普通人的稳健投资秘籍：风险平价模型指南

ounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(lineounter(line
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

# 设置绘图参数，中文显示
plt.style.use('ggplot')
plt.rcParams['font.sans-serif'] = ['SimHei']
plt.rcParams['axes.unicode_minus'] = False

def generate_market_data(n_days=1000):
    """生成具有负相关性的股债数据，模拟牛熊周期"""
    np.random.seed(42)
    dt = 1/252

    # 参数设定：[股票, 债券]
    # 股票：高收益高波动；债券：低收益低波动
    mu = np.array([0.15, 0.04])      
    vol = np.array([0.25, 0.05])     

    # 关键：设置强负相关性 (-0.4)，确保股债对冲效果
    corr_matrix = np.array([
        [1.0, -0.4],
        [-0.4, 1.0]
    ])

    # 生成相关联的随机过程
    cov = np.outer(vol, vol) * corr_matrix
    L = np.linalg.cholesky(cov)

    returns = np.zeros((n_days, 2))
    market_status = [] # 记录市场状态用于验证

    for t in range(n_days):
        # 模拟随机冲击
        shock = np.random.normal(0, 1, 2)
        correlated_shock = L @ shock

        # 人为制造"熊市"冲击 (在第300-450天模拟股灾)
        if 300 < t < 450:
            # 股票暴跌，债券因避险上涨
            current_mu = np.array([-0.30, 0.08]) 
            market_status.append('熊市')
        else:
            current_mu = mu
            market_status.append('牛市/震荡')

        returns[t] = current_mu * dt + correlated_shock * np.sqrt(dt)

    dates = pd.date_range('2020-01-01', periods=n_days, freq='B')
    df = pd.DataFrame(returns, index=dates, columns=['股票', '债券'])
    return df

def run_risk_parity_strategy(returns, lookback=60):
    """极简风险平价策略：倒数波动率加权"""
    # 1. 计算滚动波动率
    rolling_vol = returns.rolling(window=lookback).std()

    # 2. 计算风险平价权重 (权重 = 1 / 波动率)
    inv_vol = 1.0 / rolling_vol

    # 3. 归一化权重
    rp_weights = inv_vol.div(inv_vol.sum(axis=1), axis=0)

    # 4. 滞后一期应用权重
    rp_weights = rp_weights.shift(1).dropna()

    # 5. 计算策略收益
    # 对应日期的收益 * 对应日期的权重
    strategy_ret = (returns.loc[rp_weights.index] * rp_weights).sum(axis=1)

    return strategy_ret, rp_weights

def calculate_metrics(series, name):
    """计算核心评价指标"""
    total_ret = (1 + series).prod() - 1
    ann_ret = (1 + series).prod() ** (252/len(series)) - 1
    ann_vol = series.std() * np.sqrt(252)
    sharpe = (ann_ret - 0.02) / ann_vol # 假设无风险利率2%

    # 最大回撤
    cum = (1 + series).cumprod()
    dd = (cum / cum.cummax()) - 1
    max_dd = dd.min()

    return {
        '策略': name,
        '年化收益': f"{ann_ret:.1%}",
        '年化波动': f"{ann_vol:.1%}",
        '夏普比率': f"{sharpe:.2f}",
        '最大回撤': f"{max_dd:.1%}"
    }

# --- 主程序 ---
# 1. 获取数据
df_returns = generate_market_data()

# 2. 运行不同策略
# 风险平价
rp_ret, rp_weights = run_risk_parity_strategy(df_returns)

# 对比基准
stock_ret = df_returns['股票'].loc[rp_ret.index]
bond_ret = df_returns['债券'].loc[rp_ret.index]
combo_ret = stock_ret * 0.6 + bond_ret * 0.4 # 传统的60/40组合

# 3. 计算净值曲线
cum_rp = (1 + rp_ret).cumprod()
cum_stock = (1 + stock_ret).cumprod()
cum_bond = (1 + bond_ret).cumprod()
cum_6040 = (1 + combo_ret).cumprod()

# 4. 打印指标
metrics = [
    calculate_metrics(stock_ret, "纯股票"),
    calculate_metrics(bond_ret, "纯债券"),
    calculate_metrics(combo_ret, "股债60/40"),
    calculate_metrics(rp_ret, "★ 风险平价 (RP)"),
]
print(pd.DataFrame(metrics))

# 5. 绘图
plt.figure(figsize=(14, 8))

# 子图1：累计收益
ax1 = plt.subplot(2, 1, 1)
ax1.plot(cum_stock, label='纯股票 (高波高回撤)', color='gray', alpha=0.5, linestyle='--')
ax1.plot(cum_bond, label='纯债券 (低波低收益)', color='blue', alpha=0.6, linestyle=':')
ax1.plot(cum_rp, label='风险平价 (稳健增长)', color='#d62728', linewidth=2.5)
ax1.plot(cum_6040, label='60/40组合', color='green', alpha=0.6)

# 标记熊市区域
ax1.axvspan(df_returns.index[300], df_returns.index[450], color='gray', alpha=0.2, label='熊市区间')
ax1.set_title('策略累计收益对比：风险平价穿越牛熊', fontsize=14, fontweight='bold')
ax1.legend(loc='upper left')

# 子图2：动态权重变化
ax2 = plt.subplot(2, 1, 2)
ax2.stackplot(rp_weights.index, rp_weights['股票'], rp_weights['债券'], 
              labels=['股票权重', '债券权重'], colors=['#ff7f0e', '#1f77b4'], alpha=0.8)
ax2.set_title('风险平价策略的动态仓位调整', fontsize=12)
ax2.set_ylabel('资产权重')
ax2.legend(loc='center left')

plt.tight_layout()
plt.show()