Lubin Investment · Blog

Broadcom (AVGO): AI up +143%, our fundamental analysis

2026-06-16 ·

Broadcom is a solidly quality business with a 30% FCF margin and a 48% Cash ROCE. Its AI revenues jumped 143% in Q2 FY2026. But with a valuation at 94 times its free cash flow, the price already reflects massive growth. The business is excellent, the price is high.

Q2 FY2026 results: numbers that take your breath away

Broadcom published results on June 3, 2026 that the market had been eagerly awaiting. Total quarterly revenues reached 22.2 billion dollars, up 48% compared to the same quarter a year earlier. A record. Adjusted EBITDA rose 52% to 15.2 billion, representing 69% of revenue. Margins at an impressive level.

What caught everyone's attention: AI-related semiconductor revenues reached 10.8 billion dollars, up 143% year-over-year. These revenues now represent 49% of total semiconductor revenue. Broadcom is no longer a diversified company doing a little AI among many other things. AI has become its main engine.

And the Q3 guidance goes even further: Broadcom anticipates AI revenues of 16 billion dollars, representing 200% growth year-over-year. Non-GAAP EPS (earnings per share, excluding exceptional items) reached 2.44 dollars in Q2, slightly above analyst expectations of 2.45 dollars.

Our fundamental reading: a well-deserved 7 out of 10

On my site, I analyze each stock through 10 fundamental criteria. The score of 7 out of 10 for Broadcom reflects brilliant points and some areas of caution. The Cash ROCE at 48% is remarkable. It measures the return on capital the company invests, calculated on real cash rather than accounting profits. 48% is almost double the threshold I consider excellent (25%).

The FCF margin of 30% is also very solid. For every dollar of revenue, Broadcom generates 30 cents of free cash after all expenses. And cash conversion is 70%: 70% of accounting profits convert to real cash, which is reassuring about the quality of results.

The main caveat: FCF per share growth is only 7% per year over 5 years. That is decent, but disappointing for a company with a market capitalization over 1.8 trillion dollars. The other caveat: dilution of +3.5% per year (the share count increases, mechanically reducing each shareholder's stake). A criterion I always watch carefully.

Broadcom's AI bet: between diversification and concentration

Broadcom's story is one of a conglomerate built through successive acquisitions: Avago, CA Technologies, Symantec Enterprise, VMware. Each acquisition brought additional revenue streams, first enterprise software, now AI semiconductors. Hock Tan, the CEO, is a formidable capital allocator known for his ability to extract margins from every acquired asset.

Broadcom's AI strategy focuses on two pillars. First, custom ASICs (purpose-built chips) for hyperscalers who want dedicated accelerators. Second, AI networking components (Ethernet switches, optical interconnects) that equip GPU clusters. Both segments directly benefit from the explosion of AI infrastructure spending by the major cloud players.

But here is the trade-off I see clearly: Broadcom is now heavily concentrated on a few hyperscaler customers for its AI revenues. If one major client decides to bring chip production in-house or switches supplier, the revenue impact could be brutal. That is the flip side of such spectacular growth: dependence on a handful of strategic accounts.

Valuation: the difficult part of the AVGO analysis

P/FCF measures how many years of free cash flow you are paying for when you buy a stock. A P/FCF of 20 is reasonable for a quality company. A P/FCF of 94, like Broadcom's today, means you are paying 94 years of free cash flow at the current price. This multiple reflects very high growth expectations.

Our model estimates a reasonable buy price at 80.66 dollars, based on current FCF and normalized growth assumptions. The current price is 393.94 dollars, a premium of 79.5%. Broadcom is an excellent business. But at this price, everything must go exactly as planned for years.

To track Broadcom's ratios in real time, check the full analysis page on my site: lubin-investment.com/analyse/AVGO.

FAQ

Is Broadcom's AI business real or just a trend?

The numbers are very concrete: 10.8 billion in AI revenues in Q2 FY2026, +143% year-over-year. These revenues come from custom ASIC chips for hyperscalers like Google and Meta, and from networking components for GPU clusters. This is not marketing, it is real billing.

Why does Broadcom score 7 out of 10 and not higher?

The score reflects 10 objective criteria. The strengths (Cash ROCE 48%, FCF margin 30%) are offset by FCF per share growth limited to 7% per year over 5 years, moderate dilution (+3.5%/year) and net debt at 2.2 times annual FCF. 7 out of 10 remains an excellent score, but it is honest about the few areas of caution.

What is FCF and why is it more important than net income?

FCF (free cash flow) is the cash actually generated by the business after all expenses, including capital expenditures. Net income includes non-cash items (depreciation, amortization) that can be manipulated accounting-wise. FCF is harder to manipulate and better reflects the true health of the business.

What is the main risk for Broadcom in 2026?

Customer concentration. With 49% of semiconductor revenues tied to AI and dependence on a few hyperscalers for those orders, any insourcing decision or supplier change by a major client could strongly impact revenues. Execution on the 200% AI growth guidance in Q3 will be crucial for market confidence.

Should you buy AVGO or wait?

Not investment advice. But our analysis shows quality is there (7/10) and AI growth is real. Valuation at 94 times FCF leaves little room for error. Our model puts the reasonable buy price at 80.66 dollars. Whether the growth premium justifies the risk at this level is yours to decide.

Voir l'analyse AVGO sur Lubin Investment

About the author

Written by Lubin Danilo, founder of Lubin Investment. A self-taught individual investor, I have analyzed stocks through their fundamentals for several years and invest my own money with this method. I codified it into a tool that judges a company's quality and its price separately, using criteria drawn from the financial literature (Warren Buffett, Michael Mauboussin, Aswath Damodaran).