Skip to content

sueun-dev/polymarket-alpha-lab

BranchesTags

Repository files navigation

Polymarket Alpha Lab

Research repository + fully executable Python trading bot implementing 100 strategies for Polymarket prediction markets.

Research Documents

File Language Description
research/EN-polymarket-market-inefficiencies.md EN Comprehensive market inefficiency research — past cases, academic foundations, live mispricings (as of Feb 2026), strategies, risks, and 100+ cited sources across EN/KR/CN
research/EN-polymarket-top-100-strategies.md EN Top 100 trading strategies curated from 600+ internet sources (Reddit, Twitter/X, Substack, Medium, academic papers, GitHub, KR/CN sources). Ranked by Tier S/A/B/C
research/KR-polymarket-top-100-strategies.md KR Same top 100 strategies document in Korean

Strategy Tiers

Tier # Description
S #1-10 Verified alpha with documented profit records
A #11-30 Strong edge backed by data/research
B #31-70 Solid strategies with reasonable evidence
C #71-100 Experimental edge — innovative but needs validation

Trading Bot

All 100 strategies are implemented as executable Python modules with shared core infrastructure.

Quick Start

# 1. Clone
git clone https://github.com/sueun-dev/polymarket-alpha-lab.git
cd polymarket-alpha-lab

# 2. Install dependencies
pip install -r requirements.txt

# 3. Configure
cp .env.example .env
# Edit .env with your API keys (Polymarket, NOAA, Kalshi, etc.)
# Edit config.yaml to enable/disable strategies and tune parameters

# 4. Run (paper trading mode by default)
python3 main.py run

# 5. List available strategies
python3 main.py list

# 6. Run specific strategies only
python3 main.py run --strategy s01_reversing_stupidity --strategy s03_nothing_ever_happens

# 7. Dry run (scan only, no orders)
python3 main.py run --dry-run

# 8. Run backtests
python3 main.py backtest --data-dir data/historical/

# 9. Launch dashboard
streamlit run dashboard/app.py

# 10. Launch React dashboard (liquid glass UI)
# terminal 1
python3 dashboard_api.py --host 127.0.0.1 --port 8001

# terminal 2
cd dashboard-react
npm install
npm run dev

# 11. Run NOAA weather strategy (s02) with native C++ weather kernel
# Optional: compile native weather kernel
clang++ -O3 -march=native -std=c++17 -o native/s02_weather_signal_engine native/s02_weather_signal_engine.cpp

# run only s02 strategy (paper/live based on config.yaml)
python3 main.py run --strategy s02_weather_noaa

# 12. Forensic audit on closed weather markets (public-data reverse analysis)
python3 tools/weather_market_forensics.py --max-rows 12000 --max-markets 120 --min-trades 5 --report-path logs/weather_forensics_report.json

Project Structure

core/               # Shared infrastructure
  client.py         # Polymarket CLOB API wrapper (paper + live modes)
  risk.py           # Portfolio risk management (position limits, daily loss)
  kelly.py          # Kelly Criterion position sizing (Half-Kelly default)
  scanner.py        # Market scanner with filtering & anomaly detection
  base_strategy.py  # Abstract base class for all strategies
  notifier.py       # Telegram / Discord notifications
  models.py         # Pydantic data models (Market, Signal, Order, etc.)
strategies/         # 100 strategy plugins (auto-discovered)
  tier_s/           # #1-#10: Verified alpha
  tier_a/           # #11-#30: Strong edge
  tier_b/           # #31-#70: Solid strategies
  tier_c/           # #71-#100: Experimental
backtest/           # Backtesting engine (slippage simulation, Sharpe/MDD)
dashboard/          # Streamlit web dashboard
dashboard-react/    # React web dashboard (Vite)
data/               # External data collectors (NOAA, Kalshi, sentiment, etc.)
tests/              # Unit/integration test suite
research/           # Research papers/notes (EN/KR market documents)
development/        # Development planning docs
config.yaml         # Strategy parameters & risk settings
.env.example        # Environment variable template

Configuration

Trading mode — set in config.yaml:

  • paper (default): Simulated trading, no real orders
  • live: Connects to Polymarket CLOB API with real funds

Risk parameters — config.yaml:

  • max_position_pct: Max % of portfolio per market (default 10%)
  • max_daily_loss: Daily stop-loss threshold (default 5%)
  • max_open_positions: Concurrent position limit (default 20)
  • kelly_fraction: Kelly sizing fraction (default 0.5 = Half-Kelly)

Tests

python3 -m pytest tests/ -v
# Run the full test suite

How to Read the Research

  1. Start with research/EN-polymarket-market-inefficiencies.md for the research foundation
  2. Then read research/EN-polymarket-top-100-strategies.md (or KR version) for actionable strategies
  3. Each strategy includes: source, evidence, execution steps, expected edge, and key risks

Disclaimer

  • All data is as of 2026-02-27
  • This repository is for research purposes only and does not constitute investment advice
  • Prediction market trading involves risk of loss

Polymarket Alpha Lab (한국어)

Polymarket 예측 시장의 비효율성과 트레이딩 전략을 분석한 리서치 + 100개 전략을 구현한 Python 트레이딩 봇 저장소입니다.

