ORB Search-Space Atlas
The complete machine-readable summary of the ORB search-space atlas. Plain HTML, no scripts, no embedded data — everything below is the text itself, which is why a reader that fetches rather than renders gets this page.
Open the interactive atlas → 13 MB, thirteen plates, meant for a browser. A browser should have been sent there already; this link is for one that was not.
Machine-readable summary
Every figure below is generated at build time from the same artifacts the interactive plates draw from, so no script has to run to read this page. The interactive version above is the same numbers with controls on them.
| field | value |
|---|---|
| grammar | v3 (85ec7c28565c) |
| configurations | 143,640 |
| lattice shape | 9 x 6 x 19 x 10 x 14 |
| axes | or_bars, buffer, stop, entry_window, rel_volume_min |
| simulated axes | or_bars, buffer, stop |
| axes applied by selection | entry_window, rel_volume_min |
| sessions | 4,531 |
| span | 2005-01-03 to 2022-12-30 |
| warehouse artifact | 1b3b7011d25c |
| built | 2026-09-09 19:19 |
| direction | long only |
| costs | gross of costs; no commission, spread or borrow |
| position sizing | equal weight across the positions open that session |
| universe | point-in-time ranked top 500 US common stocks |
| hidden-axis reducer | MEDIAN, everywhere, on every plate and table |
| held out | 2023 onward is sealed and was not read to produce any number here |
| axis | unit | levels | values | meaning |
|---|---|---|---|---|
| or_bars | 5-min bars | 9 | 1, 2, 3, 4, 6, 9, 12, 18, 24 | opening-range length: 5 to 120 minutes from 09:30 |
| buffer | OR widths | 6 | 0.0, 0.025, 0.05, 0.1, 0.15, 0.25 | how far above the OR high a close must be to trigger |
| stop | method x multiple | 19 | 0.25x OR, 0.50x OR, 0.75x OR, 1.00x OR, 1.50x OR, 2.00x OR, 3.00x OR, 4.00x OR, no stop, 3.00x ATR, 2.20x ATR, 1.40x ATR, 0.90x ATR, 0.60x ATR, 0.40x ATR, 0.25x ATR, 0.15x ATR, 0.10x ATR, 0.07x ATR | stop methodology and its multiple, mutually exclusive: OR_WIDTH tight to loose, then NONE, then ATR loose to tight. Index order is tightness order; only the method label changes across index 8 |
| entry_window | minutes | 10 | 5, 10, 15, 20, 30, 60, 90, 150, 240, 360 | how long after the OR closes a trigger still counts |
| rel_volume_min | ratio | 14 | 0.0, 0.6, 0.8, 1.0, 1.3, 1.8, 2.5, 3.0, 4.0, 5.0, 6.0, 8.0, 10.0, 15.0 | minimum OR volume over its own trailing 60-session median; 0 = no filter. The top levels are deliberately sample-starved, to show where the effect runs out of data rather than where the grid was cut |
| term | share of total variance |
|---|---|
| rel_volume_min | 24.69% |
| entry_window | 24.62% |
| stop | 13.39% |
| or_bars | 2.32% |
| buffer | 1.77% |
| entry_window|rel_volume_min (interaction) | 19.97% |
| or_bars|entry_window (interaction) | 2.48% |
| or_bars|rel_volume_min (interaction) | 1.41% |
| stop|rel_volume_min (interaction) | 1.16% |
| buffer|entry_window (interaction) | 0.73% |
| stop|entry_window (interaction) | 0.51% |
| three-way and higher | 5.99% |
Study: is the grammar wide enough?
Every boundary that was a CUT rather than a structural floor was extended and re-measured, with each new level compared against the old boundary at identical coordinates on every other axis. A 5-minute opening range, a zero buffer, a 5-minute entry window and a zero volume filter are floors, not cuts, and cannot be extended.
| axis | verdict | per end | typical step inside the grammar | flat below | floor is structural |
|---|---|---|---|---|---|
| or_bars | clipped | upper: clipped | 0.0648 | 0.0324 | yes |
| buffer | clipped | upper: clipped | 0.0371 | 0.0185 | yes |
| stop | clipped | upper: clipped; tight: plateau | 0.0709 | 0.0355 | no |
| entry_window | plateau | upper: plateau | 0.1154 | 0.0577 | yes |
| rel_volume_min | clipped | upper: clipped | 0.0788 | 0.0394 | yes |
Study: the relative-volume axis, run past the grammar's ceiling
Verdict: STILL IMPROVING. the median Sharpe is +0.051 at 20x and +0.245 at 50x against the 15x boundary, against a typical step on this axis of 0.079, and 83% of the 10,260 paired configurations move the same way. The upper boundary is still unresolved. The boundary tested was 15.0000x; the furthest level with enough sample to evaluate was 50.0000x; levels 80x, 120x are sample-starved and their figures are reported but not relied on.
