Home

FinSight

Turn financial news sentiment into a tradable factor — and prove it isn't luck.

FinSight builds a cross-sectional, dollar-neutral long/short signal from news sentiment, backtests it with realistic transaction costs, and then subjects the result to the statistics that separate a real edge from backtest overfitting:

• Deflated Sharpe Ratio (DSR) — corrects for multiple-testing across configurations tried
• Probabilistic Sharpe Ratio (PSR) — P(true Sharpe > 0) under parameter uncertainty
• Purged + embargoed walk-forward CV — eliminates look-ahead leakage across fold boundaries
• News event study (CAR / t-stat) — independent confirmation from a second statistical angle

The pipeline is offline-first: a deterministic synthetic market embeds a known sentiment-to-return relationship so every metric has ground truth. CI runs green with no API keys and no model downloads.

Architecture

flowchart LR
    subgraph Data["Data backend (env-selectable)"]
        SYN[Synthetic market\nknown signal + AR(1) news]
        CSV[CSV / yfinance]
    end
    subgraph Sentiment["Sentiment backend"]
        LEX[Lexical\nfinance lexicon · offline]
        FB[FinBERT\noptional extra]
    end
    SYN --> NEWS[Headlines]
    CSV --> NEWS
    NEWS --> LEX --> SCORE[Daily asset sentiment]
    NEWS --> FB --> SCORE
    SCORE --> SIG[Cross-sectional z-score\n→ dollar-neutral weights]
    SIG --> BT[Backtest engine\nshift+1 timing · txn costs]
    BT --> STATS[Overfitting-aware stats\nSharpe · PSR · DSR · purged CV]
    SIG -.-> ES[Event study · CAR]

Loading

Quick start

python -m venv .venv && source .venv/bin/activate   # Windows: .venv\Scripts\activate
pip install -e ".[dev]"
pytest -q                    # 18 tests
finsight backtest            # run the strategy

Wiki pages

Page	What it covers
Architecture	Component design, sentiment backends, backtester, DSR/PSR math
Evaluation	Benchmark setup, results table, ablation, reproduce commands
Configuration	Env vars, backend matrix, .env.example
Development	Setup, test commands, adding new backends, code layout