Valuation re-scored on the new warranted-multiple anchor. The clean 60× P/E is 2.2× its ~27× warranted multiple (two-stage DCF: r 9%, g 14% near / 3% terminal) and sits above the 28× semis guardrail → deep-Expensive. High Quality (86) + Expensive → BUY → HOLD across all three horizons; the STRONG-BUY amplification is blocked (no BUY base to amplify). Deep-expensive but not a Do-Not-Buy: DNB Trigger 2(a) exempts a name with exceptional, proven, durable growth, which AVGO's AI-driven franchise qualifies for. Business Quality, drivers, technicals and targets are unchanged. Valuation score 66 → 28; entry ladder Half-Size → Wait.
All three horizon signals are unchanged (HOLD / BUY-accumulate / STRONG BUY) on a price that fell −4.3% from $376.71 to $360.45. The stock has continued its post-earnings correction and is now testing the 200-day SMA (~$361). The meaningful shift is in the macro overlay, not the franchise.
Broadcom is one of the world's largest semiconductor and infrastructure-software companies. Its chip business spans networking silicon, custom AI accelerators (XPUs/ASICs it co-designs for hyperscalers), broadband, wireless and storage connectivity; its software arm — anchored by the 2023 VMware acquisition plus mainframe and security assets — sells sticky, high-margin enterprise infrastructure on multi-year contracts. What sets it apart is a two-part moat: it is the dominant merchant designer of custom AI chips (roughly 70% of the market), letting it ride the AI data-center build-out alongside — not against — Nvidia, while VMware provides a recurring, cash-generative annuity that funds the dividend and de-levers the balance sheet. Run by Hock Tan, Broadcom pairs ~67% gross margins and ~44% operating margins with disciplined, acquisition-led capital allocation. Think of it as the 'picks-and-shovels' networking-and-custom-silicon supplier to the AI cloud, bolted onto a VMware software cash machine.
Lifecycle & sector: Growth-stage hybrid — Semiconductors (custom AI ASIC + networking) blended with Infrastructure Software (VMware/mainframe/security). Scored on semi metrics (gross margin, utilisation/cycle, ROIC) plus software recurring-revenue quality. Revenue is compounding at hypergrowth rates (+48% in Q2 FY26) while the software arm supplies a sticky annuity.
| Sub-signal | Value | Benchmark | Score | Rationale |
|---|---|---|---|---|
| Revenue trajectory | Q2 FY26 +47.9% YoY ($22.19B); FY26 consensus ~$105B (+66%) | Semis median mid-single-digit | 95 | Top-decile growth, accelerating on AI custom silicon. |
| Gross margin | 67.0% TTM | Fabless strong >60% | 88 | Software mix + design-win pricing. |
| Operating / net margin | 43.7% op / 38.8% net TTM | Elite | 90 | Clean operating earnings (see §4 earnings-quality note). |
| Cash generation | FCF/sh $6.90; FCF ~$27B TTM; FCF/OCF 0.97 | >20% FCF margin elite | 90 | Prodigious cash — funds dividend + de-lever. |
| Balance sheet | Int. coverage ~11x; current 2.24; D/E 0.74 | Healthy | 70 | VMware debt being paid down steadily. |
Average moat score ≈ 73. Wide but not impregnable — the switching-cost and cost-advantage legs are derived from the competitive read below.
| Rival / threat | Type | Share trajectory | Moat-erosion vector |
|---|---|---|---|
| Nvidia (NVDA) | Merchant GPU + networking | Coexists; AVGO gaining in custom ASIC vs merchant GPU | NVLink/Spectrum networking competes with AVGO's Ethernet/Tomahawk switching |
| AMD | Merchant AI accelerator / DC CPU | AVGO stable — different (custom vs merchant) lane, but AMD wins some GPU sockets | Cheaper merchant alternative can cap ASIC pricing at the margin |
| Marvell (MRVL) | Direct custom-ASIC rival | AVGO holds ~70% share; Marvell smaller but accelerating | The one direct merchant-ASIC competitor — pricing/design-win pressure |
| Hyperscaler in-house silicon (Google TPU, Meta MTIA/Meta Compute, Amazon Trainium) | Vertical substitution | Watch — customers are also partners; Meta Compute (Jul '26) signals insourcing intent | The genuine long-run switching-cost decay: a customer that designs its own accelerator |
→ Net effect on moat: Switching Costs trimmed to 78 and Cost Advantage to 72 for the hyperscaler-insourcing risk; competitive threat level moderate. Propagated to the §11 Bear trigger and the §12 thesis-invalidation rule.
