.jpg){kind=link}
The useful contrast in this week’s reporting was between the price of financing and the evidence of commercial demand. Microsoft had product developments; Akamai had a large customer agreement; oil and yields kept changing the conditions around both. I do not want to compress that into a generic good week or bad week for AI.
My published portfolio includes chip suppliers such as AMD, MU and AVGO, platforms such as META and GOOGL, and cloud-related exposure such as CRWV and NBIS. That is not three unrelated stories. Spending by one part of the chain can become revenue for another.
This connection is helpful when demand is strong and dangerous when I mistake linked evidence for independent confirmation. A customer’s commitment cannot be counted once as proof of the cloud operator’s demand and again as completely separate proof of the chip supplier’s demand without acknowledging the common source.
I would put the funding chain beside the holdings list: who pays, who supplies, who borrows, and who absorbs a delay. I am not assigning historical daily dollar contributions here because the current share file is not a verified transaction history for every date.
That limitation does not prevent useful research. It prevents a false precision. The week’s evidence makes the common dependency clearer, and that is worth recording even without an invented account return. This is a retrospective assessment of the published book, not a claim that a particular trade or conversation happened on the coverage date.
Aydin Ali
AEA Capital Research