Skip to main content
The Frontier File · Issue #6 · July 2 2026 · Best Quarter Since 2020

The Rocket
Went Shopping.

One concentrated frontier-tech portfolio, documented in public. Rocket Lab stopped being a launch proxy and started becoming a space platform — while the chip trade cracked and the grid quietly revealed the next bottleneck.

+65%
Q1 done · on capital
▼ off the +68% high
$8,672
Book value · from $5,270
19 / 20
Positions green
Cold Open

The thesis grew up. The sizing is still catching up.

For three months my space thesis was a halo trade: own the public names, ride the SpaceX wave, collect the spillover. This week it became something bigger. Rocket Lab — the little launch company I bought at $57 — announced it's acquiring Iridium, a profitable satellite-communications business, for eight billion dollars. The stock jumped 16%. It's no longer just a rocket. It's a consolidator, using its own elevated stock to roll up the space economy while the sector is full of small, cash-strapped public companies ripe for the taking.

The deal lit the whole complex — Planet Labs, which I also own, jumped in sympathy. Alphabet debuted in the Dow, SpaceX picked up a fresh index catalyst, and the market closed a powerful quarter with frontier assets no longer moving as one crowded trade. My self-reported book crossed a new high during the week.

At the same time, the largest U.S. power grid prepared for record demand as heat and data centers hit together. That matters because the frontier story is widening: launch companies are becoming communications platforms, and AI is becoming an electricity, cooling, transmission, storage, and materials problem. The next winners may not look like the last ones.

And then the back half of the week reminded me the run isn't a straight line. A second straight semiconductor rout knocked Micron down 5.5% Thursday, and because I still haven't trimmed it, my biggest position dragged the whole book off its high. The thesis grew up this quarter. My discipline about position size did not. That's the story.

The Thesis Board

Where my head is at

The map before the positions. firing   mixed/watching   next-wave seed   headwind.

Space Economy RKLB · PL · MDA · ASTS · SPCX
From halo trade to consolidation. RKLB's $8B Iridium buy turns it into a sector roll-up; PL +16% in sympathy; SPCX joins the Nasdaq-100 July 7. This cluster held its gains even as chips sold off — the diversification working.
Power Is the Next Bottleneck BWXT · OKLO · CENX
Record grid demand is the less glamorous half of the AI buildout. Compute now pulls power, cooling, transmission, storage, nuclear, and materials into the center of the thesis.
Concentration MU ~21% + Space ~22% ≈ 43%
THE issue of the quarter — and this week it bit. MU fell 5.5% Thursday in the chip rout and pulled the whole book down ~$100 on that one name. Two themes, nearly half the account. Both winning. Both deserve deliberate sizing instead of automatic trimming.
The Chip Rotation SMH −4.5% · MU · NVDA · ARM
Two-day semiconductor rout, Korean chips down hard (Kospi −7.9%), a "revaluation of the AI trade." Not a demand break — MU just guided to $50B/quarter — but a real rotation out of the group that carried me all quarter.
Rates / Jobs dovish shift
June payrolls came in soft — 57K vs 113K expected — breaking the hot streak and easing the hike fear that's pressured me for weeks. Yields fell to 4.41%. A genuine tailwind for my rate-sensitive names.
Iran / Geopolitics fragile but easing
The ceasefire cracked and re-mended over the past week; Qatar says talks are positive. Oil fell to $67 — the lowest since March. Disinflationary if it holds, but the headline risk never fully leaves.
Next-Wave Seeds SERV · OKLO · QBTS
Planted small, behaving. The Marvell ($230–250) and Dell ($400–420) dip entries and the HONA spin are all waiting — funded by the trims I keep promising.
The Week in One Tape

Just the numbers

MarkerLevelRead
RKLB buys Iridium $8B
~$105
+16% Mon
Dow record close
52,900
+1.14%
Nasdaq chips drag
25,833
-0.80%
MU chip rout
~$1,140
-5.5%
June jobs dovish miss
57K
vs 113K
WTI Crude Iran talks
~$67
Mar low

The split couldn't be cleaner: the Dow set a record while the Nasdaq fell, as money rotated out of the semiconductors that carried the best quarter since 2020 and into everything else. The soft jobs report is dovish — good for rates — but it landed the same day chips had their second ugly session. My book is caught in the middle: space and Dow-style names up, my oversized memory position down.

