Dear WIMM readers,
Today I am looking at the latest earnings from Broadcom, a company I owned last year and sold in October. Yes, I regret selling it at that price. But, to be fair, I still closed the position with a 30% profit, so it was not exactly a tragedy.
That said, the company remains too important to ignore, which is why I thought it was worth revisiting the thesis. My first verdict on Broadcom was buy. Now it is hold.
Below, you can see what changed… and why.
TL;DR
Broadcom guided AI chip revenue below expectations ($16B vs. $17.2B for Q3), triggering a $1.3 trillion sell-off across the AI chip supply chain, which was the worst single-day drop for the PHLX index since March 2020.
Despite the Google/Anthropic partnership, CEO Hock Tan admitted Google will diversify its chip suppliers (MediaTek, in-house design), suggesting Broadcom's TPU business is a semi-captive product line tied to a narrow customer base rather than a broad AI platform.
The miss is less an "AI demand slowdown" and more an Nvidia victory, the ecosystem that matters is Nvidia's, not custom silicon in general.
Verdict: Hold.
Q2 2026 in a nutshell:
However, this is from Bloomberg:
Broadcom Inc. shares plunged by the most in more than 16 months after the company’s forecast for sales of its artificial intelligence chips disappointed investors. AI semiconductor revenue will be $16 billion in the fiscal third quarter running through July, the company said in a statement Wednesday, falling well below analysts’ expectations of $17.2 billion on average. Chief Executive Officer Hock Tan said Broadcom will sell $56 billion worth of AI chips in the fiscal year that ends in October, also falling short of estimates.
The share price is still +11.4% YTD (as of 12th of June 2026):
However, as a result of the call, a massive sell-off began in the AI supply chain. From Reuters:
U.S.-traded chipmakers plunged on Friday, losing about $1.3 trillion in market value, with deep losses in AI heavy hitters including Nvidia, Micron Technology, and Advanced Micro Devices, as Broadcom’s weak report earlier this week reverberated across Wall Street. The PHLX chip index slumped 10.3% in its deepest one-day loss since March 2020, when the coronavirus pandemic threw global markets into a tailspin. Friday’s selloff added to losses on Thursday after Broadcom issued a quarterly report that showed demand for its custom AI chips business falling short of lofty expectations.
What happened to Broadcom? Everything should be fine, especially after their partnership with Google. Here is The Wall Street Journal:
Broadcom will develop and supply custom artificial-intelligence chips for Google and additional computing capacity to Anthropic in an expansion of the strategic collaboration between the three companies.
The chip manufacturer said Monday it will supply Google with custom Tensor Processing Units, along with networking and other components for Google’s next-generation AI data center racks through up to 2031 as part of a supply assurance agreement.
I found out in the prepared remarks of the CEO Hock Tan:
Now we also accept the fact that while we like to win every design in that program, we also accept the fact that given the growth of consumption and — development and consumption of AI compute even by our partner, Google, that we fully expect that there will be some diversity of sources for them. (NB! my emphasis with bold)
Aha! Broadcom won’t be their only partner.
Google is already known to be working with MediaTek, while also trying to move more chip-design expertise inside the company. Anthropic, for its part, almost certainly has similar incentives over the very long term. If AI infrastructure becomes the core constraint of the business, then owning more of the stack becomes strategically attractive.
That raises an interesting question → “How connected is this news to the earlier one?” Broadcom’s most important AI exposure is through TPUs. But if TPUs are mainly tied to a narrow customer base, essentially Google and Anthropic, then perhaps the market is starting to treat them less like a broad AI platform and more like a specialized, semi-captive product line. Nvidia, meanwhile, remains the default choice for almost everyone else building AI at scale.
Maybe Broadcom’s weakness is not really an “AI miss” in the usual sense, but it may be better understood as an Nvidia victory. Broadcom is not necessarily losing because demand for AI infrastructure is slowing, but because the category that matters most is not custom silicon in general, but Nvidia’s ecosystem in particular. Without more evidence, and I probably will not get that until the next earnings cycle, I would lean toward the latter interpretation.
Verdict: Hold
WIMM for companies: If you would like to offer the newsletter as a benefit to other employees in your company, please contact me directly. ([email protected])
Work with me: If you are interested in working with me on your personal portfolio, again, contact me. I’m a certified financial consultant.
Finally, “Where is my moat?” is designed for individual readers, though the occasional forward is absolutely fine.
Thanks for your support & have a wonderful day!
