July 2026 · Positioning
Portfolio Composition & Ongoing Risks
I want to open this journal the way a fund manager should — not with a victory lap, but with an honest map of what I hold and what could go wrong. Anyone can post their winners. The useful exercise is laying the whole book out on the table, naming every position by dollar amount, and naming its risks.
I started with $500 of saved earnings and gift money. The portfolio today is 32 positions across large, durable businesses, some smaller growth names, a small speculative sleeve, leveraged exposure, and index hedges, worth $34,822.51 as of this letter — up $14,593.83 (+72.14%) all-time, down $798.81 (−2.24%) today. I used to withhold dollar balances; I've dropped that policy, because a percentage without a dollar amount hides how small this still is. This is a $500 account that grew through saving and compounding, not a fund.
Portfolio composition
| Ticker | Position | Bucket | Avg. price | Return | % of book |
|---|---|---|---|---|---|
| SNDK | Sandisk Corp | Core | $483.21 | +261.01% | 10.02% |
| META | Meta Platforms | Core | $234.32 | +156.18% | 7.98% |
| AMD | Advanced Micro Devices | Core | $199.34 | +176.94% | 7.93% |
| SPY | SPDR S&P 500 ETF | Hedge | $655.24 | +14.66% | 6.47% |
| VOO | Vanguard S&P 500 ETF | Hedge | $580.93 | +18.88% | 5.95% |
| MU | Micron Technology | Core | $325.83 | +202.23% | 5.66% |
| PANW | Palo Alto Networks | Core | $160.32 | +123.01% | 5.13% |
| INTC | Intel Corporation | Core | $37.43 | +226.48% | 4.91% |
| EL | Estée Lauder Companies | Core | $69.12 | +22.83% | 4.39% |
| QQQ | Invesco QQQ Trust | Hedge | $558.28 | +29.47% | 4.15% |
| SPCX | Space Exploration Technologies (SpaceX) | Spec | $135.00 | +18.83% | 3.69% |
| TLN | Talen Energy Corporation | Growth | $314.57 | +20.10% | 3.25% |
| GOOGL | Alphabet Inc | Core | $274.57 | +33.47% | 3.16% |
| NBIL | GraniteShares 2x Long NBIS Daily ETF | Leveraged | $9.23 | +249.62% | 2.78% |
| VRT | Vertiv Holdings Co | Core | $304.90 | +4.45% | 2.74% |
| DDOG | Datadog Inc | Core | $119.27 | +114.11% | 2.20% |
| NOW | ServiceNow Inc | Core | $84.76 | +27.34% | 1.86% |
| ARM | Arm Holdings plc | Growth | $207.83 | +55.05% | 1.85% |
| CAKE | The Cheesecake Factory | Growth | $45.40 | +69.36% | 1.77% |
| AVGO | Broadcom Inc | Core | $293.47 | +27.41% | 1.44% |
| WYFI | WhiteFiber Inc | Spec | $13.33 | +166.62% | 1.43% |
| FTNT | Fortinet Inc | Core | $84.30 | +92.59% | 1.40% |
| CRWV | CoreWeave Inc | Growth | $82.94 | +4.24% | 1.24% |
| NBIS | Nebius Group N.V. | Growth | $90.12 | +136.37% | 1.22% |
| AAL | American Airlines Group | Growth | $10.18 | +74.36% | 1.17% |
| CRWD | CrowdStrike Holdings | Core | $94.76 | +110.41% | 1.15% |
| SCHD | Schwab U.S. Dividend Equity ETF | Hedge | $32.24 | +0.00% | 1.11% |
| VST | Vistra Corp | Growth | $134.71 | +16.71% | 0.90% |
| SYM | Symbotic Inc | Growth | $38.55 | +15.33% | 0.89% |
| RKLB | Rocket Lab USA | Spec | $78.59 | +18.45% | 0.80% |
| TEM | Tempus AI Inc | Spec | $49.73 | +22.06% | 0.70% |
| NFLX | Netflix Inc | Core | $79.23 | −4.05% | 0.65% |
By bucket
Sorted into the buckets from my Investment Policy Statement:
- 60.6% Core / quality compounders
- 17.7% Index hedge
- 12.3% Emerging growth
- 6.6% Speculative sleeve
- 2.8% Leveraged / derivative
Core is the anchor, and where any edge should show up — 15 positions now, up from 7, as the book has diversified across cybersecurity (PANW, FTNT, CRWD), cloud software (NOW, DDOG), and semiconductors (AVGO) alongside the original names. Index hedge is always held, ballast, now split across four funds instead of two. Emerging growth picked up a power/AI-infrastructure tilt (TLN, VST) alongside the existing names. Speculative sleeve is 6.6%, comfortably under its 25% cap, and now includes a small, deliberately limited position in SPCX (SpaceX) — more on why it's sized small below. Leveraged / derivative is 2.8%, under its 5% cap.
