95B | 24% FCF margin — excellent |
| FCF Yield | ~2.3% | ~2.8% | Low due to rich valuation |
| Capex/Revenue | ~3% | ~3% | Asset-light model |
| Cash Conversion | >100% (NI → FCF) | — | Earnings are real cash |
| Share Buybacks | ~$800M/yr | — | Consistent; shares declining ~2%/yr |
| Dividend | $0 | $0 | No dividend — all buybacks |
Cash Flow Verdict: A-tier. FCF conversion is outstanding, capex is minimal, and the company returns cash aggressively through buybacks. The only knock: no dividend, and FCF yield at 2.3% is thin at this price.
Module D: Forward Guidance ★★★★☆
- Q2 FY2026 guidance: $1.69–1.71B revenue (+30% YoY), Non-GAAP EPS $2.27–2.33
- Management raising guidance consistently — beat-and-raise pattern for 3 consecutive quarters
- AI datacenter demand called out as “multi-year tailwind”
- Defense orders characterized as “structurally higher” post-2024 geopolitical environment
Module E: Competitive Landscape ★★★★☆
- Clear #1 in RF/microwave test
- #1 or #2 in 5G/6G protocol testing
- Growing rapidly in AI datacenter interconnect test (800G/1.6T)
- PathWave software platform creating stickiness
- R&D spend at $1.5B+ creates formidable barrier to entry
- Risk: Rohde & Schwarz privately funded, willing to undercut on price
Module K: Valuation Matrix ★★★☆☆
| Method | Value per Share | vs Current ($330) |
|---|---|---|
| DCF (10% WACC, 3% terminal) | $310–$370 | -6% to +12% |
| Forward PE (35x × FY26E EPS) | $316 | -4% |
| Forward PE (40x × FY26E EPS) | $362 | +10% |
| EV/EBITDA (25x × FY26E) | $340 | +3% |
| Peer Comp (avg fwd PE) | $290–$350 | -12% to +6% |
Valuation Verdict: Fair to slightly overvalued. Most methods cluster around $310–$370. The stock sits at the upper end of fair value, priced for near-perfect execution.
Module L: Ownership & Management ★★★★☆
| Factor | Detail |
|---|---|
| Institutional Ownership | ~89% (Vanguard, BlackRock, Capital Group top holders) |
| Insider Ownership | <1% (typical for large-cap) |
| CEO Tenure | Satish Dhanasekaran since 2022 — executing well |
| Capital Allocation | Buyback-focused ($800M+/yr); selective M&A; no dilutive deals |
| Related-Party Risk | None identified |
Module O: Accounting Quality ★★★★☆
| Factor | Assessment |
|---|---|
| GAAP vs Non-GAAP Gap | Moderate — ~$0.54/share difference (stock comp, amortization) |
| D&A Policy | Conservative — intangibles amortized over 3–10 years |
| Goodwill | $3.4B (~30% of assets) — from M&A history; no recent impairments |
| Off-Balance-Sheet | Operating leases (~$350M); no unusual items |
| Audit Quality | Deloitte — clean opinions |
8. Six Perspectives Analysis
1. Quality Compounder (Buffett/Munger) — ★★★★☆
Durable advantage? YES. Wide moat via switching costs + intangible assets. 28% R&D-to-revenue ratio creates an innovation flywheel. The HP→Agilent→Keysight lineage spans 85 years of T&M leadership.
Concern: At 58x trailing PE, even Buffett would wait for a cheaper entry. Quality is unquestioned; price is the issue.
2. Imaginative Growth (Baillie Gifford/ARK) — ★★★☆☆
10x optionality? Unlikely from here (already $57B). But the software transformation (PathWave, SaaS) could re-rate the business from “hardware company” (30x PE) to “software-enabled platform” (40–50x PE) over 5 years. AI test infrastructure is an underappreciated TAM expansion.
Upside case: If software reaches 50% of revenue by 2030, PE re-rates to 45x on $15 EPS = $675 (+105%).
