The Bull Rankings scorecard — our quality-growth score is 72.3 / 100, built from three pillars each graded 0–100 against sector peers: Quality 80, Growth 93, Value 51. At today's price, our reverse-DCF read says the market is implicitly betting on about 48% a year in free-cash-flow growth sustained for a decade — a gauge of how much optimism is already in the stock.
The thesis
Broadcom is a high-quality compounder trading at a premium because the market is paying up for its unmatched ability to turn AI infrastructure demand into cash. In the quarter ended 2026-05-03, revenue grew 32.3% year over year while the profit margin sat at 38.8%. That’s the kind of operating leverage that justifies a P/E of 63.3, especially when the return on equity is 33.4% and our model’s quality-growth score stands at 72.3, with Growth the strongest pillar at 93 and Value the weakest at 51. The market isn’t rewarding AVGO for being cheap; it’s rewarding the business for being the closest thing to a tollbooth on the AI build-out.
What the business actually is
Broadcom sells the glue that holds AI data centers together. In Semiconductor Solutions, it peddles custom silicon for networking, ethernet switching and routing, fiber-optic components, and PCIe switches that shuttle data between accelerators and CPUs. In Infrastructure Software, it licenses mainframe and automation software that enterprises still can’t live without. The Semiconductor Solutions segment is the growth engine; the Infrastructure Software segment is the annuity that funds the next round of R&D. Together, they let Broadcom capture a slice of every dollar spent on AI infrastructure, whether it’s spent on chips, cables, or the software that glues them.
Why it can (or can't) keep compounding
The durability story is simple: Broadcom’s returns are so high that competitors can’t afford to play the same game. A 33.4% return on equity means the company generates more than a third of its book value in profit every year, and a 38.8% profit margin shows that scale hasn’t eroded pricing power. Our model’s signal—raising the dividend—confirms management’s confidence that these returns aren’t a one-time spike. The moat isn’t just the product portfolio; it’s the fact that designing a custom ASIC for hyperscale AI is a multi-year, multi-billion-dollar gamble few can make. Nvidia may own the GPU, but Broadcom owns the switch that connects it to the rest of the rack. That’s a position incumbents won’t cede without a fight.
The valuation question
The price already assumes AI demand stays red-hot for a decade. Our model’s reverse DCF says today’s $380 price implies free-cash-flow growth of roughly 48% a year for the next ten years. That’s more than 50% faster than the 32.3% revenue growth posted in the quarter ended 2026-05-03. The PEG ratio of 0.44 suggests the market is pricing in both high growth and high quality, but the reverse DCF is the real tell: it’s betting the company can sustain hyper-growth while the P/E sits at 63.3. Either AI capex keeps accelerating, or the multiple has to come down. There’s little room for error.
The bear case
The weakest pillar—Value at 51—is the skeptic’s best friend. A debt-to-equity of 0.74 isn’t reckless, but it’s high for a semiconductor company that’s already trading at a premium. If AI spending slows, Broadcom’s leverage magnifies the downside. The stock’s 52-week range—$281.87 to $495—shows how quickly sentiment can shift when growth stalls. The market has already knocked a chunk off a business that grew revenue 32.3% last quarter, a reminder that even the best franchises get punished when the cycle turns.
What would change our mind
Two things would flip the thesis. First, if the profit margin slips below 35%, it would signal pricing power is eroding. Second, if the dividend raise stalls, it would break the model signal that underpins the compounding story. Either would drag the quality-growth score lower and force a rethink of the 48% implied growth. Until then, Broadcom remains the tollbooth worth paying for.