Ara.AI v8: cross-sectional stock ranking, v7 frozen to legacy/ - #183
Merged
Conversation
v7 predicted each stock's absolute next-day return and had no edge out of sample (50.23% direction accuracy against a 51.44% always-up baseline) — most of a daily return is the market's, which OHLCV cannot predict. v8 predicts the cross-sectional residual instead and ranks the universe long/short, using seed-averaged gradient-boosted trees rather than a transformer: ~11s per fit instead of ~7min, no torch. Walk-forward IC is +0.013 to +0.015 (t 1.1-1.4) depending on window — positive and consistent, not statistically significant, pre-cost; both windows are in docs/ARA_V8.md. CI now runs the backtest as a publishing gate and trains daily instead of hourly. The v7 stack moves to legacy/, dispatch-only.
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
v7 predicted each stock's absolute next-day return and had no edge out of sample (50.23% direction accuracy against a 51.44% always-up baseline) — most of a daily return is the market's, which OHLCV cannot predict. v8 predicts the cross-sectional residual instead and ranks the universe long/short, using seed-averaged gradient-boosted trees rather than a transformer: ~11s per fit instead of ~7min, no torch. Walk-forward IC is +0.013 to +0.015 (t 1.1–1.4) depending on window — positive and consistent, not statistically significant, pre-cost; both windows are in
docs/ARA_V8.md. CI now runs the backtest as a publishing gate and trains daily instead of hourly. The v7 stack moves tolegacy/, dispatch-only.Verified:
pytest tests/7 passed,pytest legacy/tests/38 passed, full pipeline green in CI (3m52s).