3. Fundamental Long/Short (Tiger Cubs) — ★★★☆☆
What is mispriced? The market may be over-extrapolating AI datacenter spend. If AI capex decelerates in 2027 (which it historically does after 2–3 year build cycles), KEYS could face a revenue deceleration that compresses the multiple from 55x → 30x rapidly. Short thesis: “AI winter” for test equipment.
Counter: Defense and automotive segments provide a floor that pure-play AI stocks lack.
4. Deep Value (Klarman/Marks) — ★★☆☆☆
Margin of safety? Minimal at $330. FCF yield of 2.3% is thin. Even on FY26E base EPS of $9.04, you are paying 36.5x forward — no obvious value cushion. Would need a pullback to $240–$260 (25–28x forward) for a proper value entry.
This is NOT a deep value stock at current prices.
5. Catalyst Driven (Tepper/Ackman) — ★★★☆☆
Specific value-unlocking events?
- Catalyst 1: PathWave software platform achieving critical mass → margin expansion + recurring revenue re-rating
- Catalyst 2: Major defense contract wins (AUKUS, NATO spending)
- Catalyst 3: 6G standard setting begins 2026–2027 → new test cycle
- Anti-catalyst: AI datacenter capex slowdown; semiconductor correction
6. Macro Tactical (Druckenmiller) — ★★★☆☆
Cycle positioning? Late in the AI capex upcycle. The time to buy KEYS was 2024 (when it was $130–$160 on the semiconductor downturn). Now you are buying late-cycle at elevated multiples.
Macro risks: Interest rate uncertainty, US-China tech decoupling (Keysight has significant China exposure), AI capex deceleration.
9. Risk Matrix
| Risk | Probability | Impact | Mitigation |
|---|---|---|---|
| AI datacenter capex slowdown | 30% | HIGH | Defense + auto segments provide floor |
| PE multiple compression | 40% | MEDIUM | Strong earnings growth can absorb some compression |
| US-China tech decoupling | 25% | MEDIUM | Diversified geographically; some China revenue at risk |
| Semiconductor cycle downturn | 20% | MEDIUM | EISG segment exposed; CSG provides offset |
| Competition from R&S/Anritsu | 15% | LOW | Moat is wide; switching costs are high |
| Key executive departure | 10% | LOW | Deep bench; institutional knowledge |
10. Anti-Bias Framework (Rule 12)
| Bias | How It Could Affect This Analysis | Mitigation |
|---|---|---|
| Anchoring | Anchoring to the 140% 1-year return as “momentum” | Used historical PE bands (22–42x) as anchor, not recent price action |
| Recency | Over-weighting the AI boom narrative | Considered 2023–2024 downturn as counterbalance |
| Confirmation | Seeking data to confirm “high quality = buy” | Multiple models independently flagged valuation risk |
| Narrative | “AI will drive growth forever” | Modeled bear scenario where AI capex slows |
| Survivorship | Only looking at KEYS because it went up | Compared to Anritsu, VIAVI (lower returns) — KEYS quality IS superior |
11. Pre-Mortem Analysis
“It is 1 year later (April 2027) and you lost 40% on KEYS. What went wrong?”
| Scenario | Probability | What Happened | Early Warning Sign |
|---|---|---|---|
| AI Capex Winter | 20% | Hyperscalers cut datacenter capex by 30%; KEYS CSG revenue drops 15%; PE compresses from 55x to 25x | Hyperscaler capex guidance cuts; order cancellations in 800G test |
| Semiconductor Double-Dip | 15% | 2027 chip glut → EISG revenue drops 20%; total revenue flat → PE crush | DRAM/NAND pricing collapse; TSMC capex guidance cut |
| US-China Tech War Escalation | 15% | Export controls expanded to T&M equipment; China revenue ($700M+) at risk | New BIS entity list additions; retaliatory tariffs on US tech |
| Multiple Compression (No Catalyst) | 25% | Growth simply normalizes to 8–10%; market decides 35x is enough; stock drifts to $250 | Two consecutive quarters of slowing order growth |
| Management Misstep (Bad M&A) | 10% | Large acquisition at peak cycle ($5B+) destroys value; integration fails | Announcement of large, high-premium acquisition |
Prob-Weighted Downside Risk: 20%×40% + 15%×35% + 15%×30% + 25%×25% + 10%×40% = ~30% weighted downside is meaningful.
