如何使用Python解决风险平价分配

import pandas as pd
import numpy as np
from scipy.optimize import minimize
import matplotlib.pyplot as plt

# 读取数据
data = pd.read_csv('portfolio_returns.csv')

# 提取收益率数据
returns = data.iloc[:, 1:].values

# 计算协方差矩阵
cov_matrix = np.cov(returns.T)

# 定义风险平价函数
def risk_parity(weights, cov_matrix):
    portfolio_variance = np.dot(weights.T, np.dot(cov_matrix, weights))
    risk_contribution = np.dot(cov_matrix, weights) / np.sqrt(portfolio_variance)
    risk_parity_error = risk_contribution - 1/len(weights)
    return np.sum(np.square(risk_parity_error))

# 定义约束条件
constraints = ({'type': 'eq', 'fun': lambda x: np.sum(x) - 1})

# 定义初始权重
init_weights = np.ones(returns.shape[1]) / returns.shape[1]

# 优化权重分配
result = minimize(risk_parity, init_weights, args=(cov_matrix,), method='SLSQP', constraints=constraints)

# 输出优化结果
weights = result.x
print("优化后的权重分配：", weights)

# 可视化展示
plt.bar(range(len(weights)), weights)
plt.xticks(range(len(weights)), data.columns[1:], rotation=45)
plt.xlabel('资产')
plt.ylabel('权重')
plt.title('风险平价分配')
plt.show()