Skip to content

Feature/improvements#1

Merged
valeriouberti merged 13 commits intomainfrom
feature/improvements
Apr 30, 2026
Merged

Feature/improvements#1
valeriouberti merged 13 commits intomainfrom
feature/improvements

Conversation

@valeriouberti
Copy link
Copy Markdown
Owner

No description provided.

valeriouberti added 13 commits April 29, 2026 17:10
- Added `calibrate.py` for threshold calibration with DSR analysis, supporting various input options and output formats.
- Introduced `risk_stats.py` to implement Probabilistic Sharpe Ratio (PSR) and Deflated Sharpe Ratio (DSR) as per Bailey-Lopez methodology.
- Enhanced scoring functionality with `rank_universe` to convert absolute scores into percentile ranks for cross-sectional analysis.
- Created `index_membership.py` for point-in-time index membership lookup to address survivorship bias in backtesting.
- Updated `schema.sql` to include new table for index membership history, enabling historical index membership queries.
…ng momentum strategies with new metrics and database support
…tors

- Added `breadth.py` for market breadth calculations, including `pct_above_ma` and `breadth_series`.
- Introduced `regime_composite.py` to compute composite z-scores from HY OAS, breadth, and VIX, with classification into 5 buckets.
- Created `fred_client.py` to fetch and cache FRED series data (HY OAS, VIX, yield curve) with a daily cache mechanism.
- Developed `test_regime_daily.py` for smoke testing the regime daily composite against historical turning points.
- Updated SQL schema to include `fred_series_daily` table for storing FRED series data.
- Enhanced documentation in SIGNAL_ROADMAP.md to reflect the new features and findings from Fase B.3.
… system

- Added DECAY_MONITOR.md detailing the rationale, architecture, and findings of the decay monitor.
- Developed core functionality in decay_monitor.py, including Rolling Sharpe, CUSUM, and SPRT methods.
- Created calibration_stats.py for Brier score, reliability diagram, and AI add-value metrics.
- Introduced analyze_ai_verdicts.py for retrospective analysis of AI verdicts.
- Implemented test_decay_monitor.py for smoke testing decay detection on synthetic and real trade data.
- Updated SIGNAL_ROADMAP.md to reflect the status and findings of Fase D.
- Enhanced backtest portfolio page with survivorship correction and cross-sectional ranking options.
- Introduced threshold calibration page for probabilistic and deflated Sharpe ratios, including user inputs for various parameters.
- Implemented regime composite analysis page to visualize daily z-scores and classifications based on HY OAS, breadth, and VIX.
- Added decay monitor for strategy performance, utilizing rolling Sharpe, CUSUM, and SPRT methods to assess edge degradation.
- Created TradingView scripts for daily regime composite and ETF rotation, incorporating macro overlays and defensive switches.
… stress testing

- Added `scenario_replay.py` to perform historical scenario replay on critical market events (2008 GFC, 2020 COVID, 2022 rate shock).
- Created `SCENARIO_REPLAY.md` to document the results of the scenario tests.
- Introduced `bootstrap.py` for stationary bootstrap methods to enhance metric distributions.
- Developed `permutation_test.py` for conducting permutation tests on path-dependent metrics.
- Updated `SIGNAL_ROADMAP.md` to reflect the status of Fase E and findings from the scenario tests.
- Added `STRESS_VALIDATION.md` to outline the stress validation framework and results.
@valeriouberti valeriouberti merged commit ad08437 into main Apr 30, 2026
1 check failed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant