Broadcom Price Targets Split After Earnings Despite 221% AI Revenue Growth

Broadcom delivered the sort of AI growth that would normally produce a straightforward bullish analyst response. Fiscal third-quarter revenue rose 86% from a year earlier to $29.6 billion, while AI semiconductor revenue reached $16.7 billion. Yet the analyst tape after the report moved in both directions.

That disagreement is the story. Broadcom's long-term AI opportunity became larger, but the near-term bar was already exceptionally high. Reuters reported that the company's approximately $34.8 billion fourth-quarter revenue outlook came in slightly below the prevailing Wall Street expectation. The result was a familiar late-stage AI trade tension: stronger absolute fundamentals alongside a market that wanted an even faster near-term trajectory.

Broadcom's AI numbers are extraordinary even before the valuation debate

Broadcom said Q3 revenue increased 86% year over year. Reuters reported AI chip sales of $16.7 billion and said management now expects roughly $115 billion of AI chip revenue in fiscal 2027, up from an earlier expectation above $100 billion, with a path to approximately $230 billion in 2028.

Those forecasts position Broadcom as one of the clearest public-market beneficiaries of hyperscaler spending outside Nvidia. Its exposure is different: custom accelerators and networking silicon give customers another route to build large AI systems, particularly when hyperscalers want hardware tailored to their own workloads.

Analysts agreed on AI demand but disagreed on what it is worth

The September 3 and 4 actions show a genuinely mixed valuation response. Macquarie upgraded Broadcom from Neutral to Outperform with a $490 target. Morgan Stanley nudged its target to $505 from $502. By contrast, Truist cut its target to $520 from $550 while maintaining Buy, TD Cowen reduced its target to $475 from $500, and Evercore ISI moved to $578 from $582 while retaining Outperform, according to Benzinga. DA Davidson subsequently cut its target to $350 from $400 while maintaining a neutral stance.

At the bullish end, Cantor Fitzgerald moved to $600 from $525 and Citi later raised its target to $515 from $500. The spread is useful because it demonstrates that the argument is no longer simply 'AI demand up, stock target up.' Analysts are assigning different values to the durability, margin profile and concentration of that demand.

The near-term forecast explains why great results were not enough

Broadcom's Q4 revenue guidance of about $34.8 billion represented 93% year-over-year growth, according to the company. In isolation, that is an exceptional forecast. Against the expectations built into the stock, however, Reuters reported it was modestly below consensus.

That gap helps explain why some analysts could become more confident in the 2027 and 2028 AI opportunity while trimming a 12-month target. A price target compresses several variables into one number: earnings estimates, free cash flow, the valuation multiple and the date at which the analyst applies it. A stronger long-run revenue forecast does not force a higher target if the analyst simultaneously reduces the multiple or near-term estimates.

Customer concentration and custom silicon are the next variables to watch

Broadcom's custom-accelerator opportunity is powerful precisely because a small number of very large customers can deploy enormous amounts of capital. The same structure creates concentration risk. Changes in a hyperscaler's internal architecture, supplier mix or capex plan can move billions of dollars of expected revenue between vendors.

Competition also matters differently in custom silicon than in general-purpose accelerators. Broadcom does not need to displace Nvidia across the whole AI market to grow rapidly. It needs to retain valuable design positions with major customers and pair those accelerators with networking content. That makes design wins, customer concentration and networking attach rates more useful future indicators than a single quarterly AI growth percentage.

GMR view: the analyst split is a valuation signal, not an AI-demand warning

Global Markets Review reads the post-earnings disagreement primarily as a valuation signal. The evidence for AI demand strengthened: Broadcom increased its long-range expectations and reported another very large step-up in AI semiconductor revenue. What became harder is deciding how much of that future should be capitalised into the stock today.

That makes Broadcom a useful case study in why price-target cuts are not automatically bearish and target raises are not automatically new information. Several firms kept positive ratings while trimming targets, while others raised targets or upgraded. The common ground is strong AI demand. The disagreement is over the price investors should pay for it, the margin mix of future customers and the probability that today's enormous forecasts are fully realised.