b3ba914a8d
Optimize v7 strategy: v7.356 achieves CW 2159% + SQS 90.7 (Pareto improvement over v7.314)
...
Key changes from v7.314 baseline (CW 2012%, SQS 90.0):
- max_position_value_pct 15→25, non_a_tier_target_1_fraction 0.2→0
- max_daily_new_risk_pct 30→50 (via v7.330, CW champion 2148%)
- bullish_raised_recovery per_trade_risk_pct 0.71→0.55 (DD improvement)
- bullish_raised_recovery max_holding_days 12→10 (sweet spot, +98pp CW)
Result: v7.356 CW 2159% (+147pp), SQS 90.7 (+0.7), risk 66.2 (+2.9), robustness 94.3 (+0.5)
All metrics improved simultaneously — return increase AND DD reduction achieved.
Also includes: web UI updates, pipeline scripts, v16/v17/v18/v19 experiment pruning,
Form4 preset additions, snapshot registry updates, domain.py enhancements.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
4 months ago
e38c314a09
Add ownership/risk-off sleeves, v17-v19 experiments, and web app restructure
...
New features:
- Ownership 13D/13G residual-cash sleeve with PIT calendar and quality filters
- Risk-off alpha sleeve (GLD/DBC rotation on crisis regime signals)
- Crisis relay target in parking: evaluates before defensive relay
- Bearish symbol allocation split (bearish_alloc_pct + sgov remainder)
- Alternative defensive ETF candidate (cash_parking_defensive_alt_symbol)
- Composite eval and engine ablation tools
- experiment and overfit CLI apps
New experiments:
- v17.x series (v17.1 champion SQS 78.4; v17.5–v17.129 exhausted)
- v18.x and v19.x families from v12.8 OOT defense branch
- v7.119 composed variants (idle alpha + ownership + risk-off sleeves)
- parking_only configs: bufb, jepq, merix, regime_tiered
- empty_strategy baseline config
Web app:
- Restructured into routers/services modules (experiments, leaderboard, runs, sqs, docs)
- Ownership sleeve and risk-off sleeve controls in backtest UI
- Frontend: ComposeStrategy page, tradeSleeves lib, idle decomposition display
Research tools:
- Ownership 13D/13G probe and PIT cache builder
- Dividend capture probe and cache builder
- Insider Form4 idle alpha probe
- Alternative ETF parking probe, put-spread overlay probe
- Wikimedia low-attention and peer-relative idle alpha probes
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
4 months ago
2395a0c0c3
feat: PEAD mid-cap strategy + pipeline hardening + README cleanup
...
- Implement PEAD 7% Long+Short strategy with mid-cap universe expansion
- Add Stock Oracle screener/company clients, text sentiment features
- Enhance backtest engine: short-side execution, walk-forward CV, MFE/MAE analysis
- Harden pipeline: sequential Oracle API calls, scoring recalibration (event_quality 65%)
- Add experiment configs for 60+ strategy variants and journal tracking
- Add review/analysis CLI tools
- Remove obsolete dev/phase0-4 design documents and analysis scripts
- Clean README to reflect only implemented features (remove unbuilt adapters/engines)
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
5 months ago
9d3427e24e
fix: disable Qwen3.5 thinking mode and switch OllamaClient to sync httpx
...
- Add `think: False` and `num_ctx: 8192` to Ollama payload:
Qwen3.5 extended thinking mode generated 1300+ internal reasoning
tokens before each response, adding 30-60s latency per LLM call.
Disabling it reduces parse time from 600s timeout to ~13s.
- Rewrite OllamaClient to use sync httpx.Client inside asyncio.to_thread():
Async httpx inside an active asyncpg SQLAlchemy session context on
Python 3.13 hung indefinitely. Synchronous httpx in a thread pool
completely isolates Ollama I/O from the asyncio event loop.
- Fix filing_poller to set issuer_id/symbol_id on Document records:
Missing FK caused feature_builder to reject all events with
event_no_symbol warning. Now looks up IssuerMaster/SymbolMaster
by ticker before creating Document rows.
- Update test_llm_client to mock _sync_call instead of _client attr.
- Raise ollama_timeout default to 600s for large document processing.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
5 months ago
bbe0e31150
feat: implement Phase 3 -- LLM enrichment, labeler, review queue, dataset export
...
- libs/llm/: OllamaClient (httpx async, retry), LLMCacheStore (SHA-256 DB cache),
prompt registry (event_classifier_v1), LLMParser (cache→prompt→validate→repair)
- libs/parser/merger.py: rule+LLM canonical merge with provenance tracking,
conflict detection (both confident + disagree), should_queue_for_review()
- libs/db/models.py: LLMCallCache, ReviewItem, EventLabel 3개 ORM 모델 추가
- libs/db/migrations/versions/0002_phase3_tables.py: Phase 3 Alembic migration
- libs/labeler/: filing_time_bucket→reaction_date, 1D/3D/5D fwd return, MFE/MAE
- libs/review/queue.py: create(dedup)/resolve/list ReviewItem
- libs/export/snapshot_export.py: temporal split + Parquet + manifest.json
- apps/: label_generator, dataset_export, review CLI, gold set evaluator
- 42개 신규 테스트 추가 (unit 32 + integration 4 + replay 1) — 152/152 통과
- libs/common/config.py: OLLAMA_URL/MODEL/TIMEOUT 설정 추가
- libs/common/logging.py: bugfix — add_logger_name incompatible with PrintLoggerFactory
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
5 months ago
0471018dd8
feat: implement ACE-F v1 Phase 1 -- Stock Oracle 기반 이벤트 파이프라인
...
Stock Oracle (localhost:18001)을 단일 데이터 소스로 사용하는 미국 주식
이벤트 스윙 트레이딩 시스템의 Phase 1 구현체.
주요 구성:
- libs/oracle_client/: Stock Oracle REST 클라이언트 (filings, price, financial, fred, finra)
- libs/common/: config, logging, time_utils, ids, retries, file_store
- libs/db/: SQLAlchemy 2.0 모델 12개 + Alembic 마이그레이션 0001
- libs/parser/: 규칙 기반 파서 (텍스트 정규화, JSON Schema 검증, LLM 스텁)
- libs/features/: 시장/이벤트 피처 계산기
- apps/pipeline/: filing_poller → filing_fetcher → event_parser → feature_builder
- apps/sync/: macro_sync (FRED), short_volume_sync (FINRA), issuer_sync
- tests/: 단위 81개 + 리플레이 5개 전체 통과, lint clean
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
5 months ago
I Luk Kim