backtest/performance_metrics.py

"""Performance metrics for backtest results."""

from __future__ import annotations

import numpy as np
import pandas as pd


def annualized_return(daily_returns: pd.Series) -> float:
    return float((1 + daily_returns).prod() ** (252 / max(len(daily_returns), 1)) - 1)


def sharpe_ratio(daily_returns: pd.Series, rf: float = 0.0) -> float:
    excess = daily_returns - rf / 252
    return float(excess.mean() / (excess.std() + 1e-8) * np.sqrt(252))


def max_drawdown(equity_curve: pd.Series) -> float:
    peak = equity_curve.cummax()
    dd = equity_curve / peak - 1
    return float(dd.min())


def summarize_returns(daily_returns: pd.Series) -> dict:
    equity = (1 + daily_returns).cumprod()
    return {
        "ann_return": annualized_return(daily_returns),
        "sharpe": sharpe_ratio(daily_returns),
        "max_drawdown": max_drawdown(equity),
        "total_return": float(equity.iloc[-1] - 1) if len(equity) else np.nan,
    }