| Dimension | Read | Score |
|---|---|---|
| ROIC vs peers | Top-decile; ROE 33.4%, ROA strong (FMP ROE/ROA both 5/5) | 88 |
| Capital allocation | Disciplined, acquisition-led (VMware accretive, de-levering); dividend payout ~40% of earnings, ~0.7% yield | 80 |
| Management skin-in-the-game | Hock Tan long-tenured; SBC present but net income clean; insider ownership moderate | 60 |
totalOtherIncomeExpensesNet is −$739M (net interest expense), i.e. non-operating items are a drag, not a mark-to-market boost. There are no OpenAI/Anthropic-style equity-stake markups inflating the bottom line. nonop_pct_of_net_income ≈ 0, so no metric recomputation is required and clean_pe = reported ≈ 60x. One caveat kept explicit: the TTM effective tax rate (~3.8%) is dragged down by a one-off Q4-FY25 tax benefit (−$1.65B); FY26 quarters run ~8–10%, so a normalized-tax trailing P/E is ~64–66x — richer, not cheaper. Conclusion: earnings are clean; the ~60x trailing P/E is real operating earnings, so the valuation debate is about the multiple, not about earnings quality — and on the warranted-multiple anchor that ~60× clean multiple is 2.2× warranted (Expensive), which the forward-P/E optimism does not clear.| Multiple | Value | Reference | Read |
|---|---|---|---|
| Trailing P/E | ~60x | Own 5-yr decile ~6; mega-cap semi range | Rich — flattered slightly by low TTM tax |
| Forward P/E FY26 / FY27 | ~31x / ~18.6x | vs +66% / +30% growth | Optically softer — but forward lens, overridden by the warranted anchor |
| Forward PEG | ~0.86 | <1 = growth-cheap | Forward lens only — anchor uses disciplined g, not hypergrowth |
| EV/Revenue TTM | ~24x | >20x ceiling clause | Expensive — well above the 20x ceiling |
| EV/EBITDA TTM | ~42x | Rich | Expensive on trailing |
| FCF yield | ~1.9% | <3% = expensive anchor | Low — pays for growth |
| Horizon | Read | Score |
|---|---|---|
| Historical (25%) | AI revenue exploded — custom-AI +65% to ~$20B in FY25; Q2 FY26 AI revenue +143% YoY. A structural up-cycle. | 90 |
| Current (50%) | ~70% custom-ASIC share, record backlog, new "Jalapeño" inference chip for OpenAI. But near-term AI-capex sustainability is being questioned (Meta Compute insourcing, semi selloff Jul '26). | 74 |
| Forward (25%) | Management guides AI acceleration; custom-AI targeted toward ~$100B by FY27. Offsetting risk: a hyperscaler capex pause would hit order timing. | 72 |
Driver score = 78 → Strong Tailwind → amplification-ELIGIBLE (≥65). It does not change the fundamental pillar scores; it feeds the amplification stage. As of the 2026-07-03 re-rating, however, the base signal is HOLD (High Quality + Expensive on the warranted-multiple anchor), so there is no BUY to amplify — the STRONG-BUY amplification is blocked at every horizon despite the eligible driver and the long-horizon macro Tailwind. Thesis-invalidation floor: a sustained hyperscaler AI-capex cut / custom-silicon order push-out that breaks the AI-revenue acceleration narrative. Confidence 65% (driver is real-time and somewhat contested near-term).
Mapped to Technology (XLK) in the 2026-06-26 MacroDriver report: XLK Short U / Medium N / Long O. So macro pressure is a Headwind short-term, Neutral at the medium horizon (the amplification anchor), and a Tailwind long-term. Finder classes AVGO 'Technology (Contrarian)' — buying quality tech into a near-term macro underperform signal is contrarian, becoming trend-following over the long horizon. Regime is 'Reacceleration lead / Stagflation rising, higher-for-longer Fed' with an armed-but-not-triggering 'AI earnings-quality unwind' tail risk — the reason near-term tech pressure is negative. Medium pressure Neutral → no amplification at the medium horizon; long pressure Tailwind + driver ≥65 would have amplified a long base BUY to STRONG BUY, but after the 2026-07-03 Valuation re-rating the base signal is HOLD (High Quality + Expensive), so there is no BUY to amplify and the long-horizon STRONG-BUY amplification is blocked. Conviction 62, confidence 68%.
