Skip to main content
Research Record

Tombstones, with hashes.

Five sanitized research verdicts. Each public summary has its own content hash and points to an original pre-registration frozen in git before the corresponding verdict evidence.

How to verify

Run node scripts/verify-record.mjs from the repository root. The script recomputes each summary hash, reconstructs the original pre-registration from its declared commit, verifies its digest and git ancestry, confirms the verdict marker appeared only later, and checks the anchored record with the ots CLI.

Record entries

  1. OpenTimestamps anchor
    KILL — P1/P2/P3 · kill · 2026-07-13

    K200 variance-premium gate 0

    A short-variance harvest was tested against frozen existence, cost, regime, and source-robustness gates.

    Decisive statistic
    All three decisive prongs failed; the independent-source sign check also disagreed.
    SHA-256
    053dd02dc885a6f46b5daabc450f3eabb7341d89a6b0651322d6b97957f78c1f
    Frozen at commit
    12ff5d967f254c0603b7e278404b89be5c3bb005
    Pre-registration SHA-256
    d26ec0d4da99248baf249c496a3d563e1043a6af6a2c11666661e9edacd6344f
  2. P2 KILL / P4 PARK · closed-negative · 2026-07-09

    Fourth carry-leg study

    A proposed fourth portfolio leg was tested for standalone eligibility, frontier improvement, regime stability, and granularity.

    Decisive statistic
    The frontier-improvement prong failed and the granularity prong parked the proposal.
    SHA-256
    fe6cde7323653006718791d8a1fbe42fddcc2c5428606bc8d75c8299a817334a
    Frozen at commit
    61eb66490024ded0ba3d387f17dd7c2a218c2e2f
    Pre-registration SHA-256
    7c703d4c6bc4a302652624ed67f14128101301fed9d211c2dfe0b248a086271d
  3. PASS 2/2 → MONITOR · monitor · 2026-07-06

    Crowding stress tripwire

    A public-filings crowding measure was tested as a stress tripwire under two frozen prongs.

    Decisive statistic
    Both preregistered prongs cleared; the result remains monitor-only for at least two quarters.
    SHA-256
    f5ff5bfffba165cc3041a9a1ac90f0edc3803c39ffcf025602e0af7a6c2a013c
    Frozen at commit
    9eda7a7aa88a84571ca53c388efccb93af5a254d
    Pre-registration SHA-256
    0202235b056f6c23b401c96c98e0dbbcae49d38351f0e00b9beab6330f3814ed
  4. KILL — selectivity >1.0% · kill-split · 2026-07-06

    Price-first choke screen

    A price-first screen was tested for both selective workload and historical episode re-detection.

    Decisive statistic
    Re-detection passed 5/5, but the 1.68% fire rate exceeded the frozen 1.0% capacity cap.
    SHA-256
    f8cb6c52cdea34de1517cc53ccf16f8f6ffbff61bc8f661b1543e317942367d5
    Frozen at commit
    54d279dbda0f0ddf8cc6c9a5d52fa69b472ba73e
    Pre-registration SHA-256
    a2dbacedb7c9320cd333dfd49fbad99f5d7b1e7fccf0fef11e530158b0f8bba0
  5. KILL — gate 1 P2 · kill · 2026-07-05

    Cross-venue basis gate 1

    A cross-venue dislocation advanced through its cheap screen and then faced a separately frozen mechanism backtest.

    Decisive statistic
    The decisive P2 primary cell failed its frozen positive, day-clustered t-statistic bar.
    SHA-256
    c67c77d0b7d4ba9c3dfff31db946f98b94de4df0204c73161be1bd06a8a36812
    Frozen at commit
    6571189d97e390dbeba10356f2de5b20c759e140
    Pre-registration SHA-256
    39515d73eccb32a377e4356860928e91629510eb5a242b15143c7a03f8ed71b7

Keyboard shortcuts

deskgd findgf strategiesgs featuresge promotionsgp apiga botgb writinggw

Shortcuts are no-ops while typing in an input or textarea.