Performance vs. benchmark
My IPS benchmark is SPY, measured time-weighted with deposits stripped out — I don't have that number yet, since it requires daily portfolio-value history I haven't been tracking. What I do have, honestly labeled as what it is:
| Measure | AEA | SPY |
|---|---|---|
| All-time return (real, dollar-weighted, since the $500 start) | +72.14% | not yet comparable — no matched start date tracked |
| Illustrative 32-holding blend vs. SPY, indexed to 100 over the same window (home page chart) | +465% | +19.0% |
| Index-hedge positions held directly, real return at cost | SPY +14.66% · VOO +18.88% | — |
Ongoing risks
Three honest risks I'm tracking. One is a live rule break, though a much smaller one than last month; the other two are structural, not violations, but worth saying plainly.
- SNDK is right at my single-position cap. Policy caps any single position at 10% at cost. SNDK is 10.02% of the book — it's still my best performer by far, up 261% on cost, and growth in the position outran my rebalancing. This is a much smaller breach than the 18.51% I reported last letter, but it's still technically over the line, and my own sell discipline says drift beyond a cap means trim back to policy rather than make an exception because the position is working. That trim is still owed, just a smaller one now.
- SPCX is a small, deliberate exception to how I normally size positions. SpaceX only began trading publicly on June 12, 2026 — barely three weeks of public price history as of this letter — and I sized the position at 3.69% of the book specifically because a company with that little trading history, however well-known the underlying business, belongs in the speculative sleeve rather than core. I'm comfortable owning it small; I'm not comfortable owning it large until the stock has a real track record.
- Thematic clustering in AI infrastructure, semiconductors, and power. SNDK, AMD, MU, INTC, AVGO, PANW, FTNT, CRWD, DDOG, and NOW touch semiconductors, memory, or enterprise software tied to AI capex; NBIS, CRWV, WYFI, NBIL, TLN, and VST are AI data-center, compute, or power-generation infrastructure; META is one of the largest AI capex spenders in the market. Together that's the majority of the book leaning on one macro bet: that AI-driven compute and power demand keep growing. That's a deliberate view, but if the theme cools, most of these fall together — the NBIS and CRWV selloffs this month were a live example of exactly that correlation.
Portfolio Review & Retrospective: Post-Mortem
This section exists because a track record only means something if it includes the parts that didn't go as planned. Anyone can publish a win. Publishing a plan I didn't follow through on — and saying so plainly — is the actual test of whether this process is disciplined or just lucky.
- The Original Thesis
- Last month's letter flagged NBIL, my 2x leveraged ETF on Nebius, at 5.27% of the book — over its 5% leveraged-sleeve cap. I said I owed a resize and intended to trim it back under policy myself.
- The Reality
- I didn't execute the trim. NBIL is now 2.78% of the book, comfortably under cap — but only because the position grew slower than the rest of the portfolio over the month. The market resized it, not my discipline.
- The Lesson
- Writing that I "owe" a trim in a letter is not the same as scheduling and executing one. The actual process flaw is the missing concrete trigger: stating an intention without a specific action date lets the market's own volatility quietly resolve — or fail to resolve — a policy breach I should be resolving myself. Going forward, a cap breach gets a stated action date, not just an acknowledgment.
How I'll keep myself honest
Starting next month, every letter will report my time-weighted return vs. SPY (deposits stripped out), the speculative and leveraged sleeve weights against their caps, and at least one thing I got wrong — starting with trimming SNDK back toward its 10% cap. If the index wins over time, I'll say so and act on it.
The point of AEA isn't to prove that investing is already figured out. It's to build discipline in public, on the record, where mistakes can't be quietly edited out.
Own the risk before the market makes you.