12. Investment Recommendation
Rating: HOLD (3/5 models) / Cautious BUY (2/5 models)
Scenario-Specific Guidance
| Scenario | Action | Entry Price | 12M Target | 24M Target | Allocation |
|---|---|---|---|---|---|
| Conservative | WAIT for pullback | $240–$260 | $280 | $310 | 0% (wait) |
| Base | SMALL position | $300–$330 | $350 | $400 | 3–5% of portfolio |
| Bull | ADD on confirmation | $330+ with beat | $420 | $500 | 5–7% of portfolio |
Triggers
| Direction | Trigger | Action |
|---|---|---|
| BUY | Stock pulls back to $260–$280 (28–31x FY26E) | Start 3–5% position |
| BUY | FY26 revenue guidance raised above $7B | Add at any price <$370 |
| SELL | Trailing PE exceeds 70x with no earnings acceleration | Trim 50% |
| SELL | Hyperscaler capex guidance cut by >20% | Exit 100% |
| HOLD | Stock stays $300–$370 with in-line results | Maintain position |
Portfolio Context
For investors already holding KEYS:
- If bought below $200: Sitting on 65%+ gains. Consider trimming 30–40% to lock in profits and let remainder ride.
- If bought $250–$300: Hold for now. Set stop-loss at $260.
- If considering new entry at $330: Small position only (3%). The quality is excellent but the price is fair, not cheap.
13. Peer Comparison
| Company | Ticker | Market Cap | Fwd PE | Revenue Growth | Gross Margin | Moat |
|---|---|---|---|---|---|---|
| Keysight Technologies | KEYS | $57B | 36.5x | +20% | 65% | Wide |
| Fortive (Tektronix) | FTV | $28B | 22x | +6% | 60% | Moderate |
| Teledyne Technologies | TDY | $22B | 25x | +8% | 43% | Moderate |
| Ametek Inc | AME | $42B | 28x | +7% | 36% | Narrow |
| National Instruments (Emerson) | EMR | $72B | 24x | +4% | 52% | Moderate |
| VIAVI Solutions | VIAV | $3B | 18x | +5% | 58% | Narrow |
KEYS trades at a premium to all peers — justified by higher growth, wider moat, and better margins. But the premium gap has widened significantly in 2026.
Appendix: Key Financial Data
Income Statement Summary
| Metric | FY2022 | FY2023 | FY2024 | FY2025 | FY2026E |
|---|---|---|---|---|---|
| Revenue ($B) | 5.42 | 5.46 | 4.98 | 5.38 | 6.50 |
| YoY Growth | +10% | +1% | -9% | +8% | +21% |
| Gross Margin | 65% | 64% | 63% | 64.5% | 65.5% |
| Non-GAAP EPS | $7.08 | $6.81 | $5.68 | $6.38 | $9.04 |
Balance Sheet Summary
| Metric | FY2025 |
|---|---|
| Total Assets | $11.3B |
| Shareholders’ Equity | $5.88B |
| Total Debt | $2.78B |
| Net Debt | $0.91B |
| Cash & Investments | $1.87B |
| Debt/Equity | 0.41x |
| Current Ratio | 2.6x |
Cash Flow Summary
| Metric | FY2023 | FY2024 | FY2025 |
|---|---|---|---|
| Operating CF | $1.35B | $1.10B | $1.45B |
| Capex | $0.18B | $0.15B | $0.15B |
| Free Cash Flow | $1.17B | $0.95B | $1.30B |
| FCF Margin | 21.4% | 19.1% | 24.2% |
| Buybacks | $0.60B | $0.70B | $0.80B |
14. Deep Dive: AI/CPO Business Valuation Model
This section models Keysight’s AI and co-packaged optics (CPO) business as a standalone value driver, with sum-of-parts implications for the full company.