Source: sector-map (XLK) · Macro report 2026-06-26
Multi-timeframe: the tool confluence is bearish. Monthly (uptrend) and weekly (uptrend but rolling — weekly MACD histogram negative) sit above a deteriorating daily (weakening, support-breakdown, RSI 42→39.6, price below the 20/50-day and pressing the 200-day at ~$361) and strongly-down hourly/15-min. Classic higher-timeframe-up / lower-timeframe-down — a pullback that has not yet found a higher low.
| Sub-signal | Read | Score |
|---|---|---|
| MTF trend confluence | Bearish; daily broke support, at 200-day | 48 |
| Risk-reward | At 200-day support; stop just below ~$352–356 (≈1.5% / <1 ATR of 17.9). Tight stop possible but in a downtrend. | 50 |
| Relative strength | Sharply underperformed 1-mo (−27% off high); in line with a weak XLK | 30 |
| Macro overlay (Semis, 0.15 wt) | XLK short U — unfavourable near-term | 35 |
| Sentiment (grades + news) | 51 Buy / 7 Hold, 1 recent downgrade (Macquarie); AVGO-specific news positive (Jalapeño, 70% ASIC) but sector tape negative (AI-capex doubt) | 50 |
| Catalyst cluster | Next earnings Sep 3 (far); no cluster — calm calendar | 65 |
Weighted (MTF 30 / RR 20 / macro 15 / sentiment 18 / catalyst 17) ≈ 42. The 200-day is the pivot: hold it and a higher low sets up the Technical entry path; lose it on volume and the Bear path opens.
| Date | Event | Impact | Forecast | Previous | Relevant? | Why |
|---|---|---|---|---|---|---|
| 2026-07-06 | ISM Services PMI (Jun) | High | 54.0 | 54.5 | ⚠️ Medium | Broad growth/AI-capex sentiment gauge; not AVGO-specific |
| 2026-07-06 | ISM Non-Manufacturing Prices (Jun) | High | — | 71.3 | ⚠️ Low | Inflation read feeds the higher-for-longer Fed narrative pressuring tech multiples |
| 2026-09-03 | Broadcom Q3 FY26 earnings | High | EPS $3.22 / rev $29.2B | — | ✅ Yes | The single biggest AVGO-specific catalyst — AI revenue trajectory & guidance |
| Date | Event | Actual | Forecast | Surprise | Impact |
|---|---|---|---|---|---|
| 2026-07-02 | Non-Farm Payrolls (Jun) | 57K | 110K | −48% (below) | Cooling labour — dovish-leaning but stagflation-tinged; mixed for tech |
| 2026-07-02 | Unemployment Rate (Jun) | 4.2% | 4.3% | Better | Slightly firmer than feared |
| 2026-07-01 | ISM Manufacturing PMI (Jun) | 53.3 | 54.0 | −1.3% (below) | Softening manufacturing; second-order for AVGO |
| 2026-07-01 | ADP Employment (Jun) | 98K | 113K | −13% (below) | Labour cooling ahead of NFP |
No high-impact, AVGO-specific macro release inside the next 3 trading days (Semiconductors = Medium sensitivity, so no WAIT-for-event override applies). The relevant tape is the ongoing 'higher-for-longer Fed + AI-capex-sustainability' debate that is compressing semi multiples — a headwind to timing, not to the franchise. The decisive AVGO catalyst is Q3 FY26 earnings on Sep 3.
| Timeframe | Trend | Direction | RSI | MACD | Key S/R | Breakout | Vol |
|---|---|---|---|---|---|---|---|
| Monthly | Uptrend ↑ | Bullish | 60 | +, rising | S: 251 · R: 415 | None | — |
| Weekly | Uptrend ↑ | Neutral | 48 | −, rolling | S: 322 · R: 386/415 | None | 0.8x |
| Daily | Weakening → | Bearish | 42 | −, falling (−11.7) | S: 361/356 · R: 410/442 | Support breakdown | 0.7x |
| Hourly | Strong Down ↓ | Bearish | 39 | −, flat | S: 356 · R: 379 | — | — |
| 15-min | Strong Down ↓ | Bearish | 55 | turning up? | S: 356 · R: 367 | — | — |
| Confluence: Bearish (higher-TF up, lower-TF down — unresolved pullback) · MTF Score 48 | |||||||
Monthly and weekly remain in uptrends, but the weekly is rolling (MACD histogram negative) and the daily has broken short-term support to test the 200-day SMA at ~$361 — with hourly and 15-min in strong downtrends. This is a pullback within a larger uptrend that has not yet printed a higher low. The line in the sand is the 200-day / ~$356 swing low: a defended bounce there is the reachable Technical entry; two closes below opens the Bear path toward the weekly $290–322 support shelf.
