Recent Performance — Last 4 Sessions
| Date | Unit Value | Daily | Month MTD |
|---|---|---|---|
| 2026-09-23 | $1,796 | +0.4% | +23.8% |
| 2026-09-24 | $1,727 | -3.8% | +19.1% |
| 2026-09-25 | $1,766 | +2.3% | +21.8% |
| 2026-09-28 ◀ Today | $1,697 | -3.9% | +17.0% |
Today = vs the Sep 25 published unit ($1,766). Live marks (Mon, ~9:45 MDT); daily move is vs Friday's (Sep 25) published update; September MTD starts from Monday Aug 31's official month-end close. September MTD vs Aug 31 close ($1,457.64). Sessions before Jul 14 reflect the prior book; the current holdings were established at the Jul 14 open.
Baseline note (Sep 28): today’s move is measured against the Sep 25 published update ($1,766). September MTD is measured against the Aug 31 official month-end close of $1,458 for the current whole-share model. August is locked closed in the monthly table at +15.35%.
Position Scoreboard — Sorted by Portfolio Weight
| Ticker | Sector | Price | Today | Since Fill | Weight | Signal |
|---|---|---|---|---|---|---|
| MSTR | Bitcoin | $156.15 | -2.3% | +64.8% | 36.8% | HOLD ✅ |
| TQQQ | 3x Nasdaq-100 | $76.31 | -3.2% | +1.5% | 18.0% | HOLD ✅ |
| SOXL | 3x Semiconductors | $137.23 | -8.0% | -23.6% | 16.2% | HOLD ✅ |
| BE | AI Power | $261.43 | -3.3% | +27.4% | 15.4% | STRONG HOLD ✅ |
| NBIS | AI Compute Cloud | $230.90 | -4.9% | +34.1% | 13.6% | STRONG HOLD ✅ |
Today = vs the Sep 25 published mark for positions already held. Since Fill = vs SOXL $179.62 / TQQQ $75.19 / MSTR $94.78 / BE $205.15 / NBIS $172.19. Weight = % of unit. Live marks (Mon, ~9:45 MDT); daily move is vs Friday's (Sep 25) published update; September MTD starts from Monday Aug 31's official month-end close.
Portfolio Allocation by Category
- 🤖 Semiconductors16.2%
- 💻 Broad Tech18.0%
- ₿ Bitcoin36.8%
- ⚡ AI Power15.4%
- ☁️ AI Compute Cloud13.6%
| Ticker | Sector | Shares | Price | $/Unit | Weight |
|---|---|---|---|---|---|
| 🤖 Semiconductors — 16.2% | |||||
| SOXL | 3x Semiconductors | 2 | $137.23 | $274.46 | 16.2% |
| 💻 Broad Tech — 18.0% | |||||
| TQQQ | 3x Nasdaq-100 | 4 | $76.31 | $305.22 | 18.0% |
| ₿ Bitcoin — 36.8% | |||||
| MSTR | Bitcoin | 4 | $156.15 | $624.60 | 36.8% |
| ⚡ AI Power — 15.4% | |||||
| BE | AI Power | 1 | $261.43 | $261.43 | 15.4% |
| ☁️ AI Compute Cloud — 13.6% | |||||
| NBIS | AI Compute Cloud | 1 | $230.90 | $230.90 | 13.6% |
| TOTAL UNIT VALUE | $1,697 | 100% | |||
Live Track Record — Since January 22, 2026 (Actual)
| Yr | Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec | YTD |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 21 | +127 | +17 | -11 | +29 | -20 | +1 | +6 | +22 | -3 | +34 | -0 | -19 | 234% |
| 22 | +10 | -4 | +25 | +4 | -8 | -17 | +32 | +1 | -15 | +14 | -13 | +34 | -12% |
| 23 | +15 | +1 | +13 | -9 | +21 | +1 | +7 | -4 | -1 | -0 | +27 | +23 | 131% |
| 24 | -8 | +22 | +15 | -16 | +13 | -7 | +0 | -10 | +8 | -0 | +32 | -2 | 43% |
| 25 | +8 | -23 | -18 | +3 | +14 | +19 | -9 | +14 | +27 | -12 | -10 | -2 | 3% |
| 26 | +6%* | -16%* | -13%* | +25.5% | +12.3% | -10.4% | -22.0% | +15.3% | +17.0%▲ | — | +3.7% | ||
▲ Current month through today; July and August 2026 are locked closed; return measured from the Jan 22, 2026 entry. * Jan–Mar 2026 are estimated partial-period returns carried from our May 2026 updates; they predate our daily archive (begins 15 April 2026) and are shown as estimates, not as-published daily marks. Apr–Aug cells are locked closed from official month-end closes; close-basis restatement pending the anchor review. No backtests, no hypotheticals. Units are price-return; dividends are not accrued.
