Tesla's ability to build cars at scale could make its Robotaxi network dwarf Waymo's in the long run. Waymo still has over twice Tesla's driverless coverage: 1,874 square miles in 15 cities versus 831 in seven. Tesla's 1,928-square-mile total includes 1,097 in the Bay Area with a safety driver. Tesla just launched in Fort Worth and expanded Houston coverage more than tenfold. Coverage is not rides or revenue; each new city requires testing, approvals where required and infrastructure. x.com
A 3-4% annual dividend could help memory stocks rerate by attracting dividend investors and trillions in equity-income strategies, if the big three start paying one. Micron, Samsung Electronics and SK hynix should generate enormous operating profits in 2027 despite slower growth than in 2026. The big three will buy back at least $300-400B in 2027, possibly more, though the amount depends on the split with special dividends. x.com
Buying Lumentum would give Marvell the AI optics signal chain from DSP to laser, a combination only Broadcom can match. Lumentum's indium phosphide lasers are the tightest bottleneck and sold out through 2028; it announced a $5bn-plus co-packaged optics win. World Trade Securities estimates a $550bn combined addressable market versus $400bn for Marvell alone. The 29% selloff prices in too much damage, despite China approval, customer hedging and Marvell running fabs for the first time. x.com
Apple's iPhone upgrade cycle is just getting started because the full iPhone 18 Pro launch and $1,999 foldable have yet to arrive. iPhone sales grew 22% last quarter; the latest quarter brought $57.6bn in iPhone revenue despite just nine days of iPhone 18 Pro sales, while revenue of $114.5bn and EPS of $2.10 beat expectations. Rising memory costs still test margins, and December guidance is the next test of whether the cycle can carry into 2027. Overweight, $380 target. x.com x.com
A stock-for-stock SpaceX offer may not deliver its headline premium to Tesla holders because merger arbitrage selling could depress SpaceX shares before closing. The Marvell-Lumentum deal illustrates the risk: an initial 24% premium shrank as Marvell shares fell, leaving Lumentum below its pre-rumor price. A SpaceX-Tesla deal could take as long; even a guessed 20-25% initial premium may be whittled down. x.com x.com x.com
Buybacks by the three DRAM makers should support their shares even if the broader market falls. Buy memory, especially DRAM, as a relatively defensive AI trade, not on conviction in the DRAM cycle. Samsung and SK hynix are nearing 1x 2027 P/B, at 1.36x and 1.47x; past moves below 1x brought technical rebounds. A memory-price decline in Q3 2027 is a credible risk. New highs depend on broader sentiment, but memory has a fairly good chance of outperforming other AI trades. x.com x.com
Tesla's AI5 chip will enable unsupervised FSD and lift its robotics portfolio. Once shipped in mid-2027, vehicle sales will surge to 4M+ a year, Robotaxi will expand exponentially and Tesla revenue will reach $200B+ a year. Growing FSD subscriptions and Robotaxi revenue should push gross margin past its 25.6% peak in 2027, with $30B in gross profit projected. Current numbers are better than they appear because high-margin tax-credit revenue has nearly vanished. x.com x.com x.com x.com x.com
Intel's EMIB-T slip should be brief because a systemic process bottleneck looks likelier than a severe MediaTek design error. MediaTek's CEO said the key yield issue was identified and is being fixed with Intel and the substrate maker; the market prices the design-error case. Switching to CoWoS would take 12-18 months and require supply. With external customers and financing secured, plus improving 18A/14A yields, Intel's rise is unstoppable despite the setback. x.com x.com x.com
Bitcoin faces midterm pullback risk because the last three midterms each brought 20%-50% corrections. Unlike those elections, Bitcoin has already bottomed, so a correction from here is possible but not a new cycle low. US spot buying and the Coinbase premium are improving near $77,700, but volume is about half normal. A strong 15-minute close above $77,900 with volume would help confirm the bounce; losing $77,500 would weaken it. Macro and the dollar could still weigh on Bitcoin. x.com x.com
Tesla's case for patient investment now has firsthand FSD reliability and profitability during infrastructure building behind it. FSD has driven almost 15,000 miles without intervention in firsthand use; Tesla has record cash on hand, and very few companies can build infrastructure while turning a profit. Autonomy and Optimus still call for patience. Tesla's fundamentals look their best since investing began in 2019, and the stock looks on the verge of an explosion. x.com x.com x.com x.com x.com