{"id":3208,"date":"2025-12-18T07:31:00","date_gmt":"2025-12-17T23:31:00","guid":{"rendered":"https:\/\/www.laoyulaoyu.com\/?p=3208"},"modified":"2025-12-09T20:50:06","modified_gmt":"2025-12-09T12:50:06","slug":"%e5%88%ab%e5%86%8d%e5%8f%aa%e7%9c%8b%e6%81%90%e6%85%8c%e6%8c%87%e6%95%b0%e4%ba%86%ef%bc%81%e9%87%8f%e5%8c%96%e4%bb%a3%e7%a0%81%e8%ae%a9%e4%bd%a0%e4%bb%8evix%e6%9b%b2%e7%ba%bf%e7%9a%84","status":"publish","type":"post","link":"https:\/\/laoyulaoyu.com\/index.php\/2025\/12\/18\/%e5%88%ab%e5%86%8d%e5%8f%aa%e7%9c%8b%e6%81%90%e6%85%8c%e6%8c%87%e6%95%b0%e4%ba%86%ef%bc%81%e9%87%8f%e5%8c%96%e4%bb%a3%e7%a0%81%e8%ae%a9%e4%bd%a0%e4%bb%8evix%e6%9b%b2%e7%ba%bf%e7%9a%84\/","title":{"rendered":"\u522b\u518d\u53ea\u770b\u6050\u614c\u6307\u6570\u4e86\uff01\u91cf\u5316\u4ee3\u7801\u8ba9\u4f60\u4eceVIX\u66f2\u7ebf\u7684\u201c\u5761\u5ea6\u201d\u548c\u201c\u5f27\u5ea6\u201d\u770b\u7a7f\u725b\u718a"},"content":{"rendered":"\n

\u4f5c\u8005\uff1a\u8001\u4f59\u635e\u9c7c<\/a><\/p>\n\n\n\n

\u539f\u521b\u4e0d\u6613\uff0c\u8f6c\u8f7d\u8bf7\u6807\u660e\u51fa\u5904\u53ca\u539f\u4f5c\u8005\u3002<\/mark><\/strong><\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n
\n
\u5199\u5728\u524d\u9762\u7684\u8bdd\uff1a<\/strong>\u522b\u4eba\u6050\u614c\u6211\u8d2a\u5a6a\uff0c\u4f46\u524d\u63d0\u662f\u4f60\u5f97\u770b\u61c2\u6050\u614c\u7684\u201c\u5f62\u72b6\u201d\u3002\u522b\u518d\u50bb\u50bb\u53ea\u76ef\u4e00\u4e2aVIX\u6307\u6570\u4e86\uff0c\u90a3\u90fd\u662f\u5c0f\u513f\u79d1\u3002\u4eca\u5929\u6211\u628a\u91cf\u5316\u57fa\u91d1\u5e38\u7528\u7684VIX\u671f\u9650\u7ed3\u6784\u6a21\u578b\uff0c\u63b0\u788e\u4e86\u8bb2\u7ed9\u4f60\u542c\uff0c\u8fd8\u9644\u4e0a\u4e86\u5f00\u7bb1\u5373\u7528\u7684\u4ee3\u7801\u3002\u8fd9\u7bc7\u6587\u7ae0\uff0c\u5c31\u662f\u60f3\u8ba9\u4f60\u6bd480%\u7684\u6563\u6237\uff0c\u66f4\u65e9\u4e00\u6b65\u55c5\u5230\u5371\u9669\u548c\u673a\u4f1a\u3002<\/mark><\/pre>\n<\/blockquote>\n\n\n\n
1. \u4e3a\u4ec0\u4e48\u53ea\u770bVIX\u70b9\u4f4d\u8fdc\u8fdc\u4e0d\u591f<\/strong><\/h2>\n\n\n\n
VIX\u6307\u6570\uff08CBOE Volatility Index\uff09\u5e38\u88ab\u79f0\u4f5c\u201c\u6050\u60e7\u6307\u6570\u201d\uff0c\u5b83\u53cd\u6620\u6807\u666e500\u672a\u676530\u5929\u7684\u9690\u542b\u6ce2\u52a8\u7387\u3002\u4f46\u5355\u72ec\u4e00\u4e2a\u6570\u5b57\u53ea\u80fd\u544a\u8bc9\u4f60\u201c\u73b0\u5728\u6ce2\u52a8\u5927\u4e0d\u5927\u201d\uff0c\u5374\u65e0\u6cd5\u56de\u7b54\u4e24\u4e2a\u66f4\u5173\u952e\u7684\u95ee\u9898\uff1a<\/p>\n\n\n\n