No simulation was re-run for this: the threshold is applied by selection over outcomes the simulation already produced, so grammar v3's Plane C (0f8e339a980b) and ratio cache (e0e2c650eb92) were read unchanged.
| threshold | origin | median Sharpe | share of configurations positive | median trades per configuration | stock-days clearing it | share of all stock-days | years positive | note |
|---|---|---|---|---|---|---|---|---|
| 0x | v3 | -1.1744 | 2.8% | 531,593 | 19,345,301 | 100.0000% | 0/18 | |
| 0.6x | v3 | -1.0898 | 4.8% | 433,435 | 16,641,375 | 86.0228% | 1/18 | |
| 0.8x | v3 | -0.9742 | 7.1% | 334,428 | 13,089,124 | 67.6605% | 1/18 | |
| 1x | v3 | -0.8437 | 9.7% | 240,449 | 9,622,851 | 49.7426% | 4/18 | |
| 1.3x | v3 | -0.7253 | 11.7% | 145,791 | 5,978,999 | 30.9067% | 2/18 | |
| 1.8x | v3 | -0.5940 | 13.6% | 69,750 | 3,025,036 | 15.6371% | 3/18 | |
| 2.5x | v3 | -0.5230 | 11.4% | 31,694 | 1,496,481 | 7.7356% | 2/18 | |
| 3x | v3 | -0.4999 | 9.1% | 20,266 | 1,023,629 | 5.2914% | 3/18 | |
| 4x | v3 | -0.3767 | 12.8% | 10,380 | 572,883 | 2.9614% | 5/18 | |
| 5x | v3 | -0.2964 | 18.8% | 6,295 | 369,431 | 1.9097% | 6/18 | |
| 6x | v3 | -0.3459 | 16.9% | 4,176 | 258,177 | 1.3346% | 5/18 | |
| 8x | v3 | -0.3389 | 12.3% | 2,213 | 144,851 | 0.7488% | 5/18 | |
| 10x | v3 | -0.3393 | 8.9% | 1,362 | 90,433 | 0.4675% | 4/18 | |
| 15x | v3 | -0.2496 | 16.1% | 564 | 36,567 | 0.1890% | 5/18 | |
| 20x | new | -0.1931 | 16.3% | 297 | 18,430 | 0.0953% | 7/18 | |
| 24.5x | new | -0.1406 | 27.1% | 183 | 11,298 | 0.0584% | 8/18 | |
| 30x | new | -0.0556 | 40.8% | 114 | 6,978 | 0.0361% | 9/18 | |
| 50x | new | -0.0035 | 49.3% | 33 | 2,019 | 0.0104% | 9/18 | |
| 80x | new | 0.1864 | 71.3% | 11 | 728 | 0.0038% | 12/18 | sample-starved |
| 120x | new | 0.2358 | 76.2% | 7 | 377 | 0.0019% | 7/14 | sample-starved |
Study: does any of the structure persist?
The pooled atlas is taken apart by calendar year and by the market's overnight gap. Findings, stated as the artifact states them:
- The overnight gap is the strongest conditioning variable in the project, and it does not survive being taken apart by year.
- Persistence is not regional: the configurations that were positive in 13 of 18 years fall into disconnected pieces, and not one has an interior cell — every member has a neighbour that failed.
- No stop at all beats both OR-width and ATR stops on every persistence measure, on one common eligible sample.
- Universe size is a plateau, not a peak.
- The grid is not four stocks: the median configuration traded over a thousand distinct names.
Grammar v3, 143,640 configurations, 4,531 sessions, 2005-01-03 to 2022-12-30.
What this is not
Not a recommendation to trade an opening-range breakout. The grid loses money gross of costs across almost all of its configurations, and the subject of the project is the shape of the search space rather than the strategy in it. No level or configuration on this page is recommended, and the boundary studies deliberately classify axes by whether an effect has STOPPED, never by whether a level performed well.