X
Everything we have picked from this writer, newest first.
Stan gave the example of 2001. $CSCO had aggressively grown its backlog every yr through the Internet 1.0 buildout & the Street expected the same in 2001, but instead $CSCO’s backlog plunged. The build was over, & with that equipment demand evaporated. That will happen again. https://t.co/7yB6vDgcl8
Here’s some color, via an options desk: Late Friday & Tuesday there was substantial FLEX call buying across mult AI cos ($CRWV, $AMD, $BE, $INTC, $SNDK, $SKHY, $DRAM. Trades represented $1.1B of delta/$5.8M vega. Total option premium was ~$315M. Buyer didn’t seem price sensitive. https://t.co/qCL8RnxHbY
The massive share buybacks at $SKHY & Samsung are highly supportive to the stocks but quickly coming to an end. The Hynix buyback is 32% done & at current pace will conclude by 10/15. For Samsung the comparable numbers are 37% & 9/30. Then we see where prices shake out.
FT: Big tech giants booked a $160B+ windfall last qtr from investments in other AI companies. $GOOG, $AMZN, $NVDA, & $MSFT all had lg boosts to their pre-tax profits from ‘other income’ in theory P&L, due largely to valuation gains on equity stakes in other leading AI companies
Why does he think hedge fund analysts “missed” $NVDA? Every fund I know has been long for some part of the last 5+ years. Did he think it became the most valuable stock in the world from just him (the only hedge fund to own it) and mutual funds? Bull market genius. https://t.co/mS6HJQX61L
Similar to what we’ve seen happen at $WMT HQ, $T has found a way to cap/hold flat AI spend; a router sends each task to the cheapest model that can do the job, & open weight models that are now good enough for most of the work. Hearing this more & more as I speak to companies.
This guide tonight from $SMTC is remarkable.