AVGO 6-month daily close (Jan–Jul 2026) with 50-day SMA. Note the parabolic post-earnings spike to $495 (early June) and the ~27% correction back onto the 200-day (~$361). SMA50 has rolled over to $409 and now sits above price — the near-term trend is down; the 200-day is the support being tested.
Triggers: 200-day holds, AI custom-silicon re-accelerates (Jalapeño/OpenAI ramp, hyperscaler capex re-affirmed), FY26 revenue tracks toward ~$105B+ and the forward multiple re-rates back toward the Street high ($582). VMware cross-sell surprises up. Path back toward the June high.
Triggers: growth stays hypergrowth but decelerates as expected; forward P/E normalises to ~24x on FY27 EPS ~$19.4 → ~$470, roughly the fair-value estimate and just below consensus $498. The stock consolidates on the 200-day, then grinds up as earnings compound into the multiple. Probability-weighted centre of gravity.
Triggers (COMPETITIVE): a hyperscaler AI-capex cut or accelerated in-house-silicon shift (Meta Compute / Google TPU / Amazon Trainium) pushes out custom-ASIC orders and breaks the AI-revenue acceleration narrative; the 60× trailing multiple compresses. Two closes below the 200-day / $356 open the weekly $290–322 support shelf.
SYSTEMIC (AI-concentration de-rating): beyond the idiosyncratic capex risk, AVGO is a core AI-infrastructure-cohort name — so it also carries the market-wide tail of a whole-cohort multiple compression, deeper and faster than a company-specific stumble because the entire AI complex re-rates together off one narrative shift. At 2.2× its warranted multiple, AVGO has more air beneath it than a fairly-valued name. Trigger: a broad "AI-capex digestion / bubble" narrative takes hold and AI-beneficiary multiples compress in unison (clustered hyperscaler guidance cuts, cohort-wide de-rating) — this leg can take the multiple below the competitive-only case toward the $290–300 shelf. Falsification: hyperscaler FY26/FY27 capex guides hold or rise and cohort AI multiples stabilise — the systemic leg is then off and only the idiosyncratic competitive risk remains.
Forecast: No entry path is open (0 of 3 → Wait). The Fundamental path is now closed by the Expensive/HOLD Valuation re-rating — a name trading at 2.2× its warranted multiple is not a value entry regardless of the target-based upside. Technical group: a 200-day defence with a higher low is the nearest path — plausible within ~1–3 weeks IF $356–361 holds (Moderate confidence; a break resets it lower). A reclaim of the $409 SMA50 is far (~13% away) and Unlikely in 4–6 weeks without a catalyst. Catalyst group is dormant until Q3 earnings on 2026-09-03. So there is no entry edge today; the earliest path to a starter position is a confirmed technical bounce off the 200-day, not the fundamentals.
Forecast: Stop-Loss: LIVE-adjacent — price ($360.45) is only ~2% above the $352 stop, so a single bad session could arm it; watch the 200-day defence closely (Moderate risk near-term). Thesis Invalidation: Unlikely before Q3 earnings (Sep 3) barring a hyperscaler-capex shock. Profit-Target: far (~41% above), Unlikely in this window. Net exit action today: HOLD (no trigger live).
Position sizing not computed — the analysis carries no user allocation or portfolio role (batch run, user_allocation_pct = null). For context only: the §12 Conviction Ladder reads Wait (0 of 3 entry paths met — the Expensive/HOLD Valuation re-rating closes the Fundamental value path and the tape is unconfirmed), i.e. no entry edge today. Volatility context: beta ~1.43, daily ATR ~$17.9 (~5% of price), so a 5% book position behaves like ~7% in market-risk terms. A staggered entry (e.g. tranches at spot, at a confirmed 200-day bounce, and near $322 weekly support) reduces timing risk given the active breakdown.
{
"ticker": "AVGO",
"exchange": "NASDAQ",
"exchange_ticker": "NASDAQ:AVGO",
"isin": "US11135F1012",
"api_ticker": "AVGO",
"finder_ticker": "AVGO",
"finder_exchange": "\ud83c\uddfa\ud83c\uddf8 NASDAQ",
"section": "Technology (Contrarian)",
"country_table": "US",
"date": "2026-07-02",
"version": "v6",
"analysis_status": "on-going",
"user_context": {
"horizon": null,
"allocation_pct": null,
"portfolio_role": null
},
"user_horizon": null,
"user_allocation_pct": null,
"portfolio_role": null,
"price_at_rating": 360.45,
"lifecycle_stage": "growth",
"sector": "Technology \u2014 Semiconductors + Infrastructure Software (hybrid)",
"signal_short": "HOLD",
"signal_medium": "HOLD",
"signal_long": "HOLD",
"primary_signal": null,
"composite_short": 47,
"composite_medium": 56,
"composite_long": 63,
"quality_score": 86,
"quality_confidence": 80,
"quality_detail": {
"industry_benchmark_name": "Semis GM+Utilisation (67% GM) blended with Software recurring-revenue/Rule-of-40",
"industry_benchmark_value": 67,
"industry_benchmark_score": 88,
"moat_score": 73,
"roic_percentile_vs_peers": 88,
"capital_allocation": 80,
"management_skin_in_game": 60,
"gross_margin_ttm": 0.6696,
"net_margin_ttm": 0.3885,
"ebitda_margin_ttm": 0.5569,
"operating_margin_ttm": 0.4366,
"roe_ttm": 0.334,
"revenue_growth_latest_q_yoy": 0.479,
"revenue_growth_fy26_consensus_yoy": 0.66
},
"valuation_score": 28,
"valuation_confidence": 85,
"val_band": "expensive",
"val_multiple_basis": "clean P/E",
"actual_multiple": 60,
"warranted_multiple": 27,
"warranted_ratio": 2.2,
"discount_rate_r": 9.0,
"risk_free_10y": 4.48,
"g_near": 14,
"g_term": 3,
"sector_guardrail_pe": 28,
"valuation_detail": {
"pe_ttm": 60.1,
"forward_pe_fy26": 31.2,
"forward_pe_fy27": 18.6,
"peg_ttm": 0.46,
"forward_peg": 0.86,
"ev_ebitda_ttm": 41.9,
"ev_revenue_ttm": 24.0,
"price_to_sales_ttm": 22.7,
"price_to_book_ttm": 19.5,
"fcf_yield": 1.9,
"historical_valuation_decile": 6,
"boundary_note": "EXPENSIVE on the warranted-multiple anchor \u2014 clean P/E 60x vs a warranted ~27x (two-stage DCF: r 9.0% = 10-Y 4.48% + 4.5% ERP, g 14% near / 3% terminal), ratio 2.2x, and above the 28x Semiconductors guardrail floor. Either condition fires EXPENSIVE (score < 40); both fire. High Quality + Expensive \u2192 HOLD, STRONG-BUY amplification blocked. Deep-Expensive (ratio \u2265 2.0) but NOT Do-Not-Buy: DNB Trigger 2(a) exempts exceptional, proven, durable AI-driven growth. Analyst targets (consensus $498, all 58 above spot $360.45) inform fair value but do not clear the warranted floor."
