Hook & thesis
Broadcom (AVGO) has pulled back from its spring highs but remains the clearest hardware beneficiary of the generative AI infrastructure buildout. The company combines a fast-growing AI semiconductor business with cash-generative infrastructure software franchises and a conservative balance sheet. That mix - high operating leverage on chips plus recurring software revenue - is why a disciplined, size-controlled long here looks attractive.
In this note I give five concrete reasons to buy the dip, back them with balance-sheet and valuation context, and lay out a specific trade: entry at $355.00, stop loss at $330.00, and an initial target at $420.00
What Broadcom does and why the market should care
Broadcom is a global technology company that designs semiconductors and infrastructure software. It operates two primary segments: Semiconductor Solutions - the chips used in data centers, networking and custom AI accelerators - and Infrastructure Software - mainframe, distributed software, cyber security and storage networking. Management has increasingly leaned into bespoke AI silicon and high-margin, recurring software, creating a rare combination of scale in both hardware and software.
The market cares because hyperscalers and major cloud players are committing large volumes of AI compute and custom silicon purchases. Newsflow in recent days points to major AI platform deals and an accelerating AI semiconductor revenue ramp for Broadcom. One recent analysis in the news set referenced Broadcom's AI semiconductor revenue expanding from roughly $58B in fiscal 2026 to as much as $230B in fiscal 2028. Whether you take the midpoint or the high end of those projections, the direction of travel favors Broadcom's product set and pricing power.
Five reasons to buy the dip
- AI-specific revenue momentum - Public reporting and third-party coverage point to dramatic AI semiconductor growth. Recent commentary cites a 221% year-over-year acceleration in Broadcom's custom AI semiconductor sales. Even if growth moderates, the company is clearly benefiting from multi-year cloud hardware commitments.
- Large free cash flow and balance sheet flexibility - Broadcom generated approximately $39.4 billion in free cash flow most recently. That cash flow supports buybacks, acquisitions, and a modest dividend ($0.65 quarterly per share) while keeping leverage manageable (debt-to-equity roughly 0.6).
- High-quality profitability - Return on equity is north of 38% and return on assets about 20%, underscoring the tight economics of its franchises. Those margins make the company resilient through cycles and able to fund R&D for next-generation accelerators.
- Strong market cap scale and liquidity - With a market cap around $1.7 trillion and average daily volume north of ~28 million shares, Broadcom offers the liquidity and institutional footprint needed for meaningful position sizing without severe market impact.
- Defensive software annuity alongside cyclical chips - The Infrastructure Software segment provides recurring revenue and higher margin stability. That mix cushions the earnings profile if semiconductor demand cools temporarily.
Support from the numbers
The stock trades in the mid-$350s today (current print $357.30, previous close $347.30). Valuation metrics show a P/E in the mid-40s (around 44.6-45.7 depending on the source) and price-to-sales roughly 19.2. Enterprise value is about $1.7425 trillion with EV/EBITDA near 33.5. Those multiples are elevated versus broader tech but reflect the rapid growth and exceptional returns on capital. Importantly, the company still converts at scale: free cash flow of about $39.4B gives the firm optionality to accelerate buybacks or tuck-in M&A if management sees attractive deployments.
| Metric | Value |
|---|---|
| Market cap | $1.7056 trillion |
| P/E | ~45x |
| Free cash flow | $39.4 billion |
| Return on equity | ~38% |
| Dividend (quarterly) | $0.65 |
Valuation framing - why the multiples make sense today
At first glance the mid-40s P/E and ~33x EV/EBITDA look expensive compared with broad-market averages. But Broadcom's multiples must be read through the lens of high profit margins, accelerating AI end-market demand, and a large installed base of proprietary IP. The company is not being priced as a commodity chipmaker; it's being priced closer to an annuity-plus-growth business because bespoke AI silicon orders can be multi-year and carry sticky follow-on demand for firmware, support and complementary networking hardware.