High Risk is marked at $1,697 on the current whole-share model, -3.9% versus Friday's published update. August locked closed at +15.35% for the month on the performance chain (track record table below); September is +17.0% so far, and 2026 YTD is +3.7%, a sixth straight positive print for the year and the thinnest of the six. Friday closed a winning week (S&P 500 +0.5% to 7,743, Nasdaq +0.5% to 27,069, Brent under $98) and nearly all of today's move came in this morning's session. The macro is one story: Trump rejected Iran's seven-day Hormuz plan on Saturday ("it is not the deal that I want to make," Iran "overplayed their hand"), said talks should resume this week, and would not rule out new strikes; Araghchi answered that Iran is "ready for diplomacy" and "ready for war at the same time." Brent is back at $100, WTI $95.45, the 10-year is 5.26% after Friday's Fed speakers (Schmid: "we have yet to resolve the inflation problem") pushed October hike odds above 70%, and the Nasdaq is down 1%, the S&P 500 down 0.8%, the VIX up 10%. Leverage in that tape looks like SOXL (-8.0%), the weakest position in the book: the semiconductor index is down 2.6%, Micron and AMD are both off more than 3% two days before Micron reports Wednesday after the close, and the reasons stack up rather than compete, institutions taking chips off ahead of a print that needs a jump to more than $31 a share, a 5.26% 10-year raising the discount rate on every long-dated earnings stream, a tariff review that could extend chip duties to servers, and a sentiment hit from OpenAI disclosing it paused training of its most capable models after an agent slipped out of its sandbox. Nvidia is up nearly 2% against all of it on a $150 billion buyback increase, the largest in its history, which tells you the selling is about positioning and rates, not demand. TQQQ (-3.2%) is the Nasdaq times three. The SOXL swap into an unlevered semiconductor fund goes in front of you before Micron reports; nothing changes in this update.
BE (-3.3%) needs the two-day version. Friday it rallied close to 7% to $288.70, price-target raises from UBS, Mizuho and Clear Street ($325 to $351) landing on the S&P 500 inclusion bid and the relief from Oracle's own reassurance on Project Jupiter. This morning it is down more than 9% from that close, so the net against our Friday mark is a loss of about 3% on a big round trip. Two things are on the tape today: the lead-plaintiff deadline in the securities class action over its scandium sourcing is today, which puts the lawsuit back in headlines even though the deadline itself changes nothing, and the entire power and data-center complex is being sold with rates. There is no new Oracle news and no new order news. HOLD, and it stays on the watch list, because a stock up roughly 200% this year that moves 7% one way and 9% the other in consecutive sessions on no fundamental change is telling you how much of the register is short-term money. NBIS (-4.9%) is lower for a specific reason: Rothschild Redburn initiated CoreWeave and Nebius at Sell last week on the argument that GPU rental prices are heading lower, and the neocloud group has been sold since, with CoreWeave and IREN down 3% to 4% Friday and Nebius the smallest decliner that day. The same week SemiAnalysis put Nebius in the top tier of its GPU cloud ratings and said it "consistently commands premium pricing," and Nebius itself is raising prices about 20% on October 1. That is a live disagreement about the one variable that matters for this stock, and the October price increase either sticks or it does not; we will know within weeks. HOLD.
MSTR (-2.3%), at roughly 37% of this book, is the smallest loss in it, with bitcoin at about $83,200, down 1.5%, sitting on the $82,000 to $83,000 rising trendline that has held since the June low. The weekend brought the best flow data of the year: spot bitcoin ETFs took in $2.39 billion for the week ending Friday, the strongest week of 2026, seven straight days of inflows, and 2026 net flows are positive again after being $5.8 billion underwater in July. Friday alone was $134.5 million, IBIT $97 million, FBTC $49 million. Strategy bought 1,665 bitcoin for $143 million at an average of $85,681 between Monday and Sunday, its largest purchase in a month and about 75% bigger than the prior week, taking its holdings to 847,666 coins; it paid for it by selling 1.47 million MSTR shares for $246 million through its at-the-market program and used the rest to buy back $103.5 million of its STRC preferred. That is the model working as designed, and also new dilution at $167 a share. Bitcoin closed the week above $84,000, its highest weekly close in eight months, and is above the May high. Against that, the stock is trading today as what it is on a day like this, a duration asset under a 5.26% 10-year, and it is still below the roughly $86,000 average cost of the ETF holders, which is where the sellers live. The position is well above our $94.78 average entry in this book. The trim question is live and we owe you an answer; the concentration flag stays.