},
"nonop_pct_of_net_income": 0,
"clean_pe": 60,
"clean_peg": 0.86,
"earnings_quality_note": "CLEAN \u2014 non-operating items are a net drag (interest expense \u2212$739M in Q2 FY26), no mark-to-market equity-stake gains; net income $9.31B sits below operating income $10.87B. TTM tax rate 3.8% (one-off Q4-FY25 benefit) modestly flatters; normalized-tax trailing P/E ~64-66x. No recompute needed; clean_pe = reported.",
"timing_score": 42,
"timing_confidence": 65,
"timing_detail": {
"mtf_confluence_score": 48,
"mtf_tool_confluence": "bearish",
"trend_monthly": "uptrend",
"trend_weekly": "uptrend (rolling)",
"trend_daily": "weakening / support_breakdown",
"trend_hourly": "strong_downtrend",
"risk_reward_score": 50,
"relative_strength_vs_spy": "sharply underperforming 1m (-27% off June high); in line 3m",
"relative_strength_vs_sector": "in line with weak XLK",
"rsi_daily": 39.6,
"macd_daily": -11.74,
"sma50_daily": 409.43,
"sma200_daily": 361.45,
"atr_daily": 17.9,
"catalyst_clustering_score": 65,
"dynamic_macro_weight": 0.15,
"macro_sensitivity": "Medium (Semiconductors)"
},
"driver_score": 78,
"driver_label": "Strong Tailwind",
"driver_confidence": 65,
"driver_detail": {
"primary_driver": "AI custom-silicon (XPU/ASIC for hyperscalers) + AI networking; secondary: VMware infrastructure-software annuity",
"historical": 90,
"current": 74,
"forward": 72,
"amplification_eligible": true,
"thesis_invalidation_floor": "A sustained hyperscaler AI-capex cut / custom-silicon order push-out that breaks the AI-revenue acceleration narrative"
},
"economic_alignment_stance": "Contrarian",
"economic_alignment_conviction": 62,
"economic_alignment_pressure": "Neutral",
"economic_alignment_pressure_short": "Headwind",
"economic_alignment_pressure_long": "Tailwind",
"economic_alignment_source": "sector-map",
"economic_alignment_confidence": 68,
"macro_report_date": "2026-06-26",
"macro_sector_signal": "XLK s:U m:N l:O",
"competitive_share_trajectory": "stable",
"competitive_threat_level": "moderate",
"overall_confidence": 65,
"fair_value_est": 470.0,
"stop_loss": 352.0,
"support_levels": [
361.45,
356.0,
322.0,
289.96
],
"resistance_levels": [
409.43,
442.36,
495.0,
582.0
],
"target_price": 498.19,
"analyst_consensus_target": 498.19,
"analyst_target_high": 582,
"analyst_target_low": 400,
"analyst_target_median": 508.5,
"analyst_target_upside_pct": 38.2,
"analyst_grades_consensus": "Buy",
"analyst_bullish_pct": 87.9,
"analyst_coverage_count": 58,
"recent_upgrades_30d": 0,
"recent_downgrades_30d": 1,
"fmp_rating": "B",
"fmp_overall_score": 3,
"hard_gate_state": "caution",
"gates_triggered": [],
"gates_caution": [
"Valuation Ceiling: clean 60x is 2.2x its ~27x warranted multiple and above the 28x semis guardrail \u2192 caps High Quality + Expensive at HOLD and blocks STRONG-BUY amplification. Deep-Expensive but NOT Do-Not-Buy (DNB 2(a) exempts exceptional, proven AI-driven growth)."
],
"do_not_buy_triggers": [],
"entry_groups_met": 0,
"entry_conviction": "Wait",
"exit_groups_live": 0,
"exit_action": "Hold",
"scenario_base_target": 470,
"scenario_bull_target": 560,
"scenario_bear_target": 300,
"next_update_date": "2026-07-16",
"next_check_date": "2026-07-16",
"next_update_basis": "default +14d (no dated catalyst before Q3 FY26 earnings 2026-09-03; Semiconductors = Medium macro-sensitivity, so recurring macro is not a scheduling trigger)",
"last_updated_human": "Jul 2, 2026",
"prior_report": {
"date": "2026-06-16",
"price_at_rating": 376.71,
"signal_short": "HOLD",
"signal_medium": "BUY_ACCUMULATE",
"signal_long": "STRONG_BUY",
"quality_score": 86,
"valuation_score": 65,
"timing_score": 44,
"driver_score": 80,
"economic_alignment_pressure": "Headwind",
"economic_alignment_conviction": 62
}
}
Re-rated 2026-07-03 on the new warranted-multiple Valuation anchor. The clean 60× P/E is 2.2× its ~27× warranted multiple (r 9.0% / g 14% near, 3% terminal) and above the 28× Semiconductors guardrail → Valuation 66→28 (Expensive). High Quality + Expensive resolves to HOLD, so all three horizons move to HOLD (was HOLD / BUY-accumulate / STRONG BUY) and the STRONG-BUY amplification is blocked despite the eligible driver (78) and the long-horizon macro Tailwind. Deep-Expensive but not a Do-Not-Buy (DNB 2(a) growth exemption). Business Quality (86), Timing (42), Driver (78), Economic Alignment and all market data are unchanged from the 2 Jul run; entry ladder Half-Size→Wait.