Trading Engine

A low-latency intraday order book imbalance trading engine in Rust.

Architecture

• ingestion: Async WebSocket ingestion (Mock feed provided for demo).
• orderbook: Lock-free BTreeMap-based order book.
• features: Computes OBI (Top-1, Top-5), Microprice, Mid-price.
• signals: Rule-based signal generation (OBI > 0.6 -> Long).
• risk: Pre-trade risk checks (Max Position, Max Daily Loss).
• execution: Execution simulator (fills at Mid-price).
• launcher: Main application entry point with Prometheus metrics.

Running

Live Mode

Runs the engine with a mock data feed.

cargo run -p launcher -- --mode live

Metrics

Prometheus metrics are exposed at port 9090

Metrics include:

• order_count: Total orders executed.
• obi_value: Current Top-1 Order Book Imbalance.
• total_pnl: Cumulative profit and loss after accounting for all executed trades
• sharpe_ratio: Risk-adjusted return metric to evaluate strategy consistency.
• max_drawdown: The largest peak-to-trough decline, measuring the worst-case risk scenario.
• total_fees_paid: Total exchange/trading fees incurred (critical for high-frequency strategies).
• win_rate: The percentage of profitable trades relative to the total number of trades.
• avg_slippage: difference between the expected price and the actual execution price.

Backtest Mode (Stub)

cargo run --release -p launcher -- --mode backtest --file ./data/replay/btcusdt.csv

Workspace Structure

• common: Shared types (Price, Volume, Order).
• ingestion: Data feed connection.
• orderbook: Core data structure.
• features: Feature extraction.
• signals: Trading logic.
• risk: Risk management.
• execution: Trade simulation.