Signal DOWNGRADED HOLD → 🚫 DO NOT BUY across all three horizons — not on any business deterioration (Quality flat 86, driver flat 78 Strong Tailwind) but on price + risk: the stock rose +7.1% ($360.45 → $386.09), pushing the clean multiple deeper into Expensive (ratio 2.2× → 2.36× warranted), and DNB Trigger 2(b) now fires — deep-expensive paired with the currently-ARMED 'S&P 500 concentration / AI earnings-quality unwind' systemic tail (macro 2026-07-14), which AVGO's AI-capex leverage makes it a cohort member of. Gate 3 (Valuation Ceiling) also triggered.
Broadcom is one of the world's largest semiconductor and infrastructure-software companies, run by Hock Tan out of Palo Alto. Its chip business spans networking silicon, custom AI accelerators (XPUs/ASICs) designed for hyperscalers, wireless components (a major Apple supplier), broadband and storage connectivity. Alongside the chips, a large infrastructure-software arm — anchored by the 2023 VMware acquisition plus mainframe and security assets — sells sticky, high-margin enterprise licences and subscriptions. Broadcom's distinctive position is this pairing: it is the leading merchant designer of bespoke AI silicon for the largest cloud buyers (the main alternative to buying Nvidia's off-the-shelf GPUs) while simultaneously running one of software's most disciplined, cash-generative annuity franchises. That combination gives it ~67% gross margins, a ~$1.9T market value, and a growth profile levered directly to the AI data-centre build-out.
Lifecycle: Growth. TTM revenue $75.5B, +47.9% YoY in Q2 FY26, with net margin 38.8% and EBITDA margin 55.7% — this is a hyper-scaled compounder still accelerating on AI, not a mature cash cow. Metrics are read on the semis (cycle-aware) and software recurring-revenue lenses blended by revenue mix.
| Sub-signal | Value | Score | Note |
|---|---|---|---|
| Revenue trajectory | +47.9% YoY (Q2 FY26) | 92 | AI silicon + VMware ramp; far above semis median |
| Gross margin | 67.0% TTM | 88 | Software mix lifts blended GM well above merchant-semis norms |
| Operating margin | 43.7% TTM | 90 | Elite operating leverage; Hock-Tan cost discipline |
| Cash generation | FCF margin ~45% | 90 | FCF/NI ~0.97 — earnings convert to cash |
| Balance sheet | Net debt/EBITDA ~0.9×, int cov ~11× | 80 | VMware debt being paid down fast |
moat_score = 73. Switching-cost and cost-advantage sub-scores are trimmed from the competitive read below (stable share vs credible rivals, not runaway dominance).
| Rival | Threat type | Share trajectory | Moat-erosion vector |
|---|---|---|---|
| Nvidia (NVDA) | Merchant GPU vs custom ASIC — the substitute at the platform level | Stable — different buying motion; hyperscalers run BOTH | If merchant GPUs get cheap/abundant enough, the case for bespoke ASIC narrows at the margin |
| Marvell (MRVL) | Direct custom-silicon / ASIC competitor for the same hyperscaler programs | Stable to slightly ceding at the top end — AVGO holds the marquee accounts | The only credible second source for a custom XPU program — caps AVGO pricing power |
| Arista (ANET) / Cisco (CSCO) | AI-networking / switching — systems layer above the silicon | Stable — AVGO is the merchant silicon INSIDE many of their boxes | In-house switch ASICs by hyperscalers is the longer-run nibble |
| Hyperscaler in-house teams (Google TPU, AWS Trainium) | Vertical substitution — the customer becomes the designer | Stable — AVGO co-designs many of these, so it captures the work rather than losing it | The structural risk: a hyperscaler internalises the full stack |
ROIC & capital allocation: ROIC in the top decile of large-cap semis (~88th percentile), ROE 33.4%. Capital allocation 80 — disciplined M&A (VMware being de-levered and margin-expanded on plan), a growing dividend (payout ~40%), buybacks. Management skin-in-the-game 60 — Hock Tan's alignment is strong but SBC and the size of past deals warrant a neutral-plus, not a maximal, score.
