{"id":1157815,"date":"2025-01-13T18:33:12","date_gmt":"2025-01-13T10:33:12","guid":{"rendered":"https:\/\/docs.pingcode.com\/ask\/ask-ask\/1157815.html"},"modified":"2025-01-13T18:33:12","modified_gmt":"2025-01-13T10:33:12","slug":"%e5%a6%82%e4%bd%95%e7%94%a8python%e7%94%bb%e5%87%bavar","status":"publish","type":"post","link":"https:\/\/docs.pingcode.com\/ask\/ask-ask\/1157815.html","title":{"rendered":"\u5982\u4f55\u7528python\u753b\u51faVaR"},"content":{"rendered":"

\"\u5982\u4f55\u7528python\u753b\u51faVaR\"<\/p>\n

\u8981\u7528Python\u753b\u51faVaR\uff08Value at Risk\uff09\uff0c\u4f60\u53ef\u4ee5\u4f7f\u7528\u7edf\u8ba1\u548c\u6570\u636e\u53ef\u89c6\u5316\u5e93\u3002\u9996\u5148\uff0c\u4f60\u9700\u8981\u7406\u89e3VaR\u7684\u5b9a\u4e49\uff0c\u5b83\u662f\u91d1\u878d\u98ce\u9669\u7ba1\u7406\u4e2d\u7528\u6765\u6d4b\u91cf\u6295\u8d44\u7ec4\u5408\u6f5c\u5728\u635f\u5931\u98ce\u9669\u7684\u6307\u6807\u3002VaR\u53ef\u4ee5\u901a\u8fc7\u5386\u53f2\u6a21\u62df\u6cd5\u3001\u65b9\u5dee-\u534f\u65b9\u5dee\u6cd5\u6216\u8499\u7279\u5361\u7f57\u6a21\u62df\u6cd5\u8ba1\u7b97\u3002\u63a5\u4e0b\u6765\uff0c\u6211\u5c06\u8be6\u7ec6\u4ecb\u7ecd\u5982\u4f55\u7528\u5386\u53f2\u6a21\u62df\u6cd5\u8ba1\u7b97\u548c\u7ed8\u5236VaR\u56fe\u3002<\/strong><\/p>\n<\/p>\n

\u5386\u53f2\u6a21\u62df\u6cd5\u3001\u5f15\u5165\u6570\u636e\u3001\u8ba1\u7b97VaR\u3001\u7ed8\u5236\u56fe\u5f62<\/strong>\u662f\u4f7f\u7528Python\u753b\u51faVaR\u7684\u6838\u5fc3\u6b65\u9aa4\u3002\u672c\u6587\u5c06\u8be6\u7ec6\u63cf\u8ff0\u5386\u53f2\u6a21\u62df\u6cd5\u8ba1\u7b97VaR\u53ca\u5176\u5b9e\u73b0\u6b65\u9aa4\u3002<\/p>\n<\/p>\n

\u4e00\u3001\u5f15\u5165\u6570\u636e<\/h3>\n<\/p>\n

\u4e3a\u4e86\u8ba1\u7b97VaR\uff0c\u6211\u4eec\u9700\u8981\u5386\u53f2\u4ef7\u683c\u6570\u636e\u3002\u53ef\u4ee5\u4f7f\u7528\u91d1\u878d\u6570\u636eAPI\uff08\u5982Yahoo Finance\uff09\u83b7\u53d6\u6570\u636e\u3002\u4e0b\u9762\u7684\u4ee3\u7801\u5c55\u793a\u4e86\u5982\u4f55\u4f7f\u7528yfinance<\/code>\u5e93\u83b7\u53d6\u6570\u636e\uff1a<\/p>\n<\/p>\n

import yfinance as yf<\/p>\n
import pandas as pd<\/p>\n
\u4e0b\u8f7d\u6570\u636e<\/strong><\/h2>\n
data = yf.download("AAPL", start="2021-01-01", end="2022-01-01")<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u8fd9\u6bb5\u4ee3\u7801\u4eceYahoo Finance\u4e0b\u8f7d\u4e86\u82f9\u679c\u516c\u53f8\uff08AAPL\uff09\u7684\u80a1\u7968\u6570\u636e\u3002<\/p>\n<\/p>\n
\u4e8c\u3001\u8ba1\u7b97\u65e5\u6536\u76ca\u7387<\/h3>\n<\/p>\n
\u63a5\u4e0b\u6765\uff0c\u6211\u4eec\u9700\u8981\u8ba1\u7b97\u6bcf\u65e5\u6536\u76ca\u7387\uff1a<\/p>\n<\/p>\n
# \u8ba1\u7b97\u6bcf\u65e5\u6536\u76ca\u7387<\/p>\n
data['Returns'] = data['Adj Close'].pct_change()<\/p>\n
data = data.dropna()  # \u5220\u9664\u7f3a\u5931\u503c<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u8fd9\u91cc\u7684pct_change()<\/code>\u65b9\u6cd5\u7528\u4e8e\u8ba1\u7b97\u76f8\u90bb\u4e24\u4e2a\u6536\u76d8\u4ef7\u4e4b\u95f4\u7684\u767e\u5206\u6bd4\u53d8\u5316\uff0c\u8868\u793a\u6bcf\u65e5\u6536\u76ca\u7387\u3002<\/p>\n<\/p>\n