The Portfolio Report

The whole book — quarter one closed

MU
+254%
~21% · −5.5% today
RKLB
+83%
Bought Iridium 🚀
PL
+69%
+16% sympathy 🚀
CENX
+48%
Held on rotation
STX
+47%
Memory tailwind
BWXT
+41%
Nuclear defense
AEVA
+38%
LiDAR
GOOGL
+33%
EU fine −1%
MDA
+32%
Space systems
ASTS
+30%
Sat connect
ACHR
+27%
eVTOL
JOBY
+26%
eVTOL
NVDA
+21%
−1.4% today
SPCX
+19%
Nasdaq-100 Jul 7
AMZN
+19%
Cloud · robots
SOFI
+9%
Dovish jobs help
OKLO
+6%
Nuclear AI
QBTS
+4%
Quantum
SERV
+3%
Delivery bots
ARM
-3%
Lone red · chips

This issue's moves

Rode: RKLB's 16% Iridium pop and PL's 16% sympathy move. Got bit: MU −5.5% in the chip rout, dragging the book off its high — the un-trimmed concentration, live. Standing order, now overdue: right-size MU first, then reassess RKLB after the Iridium deal’s financing and dilution are fully understood. Rebuild cash deliberately instead of trimming two different theses by reflex. The dovish jobs tape is a calm backdrop to finally do it.

+65%
Quarter 1 · on capital
$8,672
Book value
19/20
Positions green
~43%
In 2 themes ⚠️
~4×
vs S&P 500
The Catalyst Calendar

We position around dates

Fri Jul 3
Markets closed — Independence Day observed
Holiday
Mon Jul 6/7
SPCX joins the Nasdaq-100 — mechanical index buying kicks in
Tailwind
This week
Right-size MU; reassess RKLB after the Iridium financing details settle
My move
Mon Jul 14
Q2 earnings season opens with the banks — fundamentals refresh
High
Q3 2026
Honeywell Aerospace (HONA) spin-off — the GE-style unlock
On deck
Late Jul
SOFI Q2 earnings — the coiled spring gets re-tested
High
The Hot Take

"When a company you bought as a launch proxy starts making eight-billion-dollar acquisitions, the thesis didn't just work — it changed shape. Value the company in front of you, not the one you bought in April."

I bought Rocket Lab as a SpaceX proxy — a way to own the space catalyst without the IPO. It doubled. But Monday it stopped being a proxy. Acquiring Iridium for $8 billion turns RKLB into a vertically integrated space company: launch, manufacturing, and a profitable satellite-services business throwing off recurring revenue. That's a consolidation play in an industry about to have a lot of small, cash-strapped public companies for a well-capitalized acquirer to absorb.

The mistake would be anchoring to my old reason for owning it. The new question isn't "is RKLB a good SpaceX proxy" — it's "is RKLB the company that rolls up the public space economy." If yes, it's still cheap after doubling. If it overpays and integration drags, the thesis breaks. Either way I have to value the company in front of me. The same logic cuts the other way on Micron: I own it for a memory-scarcity thesis that's fully intact — but "the thesis is right" was never the same as "the position is the right size." This week proved it.

Buddy, Real Talk

This week the concentration stopped being theoretical

For four issues I've flagged that Micron is too big and promised to trim it. For four issues I haven't. And this week the market finally collected on that procrastination: a two-day chip rout knocked MU down 5.5% on Thursday alone, and because it's still ~21% of my book, that one name pulled my whole account off its all-time high. The risk I kept describing in the abstract showed up as an actual dollar amount. That's what unmanaged concentration does — it's invisible right up until the day it isn't.

Here's the honest read: I got lucky that it bit on a −5.5% day and not a −25% one. The thesis is intact — Micron guided to $50 billion a quarter, the memory scarcity is real, and I'm not selling the position. But "don't sell it" and "don't let it be a fifth of the account" are two completely different statements, and I've been hiding behind the first to avoid doing the second. A soft jobs report just handed me a calm, dovish tape to trim into. No more "next issue." The first move is still MU, because that is the single-name concentration I already identified. RKLB deserves a fresh valuation after the Iridium deal before I treat a strong move as an automatic sell signal.

Four issues of narrating discipline. Time to be able to say I did it.

The Bottom Line

Quarter one: extraordinary, and a warning

Ninety-three days. $5,270 deployed, ~$8,672 on the board, +65% through the best market quarter since 2020 — and a week that delivered both the highlight (Rocket Lab's $8B space consolidation) and the warning (my oversized Micron dragging the book off its high in a chip rout).

Both halves of that sentence matter. The thesis is doing everything I hoped: space is maturing from halo to roll-up, memory is scarce and profitable, the frontier is real. But an extraordinary quarter is precisely when concentration is most dangerous, because a hot hand feels like a reason to stay lopsided. It isn't. The chip rotation this week was a small, cheap lesson — a −5.5% MU day, not a −25% one — and the smart response to a cheap lesson is to act on it before the expensive version arrives.

So quarter two starts with one job, and it isn't a new pick: right-size the known single-name risk, reassess the newly transformed Rocket Lab thesis, rebuild the powder, respect the rotation and the Iran headline risk, and stop confusing a great run with permission to skip the discipline. The frontier delivered. Now I protect it.

See you next Sunday, buddy. — Alex

§ Related reading
§ The frontier continues.

Get the next issue when it drops.

One sharp dispatch from the frontier: what changed, what matters, what I'm watching, what the evidence changed — and what the portfolio is doing about it.

No spam. No daily market sludge. Research notes only. Unsubscribe whenever.