Capital_Squeeze❄️

Since I’ve been asked a lot: Buy stock in several companies that make products & services that *you* believe in. Only sell if you think their products & services are trending worse. Don’t panic when the market does. This will serve you well in the long-term.

$LITE CEO Michael Hurlston addressed the 45% selloff today. On demand, he said “There’s a massive disconnect between what we’re seeing in our demand book and what the markets are showing. Our demand couldn’t be stronger.” Asked whether the market reaction gives him pause on $MRVL as the buyer, he said “We really like Marvell. We think they’ve got a lot of upside from where they are, and that’s part of the reason we took all stock.” He admitted “the timing was bad” but didn’t back off. “Our board of directors never considered anything but taking all stock,” citing diligence on Marvell’s books and design wins. On getting the deal closed, targeted for 1H28: “It’s going to be a bumpy one.” He named regulatory approval as the biggest hurdle left. $LITE $MRVL

Rosenblatt on $AAOI: "We believe each $1M of InP substrates can support ~$30mn of transceiver sales, so the $194.4M 2027 contract could underpin ~$5.8bn of annual transceiver revenue, or roughly 12mn modules at the ~$500 blended ASP implied by AAOI’s 2027 revenue and capacity targets. Over the full five-year agreement, that implies ~$29bn of transceiver revenue at constant pricing. Our estimate is consistent with AAOI’s target of a ~$471mn monthly data center transceiver run rate by mid-2027, or ~$5.7bn annualized. Substrates are therefore only ~3% of transceiver revenue, so rising substrate prices should have a limited impact on margins, and over its full term the contract can also support growth beyond 2027, including high-power lasers for ELSFP and CPO"

$LITE $MRVL Win take by Itachidata. Why would Lumentum take 100% stock if they thought the AI cycle was peaking? You'd take the cash and walk away. Instead, they're going all in. A company sold out into 2029 that still asks for stock is telling you where it thinks this is going: up and to the right. For Marvell, it's vertical integration: its DSPs and custom silicon plus Lumentum's lasers, just as optics move closer to the processor. As AI scales, interconnect becomes the priority. It's also a hedge. Marvell's crown jewel is its optical DSPs, and co-packaged optics threatens them. Move the optics onto the chip package and a lot of those DSPs disappear. The one part CPO can't remove is the laser. So Marvell is protecting itself from its own disruption by buying the light source, and paying in stock means doing it without the cash burn the market keeps punishing. To me, this is a win-win. Lumentum gets Marvell's scale and a seat next to every custom chip Marvell builds for the hyperscalers. Marvell gets to own the shift to CPO, starting from the light source. This is the real path to becoming a trillion dollar company.

*NVIDIA’S HUANG 'GAVE BLESSING' TO MARVELL-LUMENTUM MEGADEAL

The counter: Lumentum took 100% stock, not cash. A seller who thought the AI cycle was peaking would want cash and walk away. By taking stock, LITE holders become a third of the combined company. They're betting even more on AI connectivity, just with a bigger partner. Its laser supply is also reportedly sold out into 2029, and companies at a peak don't sell with that kind of backlog. They sell when they're running out of customers. The main reason in my view is that now they have one combined product Marvell makes the electrical chips (signal processors, switching and custom AI chips), and Lumentum makes the lasers and optical parts. AI data centers increasingly want the chip and the optics built together, especially as optics move closer to the processor. One company owning both cuts out the back-and-forth between suppliers. Owning both the chip and the optics lets it offer AI customers a complete package that no single competitor can match. Its more than $1B in cost savings comes from producing that package together, which lowers its costs and lets it compete harder on price. Once a hyperscaler designs that combined product into its data centers, switching suppliers gets costly and slow. Speed is the edge. AI hardware now moves on a yearly cycle, and whoever delivers connectivity first gets designed into the next system. Owning both the chip and the laser shortens every step

JUST IN: UAE signs 4 deals with OpenAI to become the world’s first "fully AI-native government"

South Korea plans to raise government AI investment to 21.3 trillion won next year, nearly twice the current level.

YOU KNOW WHAT IT MEANS WHEN AN AI COMPANY CUTS OVER A QUARTER OF ITS WORKFORCE, RIGHT?

More than half of the Fortune 500 already qualifies as a significant customer, but Anthropic’s revenue from those companies averages less than 0.1% of their expenses. Its ten largest Fortune 500 customers spend more than eight times the broader cohort average while still spending less than 1% of expenses.

$LITE $COHR $AAOI $SIVE $CRDO WOW. @JensenHuang just confirmed the optical supercycle. “Scale up is going to become optical.” 15x volume. Orders of magnitude more optics. Copper is tapped out inside the rack. The bottleneck is now light. > $LITE: NVIDIA already wrote the $2B check. Sold out. > $COHR: Same deal. Capacity locked through 2028. > $AAOI: 800G/1.6T modules into a wall of demand. > $SIVE: Pure InP laser arrays for CPO. Built for this. > $CRDO: AECs today, optics + SiPho tomorrow. Jensen just named the volume. These names own the capacity. Position accordingly, be smart. We are long and bullish many of the stocks mentioned here. NFA.

It’s a great tax strategy if you need a write off to hold CCXI into the merger

$CCXI stock was 🟢 by less than 1% for over 30 min during regular market hours. @agilityrobotics should PR that. One of their biggest milestones so far.