리서치 문서

파일 언어 설명
research/EN-polymarket-market-inefficiencies.md EN 시장 비효율성 종합 연구 — 과거 사례, 학술 근거, 라이브 미스프라이싱 (2026년 2월 기준), 전략, 리스크, 영/한/중 100+개 출처
research/EN-polymarket-top-100-strategies.md EN 인터넷 600+개 글에서 엄선한 Top 100 트레이딩 전략 (영어 버전)
research/KR-polymarket-top-100-strategies.md KR 동일한 Top 100 전략 문서 (한국어 버전)

전략 티어

티어 번호 설명
S #1-10 검증된 알파 — 실제 수익 기록 존재
A #11-30 강한 엣지 — 데이터/연구 기반
B #31-70 견고한 전략 — 합리적 증거 존재
C #71-100 실험적 엣지 — 혁신적이나 검증 필요

트레이딩 봇

100개 전략이 모두 실행 가능한 Python 모듈로 구현되어 있습니다.

빠른 시작

# 1. 클론
git clone https://github.com/sueun-dev/polymarket-alpha-lab.git
cd polymarket-alpha-lab

# 2. 의존성 설치
pip install -r requirements.txt

# 3. 설정
cp .env.example .env
# .env에 API 키 입력 (Polymarket, NOAA, Kalshi 등)
# config.yaml에서 전략 활성화/비활성화 및 파라미터 조정

# 4. 실행 (기본: 페이퍼 트레이딩)
python3 main.py run

# 5. 전략 목록 확인
python3 main.py list

# 6. 특정 전략만 실행
python3 main.py run --strategy s01_reversing_stupidity --strategy s03_nothing_ever_happens

# 7. 드라이런 (스캔만, 주문 없음)
python3 main.py run --dry-run

# 8. 백테스트 실행
python3 main.py backtest --data-dir data/historical/

# 9. 대시보드 실행
streamlit run dashboard/app.py

# 10. React 대시보드 실행 (liquid glass UI)
# 터미널 1
python3 dashboard_api.py --host 127.0.0.1 --port 8001

# 터미널 2
cd dashboard-react
npm install
npm run dev

# 11. NOAA 날씨 전략(s02) 네이티브 C++ 커널 실행
# 선택: 네이티브 날씨 커널 빌드
clang++ -O3 -march=native -std=c++17 -o native/s02_weather_signal_engine native/s02_weather_signal_engine.cpp

# s02 전략만 실행 (config.yaml의 paper/live 모드 따름)
python3 main.py run --strategy s02_weather_noaa

# 12. 종료된 날씨 시장 포렌식 분석 (공개 데이터 기반 역추적)
python3 tools/weather_market_forensics.py --max-rows 12000 --max-markets 120 --min-trades 5 --report-path logs/weather_forensics_report.json

설정

트레이딩 모드 — config.yaml:

  • paper (기본): 시뮬레이션, 실제 주문 없음
  • live: Polymarket CLOB API 연결, 실제 자금 사용

리스크 파라미터 — config.yaml:

  • max_position_pct: 마켓당 최대 포트폴리오 비율 (기본 10%)
  • max_daily_loss: 일일 손실 한도 (기본 5%)
  • max_open_positions: 동시 포지션 한도 (기본 20개)
  • kelly_fraction: 켈리 비율 (기본 0.5 = Half-Kelly)

테스트

python3 -m pytest tests/ -v
# 전체 테스트 스위트 실행

읽는 방법

  1. research/EN-polymarket-market-inefficiencies.md로 리서치 기반 파악
  2. research/KR-polymarket-top-100-strategies.md (또는 EN 버전)으로 실행 가능한 전략 확인
  3. 각 전략에는 출처, 증거, 실행 방법, 예상 엣지, 핵심 리스크 포함

참고

  • 모든 데이터는 2026-02-27 기준입니다
  • 본 저장소는 연구 목적이며 투자 조언이 아닙니다
  • 예측 시장 트레이딩에는 손실 위험이 수반됩니다

About

Exhaustive research on Polymarket prediction market inefficiencies — past exploits, academic foundations, live mispricings, trading strategies, and future alpha. 100+ sources across EN/KR/CN.

Topics

python research arbitrage prediction-markets bilingual quantitative-research alpha-research polymarket

Resources

Readme

Code of conduct

Code of conduct

Contributing

Contributing

Security policy

Security policy
Activity

Stars

4 stars

Watchers

0 watching

Forks

2 forks
Report repository

Releases

No releases published

Packages

 
 
 

Contributors

Languages