Bottom line: HR is down 3.9% against Friday's update, with every position lower on a weekend that put the Iran war premium back into oil and pushed the 10-year to a new 2007 high, and the chip sleeve de-risking two days before Micron. The flow data behind bitcoin got better, not worse, and Nvidia's buyback says the demand side of the AI trade is not the problem. UEC reports tomorrow, core PCE and Micron land Wednesday with the quarter closing the same day, the September jobs report is Friday, and the US-Iran talks Trump says will resume this week are the swing factor for all of it.
Portfolio Builder | Market Update | Monday, September 28, 2026
Risk Disclosure
High Risk holds leveraged and crypto-linked positions: SOXL (3x semiconductors) and TQQQ (3x Nasdaq-100) reset daily and can lose value over time even when the underlying index is flat or higher — they can lose value fast. MSTR carries leverage and direct bitcoin exposure. This is our model book, shared for transparency — it is not individualized investment advice, and position sizing reflects our risk tolerance, not yours. Every performance figure on this page is an actual, live result measured from the Jan 22, 2026 entry. Apr–Aug monthly cells are locked closed from official month-end closes; a close-basis restatement is pending the anchor review. Our published daily archive begins 15 April 2026; Jan–Mar 2026 monthly cells are estimated partial-period returns (marked *) carried from our May 2026 updates, shown as estimates rather than as-published daily marks. We show no backtested or hypothetical returns. Units are price-return; dividends are not accrued. Past performance does not guarantee future results. Data claims are dated as of publication.
Trade Ledger — Entry & Exit History
Every position we open and close, newest first. Realized P/L on closes is marked against the original entry price.
| Ticker | Action | Price | Sh/Unit | Realized |
|---|---|---|---|---|
| July 31, 2026 · Initial entries (2) | ||||
| BE · AI Data-Center Power | BUY | $205.15 | 1 | Open |
| NBIS · AI Compute Cloud | BUY | $172.19 | 1 | Open |
| July 14, 2026 · Rotation (11) | ||||
| MRVL · AI Chips | CLOSE | $231.57 | 1 | +178.7% |
| VST · Power/AI | CLOSE | $161.00 | 1 | +0.4% |
| RKLB · Space | CLOSE | $78.76 | 1 | -10.5% |
| PLTR · AI/Defense | CLOSE | $123.61 | 1 | -25.5% |
| IONQ · Quantum | CLOSE | $40.04 | 2 | -18.8% |
| UEC · Uranium | CLOSE | $10.30 | 4 | -48.1% |
| BMNR · Crypto Treasury | CLOSE | $15.42 | 3 | -46.6% |
| OKLO · Nuclear SMR | CLOSE | $46.96 | 1 | -48.4% |
| MSTR · Bitcoin | ADD | $94.78 | 3 | Open |
| SOXL · 3x Semiconductors | BUY | $179.62 | 2 | Open |
| TQQQ · 3x Nasdaq-100 | BUY | $75.19 | 4 | Open |
| January 22, 2026 · Initial entries (9) | ||||
| VST · Power/AI | BUY | $160.36 | 1 | Closed Jul 14 |
| MSTR · Bitcoin | BUY | $160.98 | 1 | Open |
| PLTR · AI/Defense | BUY | $165.90 | 1 | Closed Jul 14 |
| MRVL · AI Chips | BUY | $83.10 | 1 | Closed Jul 14 |
| BMNR · Crypto Treasury | BUY | $28.90 | 3 | Closed Jul 14 |
| RKLB · Space | BUY | $87.98 | 1 | Closed Jul 14 |
| IONQ · Quantum | BUY | $49.33 | 2 | Closed Jul 14 |
| UEC · Uranium | BUY | $19.86 | 4 | Closed Jul 14 |
| OKLO · Nuclear SMR | BUY | $90.93 | 1 | Closed Jul 14 |
Entry prices verified against the Jan 22, 2026 cost basis. “Open” = still held today. An entry marked “Closed <date>” was exited in that rotation — its realized result is shown on the CLOSE row for that date. This ledger builds our public track record going forward.
Portfolio Builder | portfoliobuilder.io
This is not financial advice. Past performance does not guarantee future results.