Earnings are clean. TTM net income $29.3B sits below TTM operating income $32.9B — non-operating items are a net drag (interest expense), not a boost. There are no mark-to-market gains on private-AI stakes inflating the number, so nonop = 0% and the 64× is a clean multiple, not an artefact. (The TTM 3.8% effective tax rate — a Q4-FY25 one-off benefit — actually flatters trailing EPS; on a normalised ~14% tax the trailing P/E would be ~70×, richer still.)
| Relative lens (orders within the band, can't override it) | Read | Score |
|---|---|---|
| Sector median (semis fwd P/E ~24×) | Above median even on forward | 32 |
| Own 5-yr historical decile | 6th decile — rich but not unprecedented for AVGO | 45 |
| PEG (fwd) | ~0.9 on hyper-growth EPS — the one relative bright spot | 58 |
| Analyst consensus target | Consensus $499, all 59 above spot; +29% implied upside | 72 |
Implied-growth read (narrative colour, not the score): at $386 the market embeds materially more than our disciplined 14% near-term estimate — the price is pricing an uninterrupted multi-year AI-silicon ramp. Our anchor says the fundamentals warrant ~27×; the tape is at ~64×.
FMP rating B (score 3): ROE/ROA sub-scores maxed (5/5); P/E and P/B sub-scores at the floor (1/1) — an independent confirmation of exactly this split: elite business, expensive stock.
Primary driver: the hyperscaler AI-capex cycle — specifically demand for bespoke accelerators (XPUs/ASICs) and AI networking silicon. Secondary: the VMware infrastructure-software annuity, which is counter-cyclical and rate-insensitive.
| Horizon | Read | Score |
|---|---|---|
| Historical (25%) | AI-capex has been a powerful, accelerating tailwind for 2+ years; AVGO AI revenue compounding | 90 |
| Current (50%) | Hyperscaler capex still rising; the new Apple $30B multi-year chip deal (2026-07-08) is a fresh order-book data point; but the macro 'AI concentration' tail is armed and mega-cap Q2 guides (late Jul) are the live swing | 76 |
| Forward (25%) | Consensus expects the AI build-out to continue, but the falsifier — a hyperscaler capex cut / order push-out — is exactly the armed tail's trigger | 72 |
driver_score = 78 — Strong Tailwind, amplification-eligible (≥65). But amplification only intensifies a BUY or SELL base; here the base is HOLD (great business, wrong price) and the final signal is DO NOT BUY, so the tailwind is moot — it does not, and cannot, rescue the signal. This is the intended asymmetry: a strong driver never overrides a Do-Not-Buy on price.
Thesis-invalidation floor: a sustained hyperscaler AI-capex cut or a custom-silicon order push-out that breaks the AI-revenue acceleration narrative. That is not merely a distant tail — it is the live trigger of the armed systemic de-rating catalyst, which is why it caps the amplification role and feeds the Bear case.
AVGO inherits Technology (XLK). The latest macro report (2026-07-14) rates XLK short N / medium O / long O — an improvement from the prior report's s:U / m:N / l:O (the soft-CPI reaction eased the near-term multiple headwind). So Economic Alignment is a mild Tailwind medium/long and Neutral short. Stance is Contrarian because we hold a DO-NOT-BUY on a sector the macro rates Outperform medium/long — the disagreement is on price/concentration risk, not on the sector's fundamentals. Crucially, the macro report also carries the 'S&P 500 concentration / AI earnings-quality unwind' tail as ARMED (stamp 2026-07-14), with SPY-vs-RSP re-widening as the breadth tell — that armed tail is what turns AVGO's rich multiple into a DO-NOT-BUY. Amplification is moot (base = HOLD).
Source: sector-map · Macro report 2026-07-14
See the multi-timeframe table above. Higher timeframes (monthly/weekly) remain in uptrends, but the daily is weakening below a falling 50-day ($405) and intraday has broken down. Price at $386 sits between the 200-day ($363, support) and the 50-day ($405, resistance) — no favourable risk-reward location. RSI daily 52 (neutral), MACD daily still negative but histogram turning up. Relative strength: recovered in line with XLK (short N) after the bounce off $360.
