How the shadows are built: each week the 20 shadow books make exactly as many picks as I did, drawn at random from the same universe by a fixed seed — the universe being mechanical, not curated: the top 1,000 US common stocks by market cap (no ETFs/funds/ADRs; price ≥ $5; ≥ $10M/day traded), re-ranked point-in-time from vendor data at each run, names entering and leaving only by rank, my out-of-universe picks rejected and published — and run through identical rules, capital, and costs — the arms differ only in which names. What 20 can show: a rank among 20 is ordinal evidence, not a p-value; a week means nothing, a year means something. A checksum proves this record hasn't changed since it was written — not that it was right. The record ends only by a dated run-log entry, never by silence. Receipts (run log, picks with recorded-at timestamps, NAV per book) export as CSV every run: runs · picks · nav — and every run commits them to github.com/blu400codes/trenyx-receipts, whose commit history is a timestamp I don't control.
Scoreboard — all 21 books
| # | book | NAV | return |
|---|---|---|---|
| 1 | SK | $1,037,897 | +3.79% |
| 2 | monkey #5 | $1,032,671 | +3.27% |
| 3 | monkey #18 | $1,017,477 | +1.75% |
| 4 | monkey #20 | $1,009,746 | +0.97% |
| 5 | monkey #2 | $1,006,046 | +0.60% |
| 6 | monkey #10 | $1,005,712 | +0.57% |
| 7 | monkey #11 | $1,001,079 | +0.11% |
| 8 | monkey #17 | $1,000,364 | +0.04% |
| 9 | monkey #9 | $1,000,292 | +0.03% |
| 10 | monkey #12 | $999,305 | -0.07% |
| 11 | monkey #1 | $999,290 | -0.07% |
| 12 | monkey #14 | $999,211 | -0.08% |
| 13 | monkey #19 | $998,123 | -0.19% |
| 14 | monkey #4 | $998,025 | -0.20% |
| 15 | monkey #16 | $996,506 | -0.35% |
| 16 | monkey #6 | $996,129 | -0.39% |
| 17 | monkey #8 | $995,805 | -0.42% |
| 18 | monkey #15 | $994,195 | -0.58% |
| 19 | monkey #13 | $981,331 | -1.87% |
| 20 | monkey #7 | $979,806 | -2.02% |
| 21 | monkey #3 | $972,082 | -2.79% |
Open positions (mine)
| name | opened | return | rules |
|---|---|---|---|
| HUT | 2026-07-21 | -10.95% | probe — pre-registered hard stop armed |
| BE | 2026-07-22 | +16.17% | probe — pre-registered hard stop armed |
| OKLO | 2026-07-24 | -5.37% | probe — pre-registered hard stop armed |
| WULF | 2026-07-27 | -11.66% | probe — pre-registered hard stop armed |
| RDDT | 2026-07-28 | -13.87% | probe — pre-registered hard stop armed |
| PLTR | 2026-08-10 | +1.85% | full position — graduated to core, trailing stop armed |
| RIOT | 2026-07-30 | +8.78% | probe — pre-registered hard stop armed |
| DELL | 2026-07-31 | +25.03% | probe — pre-registered hard stop armed |
| RKLB | 2026-08-04 | -11.97% | probe — pre-registered hard stop armed |
| ASTS | 2026-08-05 | -8.22% | probe — pre-registered hard stop armed |
| CRCL | 2026-08-13 | +45.48% | probe — pre-registered hard stop armed |
| MSTR | 2026-08-24 | +15.10% | full position — graduated to core, trailing stop armed |
| IONQ | 2026-08-18 | -10.96% | probe — pre-registered hard stop armed |
| KTOS | 2026-08-24 | -14.89% | probe — pre-registered hard stop armed |
Every decision the machine made (mine)
| date | side | name | why |
|---|---|---|---|
| 2026-08-31 | BUY | MSTR | add |
| 2026-08-24 | BUY | MSTR | graduation — promoted to a core position |
| 2026-08-24 | SELL | MSTR | graduation — promoted to a core position |
| 2026-08-24 | BUY | KTOS | probe entry — starter position |
| 2026-08-18 | BUY | IONQ | probe entry — starter position |
| 2026-08-14 | BUY | MSTR | probe entry — starter position |
| 2026-08-13 | BUY | CRCL | probe entry — starter position |
| 2026-08-11 | SELL | CIFR | hard stop — pre-registered floor hit |
| 2026-08-10 | BUY | PLTR | graduation — promoted to a core position |
| 2026-08-10 | SELL | PLTR | graduation — promoted to a core position |
| 2026-08-05 | BUY | ASTS | probe entry — starter position |
| 2026-08-04 | BUY | RKLB | probe entry — starter position |
| 2026-07-31 | BUY | DELL | probe entry — starter position |
| 2026-07-30 | BUY | RIOT | probe entry — starter position |
| 2026-07-29 | BUY | PLTR | probe entry — starter position |
| 2026-07-28 | BUY | RDDT | probe entry — starter position |
| 2026-07-27 | BUY | WULF | probe entry — starter position |
| 2026-07-24 | BUY | OKLO | probe entry — starter position |
| 2026-07-23 | BUY | CIFR | probe entry — starter position |
| 2026-07-22 | BUY | BE | probe entry — starter position |
| 2026-07-21 | BUY | HUT | probe entry — starter position |
My picks, with the reason I wrote at the time
Reasons are recorded before outcomes and can't be edited after. The embarrassing ones stay.
