26 published issues · updated weekly
An AI-infrastructure selloff collides with resilient orders, local data-center restrictions, rising financing costs and new company results. Tower, Coherent, BWXT and the question of who keeps the money.
Microsoft could fund its infrastructure spending from operating cash; Alphabet and Amazon could not. AI hardware suppliers are increasingly taking on financing risk themselves while the Frontier File assesses a large PSQ hedge.
The Nasdaq-100 fell less than one percent, yet the Frontier File portfolio lost 3.48%. SpaceX, Meta and Tesla show three different ways to finance the AI buildout while the PSQ hedge faces a crucial test.
TSMC and ASML show AI infrastructure demand is still real. CoreWeave's debt raises the financing question. A reconstructed defensive portfolio avoids part of a brutal week but leaves a nearly 30% PSQ hedge that needs an exit plan.
Lumentum says its optical capacity is sold out through early 2029. The Treasury borrowed at its highest 10-year auction yield since 2000, while the Fed discussed AI-related borrowing. The demand looks real. So does the invoice.
Micron printed. BWXT opened the books. Rocket Lab sold twenty launches. Jobs rolled over. Bond yields hit a 24-year high. Then the homework stopped being research and became an allocation.
Nuclear has my capital. Memory has my attention. Now optics wants an interview. Tower, Coherent, and Ciena sell into the same broad need—but offer very different ways to get paid.
The Fed raised rates. Rocket Lab raised capital. And I caught myself wondering whether the portfolio had become too safe. The next dollar needs a better argument than the last green candle.
Oracle proved the AI build is real—and showed what it takes to finance. Rocket Lab flew again and found another piece of the space stack it can sell today. Inflation firmed, oil moved above $107, Treasury yields nearly touched 5%, and my portfolio learned the next lesson after proof: time itself has a price.
Nvidia and Broadcom confirmed the AI build. Rocket Lab flew. Planet printed a better business. The market still reserved the right to pay scarcity and punish timing. Receipts arrived. The rent stayed due.
Nvidia proved the AI build is still accelerating. Jackson Hole reminded the market that real does not mean cheap. The technology passed. The discount rate did not.
The long bond hit 2007 territory, semiconductors cracked, and the portfolio’s latest moves made one thing clear: the future still works — the market is just charging more to wait for it.
The companies are getting bigger. The questions are getting harder. Scale is proof with new failure modes.
Proof is becoming abundant. The technology argument is getting easier. The economic argument is getting harder.
Six portfolio companies report before next Monday, one jobs print can move the discount rate, and BWXT may have the cleanest chance to earn more weight.
Intel confirmed the demand. The market questioned the economics. Now the four largest technology spenders face the same test.
Inflation cooled, yields eased, and the market opened the door—but execution decided which frontier companies deserved to walk through.
The market keeps climbing despite rising yields, geopolitical risk, and a hawkish Fed because AI infrastructure demand remains louder than the wall of worry.
After five issues of promising to trim Micron, Alex finally did it — then rolled the proceeds into SpaceX and traded one kind of concentration for another.
The chip trade cooled, the grid hit record demand, and the AI thesis began moving down the stack from silicon into power, cooling, transmission, and materials.
Rocket Lab stopped being a launch proxy and started becoming a space platform, while chips rotated lower and the power grid revealed the next AI bottleneck.
Micron won the week while tech lost it — a live case study in the Memory Wedge, concentration risk, and why owning the scarce thing still requires discipline.
A near-sale, a Micron blowout, and the discipline of holding the thesis while managing the size. Day 85, +58% on capital, 20/20 positions green.
One retail account, one frontier-tech thesis, documented in public. A hawkish Fed surprise on Wednesday, a one-day correction, and by Thursday's close the dip-buyers — and a signed peace deal — had already erased it.
One retail account, one frontier-tech thesis, documented in public. This week the two biggest overhangs of the whole journey resolved in four days.
Opening dispatch: AI compute, space, nuclear, quantum, fintech rails.