14.1 AI/Optical Test Revenue Breakdown & Forecast
Keysight does not break out AI/CPO revenue separately. The following is a bottoms-up estimate based on segment disclosures, earnings call commentary, and market sizing data.
| Revenue Line | FY25E | FY26E | FY27E | FY28E |
|---|---|---|---|---|
| 800G pluggable test | $400M | $500M | $550M | $500M |
| 1.6T transceiver test | $50M | $200M | $450M | $700M |
| CPO module test | $20M | $50M | $200M | $500M |
| Software (PathWave, subscriptions) | $130M | $150M | $200M | $300M |
| Total AI/Optical Revenue | $600M | $900M | $1,400M | $2,000M |
| YoY Growth | — | +50% | +56% | +43% |
14.2 Margin Profile — Hardware vs Software
A common error is applying the ~80% software margin to the entire AI business. The reality:
| Component | Revenue Mix (FY28E) | Gross Margin | Gross Profit |
|---|---|---|---|
| Hardware (test instruments) | $1,400M (70%) | ~60% | $840M |
| Software (PathWave, subscriptions) | $600M (30%) | ~82% | $492M |
| Blended | $2,000M | 66.6% | $1,332M |
14.3 Full P&L Walk: AI/Optical Business (FY28E)
| Line Item | Amount | Margin |
|---|---|---|
| Revenue | $2,000M | 100% |
| Gross Profit | $1,332M | 66.6% |
| R&D (allocated ~20%) | ($400M) | |
| SG&A (allocated ~12%) | ($240M) | |
| Operating Profit (EBIT) | $692M | 34.6% |
| Tax (~21%) | ($145M) | |
| Net Income | $547M | 27.4% |
Key takeaway: Net margin on the AI business is ~27%, not 80%. The 80% figure applies only to incremental software gross margin — R&D, SG&A, and taxes consume the rest.
14.4 AI/CPO Business — What Is It Worth?
| PE Multiple | AI Biz Value | % of Current $57B Cap |
|---|---|---|
| 30x (conservative) | $16.4B | 29% |
| 35x (base) | $19.1B | 34% |
| 40x (growth premium) | $21.9B | 38% |
| 50x (aggressive) | $27.4B | 48% |
14.5 Sum-of-Parts Valuation (FY28E)
| Segment | Revenue | Net Income | PE | Value |
|---|---|---|---|---|
| AI/Optical test | $2,000M | $547M | 35x | $19.1B |
| Core T&M (non-AI CSG + EISG) | $5,000M | $750M | 25x | $18.8B |
| Total | $7,000M | $1,297M | ~29x blended | $37.9B |
| Current Market Cap | — | — | — | $57.0B |
| Premium / (Discount) | — | — | — | ($19.1B) = -33% |
Implication: At $330/share, the market is already pricing in FY28E bull-case execution PLUS a sustained 40–45x blended PE (growth premium). Even on optimistic AI/CPO assumptions, fair value on FY28E numbers is ~$38B — roughly 33% below today’s price.
