Hook & thesis
Broadcom just delivered another blowout quarter: revenue of $29.59 billion and adjusted EPS of $3.32, with AI semiconductor sales surging 221% year-over-year to $16.7 billion. Yet the stock is trading sharply lower from the prior close after management's Q4 revenue guide of $34.8 billion came in a hair below consensus. That knee-jerk reaction opens a tactical buying opportunity. The business that actually matters for valuation - AI accelerators plus sticky infrastructure software - is accelerating, and Broadcom's balance sheet and free cash flow provide the fuel for aggressive R&D, partnerships, and M&A that can compound EPS toward $30 over the next several years.
This is a trade idea, not a sermon. The plan: buy into the guidance dip, set a strict stop, and ride a potential re-rating as AI revenue continues to compound and software margins expand. Entry $350, stop $320, target $525. Long-term thesis: Broadcom's AI mix plus high-margin infrastructure software could lift adjusted EPS toward $30 if current growth sustains and the company executes on pricing and product ramps.
Why the market should care - business explained
Broadcom is two businesses under one roof: Semiconductor Solutions and Infrastructure Software. The former supplies custom AI accelerators and networking silicon that feed hyperscale data centers; the latter is a high-margin, subscription-like business spanning mainframe, security, and storage software. The combination is potent: hardware captures fast-growing AI demand, software provides recurring revenues and margin stability.
Q3 results highlight the shift. Total revenue was $29.59 billion, but the AI chip line alone reached $16.7 billion and grew 221% year-over-year. That single line now represents a sizable share of total revenue and is the primary fundamental driver of earnings upside. Meanwhile, Broadcom generates meaningful free cash flow - $32.762 billion on the most recent figures - that can be redeployed into R&D, capacity expansion, and strategic tuck-ins that accelerate EPS growth.
Key numbers to anchor the thesis
| Metric | Value |
|---|---|
| Last quarter revenue | $29.59B |
| AI semiconductor revenue (Q3) | $16.7B (+221% YoY) |
| Adjusted EPS (Q3) | $3.32 |
| Trailing EPS | $6.16 |
| Free cash flow (most recent) | $32.762B |
| Market cap | $1.68T |
| Enterprise value | $1.79T |
| 52-week range | $289.96 - $495.00 |
| Current price | $353.13 |
| Dividend (quarterly) | $0.65 (yield ~0.69%) |
Valuation framing
At roughly $353 per share and a TTM EPS around $6.16, Broadcom trades near a mid-to-high 50s P/E on the current run rate. That multiple reflects investor willingness to pay for episodic AI upside and sticky software cash flows. But valuation becomes much more attractive as AI revenue scales. A plausible execution scenario: AI-related revenue continues to grow at multi-hundred percent rates this year before normalization, software margins expand modestly, and the company repurchases shares with meaningful free cash flow. If adjusted EPS can scale from the current run-rate toward a multi-digit number - management and the market could reasonably model a path to $30 over a multi-year horizon - the multiple required to reach meaningful upside is not extreme. For instance, $30 EPS with a 18x-20x multiple implies a stock price between $540 and $600, and a re-rating to 25x would justify $750. My target of $525 assumes partial realization of this path and a re-rating from the mid-50s P/E to the low-20s as growth stabilizes and operating leverage kicks in.
Catalysts to watch (2-5)
- Continued AI order flow and ramp beyond the current Q - any quarter where AI revenue exceeds the current $16.7B figure by a wide margin will validate long-term EPS upside.
- Guidance re-acceleration - if management's next guide reinstates and raises revenue expectations above $34.8B for Q4, sentiment should flip fast.
- Margin expansion in Infrastructure Software through cross-selling and price realization, which would materially lift operating leverage.
- Large hyperscaler design wins and multi-year contracts for custom accelerators that lock in demand and improve revenue visibility.
- Share repurchases funded by the $32.8B free cash flow figure - aggressive buybacks accelerate EPS per share.
Trade plan (actionable)
Entry: $350.00
Stop loss: $320.00
Target: $525.00
Horizon: long term (180 trading days). Why this horizon? The thesis relies on continued AI revenue acceleration and either a sentiment reset or an operational update from management. Those catalysts are multi-week to multi-month events: revenue recognition, product ramps, and margin expansion take time to appear in financials. A 180-trading-day window gives time for quarterly cadence and multiple data points (order intake, bookings, margin trends) to confirm the thesis.
Risk management: size the position so that the $30 per-share downside to the stop corresponds to an acceptable dollar loss. Re-evaluate position at each quarterly print or if short-term technicals deteriorate sharply (for example, RSI persistently below 30 with accelerating volume on down-days).
Risks and counterarguments
Broadcom is not a risk-free idea. Below are the main risks, plus one direct counterargument to the bullish thesis.
- Guidance disappointments can persist. The most immediate risk is a multi-quarter cycle where management understates revenue potential and the market demands visible, near-term guidance beats before re-rating. If guidance continues to undershoot expectations, multiple contraction could offset fundamental growth.
- AI demand could normalize faster than expected. If hyperscalers saturate capacity or diversify to competitor silicon, the 221% YoY growth in AI revenue could decelerate quickly, removing the primary earnings catalyst.
- Competition and pricing pressure. NVIDIA and other players (plus potential new entrants) could force price erosion on accelerators or accelerate feature parity, squeezing gross margins.
- Macro & capital cycle risk. Data center capex is cyclical. A macro slowdown or tightening in hyperscaler budgets would hit Broadcom's hardware top line first and most directly.
- Execution on software integration and margins. Infrastructure Software contributes stickiness, but if Broadcom cannot expand margins or cross-sell effectively, EPS leverage will be reduced.
Counterargument: The market is right to be cautious. A company trading at a mid-to-high 50s P/E needs consistent, visible evidence that AI growth will be durable and margin-enhancing. If AI growth reverts to lower rates or competitive pricing squeezes hardware margins, Broadcom could underperform materially despite superior recent results.
What would change my mind
I will scale into the trade only while the following conditions hold: AI revenue continuation above current levels, margin expansion in software, and disciplined capital allocation (buybacks or targeted M&A that improve moat and margins). Conversely, I would exit the position and revise the thesis if management signals a meaningful slowdown in order intake, if AI revenue growth drops below high-single-digits YoY, or if competitive dynamics force sustained margin compression. Missed guidance accompanied by weakening demand indicators (book-to-bill deterioration, cancelled orders) would force a reassessment.
Conclusion
Broadcom's selloff on slightly softer guidance is a classic market overreaction to a near-term number while underweighting the structural shift in AI spending and the defensive sauce of infrastructure software. With $16.7B in AI revenue for the quarter, free cash flow north of $32B, and a balance sheet that supports buybacks and product investment, a path to $30 in EPS is plausible if execution continues and the market recognizes it. The recommended trade - buy at $350, stop $320, target $525 over 180 trading days - captures the asymmetric upside while keeping downside defined.
Key monitorables: upcoming quarterly guidance, AI revenue trajectory, software margin trends, large hyperscaler contract announcements, and free cash flow allocation decisions.