The study store did not answer; this is the study as last read.
DESCRIPTIVEWhich habits in the journal's posted 0DTE trades did better? (descriptive)
Among the 838 posted 0DTE SPY, QQQ and IWM trades of Zero to Profit Vol. II from January 2024 to January 2026, replayed on real quotes, which habits did better: entry time, call or put, with or against a 20-day trend, first entry or later, symbol, or a posted price inside the autopilot's premium band?
Descriptive, not a test. Followers who bought a minute after each post lost money in almost every slice; at the journal's own posted prices most slices were positive. Two findings stood out across slices: half off at +100% with everything out at -50% beat the journal's own exits by about 10 points per trade (follower -4.7% vs -14.7%; at his price +14.9% vs +3.7%), and posted prices inside the premium band did better than those outside it (follower -1.9% vs -8.5% with half off at +100% and all out at -50%). With that exit, the 09:36-09:45 entry window was the only first-hour window above zero for a follower (+5.7%, standard error 9.1; -3.0% under the journal's own exits), entries at or before 09:35 looked good at his price and poor for a follower, and trades on the side of a 20-day trend did better than those against it, which the journal took more often. Most differences sit within two standard errors.
Descriptive and retrospective: the record was written after the results, nothing is a counted trial, and 80 mean estimates were read with no correction. It is a prior look for the autopilot variants study and the journal's 2026 test, which test its suggestions: the 2026 test on trades it did not use, the variants study with a discovery window it did not use and a validation window that overlaps it.
Backtested by QS Research
Pre-registration
No pre-registration existed. A retrospective record written on 2026-11-22 states the data (an uncounted re-run of the published Vol. II replay whose report matches the published one byte for byte, on the same pinned 1-minute option-quote snapshot, and 1-minute SPY bars), the scope (0DTE SPY, QQQ and IWM entries from 2024-01-01 that filled), the two prices (follower and at his price), the two exits (the journal's own and half at +100%, all at -50%), the eight slicings and the trend rule, and the prior looks. The scripts are kept outside the repository and cited by sha256.
Registered 2026-11-22Trials · 0 trials · 0 passed · judged at N = 80 trials; 2972 trials run in total
| Trial | Verdict | |
|---|---|---|
| All | RESULT | not counted |
| 2024 | RESULT | not counted |
| 2025 | RESULT | not counted |
| 2026 (to 12 Jan) | RESULT | not counted |
| Entry ≤09:35 ET | RESULT | not counted |
| Entry 09:36-09:45 | RESULT | not counted |
| Entry 09:46-10:30 | RESULT | not counted |
| Entry after 10:30 | RESULT | not counted |
| Calls | RESULT | not counted |
| Puts | RESULT | not counted |
| With trend-1 | RESULT | not counted |
| Against trend-1 | RESULT | not counted |
| No trend (mixed) | RESULT | not counted |
| First in-scope 0DTE entry of the day | RESULT | not counted |
| Later entries (adds, second trades) | RESULT | not counted |
| SPY | RESULT | not counted |
| QQQ | RESULT | not counted |
| IWM | RESULT | not counted |
| Posted price in the band | RESULT | not counted |
| Outside the band | RESULT | not counted |
Limitations
- Nothing was pre-registered: the record was written after the results, and 80 mean estimates were looked at with no correction.
- At his price is a best case no follower or autopilot can count on, and under exit B it is approximate because the replay's triggers key off the follower's fill.
- 2026 is 21 trades; QQQ (51, mostly 2024) and IWM (38) are small; trend-1 uses SPY's closes for every symbol.
- 2023 is left out of scope; its index-0DTE trades (21) lost 45.7% with the journal's exits and 17.2% with half off at +100% and all out at -50%, as a follower.
- 153 of the trades fall inside the dates of the entry-time study's holdout; only the journal's own contracts are read, and no rule of that study is evaluated.
The numbers: every trial's checks, metrics, curve and yearly results, and the provenance
1 published version · current 3ec1990f1525e605, published 2026-11-22