We are starting to see rotation on the market from chips into software 🔄 The best current examples are Broadcom and Snowflake.
Broadcom reported objectively very strong results yesterday. Revenue reached $29.6bn (+86% YoY), GAAP EPS $2.68 and AI semiconductor revenue jumped to $16.7bn, more than triple year-over-year. Despite that, the stock is down about 5%, and it is already down about 27% from its highs. At the same time, AMD (-21%), ASML (-17%), LRCX (-33%) and others are significantly below their ATHs.
The reason is guidance for the next quarter of $34.8bn, which was not enough for high market expectations (the most optimistic analyst expectations were around USD 35.0 to 35.1 billion). With chips, it is apparently no longer enough to be good; you have to beat already extremely high expectations.
We see a completely opposite story with Snowflake. Product revenue grew 37% to $1.49bn, growth has even been accelerating for a third consecutive quarter, and the company also raised its full-year product revenue outlook from $5.84bn to $6.07bn. The stock is reacting with a gain of about 20%. A large part of the acceleration is coming precisely from AI products.
Salesforce reported similarly strong results last week. The stock soared more than 20% after the results, and ServiceNow, Adobe and other software companies rose along with it.
The market is thus gradually moving away from the "AI will destroy SaaS" narrative and starting to reward companies that can actually monetize AI. This is also confirmed by the August performance of the software ETF IGV, which rose by about 16%, while chip-focused SOXX only about 1%.
In my view, we are not seeing capital leaving AI. Rather a shift - from hardware, where expectations and valuations are already extremely high, toward software, where investors are finally getting proof that AI may not be a threat, but another source of growth.
Of course, this does not have to mean the end of the chip cycle. On the contrary, TSMC announced this week that its need for manufacturing equipment has grown to as much as 1.9 times the original estimate over roughly 6 months, while the company is simultaneously starting nearly 20 fabs around the world and still cannot keep up with AI demand. That is a very strong signal that fundamental demand for chips and the entire manufacturing infrastructure remains extremely strong.
But the relative strength of the market has been visibly shifting toward software in recent weeks. What do you think, will this rotation last in the coming months, or will capital return to chip stocks? 👇
