September 2026 · Interim Note
The Book Fell $816.28 Today. SanDisk, Micron and AMD Did Most of the Damage.
The book closed at $33,318.75 today, down $816.28 from yesterday’s $34,135.02 — the hardest single-day loss I’ve logged since I started tracking daily book impact. Twenty-seven of my 32 positions fell. SanDisk did the most damage at −$143.16, followed by Micron (−$100.72), AMD (−$87.48), my leveraged Nebius ETF NBIL (−$87.30) and Intel (−$82.88). None of it was company-specific: a hot wholesale-inflation print and a spike in oil, covered in full in today’s News piece, pushed Fed rate-hike odds to 70% and sent almost every rate-sensitive name I own down together. Only five positions closed green, led by Palo Alto Networks at a modest +$16.95. No trades.
Where the $816.28 actually came from
Figure 1 · Cumulative dollar book impact, September 10
Nine names and one "everything else" bucket walk the book down to a net loss of $816.28
The nine largest single-day dollar movers in the 32-position book, September 9 to September 10, 2026 close, plus the remaining 23 positions combined and the net across all 32. Share counts × the change in each position's closing price, using my own share counts and cost basis. Source: Massive Market Data.
This is a different shape of day than the one two sessions ago, when Meta's Muse launch and Vertiv's unexplained drop offset each other almost to the dollar. Today there was no offsetting story — eight of my nine largest movers fell together on the same macro trigger, and the ninth (Palo Alto Networks) barely moved. Today’s Opinion piece goes further into what it means that this many nominally different positions — a memory chipmaker, a leveraged single-stock ETF, an AI cloud provider, a data-center cooling company — all moved on one interest-rate number.
| Position | Close | Day | Book impact | vs. my cost |
|---|---|---|---|---|
| SanDisk (SNDK) | $1,692.59 | −4.06% | −$143.16 | +250.28% |
| Micron (MU) | $977.41 | −4.90% | −$100.72 | +199.98% |
| AMD | $503.60 | −3.36% | −$87.48 | +152.63% |
| GraniteShares 2x Nebius (NBIL) | $25.26 | −10.33% | −$87.30 | +173.67% |
| Intel (INTC) | $100.32 | −5.57% | −$82.88 | +168.02% |
| Vertiv (VRT) | $248.13 | −5.61% | −$44.28 | −18.62% |
| Meta Platforms (META) | $644.38 | −1.42% | −$43.11 | +175.00% |
| Estée Lauder (EL) | $96.47 | −2.33% | −$41.40 | +39.57% |
| CoreWeave (CRWV) | $89.12 | −6.13% | −$29.10 | +7.45% |
| Talen Energy (TLN) | $311.54 | −2.54% | −$24.39 | −0.96% |
| Nebius (NBIS) | $228.11 | −5.09% | −$24.48 | +153.12% |
| Arm Holdings (ARM) | $254.18 | −3.80% | −$20.10 | +22.30% |
| Cheesecake Factory (CAKE) | $101.16 | −2.36% | −$19.60 | +122.82% |
| Palo Alto Networks (PANW) | $338.49 | +1.01% | +$16.95 | +111.13% |
The position that took the biggest hit
SanDisk remains my best-performing position by far — up 250.28% against my $483.21 average cost on 2 shares even after today's drop — so a 4.06% pullback here reads as arithmetic on a name that has roughly quadrupled, not a thesis problem. The same is true of Micron (+199.98% vs. cost) and AMD (+152.63%). The more useful question isn't whether any single one of these names is still working; it's how much of my book is now moving as one correlated bet on Fed policy rather than eight or nine independent theses. The table below sizes that concentration directly, using today's own real, dated instrument moves as two of the scenario rows rather than guessed numbers.
| If my 8 rate-sensitive positions (SNDK, MU, AMD, NBIL, INTC, VRT, CRWV, NBIS — $11,666.49 combined) move… | Combined value | Total book value | vs. today |
|---|---|---|---|
| Unchanged | $11,666.49 | $33,318.75 | — |
| Another SMH-sized move (−2.44%, today's semiconductor-index decline) | $11,381.83 | $33,034.09 | −$284.66 |
| Another SLV-sized move (−5.30%, today's silver decline) | $11,048.17 | $32,700.43 | −$618.32 |
| Illustrative stress case (−10%) | $10,499.84 | $32,152.10 | −$1,166.65 |
Assumptions, stated plainly. The $11,666.49 combined value is my own real position value for these eight names at today's close, computed from my own share counts. The first two move-size scenarios reuse today's own real, dated instrument changes — the SMH semiconductor-index decline and the SLV silver decline, both from Massive Market Data — as illustrative magnitudes, not as a claim that either instrument predicts these stocks specifically. The 10% stress case is not sourced to any analyst estimate; it exists purely to show the dollar sensitivity at a round number. None of these are price forecasts, and Friday's CPI print, flagged in today's News piece as the next data point ahead of the Fed's meeting, could move this basket in either direction.
No trades today
The Blackout Rule and Holding Period exist for exactly this kind of session: a broad, macro-driven decline across positions I've already underwritten is not new information about any individual business until I actually have some. Today's move was explained by one dated macro story affecting the whole rate-sensitive cohort at once, not by anything specific to SanDisk, Micron, AMD or any other name in it. I made no trades.