Who I am as an investor
I’m Alex Oklesson, a self-directed retail investor running a concentrated frontier-technology portfolio with a strong growth and asymmetric-upside orientation.
I write The Frontier File, a weekly public record of a real account: the wins, the misses, the trims, the date mistakes, and the reasoning behind all of it. The goal is not to cosplay Wall Street. The goal is to build a sharper investing brain in public.
Working style: bold calls over mushy caveats, catalysts over vibes, and a permanent allergy to corporate-speak.
The three rules
1Seasoning, not the steak
Sub-$5 names and pre-revenue moonshots can live in the book, but they do not get to drive the bus. Size them small, expect some to disappoint, and never confuse a lottery ticket with dinner.
2Catalysts over vibes
I want dated catalysts: earnings, trial readouts, regulatory milestones, IPO windows, contract cycles, or sector events I can circle on a calendar.
3Cheap ≠ distressed
A beaten-down quality company is one thing. A company with going-concern smoke, debt problems, and cash running on fumes is a falling knife wearing a fake mustache.
How I build the book
Core
Profitable large-caps, diversified revenue, strong balance sheets. These can carry larger allocations.
FoundationGrowth
Strong revenue growth, volatility, and catalysts. Sized by conviction and near-term setup.
Main engineSpeculative
Binary outcomes, dilution risk, pre- or minimal revenue. Keep them small and honest.
High riskMoonshot
Frontier themes on a 3–5 year horizon. Research positions first; build only when the proof improves.
Seed layerDefault allocation language
- Concentration over decoration: fewer names with clearer reasons beats collecting tickers like Pokémon with earnings calls.
- 75/25 or 80/20 aggressive splits: growth and conviction get the weight; speculative names stay intentionally undersized.
- Trims are not betrayal: a position that grows too large gets trimmed because sizing is risk control, not a breakup text.
- Exclusions matter: every serious report should say what did not make the cut and why.
Portfolio universe
Latest fully reconciled published mark: the account started at $5,270 and stood at $8,904 on Day 104 (July 13, 2026) — a gain of $3,634, or +69.0%. These are historical published figures, not a live brokerage total: the last fully reconciled mark is July 13, 2026, and the position list below was published with Issue #12 on August 1, 2026. Nothing here updates in real time.
Published position list · 21 positions as of August 1, 2026
Core · 7
MSFT 0.5556 · GOOGL 2 · NVDA 3 · MU 0.79 · STX 1 · BWXT 1 · AMZN 1
FoundationGrowth · 7
ACHR 110 · JOBY 30 · RKLB 5 · PL 12 · MDA 9 · SOFI 10 · ASTS 2
Main engineSpeculative · 6
SPCX 1 · CENX 5 · OKLO 2 · SERV 4.5 · AEVA 4 · INTC ≈0.965
High riskMoonshot · 1
QBTS 16
Seed layerRecent moves
- Microsoft added with $250.00 of cash at the July 31, 2026 market open at a $450.00 opening price, producing 0.5556 shares (about $258.18 by Friday's close).
- Intel added, now roughly 0.965 shares.
- ARM fully exited.
- Serve Robotics reduced from 9 shares to 4.5 shares.
- Micron reduced from 1.58 shares to 0.79 shares.
- SPCX reduced to 1 share.
Current cash: not fully reconciled.
Watchlist universe by theme — tracked names, not all currently owned
| Theme | Names | Lens |
|---|---|---|
| AI / Semis | NVDA, MU, MSFT, AMZN, GOOGL, ARM, AMD, PLTR | The infrastructure backbone. |
| Space economy | RKLB, ASTS | Public proxies around the SpaceX halo. |
| Quantum | IONQ, RGTI, QBTS | Real theme, dangerous sizing. |
| eVTOL / Air mobility | ACHR, JOBY | FAA milestones, pre-revenue patience. |
| Fintech rails | SOFI, HOOD, COIN, OWL | Transaction and access infrastructure. |
| Nuclear AI power | OKLO, BWXT, VST, NNE | The energy layer under compute demand. |
| Physical AI | SERV, SYM, RXRX | Robots leaving the demo stage. |
| Speculative biotech | CMPS, MNMD/DFTX, ATAI | Seasoning only. Binary means binary. |
Mega-themes I’m tracking
🟢 Cash-Flow Infrastructure
MSFT · GOOGL · AMZN · NVDA · BWXT — companies that can finance their own growth and hold hard-to-replicate infrastructure positions. The strongest direction in the book.
Firing now🟢 Memory & Storage
MU · STX — the bottleneck thesis is intact, and both positions are right-sized after the Micron trim.
Right-sized🟡 Space Stack
SPCX · RKLB · PL · MDA · ASTS — this week separates operating proof from thematic excitement. Every company gets graded separately; no basket-level decision.
Catalyst-heavy🟡 Electric Aviation
ACHR · JOBY — Archer needs a concentration rule; Joby is the steadier counterweight. Certification still matters more than the tape.
Patience trade🔴 Quantum & Physical AI
QBTS · SERV · AEVA — all three face near-term proof tests. Small positions remain the correct positions until revenue and unit economics catch up.
Binary risk🔵 Nuclear
BWXT · OKLO — BWXT is the operating nuclear position with real cash flow; Oklo is the small asymmetric seasoning. Keep the dream small.
Operating + optionality🔵 Industrial & Materials
CENX — exposure to the physical inputs underneath the build-out.
Supporting layer🟡 Turnaround
INTC — a foundry proof story sized as speculation, not conviction.
Show meThe calendar beats the campfire story
This strategy only works if dates are treated like load-bearing beams. A catalyst that is “soon-ish” is not a catalyst. It is a vibe in a trench coat.
| Date | Event | What matters | Current action / risk |
|---|---|---|---|
| Mon · Aug 3 | BWXT earnings, after the close | Guidance, backlog, margins, cash conversion, and Precision Components integration. | Most actionable report of the week. Possible ~$100 add only if the numbers hold. |
| Tue · Aug 4 | SpaceX earnings after the close · JOLTS at 10:00 a.m. ET | Starlink economics, launch cadence, Starship spending, cash generation. | Hold 1 share. No add into the print. |
| Wed · Aug 5 | Aeva earnings, after the close | Commercial nominations, production timing, industrial demand, liquidity. | No add. High binary risk. |
| Thu · Aug 6 | D-Wave before the open · Serve Robotics after the close · productivity data at 8:30 a.m. ET | QBTS: bookings conversion and revenue. SERV: utilization, revenue per robot, cash burn, Diligent integration. | Hold small on both. Double binary day. |
| Fri · Aug 7 | MDA Space before the open · July U.S. employment report at 8:30 a.m. ET | Backlog, satellite margins, robotics demand, financing. The jobs print can move the discount rate under everything. | No action. Macro watch. |
| Mon · Aug 10 | Archer Aviation and Rocket Lab earnings, after the close | ACHR: certification, production, cash. RKLB: Neutron, margins, backlog, Iridium financing. | Possible ACHR trim on an unsupported spike. |
Hot take I stand by: the bubble bears are not automatically wrong about valuation. I think many are early on timing. I get cautious when orders start getting cancelled, not while the capex firehose is still open.
Instructions for a new research assistant
If you are reading this in a fresh chat: I’m Alex, an aggressive frontier-tech retail investor. Default to bold, catalyst-driven, tier-aware analysis in a casual voice. Honor the three rules: seasoning not steak, catalysts over vibes, cheap does not mean buyable.
- Verify live prices, market caps, earnings dates, analyst targets, macro data, and filing status before using them.
- Use Core / Growth / Speculative / Moonshot tiers when modeling allocations.
- Explain exclusions, not just picks.
- For daily or weekly reports, use styled HTML. For monthly deep dives, use formatted Word documents.
- Give the call. Do not hide behind a buffet of caveats.