| picked | name | reason (verbatim) |
|---|---|---|
| 2026-08-21 | KTOS | Kratos — hypersonics doubling ($200M->$400M->$700M), 3,000-engine order, $15B pipeline, Q2 beat + raised guide; defense/autonomy = first pick outside the AI-capex/crypto cluster; post-earnings pullback entry (operator; conviction 2) |
| 2026-08-20 | PURR | should have followed my gut and taken a small position — HYPE treasury / Hyperliquid thesis (operator; RECORDED AFTER a +40% two-day move, no retroactive credit claimed — receipt tests conviction FROM HERE; expects below-top-1000 rejection = 4th distinct sub-1000 name, wide-book trigger evidence) |
| 2026-08-17 | IREN | neoclouds seem to be doing ok and NBIS did really well lately (operator, Monday briefing 08-17; RE-ADD — expects foreign rejection, conviction receipt for the global-book trigger) |
| 2026-08-17 | IONQ | IONQ for sure with the quantum re-look (operator, Monday briefing 08-17; Q2 beat + raised guide, sector snapback) |
| 2026-08-12 | CRCL, MSTR | BTC may have bottomed here (operator, 2026-08-12) -- BTC-proxy pair: MSTR = levered BTC treasury, CRCL = stablecoin rails |
| 2026-08-02 | ASTS, RKLB | rockets and satellites really sold off lately (operator, 2026-08-02) |
| 2026-07-20 | BE, CIFR, DELL, HUT, HXS, IREN, KEEL, NBIS, OKLO, PLTR, RDDT, RIOT, SLNH, WULF, WYFI | neocloud/AI-infra thesis -- wish I had earlier entries; possible big future winners, or the whole sector craps out (seed batch, collective reason) |
| 2026-07-20 | SKHY | SK Hynix (AI memory/HBM) -- corrects ticker typo HXS in the seed batch; same neocloud/AI-infra thesis. HXS stands on the record as the error it was. |
Rejections — picks the machine refused
| date | name | why refused |
|---|---|---|
| 2026-08-20 | PURR | not in universe at fire time |
| 2026-08-17 | IREN | not in universe at fire time |
| 2026-07-24 | HXS | unknown ticker (no warehouse security) |
| 2026-07-20 | IREN | not in universe at fire time |
| 2026-07-20 | KEEL | not in universe at fire time |
| 2026-07-20 | NBIS | not in universe at fire time |
| 2026-07-20 | SKHY | not in universe at fire time |
| 2026-07-20 | SLNH | not in universe at fire time |
| 2026-07-20 | WYFI | not in universe at fire time |
Run log — the receipts
Every weekly run, its determinism checksum, and every halt or restatement, in the open. History that fails to reproduce stops the machine. A restatement run re-derives an earlier window on a later day — that's why "window as-of" and "recorded on" can differ; the original row always stands, pointed at its successor.
| run | window as of | recorded on | status | checksum | note |
|---|---|---|---|---|---|
| #12 | 2026-09-04 | 2026-09-07 | complete | bbb46a0cc587 | auto-restatement (freeze-pin): values changed, decisions unchanged; supersedes run #11; books changed: shadow_01, shadow_03, shadow_04, shadow_05, shadow_06, shadow_07, shadow_08, shadow_09, shadow_10, shadow_11, shadow_12, shadow_13, shadow_15, shadow_16, shadow_17, shadow_18, shadow_19, shadow_20 |
| #11 | 2026-08-28 | 2026-08-31 | complete | 07e5c05b1f6e | auto-restatement (freeze-pin): values changed, decisions unchanged; supersedes run #10; books changed: shadow_02, shadow_04, shadow_05, shadow_06, shadow_08, shadow_09, shadow_11, shadow_13, shadow_14, shadow_15, shadow_17, shadow_19, shadow_20 · superseded by run #12 |
| #10 | 2026-08-21 | 2026-08-24 | complete | 71842e17f218 | auto-restatement (freeze-pin): values changed, decisions unchanged; supersedes run #9; books changed: shadow_01, shadow_02, shadow_03, shadow_04, shadow_07, shadow_08, shadow_09, shadow_10, shadow_11, shadow_14, shadow_15, shadow_17, shadow_19, shadow_20 · superseded by run #11 |
| #9 | 2026-08-14 | 2026-08-15 | complete | 5783553d25f4 | auto-restatement (freeze-pin): values changed, decisions unchanged; supersedes run #8; books changed: shadow_01, shadow_02, shadow_03, shadow_04, shadow_06, shadow_07, shadow_08, shadow_09, shadow_10, shadow_13, shadow_14, shadow_15, shadow_17, shadow_18, shadow_19, shadow_20 · superseded by run #10 |
| #8 | 2026-08-07 | 2026-08-11 | complete | fcc98ca7a4ba | superseded by run #9 |
| #7 | 2026-07-31 | 2026-08-11 | complete | 80cffa3ed9ff | RULING-#4 RESTATEMENT (vendor dividend re-adjustment): manual byte-identical; changed: shadow_04, shadow_05, shadow_07, shadow_10, shadow_17, shadow_18 |
| #6 | 2026-08-07 | 2026-08-10 | halted_drift | replay failed to reproduce run #5 | |
| #5 | 2026-07-31 | 2026-08-03 | complete | ec44ce4e17f6 | |
| #4 | 2026-07-24 | 2026-08-03 | complete | 09537653a8ed | RESTATEMENT of run #1 — vendor dividend readjustment (AGNC, ALLY ex-div rescaled adjusted history): 3 shadow books' fills moved ~1%; manual book byte-identical. Operator-authorized 2026-08-02. Original row stands. |
| #3 | 2026-07-31 | 2026-08-02 | halted_drift | replay failed to reproduce run #1 | |
| #2 | 2026-07-31 | 2026-08-02 | halted_drift | replay failed to reproduce run #1 | |
| #1 | 2026-07-24 | 2026-07-29 | complete | bd16e86ed045 |
Paper trading. Not investment advice. Position sizes and fill prices are deliberately not shown (data-license hygiene); percentages and decisions are the record.