Architecture 🏗️

QBT-Lite provides two engines and a modular set of components.

Data → Strategy → Engine → Broker → Portfolio → Metrics → Report → Visualization

Engines

• Vectorized Engines: qbt/core/engine.py (single-asset), qbt/core/engine_multi.py (multi-asset)
• Event-Driven Engine: qbt/core/event_engine.py (intraday with events)

Events (from event_engine.py)

• MarketEvent(timestamp: pd.Timestamp, bar: pd.Series, symbol: str)
• FillEvent(timestamp: pd.Timestamp, symbol: str, side: str, qty: int, price: float, fee: float)

Broker

• qbt/core/broker.py: transact(order) applies commission() and slippage.

Portfolio

• on_fill(fill), mark_to_market(last_prices), equity_series().

Metrics

• performance_from_nav, compute_drawdown (core)
• trade_stats_from_fills (analytics)

Reports

• qbt/report/report.py writes CSV/MD and saves equity.png & drawdown.png.

CLI

• qbt/cli.py exposes the backtester to the command line with common options.