Put simply: you are paying for durable cash flow and high returns on capital today, plus optional upside if Broadcom captures more of the custom AI silicon market. If AI silicon revenue scales toward the high-end projections being discussed, current multiples will look modest. If AI demand disappoints, those multiples could compress quickly — which is why position sizing and a stop are essential.
Trade plan (actionable)
Direction: Long
Entry: $355.00
Stop loss: $330.00
Initial target: $420.00
Time horizon: long term (180 trading days) - plan to hold through multiple ordering cycles and at least one major quarterly results season. The 180-day window lets you capture sequential product ramps, cloud customer order cadence and potential re-rating if management reports sustained AI revenue expansion.
Why these levels? Entry at $355 sits slightly below the current trade and recent short-term moving averages, giving room for intra-day volatility. The stop at $330 limits downside and sits clear of the 52-week low range ($289.96) while allowing the trade some breathing room for market noise. The $420 target is a reasonable first upside objective toward a re-rating and reflects a mid-cycle recovery rather than chasing the 52-week high; if catalysts accelerate, I would scale targets higher or add a second target around $480.
Catalysts to watch (2-5)
- Stronger-than-expected AI chip revenue in upcoming quarter(s) - sequential growth and margin expansion from custom accelerators.
- New multi-year supply/compute commitments from hyperscalers (public announcements or disclosed customer wins).
- Further evidence of cross-selling software into large enterprise customers or improved renewal dynamics in the Infrastructure Software segment.
- Management commentary indicating favorable ASP pricing or extended lead times for next-gen accelerators.
Risks and counterarguments
Below are the main reasons this trade can go wrong and what would change my view.
- AI demand disappointment - If hyperscalers slow procurement or AI model economics change reducing hardware intensity, revenue growth could fall short and multiples could compress quickly.
- Customer concentration & competition - Heavy dependence on a handful of cloud customers for large, bespoke orders creates execution risk and pricing pressure. Competitors (including custom silicon from hyperscalers themselves) could win share.
- Valuation vulnerability - With P/E in the mid-40s and EV/EBITDA north of 30, the stock is vulnerable to multiple compression in weak macro or risk-off markets.
- Execution & supply chain - Supply constraints, yield problems on new accelerators, or delays in ramping manufacturing could push out revenue and margin gains.
- Short-term technical pressure - Technicals show bearish MACD momentum and RSI around mid-40s; momentum could continue to weigh on the stock near-term and trigger stops.
Counterargument: A valid counterargument is that Broadcom is already priced for perfection on AI adoption. If the market decides many AI workloads move more to on-prem or different architectures that do not favor Broadcom's designs, today's multiples will be too generous. That's why I cap the initial target at $420 and keep a tight stop at $330. The trade is not a buy-and-forget—it's a thesis that requires confirming revenue and margin signals.
What would change my mind
I would materially downgrade this bullish view if one or more of the following occur: confirmed multi-quarter revenue deceleration in AI semiconductors, meaningful loss of a major hyperscaler customer or contract, a sharp increase in leverage or a sudden deterioration in free cash flow conversion. Conversely, I would become more bullish if Broadcom reports consecutive quarters of 100%+ year-over-year AI semiconductor growth, sustained margin expansion, and management signals higher buyback activity funded by cyclical cash flow.
Conclusion
Broadcom's pullback presents an actionable buying opportunity for investors who believe the AI infrastructure cycle will remain strong over the next 6-9 months. The company's mix of high free cash flow, high returns on equity, and exposure to bespoke AI silicon gives it asymmetric upside if the cloud order environment holds. Use disciplined sizing, an explicit stop at $330, and a 180-trading-day horizon to capture the next leg of the cycle while limiting downside risk.
Key tactical reminders: keep position size aligned to risk tolerance, watch quarterly AI revenue cadence closely, and be ready to adjust targets if customer wins or macro conditions surprise to the upside or downside.