14.6 CPO Test Competitive Landscape
| Vendor | Market Share (800G/1.6T/CPO) | Key Advantage | Weakness |
|---|---|---|---|
| Keysight 🥇 | ~25–30% | Proprietary ASIC in DCA-M oscilloscopes; FlexOTO platform; first to ship 1.6T test (Mar 2025); IEEE 802.3df standards committee seat | Premium priced |
| VIAVI 🥈 | ~25% | Broad portfolio (BERT, OSA); strong in R&D + production test | Less cutting-edge CPO focus |
| Anritsu 🥉 | ~10–15% | Protocol/signal integrity; strong in Asia/Japan | Smaller scale |
| Tektronix (Fortive) | ~10% | Signal integrity, oscilloscopes; strong in US labs | Niche; weaker in production |
| EXFO / Others | ~5–10% | Fiber/network testing | Not CPO leaders |
14.7 Keysight’s Specific Moats in CPO Test
| Moat | Detail |
|---|---|
| Proprietary ASIC/MMIC | Custom silicon in DCA-M oscilloscopes — competitors cannot replicate easily |
| FlexOTO platform | Multi-channel automated test; scales from 800G → 1.6T → 3.2T on same platform |
| Standards influence | Keysight sits on IEEE 802.3df committee — they help WRITE the test specifications |
| First-mover on 1.6T | Shipped 1.6T test solutions March 2025; competitors are 6–12 months behind |
| Installed base lock-in | Customers already using Keysight for 400G/800G — natural upgrade path to CPO |
14.8 CPO Market Sizing (Test Equipment TAM)
| Metric | Value | Source |
|---|---|---|
| SiPh & CPO T&M equipment market (2025) | ~$1.36B | BusinessWire / ResearchAndMarkets |
| SiPh & CPO T&M equipment market (2032E) | ~$2.04B | CAGR ~6% |
| CPO component market (2025) | ~$95M | Adtek Fiber |
| CPO component market (2034E) | ~$1.05B | CAGR 30%+ |
| LPO + CPO combined (2026E) | ~$10B | LightCounting |
| Ethernet transceiver unit growth | 60M → 120M units (2025–2029) | Keysight blog |
14.9 Implied Entry Prices for AI/CPO Thesis
| Entry Price | Implied Blended PE (on FY28E $1.3B NI) | Risk/Reward |
|---|---|---|
| $330 (today) | 43x | Priced for perfection — no margin of safety |
| $280 | 37x | Fair if AI/CPO accelerates beyond base case |
| $240 | 31x | Good entry — margin of safety on AI thesis |
| $200 | 26x | Deep value — assumes temporary market dislocation |
15. 800G/1.6T Test Market — Bottoms-Up TAM & Keysight Revenue Model
This section corrects earlier CPO-centric framing. The real AI revenue driver for Keysight is 800G/1.6T transceiver test equipment, not CPO specifically.
15.1 Transceiver Shipment Volume Forecast (Cross-Checked)
| Year | 800G Units (M) | 1.6T Units (M) | Total Hi-Speed (M) | Source |
|---|---|---|---|---|
| 2024 | 10–12 | ~0 | ~12 | Cignal AI, TrendForce |
| 2025 | 18–20 | <1 | ~20 | LightCounting, Cignal AI |
| 2026 | 50+ | 2–3 | ~53 | TrendForce (2.6x jump) |
| 2027 | 55+ | 2–4 | ~58 | LightCounting |
| 2028 | 60+ | 5+ | ~65 | LightCounting |
15.2 800G/1.6T Module Market ($)
| Year | 800G Market | 1.6T Market | Total | Source |
|---|---|---|---|---|
| 2024 | $1.2–1.8B | ~$0 | ~$1.5B | WiseGuy, Verified Market |
| 2025 | $2.1–3.8B | <$0.5B | ~$3.5B | WiseGuy, Dataintelo |
| 2026 | $4.3–4.5B | $1–2B | ~$6B | Interpolated @17-19% CAGR |
| 2027 | ~$5B | $3–4B | ~$8B | Projected |
| 2028 | ~$5B (plateau) | $5–7B | ~$11B | 1.6T overtakes 800G |
15.3 Test Equipment TAM = 6–10% of Module Market (Industry Rule)
Multiple sources confirm test equipment spending runs 6–10% of module market value. Test cost per 800G module is approximately $25–50.