\u4e09\u3001\u8ba1\u7b97VaR<\/h3>\n<\/p>\n
\u6211\u4eec\u5c06\u4f7f\u7528\u5386\u53f2\u6a21\u62df\u6cd5\u6765\u8ba1\u7b97VaR\uff0c\u5373\u4ece\u5386\u53f2\u6536\u76ca\u7387\u4e2d\u76f4\u63a5\u4f30\u8ba1\u51fa\u7279\u5b9a\u7f6e\u4fe1\u6c34\u5e73\u4e0b\u7684\u98ce\u9669\u503c\u3002<\/p>\n<\/p>\n
import numpy as np<\/p>\n
\u8ba1\u7b97VaR<\/strong><\/h2>\n
confidence_level = 0.95<\/p>\n
var = np.percentile(data['Returns'], (1 - confidence_level) * 100)<\/p>\n
print(f'VaR(95%): {var}')<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u5728\u8fd9\u6bb5\u4ee3\u7801\u4e2d\uff0c\u6211\u4eec\u4f7f\u7528np.percentile<\/code>\u51fd\u6570\u8ba1\u7b9795%\u7f6e\u4fe1\u6c34\u5e73\u4e0b\u7684VaR\u3002<\/p>\n<\/p>\n
\u56db\u3001\u7ed8\u5236VaR\u56fe<\/h3>\n<\/p>\n
\u6700\u540e\uff0c\u6211\u4eec\u5c06\u7ed8\u5236\u6536\u76ca\u7387\u5206\u5e03\u56fe\uff0c\u5e76\u5728\u56fe\u4e2d\u6807\u8bb0\u51faVaR\u503c\u3002<\/p>\n<\/p>\n
import matplotlib.pyplot as plt<\/p>\n
\u7ed8\u5236\u76f4\u65b9\u56fe<\/strong><\/h2>\n
plt.hist(data['Returns'], bins=50, alpha=0.75, color='blue', edgecolor='black')<\/p>\n
\u6807\u8bb0VaR<\/strong><\/h2>\n
plt.axvline(x=var, color='red', linestyle='--', linewidth=2, label=f'VaR(95%): {var:.2%}')<\/p>\n
plt.xlabel('DAI<\/a>ly Returns')<\/p>\n
plt.ylabel('Frequency')<\/p>\n
plt.title('Histogram of Daily Returns')<\/p>\n
plt.legend()<\/p>\n
plt.show()<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u8fd9\u6bb5\u4ee3\u7801\u4f7f\u7528matplotlib<\/code>\u5e93\u7ed8\u5236\u4e86\u6bcf\u65e5\u6536\u76ca\u7387\u7684\u76f4\u65b9\u56fe\uff0c\u5e76\u7528\u7ea2\u8272\u865a\u7ebf\u6807\u8bb0\u51faVaR\u503c\u3002<\/p>\n<\/p>\n
\u4e94\u3001\u603b\u7ed3<\/h3>\n<\/p>\n
\u901a\u8fc7\u4e0a\u8ff0\u6b65\u9aa4\uff0c\u6211\u4eec\u6210\u529f\u5730\u4f7f\u7528Python\u8ba1\u7b97\u5e76\u7ed8\u5236\u4e86VaR\u3002\u5386\u53f2\u6a21\u62df\u6cd5\u662f\u8ba1\u7b97VaR\u7684\u4e00\u79cd\u76f4\u63a5\u800c\u6709\u6548\u7684\u65b9\u6cd5\uff0c\u4f46\u4e5f\u5b58\u5728\u4e00\u4e9b\u5c40\u9650\u6027\uff0c\u4f8b\u5982\u5b83\u5047\u8bbe\u672a\u6765\u6536\u76ca\u7387\u5206\u5e03\u4e0e\u5386\u53f2\u76f8\u540c\u3002\u4e3a\u4e86\u63d0\u9ad8\u51c6\u786e\u6027\uff0c\u53ef\u4ee5\u8003\u8651\u7ed3\u5408\u5176\u4ed6\u65b9\u6cd5\u5982\u65b9\u5dee-\u534f\u65b9\u5dee\u6cd5\u548c\u8499\u7279\u5361\u7f57\u6a21\u62df\u6cd5\u3002<\/p>\n<\/p>\n
\u516d\u3001\u65b9\u5dee-\u534f\u65b9\u5dee\u6cd5\u8ba1\u7b97VaR<\/h3>\n<\/p>\n
\u65b9\u5dee-\u534f\u65b9\u5dee\u6cd5\u662f\u4e00\u79cd\u57fa\u4e8e\u5047\u8bbe\u6536\u76ca\u7387\u670d\u4ece\u6b63\u6001\u5206\u5e03\u7684VaR\u8ba1\u7b97\u65b9\u6cd5\u3002\u5b83\u5229\u7528\u6536\u76ca\u7387\u7684\u5747\u503c\u548c\u6807\u51c6\u5dee\u6765\u4f30\u8ba1VaR\u3002\u4e0b\u9762\u4ecb\u7ecd\u5982\u4f55\u4f7f\u7528\u8fd9\u79cd\u65b9\u6cd5\u8ba1\u7b97VaR\u3002<\/p>\n<\/p>\n
1\u3001\u8ba1\u7b97\u5747\u503c\u548c\u6807\u51c6\u5dee<\/h4>\n<\/p>\n
\u9996\u5148\uff0c\u6211\u4eec\u9700\u8981\u8ba1\u7b97\u6536\u76ca\u7387\u7684\u5747\u503c\u548c\u6807\u51c6\u5dee\uff1a<\/p>\n<\/p>\n
mean = data['Returns'].mean()<\/p>\n
std_dev = data['Returns'].std()<\/p>\n
<\/code><\/pre>\n<\/p>\n
2\u3001\u8ba1\u7b97VaR<\/h4>\n<\/p>\n
\u63a5\u4e0b\u6765\uff0c\u57fa\u4e8e\u6b63\u6001\u5206\u5e03\u5047\u8bbe\u8ba1\u7b97VaR\uff1a<\/p>\n<\/p>\n
from scipy.stats import norm<\/p>\n
\u8ba1\u7b97VaR<\/strong><\/h2>\n
var_95 = norm.ppf(1 - confidence_level) * std_dev + mean<\/p>\n
print(f'VaR(95%): {var_95}')<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u5728\u8fd9\u6bb5\u4ee3\u7801\u4e2d\uff0c\u6211\u4eec\u4f7f\u7528scipy.stats<\/code>\u5e93\u7684norm.ppf<\/code>\u51fd\u6570\u8ba1\u7b97\u6b63\u6001\u5206\u5e03\u7684\u5206\u4f4d\u6570\u3002<\/p>\n<\/p>\n
\u4e03\u3001\u8499\u7279\u5361\u7f57\u6a21\u62df\u6cd5\u8ba1\u7b97VaR<\/h3>\n<\/p>\n
\u8499\u7279\u5361\u7f57\u6a21\u62df\u6cd5\u901a\u8fc7\u6a21\u62df\u5927\u91cf\u672a\u6765\u53ef\u80fd\u7684\u6536\u76ca\u7387\u8def\u5f84\u6765\u4f30\u8ba1VaR\u3002\u5b83\u4e0d\u4f9d\u8d56\u4e8e\u6536\u76ca\u7387\u7684\u5206\u5e03\u5047\u8bbe\uff0c\u56e0\u6b64\u66f4\u52a0\u7075\u6d3b\u3002\u4e0b\u9762\u4ecb\u7ecd\u5982\u4f55\u5b9e\u73b0\u8fd9\u79cd\u65b9\u6cd5\u3002<\/p>\n<\/p>\n
1\u3001\u751f\u6210\u6a21\u62df\u8def\u5f84<\/h4>\n<\/p>\n
\u9996\u5148\uff0c\u6211\u4eec\u751f\u6210\u672a\u6765\u6536\u76ca\u7387\u7684\u6a21\u62df\u8def\u5f84\uff1a<\/p>\n<\/p>\n
num_simulations = 10000<\/p>\n
simulation_days = 252<\/p>\n
\u6a21\u62df\u672a\u6765\u6536\u76ca\u7387<\/strong><\/h2>\n
simulated_returns = np.random.normal(mean, std_dev, (simulation_days, num_simulations))<\/p>\n
<\/code><\/pre>\n<\/p>\n
2\u3001\u8ba1\u7b97\u672a\u6765\u4ef7\u683c\u8def\u5f84<\/h4>\n<\/p>\n
\u63a5\u4e0b\u6765\uff0c\u57fa\u4e8e\u6a21\u62df\u7684\u6536\u76ca\u7387\u8ba1\u7b97\u672a\u6765\u4ef7\u683c\u8def\u5f84\uff1a<\/p>\n<\/p>\n
# \u521d\u59cb\u4ef7\u683c<\/p>\n
initial_price = data['Adj Close'][-1]<\/p>\n
\u8ba1\u7b97\u672a\u6765\u4ef7\u683c\u8def\u5f84<\/strong><\/h2>\n
price_paths = initial_price * np.exp(np.cumsum(simulated_returns, axis=0))<\/p>\n
<\/code><\/pre>\n<\/p>\n
3\u3001\u8ba1\u7b97VaR<\/h4>\n<\/p>\n
\u6700\u540e\uff0c\u57fa\u4e8e\u6a21\u62df\u7684\u4ef7\u683c\u8def\u5f84\u8ba1\u7b97VaR\uff1a<\/p>\n<\/p>\n
# \u8ba1\u7b97\u672b\u671f\u4ef7\u683c\u5206\u5e03<\/p>\n
ending_prices = price_paths[-1, :]<\/p>\n
\u8ba1\u7b97\u6536\u76ca\u7387\u5206\u5e03<\/strong><\/h2>\n
ending_returns = ending_prices \/ initial_price - 1<\/p>\n
\u8ba1\u7b97VaR<\/strong><\/h2>\n
var_mc = np.percentile(ending_returns, (1 - confidence_level) * 100)<\/p>\n
print(f'VaR(95%): {var_mc}')<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u516b\u3001\u6bd4\u8f83\u4e0d\u540c\u65b9\u6cd5\u7684VaR\u7ed3\u679c<\/h3>\n<\/p>\n
\u901a\u8fc7\u4e0a\u8ff0\u6b65\u9aa4\uff0c\u6211\u4eec\u5206\u522b\u4f7f\u7528\u5386\u53f2\u6a21\u62df\u6cd5\u3001\u65b9\u5dee-\u534f\u65b9\u5dee\u6cd5\u548c\u8499\u7279\u5361\u7f57\u6a21\u62df\u6cd5\u8ba1\u7b97\u4e86VaR\u3002\u53ef\u4ee5\u5bf9\u6bd4\u4e0d\u540c\u65b9\u6cd5\u8ba1\u7b97\u51fa\u7684VaR\u503c\uff0c\u4ee5\u8bc4\u4f30\u5176\u5dee\u5f02\u548c\u9002\u7528\u6027\u3002<\/p>\n<\/p>\n
print(f'Historical VaR(95%): {var}')<\/p>\n
print(f'Parametric VaR(95%): {var_95}')<\/p>\n
print(f'Monte Carlo VaR(95%): {var_mc}')<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u4e5d\u3001\u7efc\u5408\u8bc4\u4f30VaR\u65b9\u6cd5<\/h3>\n<\/p>\n
\u4e0d\u540c\u7684VaR\u8ba1\u7b97\u65b9\u6cd5\u5404\u6709\u4f18\u7f3a\u70b9\u3002\u5386\u53f2\u6a21\u62df\u6cd5<\/strong>\u7b80\u5355\u76f4\u63a5\uff0c\u4f46\u4f9d\u8d56\u4e8e\u8fc7\u53bb\u7684\u6536\u76ca\u7387\u5206\u5e03\uff1b\u65b9\u5dee-\u534f\u65b9\u5dee\u6cd5<\/strong>\u5047\u8bbe\u6536\u76ca\u7387\u670d\u4ece\u6b63\u6001\u5206\u5e03\uff0c\u8ba1\u7b97\u7b80\u4fbf\uff0c\u4f46\u5bf9\u6781\u7aef\u503c\u4e0d\u654f\u611f\uff1b\u8499\u7279\u5361\u7f57\u6a21\u62df\u6cd5<\/strong>\u7075\u6d3b\u4e14\u4e0d\u4f9d\u8d56\u5206\u5e03\u5047\u8bbe\uff0c\u4f46\u8ba1\u7b97\u590d\u6742\u4e14\u5bf9\u6a21\u62df\u6b21\u6570\u654f\u611f\u3002<\/p>\n<\/p>\n
\u5341\u3001\u98ce\u9669\u7ba1\u7406\u4e2d\u7684VaR\u5e94\u7528<\/h3>\n<\/p>\n
VaR\u4e0d\u4ec5\u7528\u4e8e\u8ba1\u7b97\u6f5c\u5728\u635f\u5931\uff0c\u8fd8\u53ef\u4ee5\u7528\u4e8e\u8bbe\u5b9a\u98ce\u9669\u9650\u989d\u3001\u8bc4\u4f30\u8d44\u672c\u9700\u6c42\u548c\u5236\u5b9a\u5bf9\u51b2\u7b56\u7565\u3002\u5728\u5b9e\u9645\u5e94\u7528\u4e2d\uff0c\u91d1\u878d\u673a\u6784\u901a\u5e38\u7ed3\u5408\u4f7f\u7528\u591a\u79cd\u65b9\u6cd5\uff0c\u4ee5\u5168\u9762\u8bc4\u4f30\u98ce\u9669\u5e76\u5236\u5b9a\u76f8\u5e94\u7684\u98ce\u9669\u7ba1\u7406\u63aa\u65bd\u3002<\/p>\n<\/p>\n
\u5341\u4e00\u3001\u98ce\u9669\u9650\u989d\u8bbe\u5b9a<\/h3>\n<\/p>\n
\u901a\u8fc7VaR\uff0c\u91d1\u878d\u673a\u6784\u53ef\u4ee5\u8bbe\u5b9a\u98ce\u9669\u9650\u989d\uff0c\u4ee5\u63a7\u5236\u5728\u7279\u5b9a\u7f6e\u4fe1\u6c34\u5e73\u4e0b\u7684\u6f5c\u5728\u635f\u5931\u3002\u4f8b\u5982\uff0c\u53ef\u4ee5\u8bbe\u5b9a\u67d0\u6295\u8d44\u7ec4\u5408\u7684\u65e5VaR\u4e0d\u8d85\u8fc7\u4e00\u5b9a\u91d1\u989d\uff0c\u4ece\u800c\u9650\u5236\u6295\u8d44\u98ce\u9669\u3002<\/p>\n<\/p>\n
\u5341\u4e8c\u3001\u8bc4\u4f30\u8d44\u672c\u9700\u6c42<\/h3>\n<\/p>\n
VaR\u8fd8\u53ef\u4ee5\u7528\u4e8e\u8bc4\u4f30\u8d44\u672c\u9700\u6c42\uff0c\u786e\u4fdd\u91d1\u878d\u673a\u6784\u6709\u8db3\u591f\u7684\u8d44\u672c\u5e94\u5bf9\u6f5c\u5728\u635f\u5931\u3002\u76d1\u7ba1\u673a\u6784\u901a\u5e38\u8981\u6c42\u91d1\u878d\u673a\u6784\u6839\u636eVaR\u8ba1\u7b97\u8d44\u672c\u5145\u8db3\u7387\uff0c\u4ee5\u7ef4\u62a4\u91d1\u878d\u7a33\u5b9a\u3002<\/p>\n<\/p>\n
\u5341\u4e09\u3001\u5236\u5b9a\u5bf9\u51b2\u7b56\u7565<\/h3>\n<\/p>\n
\u901a\u8fc7VaR\uff0c\u6295\u8d44\u8005\u53ef\u4ee5\u5236\u5b9a\u5bf9\u51b2\u7b56\u7565\uff0c\u4ee5\u51cf\u5c11\u6f5c\u5728\u635f\u5931\u3002\u4f8b\u5982\uff0c\u53ef\u4ee5\u901a\u8fc7\u671f\u6743\u3001\u671f\u8d27\u7b49\u884d\u751f\u54c1\u5bf9\u51b2\u98ce\u9669\uff0c\u4ece\u800c\u964d\u4f4eVaR\u503c\u3002<\/p>\n<\/p>\n
\u5341\u56db\u3001VaR\u7684\u5c40\u9650\u6027<\/h3>\n<\/p>\n
\u5c3d\u7ba1VaR\u5728\u98ce\u9669\u7ba1\u7406\u4e2d\u5e7f\u6cdb\u5e94\u7528\uff0c\u4f46\u5b83\u4e5f\u5b58\u5728\u4e00\u4e9b\u5c40\u9650\u6027\u3002\u9996\u5148\uff0cVaR\u5ffd\u7565\u4e86\u8d85\u8fc7\u7f6e\u4fe1\u6c34\u5e73\u7684\u6781\u7aef\u635f\u5931\uff0c\u5373\u5c3e\u90e8\u98ce\u9669\u3002\u5176\u6b21\uff0cVaR\u5047\u8bbe\u5e02\u573a\u6761\u4ef6\u7a33\u5b9a\uff0c\u5ffd\u7565\u4e86\u5e02\u573a\u5267\u53d8\u7684\u53ef\u80fd\u6027\u3002\u6b64\u5916\uff0cVaR\u4f9d\u8d56\u5386\u53f2\u6570\u636e\uff0c\u53ef\u80fd\u4f4e\u4f30\u672a\u6765\u98ce\u9669\u3002<\/p>\n<\/p>\n
\u5341\u4e94\u3001\u6539\u8fdbVaR\u7684\u63aa\u65bd<\/h3>\n<\/p>\n
\u4e3a\u4e86\u5f25\u8865VaR\u7684\u5c40\u9650\u6027\uff0c\u53ef\u4ee5\u7ed3\u5408\u4f7f\u7528\u5176\u4ed6\u98ce\u9669\u6307\u6807\uff0c\u5982\u9884\u671f\u77ed\u7f3a\uff08Expected Shortfall\uff09\u548c\u6761\u4ef6VaR\uff08Conditional VaR\uff09\u3002\u8fd9\u4e9b\u6307\u6807\u8003\u8651\u4e86\u8d85\u51faVaR\u7684\u6781\u7aef\u635f\u5931\uff0c\u66f4\u5168\u9762\u5730\u8bc4\u4f30\u98ce\u9669\u3002<\/p>\n<\/p>\n
\u5341\u516d\u3001\u4f7f\u7528Python\u8ba1\u7b97\u548c\u7ed8\u5236\u9884\u671f\u77ed\u7f3a<\/h3>\n<\/p>\n
\u9884\u671f\u77ed\u7f3a\uff08ES\uff09\u662fVaR\u7684\u6539\u8fdb\u7248\uff0c\u5b83\u6d4b\u91cf\u8d85\u8fc7VaR\u7684\u5e73\u5747\u635f\u5931\u3002\u4e0b\u9762\u4ecb\u7ecd\u5982\u4f55\u4f7f\u7528Python\u8ba1\u7b97\u548c\u7ed8\u5236ES\u3002<\/p>\n<\/p>\n
1\u3001\u8ba1\u7b97ES<\/h4>\n<\/p>\n
\u57fa\u4e8e\u5386\u53f2\u6536\u76ca\u7387\u8ba1\u7b97ES\uff1a<\/p>\n<\/p>\n
# \u8ba1\u7b97ES<\/p>\n
es = data['Returns'][data['Returns'] <= var].mean()<\/p>\n
print(f'ES(95%): {es}')<\/p>\n
<\/code><\/pre>\n<\/p>\n
2\u3001\u7ed8\u5236ES\u56fe<\/h4>\n<\/p>\n
\u5728\u6536\u76ca\u7387\u5206\u5e03\u56fe\u4e2d\u6807\u8bb0ES\u503c\uff1a<\/p>\n<\/p>\n
# \u7ed8\u5236\u76f4\u65b9\u56fe<\/p>\n
plt.hist(data['Returns'], bins=50, alpha=0.75, color='blue', edgecolor='black')<\/p>\n
\u6807\u8bb0VaR\u548cES<\/strong><\/h2>\n
plt.axvline(x=var, color='red', linestyle='--', linewidth=2, label=f'VaR(95%): {var:.2%}')<\/p>\n
plt.axvline(x=es, color='green', linestyle='--', linewidth=2, label=f'ES(95%): {es:.2%}')<\/p>\n
plt.xlabel('Daily Returns')<\/p>\n
plt.ylabel('Frequency')<\/p>\n
plt.title('Histogram of Daily Returns')<\/p>\n
plt.legend()<\/p>\n
plt.show()<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u5341\u4e03\u3001\u6761\u4ef6VaR\u7684\u8ba1\u7b97<\/h3>\n<\/p>\n
\u6761\u4ef6VaR\uff08CVaR\uff09\u662f\u53e6\u4e00\u79cd\u8bc4\u4f30\u6781\u7aef\u98ce\u9669\u7684\u6307\u6807\uff0c\u8868\u793a\u5728\u8d85\u8fc7VaR\u7684\u6761\u4ef6\u4e0b\u7684\u5e73\u5747\u635f\u5931\u3002\u53ef\u4ee5\u901a\u8fc7\u4ee5\u4e0b\u65b9\u6cd5\u8ba1\u7b97CVaR\uff1a<\/p>\n<\/p>\n
# \u8ba1\u7b97CVaR<\/p>\n
cvar = data['Returns'][data['Returns'] <= var].mean()<\/p>\n
print(f'CVaR(95%): {cvar}')<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u5341\u516b\u3001\u7ed3\u5408\u591a\u79cd\u6307\u6807\u8fdb\u884c\u98ce\u9669\u8bc4\u4f30<\/h3>\n<\/p>\n
\u5728\u5b9e\u9645\u5e94\u7528\u4e2d\uff0c\u53ef\u4ee5\u7ed3\u5408\u4f7f\u7528VaR\u3001ES\u548cCVaR\u7b49\u591a\u79cd\u6307\u6807\uff0c\u5168\u9762\u8bc4\u4f30\u98ce\u9669\u3002\u8fd9\u4e9b\u6307\u6807\u4e92\u4e3a\u8865\u5145\uff0c\u53ef\u4ee5\u5e2e\u52a9\u91d1\u878d\u673a\u6784\u66f4\u51c6\u786e\u5730\u8bc6\u522b\u548c\u7ba1\u7406\u98ce\u9669\u3002<\/p>\n<\/p>\n
\u5341\u4e5d\u3001VaR\u5728\u4e0d\u540c\u5e02\u573a\u7684\u5e94\u7528<\/h3>\n<\/p>\n
VaR\u4e0d\u4ec5\u9002\u7528\u4e8e\u80a1\u7968\u5e02\u573a\uff0c\u8fd8\u53ef\u4ee5\u5e94\u7528\u4e8e\u503a\u5238\u3001\u5916\u6c47\u3001\u5927\u5b97\u5546\u54c1\u7b49\u4e0d\u540c\u5e02\u573a\u3002\u9488\u5bf9\u4e0d\u540c\u5e02\u573a\u7684\u7279\u70b9\uff0c\u53ef\u4ee5\u8c03\u6574VaR\u7684\u8ba1\u7b97\u65b9\u6cd5\u548c\u53c2\u6570\uff0c\u4ee5\u63d0\u9ad8\u51c6\u786e\u6027\u3002<\/p>\n<\/p>\n
\u4e8c\u5341\u3001\u672a\u6765VaR\u7684\u53d1\u5c55\u65b9\u5411<\/h3>\n<\/p>\n
\u968f\u7740\u91d1\u878d\u5e02\u573a\u7684\u4e0d\u65ad\u53d1\u5c55\uff0cVaR\u7684\u8ba1\u7b97\u65b9\u6cd5\u548c\u5e94\u7528\u4e5f\u5728\u4e0d\u65ad\u6f14\u8fdb\u3002\u672a\u6765\uff0c\u7ed3\u5408\u673a\u5668\u5b66\u4e60<\/a>\u548c\u5927\u6570\u636e\u6280\u672f\u7684VaR\u8ba1\u7b97\u65b9\u6cd5\u5c06\u8d8a\u6765\u8d8a\u53d7\u5230\u5173\u6ce8\u3002\u8fd9\u4e9b\u65b0\u6280\u672f\u53ef\u4ee5\u66f4\u51c6\u786e\u5730\u9884\u6d4b\u98ce\u9669\uff0c\u5e2e\u52a9\u91d1\u878d\u673a\u6784\u5236\u5b9a\u66f4\u6709\u6548\u7684\u98ce\u9669\u7ba1\u7406\u7b56\u7565\u3002<\/p>\n<\/p>\n
\u4e8c\u5341\u4e00\u3001\u673a\u5668\u5b66\u4e60\u5728VaR\u8ba1\u7b97\u4e2d\u7684\u5e94\u7528<\/h3>\n<\/p>\n
\u673a\u5668\u5b66\u4e60\u6280\u672f\u53ef\u4ee5\u7528\u4e8e\u9884\u6d4b\u672a\u6765\u6536\u76ca\u7387\u5206\u5e03\uff0c\u4ece\u800c\u63d0\u9ad8VaR\u7684\u51c6\u786e\u6027\u3002\u4f8b\u5982\uff0c\u53ef\u4ee5\u4f7f\u7528\u56de\u5f52\u6a21\u578b\u9884\u6d4b\u6536\u76ca\u7387\uff0c\u6216\u8005\u4f7f\u7528\u6df1\u5ea6\u5b66\u4e60\u6a21\u578b\u6a21\u62df\u590d\u6742\u7684\u5e02\u573a\u884c\u4e3a\u3002<\/p>\n<\/p>\n
\u4e8c\u5341\u4e8c\u3001\u4f7f\u7528Python\u5b9e\u73b0\u673a\u5668\u5b66\u4e60VaR<\/h3>\n<\/p>\n
\u4e0b\u9762\u4ecb\u7ecd\u5982\u4f55\u4f7f\u7528Python\u548c\u673a\u5668\u5b66\u4e60\u6280\u672f\u8ba1\u7b97VaR\u3002<\/p>\n<\/p>\n
1\u3001\u6570\u636e\u9884\u5904\u7406<\/h4>\n<\/p>\n
\u9996\u5148\uff0c\u51c6\u5907\u6570\u636e\u5e76\u8fdb\u884c\u9884\u5904\u7406\uff1a<\/p>\n<\/p>\n
from sklearn.preprocessing import StandardScaler<\/p>\n
