backtester.py

"""백테스팅 엔진 - 포트폴리오 시뮬레이션 및 성과 지표"""

import numpy as np
import pandas as pd
import config


def run_backtest(
    df: pd.DataFrame,
    signals_df: pd.DataFrame,
    initial_capital: float = config.INITIAL_CAPITAL,
    commission_pct: float = config.COMMISSION_PCT,
) -> dict:
    """
    Long-only 백테스터: BUY → 전량 매수, SELL → 전량 매도, HOLD → 유지.
    """
    print(f"[백테스트] 초기자본 ${initial_capital:,.0f}, 수수료 {commission_pct*100:.1f}%")

    cash = initial_capital
    shares = 0
    portfolio_values = []
    trades = []
    entry_price = 0.0
    entry_date = None

    for i in range(len(df)):
        date = df.index[i]
        close = df.iloc[i]["Close"]
        position = signals_df.iloc[i]["Position"]

        # 매수: 현금 보유 중 + BUY 신호
        if position == "BUY" and shares == 0:
            cost = cash * (1 - commission_pct)
            shares = cost / close
            entry_price = close
            entry_date = date
            cash = 0.0

        # 매도: 주식 보유 중 + SELL 신호
        elif position == "SELL" and shares > 0:
            proceeds = shares * close * (1 - commission_pct)
            trade_return = (close - entry_price) / entry_price * 100
            trades.append({
                "entry_date": entry_date,
                "exit_date": date,
                "entry_price": entry_price,
                "exit_price": close,
                "return_pct": trade_return,
                "profit": proceeds - (shares * entry_price),
            })
            cash = proceeds
            shares = 0

        # 포트폴리오 가치
        portfolio_value = cash + shares * close
        portfolio_values.append({"Date": date, "Value": portfolio_value})

    portfolio_df = pd.DataFrame(portfolio_values).set_index("Date")
    pv = portfolio_df["Value"]

    # 성과 지표 계산
    total_return = (pv.iloc[-1] - initial_capital) / initial_capital * 100
    trading_days = len(pv)
    years = trading_days / config.TRADING_DAYS_PER_YEAR
    annual_return = ((pv.iloc[-1] / initial_capital) ** (1 / years) - 1) * 100 if years > 0 else 0

    daily_returns = pv.pct_change().dropna()
    sharpe = calculate_sharpe_ratio(daily_returns)
    max_dd = calculate_max_drawdown(pv)

    # Buy & Hold 비교
    bh_return = (df.iloc[-1]["Close"] - df.iloc[0]["Close"]) / df.iloc[0]["Close"] * 100

    # 승률
    if trades:
        wins = [t for t in trades if t["return_pct"] > 0]
        losses = [t for t in trades if t["return_pct"] <= 0]
        win_rate = len(wins) / len(trades) * 100
        gross_profit = sum(t["profit"] for t in wins) if wins else 0
        gross_loss = abs(sum(t["profit"] for t in losses)) if losses else 1
        profit_factor = gross_profit / gross_loss if gross_loss > 0 else float("inf")
    else:
        win_rate = 0
        profit_factor = 0

    result = {
        "total_return_pct": round(total_return, 2),
        "annual_return_pct": round(annual_return, 2),
        "sharpe_ratio": round(sharpe, 2),
        "max_drawdown_pct": round(max_dd, 2),
        "total_trades": len(trades),
        "win_rate": round(win_rate, 1),
        "profit_factor": round(profit_factor, 2),
        "buy_and_hold_return_pct": round(bh_return, 2),
        "portfolio_values": pv,
        "trades": trades,
        "years": round(years, 1),
    }
    return result


def calculate_sharpe_ratio(
    daily_returns: pd.Series,
    risk_free_rate: float = config.RISK_FREE_RATE,
) -> float:
    """연간화 샤프 비율"""
    if daily_returns.std() == 0:
        return 0.0
    excess_return = daily_returns.mean() * config.TRADING_DAYS_PER_YEAR - risk_free_rate
    annual_std = daily_returns.std() * np.sqrt(config.TRADING_DAYS_PER_YEAR)
    return excess_return / annual_std


def calculate_max_drawdown(portfolio_values: pd.Series) -> float:
    """최대 낙폭 (%)"""
    running_max = portfolio_values.cummax()
    drawdown = (portfolio_values - running_max) / running_max * 100
    return drawdown.min()


def print_summary(result: dict) -> None:
    """백테스트 결과 콘솔 출력"""
    print("\n" + "=" * 60)
    print("  QuantMiroFish 백테스트 결과")
    print("=" * 60)
    print(f"  분석 기간        : {result['years']}년")
    print(f"  총 수익률         : {result['total_return_pct']:+.2f}%")
    print(f"  연간 수익률       : {result['annual_return_pct']:+.2f}%")
    print(f"  샤프 비율         : {result['sharpe_ratio']:.2f}")
    print(f"  최대 낙폭 (MDD)   : {result['max_drawdown_pct']:.2f}%")
    print("-" * 60)
    print(f"  총 거래 횟수      : {result['total_trades']}회")
    print(f"  승률              : {result['win_rate']:.1f}%")
    print(f"  수익/손실 비율    : {result['profit_factor']:.2f}")
    print("-" * 60)
    print(f"  Buy & Hold 수익률 : {result['buy_and_hold_return_pct']:+.2f}%")
    alpha = result['total_return_pct'] - result['buy_and_hold_return_pct']
    print(f"  초과 수익 (Alpha) : {alpha:+.2f}%")
    print("=" * 60)