| Year | Module Market | Test TAM (6-10%) | Midpoint |
|---|---|---|---|
| 2024 | $1.5B | $90–150M | $120M |
| 2025 | $3.5B | $210–350M | $280M |
| 2026 | $6B | $360–600M | $480M |
| 2027 | $8B | $480–800M | $640M |
| 2028 | $11B | $660M–1.1B | $880M |
15.4 Keysight’s Share (25–30% of Test TAM)
| Year | Test TAM | KEYS @25% | KEYS @30% | Midpoint |
|---|---|---|---|---|
| 2025 | $280M | $70M | $84M | $77M |
| 2026 | $480M | $120M | $144M | $132M |
| 2027 | $640M | $160M | $192M | $176M |
| 2028 | $880M | $220M | $264M | $242M |
15.5 But the Real Story Is MUCH Bigger Than Optical Test
800G/1.6T optical transceiver test is ~$130–240M for Keysight by FY28. But the total wireline/AI shift in Commercial Communications is driving from ~$2.3B → $4.0B (+74%) over 3 years. This includes:
| Revenue Category | Est. FY26 | Notes |
|---|---|---|
| 800G/1.6T optical transceiver test | ~$130M | Bottoms-up from TAM model above |
| PCIe Gen6 / CXL test | ~$200M | AI server interconnect |
| Signal integrity / 224G SerDes | ~$300M | Every AI chip needs SerDes validation |
| AI network emulation (Spirent) | ~$200M | Acquired capability |
| 5G/6G wireless test | ~$500M | Still the largest sub-segment |
| Other wireline/protocol test | ~$400M | Ethernet, InfiniBand, etc. |
| Software/PathWave | ~$250M | Growing recurring revenue |
| Total Commercial Comms | ~$2.0B | Q1 FY26 run-rate: ~$3.0B annualized |
15.6 Full Company Model (Grounded, FY26–FY28)
| Metric | FY25A | FY26E | FY27E | FY28E |
|---|---|---|---|---|
| Total Revenue | $5.38B | $6.50B | $7.30B | $8.10B |
| CSG Revenue | $3.42B | $4.50B | $5.10B | $5.70B |
| EISG Revenue | $1.56B | $2.00B | $2.20B | $2.40B |
| Non-GAAP Op Margin | 28.5% | 30% | 31.5% | 33% |
| Non-GAAP EPS | $6.38 | $9.04 | $11.03 | $13.27 |
| PE @$330 | 51.7x | 36.5x | 29.9x | 24.9x |
15.7 Multi-Model Probability-Weighted Consensus
Three independent models (Conservative/Klarman, Growth/ARK, Cycle-Aware/Druckenmiller) weighted 30%/30%/40%:
| Year | Weighted EPS | PE @$330 |
|---|---|---|
| FY26E | $9.63 | 34.3x |
| FY27E | $11.27 | 29.3x |
| FY28E | $11.84 | 27.9x |
| FY29E | $12.93 | 25.5x |
Key risk: The Cycle-Aware model assigns 40% probability to a mid-cycle AI correction in FY28, which would compress EPS to ~$7.60 (PE 43x at $330). This tail risk is why the weighted FY28 EPS ($11.84) is lower than the base case ($13.27).
16. Q1 FY2026 10-Q Deep Dive & Earnings Spike Analysis
On February 24, 2026, KEYS spiked 23% in a single day. This section explains why.
16.1 The Beat (Feb 23, 2026 — Quarter Ended Jan 31, 2026)
| Metric | Consensus | Actual | Beat |
|---|---|---|---|
| Revenue | $1.54B | $1.60B | +$60M (+4%) |
| Non-GAAP EPS | $2.00–2.04 | $2.17 | +$0.13–0.17 (+8%) |
| Orders | — | $1.645B (+30% YoY) | Record |
16.2 Five Catalysts That Drove the 23% Spike
Catalyst 1: Orders > Revenue = Acceleration Signal
- Orders grew +30% YoY vs revenue +23% → Book-to-bill of 1.03x
- Orders exceeding revenue means backlog is BUILDING, not depleting
- The market had feared post-recovery deceleration; +30% orders killed that narrative
Catalyst 2: Wireline > Wireless — A Structural First
- First time in company history that wireline orders surpassed wireless
- Wireline = AI datacenter (800G/1.6T Ethernet, optical interconnects)
- Market re-classified KEYS from “5G test company” to “AI infrastructure platform”
- AI-specific customer count doubled in the past year
Catalyst 3: Guidance Raised Well Above Consensus
| Metric | Q2 FY26 Guidance | Implied YoY |
|---|---|---|
| Revenue | $1.69–1.71B | +30% |
| Non-GAAP EPS | $2.27–2.33 | +35% |
| Full-year | >20% revenue & earnings growth | Raised |
Classic beat-and-raise — the market’s favorite signal for growth compounders.