Catalyst clustering 68 (calm near-term — no name-specific event inside 14 days). The one timing note that matters: no Technical or Catalyst entry group is met, so even the base HOLD would carry the short technical-confirmation caveat — but this is moot under the Do-Not-Buy.
| Date | Event | Impact | Forecast | Previous | Relevant? | Why |
|---|---|---|---|---|---|---|
| 2026-07-16 | US Retail Sales (Jun) | Medium | +0.3% MoM | prior soft | ⚠️ Medium | Consumer read; indirect for semis via demand |
| 2026-07-29 | FOMC Rate Decision | High | Hold | Hold | ⚠️ Medium | Rate path drives long-duration/growth multiples — but Semis = Medium macro-sensitivity, so not a scheduling trigger |
| 2026-07-30 | Q2 GDP (Advance) | High | decelerating | prior | ⚠️ Medium | Growth read into the AI-capex demand backdrop |
| late Jul | Mega-cap AI Q2 earnings | High | — | — | ✅ Yes | Hyperscaler capex guides are the direct AVGO driver AND the armed-tail trigger |
| 2026-09-03 | AVGO Q3 FY26 earnings | High | EPS $3.22, rev $29.2B | — | ✅ Yes | Broadcom's own print — the next name-specific catalyst |
| Date | Event | Actual | Forecast | Surprise | Impact |
|---|---|---|---|---|---|
| 2026-07-08 | Apple–Broadcom $30B chip deal | positive | — | large multi-year order | Positive — stock jumped; concrete demand data point |
| 2026-07-07 | Erste Group downgrade Buy→Hold | negative | — | valuation concerns | Negative — a fresh sell-side de-rate on exactly the price concern |
| 2026-07-15 | PPI (Jun) | neutral | +0.2% MoM | in line | Neutral for semis |
No high-impact name-specific event inside 14 days — AVGO earnings are 2026-09-03. The relevant near-term catalyst cluster is the late-July mega-cap AI earnings, which are the direct driver AND the armed-tail trigger. Semiconductors are Medium macro-sensitivity, so recurring macro (FOMC/CPI) does not drive the schedule.
| Timeframe | Trend | Direction | RSI | MACD | Key S/R | Breakout | Vol |
|---|---|---|---|---|---|---|---|
| Monthly | Uptrend ↑ | Bullish | 64 | +, rising | S: 138 / R: 415 | Resistance breakout | 0.4x |
| Weekly | Uptrend ↑ | Bullish | 54 | -, flat | S: 322 / R: 495 | None | 0.5x |
| Daily | Weakening → | Neutral | 52 | -, turning up | S: 356/363 / R: 405 | Below SMA50 | 0.6x |
| Hourly | Weakening → | Bearish | 38 | -, falling | S: 382 / R: 399 | Support breakdown | 0.1x |
| 15-min | Downtrend ↓ | Bearish | 35 | -, negative | S: 387 / R: 399 | Support breakdown | 2.4x |
| Confluence: Mixed / rolling over near-term inside a larger uptrend · MTF Score 50 | |||||||
Monthly and weekly trends remain solidly bullish — the secular uptrend is intact and the June low ($360) held, price recovering to ~$386. But daily has slipped below the 50-day ($405, now falling) and intraday timeframes have turned bearish with a support breakdown on the day. This is a stock consolidating below its 50-day after a sharp recovery, not a clean breakout. Price sits between the 200-day ($363) support and the 50-day ($405) resistance — a no-man's-land that offers no timing edge for a new entry. Timing score 44 (weak/mixed). For the record: even if timing turned bullish, the valuation Do-Not-Buy stands — technicals cannot rescue a 2.36×-warranted multiple into an armed de-rate.
AVGO recovered off the $360 June low to ~$386 but sits below a falling 50-day ($405). Above the 200-day ($363). No timing edge; valuation DO NOT BUY governs regardless.
The AI-capex cycle keeps accelerating with no hyperscaler air-pocket, custom-silicon share holds, VMware over-delivers on margin, and the concentration tail disarms as breadth broadens (RSP catches SPY). Multiple stays rich (~55–60× on rising EPS) and consensus targets ($499–582) are met. In this world the stock works — but you'd have overridden a hard price/risk signal to own it.
AI demand stays healthy, EPS compounds toward FY26/FY27 consensus, and the multiple grinds modestly lower off 64× as growth normalises — price drifts up toward the analyst-consensus zone but with high path-volatility around each mega-cap guide. Roughly the sell-side consensus centre of gravity. Even here, buying at $386 (64× clean) offers a poor risk-reward vs the bear.
The ARMED cohort de-rating fires: a hyperscaler guides AI-capex down, a custom-silicon program slips, or a broad AI private-valuation markdown hits the index — and AVGO's clean ~64× compresses toward its ~27× warranted multiple. On roughly flat TTM EPS (~$6.00), a multiple move from ~64× to ~35× lands near $210 — a ~45% drawdown, deeper than any company-specific story. A live competitive leg (Marvell wins a marquee program, or a hyperscaler internalises a stack) reinforces it. This is the cohort-inherited bear, not just AVGO's own risk — and it is the reason for the Do-Not-Buy.