from sklearn.model_selection import train_test_split<\/p>\n
\u63d0\u53d6\u7279\u5f81\u548c\u6807\u7b7e<\/strong><\/h2>\n
X = data.drop(columns=['Returns'])<\/p>\n
y = data['Returns']<\/p>\n
\u6807\u51c6\u5316\u6570\u636e<\/strong><\/h2>\n
scaler = StandardScaler()<\/p>\n
X_scaled = scaler.fit_transform(X)<\/p>\n
\u5212\u5206\u8bad\u7ec3\u96c6\u548c\u6d4b\u8bd5\u96c6<\/strong><\/h2>\n
X_train, X_test, y_train, y_test = train_test_split(X_scaled, y, test_size=0.2, random_state=42)<\/p>\n
<\/code><\/pre>\n<\/p>\n
2\u3001\u8bad\u7ec3\u56de\u5f52\u6a21\u578b<\/h4>\n<\/p>\n
\u63a5\u4e0b\u6765\uff0c\u8bad\u7ec3\u4e00\u4e2a\u56de\u5f52\u6a21\u578b\u9884\u6d4b\u6536\u76ca\u7387\uff1a<\/p>\n<\/p>\n
from sklearn.linear_model import LinearRegression<\/p>\n
\u8bad\u7ec3\u56de\u5f52\u6a21\u578b<\/strong><\/h2>\n
model = LinearRegression()<\/p>\n
model.fit(X_train, y_train)<\/p>\n
\u9884\u6d4b\u6536\u76ca\u7387<\/strong><\/h2>\n
y_pred = model.predict(X_test)<\/p>\n
<\/code><\/pre>\n<\/p>\n
3\u3001\u8ba1\u7b97VaR<\/h4>\n<\/p>\n
\u57fa\u4e8e\u9884\u6d4b\u7684\u6536\u76ca\u7387\u8ba1\u7b97VaR\uff1a<\/p>\n<\/p>\n
# \u8ba1\u7b97\u9884\u6d4b\u6536\u76ca\u7387\u7684VaR<\/p>\n
var_ml = np.percentile(y_pred, (1 - confidence_level) * 100)<\/p>\n
print(f'Machine Learning VaR(95%): {var_ml}')<\/p>\n
<\/code><\/pre>\n<\/p>\n
\u4e8c\u5341\u4e09\u3001\u603b\u7ed3\u4e0e\u5c55\u671b<\/h3>\n<\/p>\n
\u901a\u8fc7\u672c\u6587\u7684\u4ecb\u7ecd\uff0c\u6211\u4eec\u8be6\u7ec6\u63cf\u8ff0\u4e86\u5982\u4f55\u4f7f\u7528Python\u8ba1\u7b97\u548c\u7ed8\u5236VaR\uff0c\u5305\u62ec\u5386\u53f2\u6a21\u62df\u6cd5\u3001\u65b9\u5dee-\u534f\u65b9\u5dee\u6cd5\u548c\u8499\u7279\u5361\u7f57\u6a21\u62df\u6cd5\u7b49\u591a\u79cd\u65b9\u6cd5\u3002\u6b64\u5916\uff0c\u6211\u4eec\u8fd8\u63a2\u8ba8\u4e86\u9884\u671f\u77ed\u7f3a\u3001\u6761\u4ef6VaR\u548c\u673a\u5668\u5b66\u4e60\u5728VaR\u8ba1\u7b97\u4e2d\u7684\u5e94\u7528\u3002\u672a\u6765\uff0c\u968f\u7740\u6280\u672f\u7684\u4e0d\u65ad\u8fdb\u6b65\uff0cVaR\u7684\u8ba1\u7b97\u65b9\u6cd5\u548c\u5e94\u7528\u5c06\u66f4\u52a0\u591a\u6837\u5316\u548c\u667a\u80fd\u5316\uff0c\u4e3a\u91d1\u878d\u98ce\u9669\u7ba1\u7406\u63d0\u4f9b\u66f4\u52a0\u7cbe\u51c6\u548c\u5168\u9762\u7684\u652f\u6301\u3002<\/p>\n<\/p>\n
\u76f8\u5173\u95ee\u7b54FAQs\uff1a<\/strong><\/h2>\n
 \u4f7f\u7528Python\u7ed8\u5236VaR\u9700\u8981\u54ea\u4e9b\u6b65\u9aa4\uff1f<\/strong>
\u7ed8\u5236VaR\uff08Value at Risk\uff09\u901a\u5e38\u9700\u8981\u51e0\u4e2a\u6b65\u9aa4\u3002\u9996\u5148\uff0c\u60a8\u9700\u8981\u51c6\u5907\u597d\u5386\u53f2\u6570\u636e\uff0c\u901a\u5e38\u662f\u8d44\u4ea7\u7684\u6536\u76ca\u7387\u3002\u63a5\u4e0b\u6765\uff0c\u8ba1\u7b97\u8fd9\u4e9b\u6536\u76ca\u7387\u7684\u5206\u5e03\uff0c\u60a8\u53ef\u4ee5\u4f7f\u7528\u7edf\u8ba1\u65b9\u6cd5\u5982\u6b63\u6001\u5206\u5e03\u6216\u5386\u53f2\u6a21\u62df\u65b9\u6cd5\u3002\u7136\u540e\uff0c\u786e\u5b9a\u60a8\u5e0c\u671b\u4f7f\u7528\u7684\u7f6e\u4fe1\u6c34\u5e73\uff0c\u4f8b\u598295%\u621699%\u3002\u6700\u540e\uff0c\u4f7f\u7528Matplotlib\u7b49\u53ef\u89c6\u5316\u5e93\u7ed8\u5236VaR\u66f2\u7ebf\uff0c\u663e\u793a\u4e0d\u540c\u7f6e\u4fe1\u6c34\u5e73\u4e0b\u7684\u98ce\u9669\u503c\u3002<\/p>\n