Catalyst 4: Massive Simultaneous Analyst Upgrades
| Analyst | Old Target | New Target | Change |
|---|---|---|---|
| UBS | $230 | $340 | +48% |
| Goldman Sachs | $243 | $322 | +33% |
| JPMorgan | ~$250 | $300 | +20% |
| Citigroup | $282 | $320 | +13% |
| Susquehanna | $225 | $300 | +33% |
| Morgan Stanley | $227 | $268 | +18% |
Six simultaneous upgrades of 18–48% triggers a forced re-pricing across buy-side models.
Catalyst 5: Spirent Integration Narrative
- $1.5B buyback authorization announced — signals confidence
- Management guided Spirent margins to improve through FY26 — temporary dilution understood
- Spirent adds $375M+ revenue to AI/network test TAM
16.3 GAAP vs Non-GAAP — Why the Market Looked Through GAAP Weakness
| GAAP Concern | Why Market Ignored It |
|---|---|
| GAAP GM fell 100bps to 62.2% | Non-GAAP GM was 66.7% — market trades on non-GAAP |
| GAAP op margin fell 130bps | Known Spirent dilution — pre-announced and temporary |
| $83M tax benefit flattered GAAP NI | Market focused on Non-GAAP $2.17 which excludes one-time items |
| Intangible amortization doubled ($67M) | Expected from $1.56B Spirent acquisition goodwill |
16.4 10-Q Key Financial Data (Quarter Ended Jan 31, 2026)
Income Statement (GAAP, $M)
| Line Item | Q1 FY26 | Q1 FY25 | YoY |
|---|---|---|---|
| Product Revenue | $1,225 | $983 | +25% |
| Services Revenue | $375 | $315 | +19% |
| Total Revenue | $1,600 | $1,298 | +23% |
| Total COGS | $605 | $478 | |
| Gross Profit | $995 | $820 | +21% |
| R&D | $303 | $249 | +22% |
| SG&A | $447 | $361 | +24% |
| Income from Operations | $248 | $218 | +14% |
| GAAP Net Income | $281 | $169 | +66% |
| GAAP Diluted EPS | $1.63 | $0.97 | +68% |
| Non-GAAP EPS | $2.17 | $1.82 | +19% |
Balance Sheet Highlights ($M, Jan 31, 2026)
| Item | Jan 31, 2026 | Oct 31, 2025 |
|---|---|---|
| Cash & Equivalents | $2,195 | $1,890 |
| Total Assets | $11,481 | $11,301 |
| Long-term Debt | $2,534 | $2,534 |
| Goodwill | $3,474 | $3,424 |
| Deferred Revenue (current) | $729 | $652 (+12%) |
| Stockholders’ Equity | $6,205 | $5,881 |
Cash Flow ($M, Q1)
| Item | Q1 FY26 | Q1 FY25 |
|---|---|---|
| Operating Cash Flow | $441 | $378 |
| Capex | ($34) | ($32) |
| Free Cash Flow | $407 | $346 |
| Share Buybacks | ($87) | ($75) |
16.5 Signal Strength Assessment
| Signal | Reading | Importance |
|---|---|---|
| Orders +30% (> revenue +23%) | Strongly bullish — growth accelerating | ★★★★★ |
| Wireline > wireless (structural first) | AI infrastructure recomposition | ★★★★★ |
| Guidance raised above consensus | Beat-and-raise cycle intact | ★★★★★ |
| Non-GAAP EPS beat by 8% | Execution confirmed | ★★★★☆ |
| Deferred revenue +12% | Backlog building | ★★★★☆ |
| FCF $407M (25% margin) | Cash generation excellent | ★★★★☆ |
| GAAP margins declining | Temporary — Spirent integration | ★★☆☆☆ |
| $83M one-time tax benefit | Non-issue for non-GAAP investors | ★☆☆☆☆ |
Data as of April 14, 2026. Not investment advice. All models may contain errors — verify independently. Report generated using the Day1Global tech-earnings-deepdive framework.