Forecast: ENTRY — Fundamental group: valuation returns to fair value. → FORECAST: only on a large price decline (≈−17% to the anchor fair-value ~$320, i.e. 27× warranted on FY-forward EPS) OR a step-change in warranted EPS that the price stops out-running. BASIS: clean P/E 64× vs 27× warranted; the gap closes via multiple compression, not more growth (growth is already generous in g_near). CONFIDENCE: Moderate that it eventually closes; the trigger is the armed tail, which is event-dependent, not time-projectable. ENTRY — Technical group: daily reclaim of the 50-day ($405) on volume OR a tested higher-low bounce off $360–363. → FORECAST: reclaim needs +5% and a volume day; catalyst-dependent (late-July mega-cap guides / Sep AVGO print). CONFIDENCE: Low near-term — daily is below a falling 50-day. NOTE: no technical or fundamental entry is actionable while the Do-Not-Buy stands — the rules are shown for the watch-list, not to license a buy.
Forecast: No exit is live — this is a DO-NOT-BUY on price, not a sell of a deteriorating business. For an existing holder, the operative watch is the Thesis-Invalidation set: a hyperscaler AI-capex guide-down (late-July mega-cap earnings) is the first place the armed tail would show, and a close below $353 for two days would confirm the technical break.
{
"ticker": "AVGO",
"exchange_ticker": "NASDAQ:AVGO",
"isin": "US11135F1012",
"date": "2026-07-16",
"version": "v6",
"analysis_status": "on-going",
"price_at_rating": 386.09,
"signal_short": "DO NOT BUY",
"signal_medium": "DO NOT BUY",
"signal_long": "DO NOT BUY",
"primary_signal": "DO NOT BUY",
"quality_score": 86,
"valuation_score": 26,
"timing_score": 44,
"driver_score": 78,
"moat_score": 73,
"lifecycle_stage": "growth",
"clean_pe": 64,
"clean_peg": 0.9,
"nonop_pct_of_net_income": 0,
"warranted_multiple": 27,
"actual_multiple": 64,
"warranted_ratio": 2.36,
"val_band": "expensive",
"competitive_share_trajectory": "stable",
"competitive_threat_level": "moderate",
"economic_alignment_stance": "Contrarian",
"economic_alignment_pressure": "Neutral",
"economic_alignment_conviction": 60,
"macro_report_date": "2026-07-14",
"hard_gate_state": "triggered",
"gates_triggered": [
"Gate 3 \u2014 Valuation Ceiling (2.36\u00d7 warranted, above 28\u00d7 semis guardrail)"
],
"do_not_buy_triggers": [
"Trigger 2(b) \u2014 deep-expensive + armed AI-concentration systemic tail; AVGO is an AI-capex-levered cohort member; no growth exemption on arm (b)"
],
"short_entry_confirmed": false,
"entry_groups_met": 0,
"entry_conviction": "Wait",
"exit_groups_live": 0,
"exit_action": "Hold",
"fair_value_est": 320.0,
"stop_loss": 353.0,
"target_price": 499.4,
"scenario_bull_target": 560,
"scenario_base_target": 460,
"scenario_bear_target": 210,
"next_update_date": "2026-07-30",
"next_update_basis": "default +14d (earnings 2026-09-03 is >14d out; Semis Medium macro-sensitivity so FOMC not a trigger)",
"overall_confidence": 65,
"currency": "USD",
"company": "Broadcom Inc."
}
Signal flipped HOLD → DO NOT BUY across all three horizons versus the 2026-07-02 report — NOT on any deterioration in the business (Quality unchanged at 86, driver still a Strong Tailwind at 78) but because (a) price rose $360→$386, pushing the clean multiple deeper into Expensive (2.36× warranted), and (b) DNB Trigger 2(b) was adjudicated afresh: deep-expensive + the currently-ARMED 'S&P 500 concentration / AI earnings-quality unwind' systemic tail (macro 2026-07-14), for which AVGO is a genuine AI-capex-levered cohort member. Arm 2(b) carries no exceptional-growth exemption (that belongs to 2a, which held HOLD at $360). Gate 3 (Valuation Ceiling) is also triggered. XLK signal improved (s:U→N, m:N→O). Next update 2026-07-30 (+14d).