Suddenly realized a point: Meta's current market cap is only 1.46 trillion, even Microsoft is only 2.8 trillion, while TSMC has risen from 1.5 trillion last year to 2.4 trillion now.

Buying the "Magnificent Seven" in the past few years was essentially buying "certainty." If the world were to develop "steadily" like this, the Magnificent Seven would indeed be the "most reliable business logic" of the future.

But after the emergence of AI, this logic has begun to "loosen," and the world has started to become "uncertain" because of AI. All business logic is being "repriced." Among them, the Magnificent Seven once evaporated about "2 trillion" US dollars in June, with a median drop of about "9.7%," while the median of other S&P stocks was "+0.3%."

On the other hand, Micron once surged over 260% this year, and SK hynix's year-to-date gains have also been very exaggerated. These "non-Magnificent Seven" stocks like MRVL, AVGO, and TSM are starting to be "repriced."

The market may be undergoing a change: from "old certainty" flowing to "new bottlenecks."

The Magnificent Seven are still very strong.

But the "certainty premium" they exclusively enjoyed in the past may now be being shared by these assets:

Newly listed giants and future AI labs: SpaceX, OpenAI, Anthropic, etc.

AI infrastructure bottlenecks: Micron, SK hynix, AVGO, MRVL, TSMC, etc.

Berkshire Hathaway buying Google is also very interesting: traditional capital is starting to take over "old certainty," while growth capital may be looking for the next batch of "new certainty."

"Certainty" hasn't disappeared; it's just that "certainty" is changing locations and may even continue to flow out in the future.

Certainty used to belong to the Magnificent Seven.

The certainty now may be the scarcest link within the "new uncertainty."

$Marvell Tech(MRVL.US)$Micron Tech(MU.US)$XL2CSOPHYNIX(07709.HK)$Roundhill Memory ETF(DRAM.US)$Taiwan Semiconductor(TSM.US)$Qualcomm(QCOM.US)$AMD(AMD.US)