\u5728Python\u4e2d\u6709\u54ea\u4e9b\u5e93\u53ef\u4ee5\u7528\u6765\u8ba1\u7b97VaR\uff1f<\/strong>
\u5728Python\u4e2d\uff0c\u60a8\u53ef\u4ee5\u4f7f\u7528\u591a\u4e2a\u5e93\u6765\u8ba1\u7b97VaR\u3002\u5e38\u7528\u7684\u5e93\u5305\u62ecNumPy\u548cPandas\uff0c\u7528\u4e8e\u6570\u636e\u5904\u7406\u548c\u8ba1\u7b97\u6536\u76ca\u7387\uff0cSciPy\u7528\u4e8e\u7edf\u8ba1\u5206\u6790\uff0c\u800cMatplotlib\u548cSeaborn\u5219\u53ef\u7528\u4e8e\u6570\u636e\u53ef\u89c6\u5316\u3002\u6b64\u5916\uff0c\u4e13\u95e8\u7684\u91d1\u878d\u5206\u6790\u5e93\u5982QuantLib\u548cPyPortfolioOpt\u4e5f\u63d0\u4f9b\u4e86VaR\u8ba1\u7b97\u7684\u529f\u80fd\uff0c\u80fd\u591f\u5e2e\u52a9\u60a8\u66f4\u9ad8\u6548\u5730\u8fdb\u884c\u98ce\u9669\u7ba1\u7406\u3002<\/p>\n
VaR\u7684\u8ba1\u7b97\u65b9\u6cd5\u6709\u54ea\u4e9b\uff1f<\/strong>
VaR\u7684\u8ba1\u7b97\u65b9\u6cd5\u6709\u51e0\u79cd\uff0c\u6700\u5e38\u89c1\u7684\u5305\u62ec\u5386\u53f2\u6a21\u62df\u6cd5\u3001\u65b9\u5dee-\u534f\u65b9\u5dee\u6cd5\u548c\u8499\u7279\u5361\u7f57\u6a21\u62df\u3002\u5386\u53f2\u6a21\u62df\u6cd5\u57fa\u4e8e\u8fc7\u53bb\u7684\u5e02\u573a\u6570\u636e\uff0c\u76f4\u63a5\u8ba1\u7b97\u635f\u5931\u5206\u5e03\uff1b\u65b9\u5dee-\u534f\u65b9\u5dee\u6cd5\u5047\u8bbe\u6536\u76ca\u7387\u5448\u6b63\u6001\u5206\u5e03\uff0c\u5229\u7528\u5747\u503c\u548c\u6807\u51c6\u5dee\u6765\u4f30\u8ba1VaR\uff1b\u8499\u7279\u5361\u7f57\u6a21\u62df\u5219\u901a\u8fc7\u968f\u673a\u751f\u6210\u5927\u91cf\u53ef\u80fd\u7684\u5e02\u573a\u60c5\u666f\u6765\u8ba1\u7b97\u6f5c\u5728\u635f\u5931\u3002\u8fd9\u4e9b\u65b9\u6cd5\u5404\u6709\u4f18\u7f3a\u70b9\uff0c\u9009\u62e9\u9002\u5408\u81ea\u5df1\u9700\u6c42\u7684\u65b9\u6cd5\u975e\u5e38\u91cd\u8981\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"\u8981\u7528Python\u753b\u51faVaR\uff08Value at Risk\uff09\uff0c\u4f60\u53ef\u4ee5\u4f7f\u7528\u7edf\u8ba1\u548c\u6570\u636e\u53ef\u89c6\u5316\u5e93\u3002\u9996\u5148\uff0c\u4f60\u9700\u8981\u7406\u89e3VaR […]","protected":false},"author":3,"featured_media":0,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[1],"tags":[],"acf":[],"_links":{"self":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/1157815"}],"collection":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/comments?post=1157815"}],"version-history":[{"count":0,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/1157815\/revisions"}],"wp:attachment":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/media?parent=1157815"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/categories?post=1157815"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/tags?post=1157815"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}