docs: record I5e CSI300 generalization result - #54
Merged
Conversation
Docs-only progress sync after merging PR #53 (I5e MMP quintile CSI300 independent generalization). Records the negative result: the I5d CSI500 Q1->Q5 MMP monotonicity does NOT generalize to CSI300 over the same 2021-06-01..2026-05-31 window (Spearman -0.5000, Q5 high-MMP worst, synthetic Q5-Q1 cumulative -26.77%) -> FAILS / not robust; pause MMP promotion until disjoint-window / third-universe evidence. - merged-PR list += #53 - new I5e progress bullet (EXPLORATORY, config-only, covered=448/481, stk_mins_live_calls=0, NAV/annual/Spearman, caveats preserved) - quality gate 677 -> 678 passed; 17 -> 18 configs; phase0 0.9600/0.8408 unchanged - roadmap: MMP framed conservatively (not robust/tradable); D6 / endpoint schema registry remain deferred No code/config/test changes; AGENTS.md kept byte-identical to CLAUDE.md.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Docs-only progress sync after merging PR #53 (P-I5e MMP quintile CSI300
independent generalization). No code / config / test changes. AGENTS.md is kept
byte-identical to CLAUDE.md (
cmp -sclean).Records the negative research result: the accepted I5d CSI500 Q1→Q5 MMP
monotonicity does NOT generalize to CSI300 over the same 2021-06-01..2026-05-31
window — Spearman −0.5000 (vs +1.0000 on CSI500), Q5 (high MMP) is the worst
group, synthetic Q5−Q1 cumulative −26.77%. Verdict: FAILS / not robust;
roadmap now pauses MMP promotion until disjoint-window / third-universe evidence.
Changes (2 files):
#53（P-I5e ... 负结果）(CSI300 448/481 = 93.1% minute coverage, no CSI1000 fallback, window not
shortened),
stk_mins_live_calls=0, Q1→Q5 NAV/annual, Spearman, caveats preserved677 → 678 passed(the +1 istest_cache_config.py'sconfig/*.yamlglob auto-covering the new config, not a new test);
17 → 18 configs; phase00.9600 / 0.8408unchangedregistry remain deferred
I5e is EXPLORATORY; nothing here frames MMP as performance/tradable evidence.