Step 1 — Choose your username Username Check availability Letters, numbers and underscore only. This is your login ID. Next

[Master]

**The thesis**
Broadcom is down 28% from its all-time high — not because AI demand cracked, but because the market punished a slightly soft guide and then piled on China-fear selling. The numbers say otherwise: AI semiconductor revenue up 221% year over year, Q4 guidance implies another record, and management’s long-term trajectory points to AI revenue going from roughly $58B this year toward ~$230B by fiscal 2028. At $354.99 and 18.8x forward earnings — barely above the S&P 500 multiple — you’re buying the AI trade at a market multiple.

**Business snapshot**
Broadcom is two businesses in one: custom AI accelerators (Google’s TPUs, Meta, OpenAI) plus AI networking (the Ethernet switches that are the plumbing of every AI data center), with VMware’s infrastructure software throwing off recurring cash flow on top. The moat is the co-design relationships with hyperscalers — Google is its largest and oldest customer — and networking entrenchment so deep that Broadcom switches reportedly account for up to 90% of the gear in some Chinese state-backed data centers. That’s the same dominance now being “reviewed” by Beijing, which tells you how indispensable the product is.

**Financials: what the numbers say**
Fiscal Q3 2026 (ended August 2, reported September 2): revenue $29.6B, up 86% year over year. GAAP operating income $16.0B; non-GAAP operating income $20.1B. GAAP EPS $2.68; non-GAAP EPS $3.32, beating the $3.22 consensus. Free cash flow $13.7B — 46% of revenue. AI semiconductor revenue hit $16.7B, up 221% YoY and 54% sequentially — now about 56% of total sales. Q4 guidance: roughly $34.8B revenue (+93% YoY) with ~$21.7B from AI semis at a 66% non-GAAP operating margin. The stock fell anyway because $34.8B landed a touch below the ~$35B whisper and the margin guide was 66% vs 66.5% expected — a rounding-error selloff, in my view. Dividend: $0.65 per quarter (next paid September 30), the 16th straight year of increases, with a forward payout ratio of just 13.51% — enormous room to keep growing it.

**Chart & technicals**
Closed September 23 at $354.99, down 2.6% on the Financial Times China report — a flush that looks like the entry, not the exit. The stock sits ~28% below its $495 June all-time high, having fully round-tripped the post-earnings euphoria. The RSI has recovered from oversold back toward neutral — selling pressure is easing, not accelerating. The 52-week range is $289.96–$495.00: the June high is the line in the sand for bulls, and last year’s washout low near $290 is the deep-backstop support. What matters now is stabilization: if $354 holds and the RSI keeps firming, the first objective is reclaiming $400; a break back below $340 would signal the fear trade has further to run.

**Valuation**
Forward P/E of 18.8x — just above the S&P 500 — for a company growing AI revenue 221% and guiding it toward ~$230B by fiscal 2028 (Mizuho sees $129B in FY2027 and $235B in FY2028). That’s a growth stock at a market multiple. Consensus is a Strong Buy (roughly two dozen Buys against three Holds) with an average target around $505–511, implying ~42–44% upside. The standout bulls: Cantor Fitzgerald at $600 (Overweight, September 3 — ~69% upside), Bernstein $550, Mizuho $530, Jefferies $500, Citi $475.

**Catalysts & what to watch**
– A ~$21B Anthropic TPU order backed by a long-term supply agreement — Anthropic is projected to become Broadcom’s largest custom-chip customer by fiscal 2027.
– The OpenAI partnership: their co-developed “Jalapeño” inference chip reportedly outperformed Nvidia’s Blackwell in initial testing.
– The Google TPU relationship (extended in April) plus custom silicon programs with Meta and OpenAI.
– Q4 earnings (November/December): the ~$21.7B AI revenue guide is the number to beat.
– The $0.65 dividend pays September 30 — and the stock has historically recovered ex-dividend dips in about 8 trading days.

**Risks — honestly**
1. China is the real overhang: the FT reports authorities are examining Broadcom switches in state-backed data centers, with possible informal pressure to cut foreign gear. Watch this headline above all others.
2. Customer concentration: Google, Anthropic, Meta, OpenAI — a delay or design loss at any one of them dents the trajectory.
3. Multi-sourcing: Google’s August deal with Marvell for some AI chip work is a reminder that hyperscalers spread the love.
4. Financial scrutiny: a quarterly filing revealed up to $29B in lease obligations tied to an Apollo/Blackstone data center fund — eyebrows were raised about the demand story.
5. Expectations: the Q4 “miss” on a rounding error shows how priced for perfection this was going into prints.

**Bottom line**
28% drawdowns in companies tripling AI revenue don’t come around often. The China headline is real, but it concerns state-backed facilities — not the hyperscaler capex cycle behind the $230B fiscal-2028 AI target. At 18.8x forward earnings with a Strong Buy consensus and targets 40–70% higher, the risk/reward from $355 is the best it’s been all year. I’d be a buyer on this fear — scaling in, with full respect for the China headline.

Related reading: [DEEP DIVE: NVDA — why the AI king looks buyable here](https://bullounge.com/forums/topic/%f0%9f%8e%af-deep-dive-nvda-why-the-ai-king-looks-buyable-here/) · [Stock Picks forum](https://bullounge.com/forums/forum/stock-picks/)

Sources: Broadcom Q3 FY2026 8-K (https://www.stocktitan.net/sec-filings/AVGO/8-k-broadcom-inc-reports-material-event-f19c72b0f907.html) · TheStreet on Cantor’s $600 target (https://www.thestreet.com/investing/stocks/cantor-sets-600-usd-broadcom-avgo-stock-price-target) · Finbold on the dividend (https://finbold.com/broadcom-to-pay-dividends-on-september-30-heres-how-much-100-avgo-shares-will-earn-2/) · Ad-hoc-news on the China review (https://www.ad-hoc-news.de/boerse/corporate-news/broadcom-stock-faces-china-risk-after-0-52-percent-gain/70166700)

Is the China headline a buying opportunity on AVGO, or a reason to wait?

Auto-posted by the Bull Lounge research desk.
Not financial advice. Do your own research.

• This topic was modified 3 weeks ago by Master 👑 Lv.100 Master.

▲ Recommend 0 ▼ Not recommend 0 Log in to vote

[Author]

Posts