    \n
  • \u6ce2\u52a8\u4e3b\u8981\u96c6\u4e2d\u5728\u77ed\u671f\u8fd8\u662f\u4f1a\u6301\u7eed\u5230\u4e2d\u957f\u671f\uff1f<\/li>\n\n\n\n
  • \u5e02\u573a\u5f53\u524d\u662f\u201c\u6b63\u5e38\u5b9a\u4ef7\uff0c\u8fd8\u662f\u6781\u7aef\u5c3e\u90e8\u98ce\u9669\uff0c\u8fd8\u662f\u5df2\u7ecf\u5728\u6050\u614c\u4e2d\u8fc7\u5ea6\u53cd\u5e94\uff1f<\/li>\n<\/ul>\n\n\n\n
    \u7b54\u6848\u85cf\u5728VIX\u671f\u8d27\u7684\u671f\u9650\u7ed3\u6784\uff08Term Structure\uff09\u91cc\u3002\u4e0d\u540c\u5230\u671f\u6708\u7684VIX\u671f\u8d27\u4ef7\u683c\u8fde\u6210\u4e00\u6761\u66f2\u7ebf\uff0c\u5f62\u72b6\u901a\u5e38\u5206\u4e3a\u4e09\u79cd\u5178\u578b\u72b6\u6001\uff1a<\/p>\n\n\n\n
    \u72b6\u6001<\/th>\u82f1\u6587\u672f\u8bed<\/th>\u66f2\u7ebf\u7279\u5f81<\/th>\u5386\u53f2\u5360\u6bd4\uff082013-2024\uff09<\/th>\u5178\u578b\u573a\u666f<\/th><\/tr>
    \u6b63\u5411\u5e02\u573a<\/td>Contango<\/td>\u8fd1\u6708\uff1c\u8fdc\u6708\uff0c\u5411\u4e0a\u503e\u659c<\/td>\u7ea685%<\/td>\u5e38\u6001\uff0c\u7f8e\u80a1\u6162\u725b<\/td><\/tr>
    \u5012\u6302\u5e02\u573a<\/td>Backwardation<\/td>\u8fd1\u6708\uff1e\u8fdc\u6708\uff0c\u5411\u4e0b\u503e\u659c<\/td>\u7ea68%<\/td>2020\u75ab\u60c5\u30012008\u5371\u673a\u30012024.8\u79d1\u6280\u6050\u614c<\/td><\/tr>
    \u5e73\u5766\/\u8fc7\u6e21<\/td>Flat<\/td>\u51e0\u4e4e\u6c34\u5e73<\/td>\u7ea67%<\/td>\u653f\u7b56\u8f6c\u5411\u671f<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n
    \u6b63\u5411\u5e02\u573a\uff1a<\/strong>\u6574\u6761\u66f2\u7ebf\u770b\u4e0a\u53bb\u662f\u6a59\u8272\u8fd8\u662f\u84dd\u8272\uff0c\u662f\u6574\u4f53\u4e0a\u79fb\u8fd8\u662f\u4e0b\u964d\uff0c\u8bf4\u660e\u5927\u5bb6\u5bf9\u672a\u6765\u7684\u6bcf\u4e00\u5929\u7684\u7126\u8651\u7a0b\u5ea6\uff0c\u8fd9\u662f\u6700\u76f4\u89c2\u7684\u6050\u614c\u7a0b\u5ea6\u3002<\/p>\n\n\n\n
    \"\"<\/figure>\n\n\n\n
    \u5012\u6302\u5e02\u573a\uff1a<\/strong>\u8fd9\u4e2a\u6307\u6807\u6700\u5fae\u5999\uff0c\u4e5f\u6700\u6709\u8da3\u3002\u5b83\u53cd\u6620\u4e86\u5e02\u573a\u5bf9\u201c\u7279\u5b9a\u65f6\u95f4\u70b9\u201d\u7684\u62c5\u5fe7\u3002<\/p>\n\n\n\n
    \"\"\/<\/figure>\n\n\n\n
    \u5e73\u5766\/\u8fc7\u6e21\uff1a<\/strong>\u66f2\u7ebf\u662f\u5411\u4e0a\u8fd8\u662f\u5411\u4e0b\uff0c\u76f4\u63a5\u51b3\u5b9a\u4e86\u5e02\u573a\u662f\u5426\u8f6c\u5411\uff0c\u4e5f\u51b3\u5b9a\u4e86\u4f60\u7684\u201c\u5c55\u671f\u6210\u672c\u201d\u662f\u6b63\u662f\u8d1f\u3002<\/p>\n\n\n\n
    \"\"\/<\/figure>\n\n\n\n
    \u7b80\u5355\u8bf4\uff0cVIX\u544a\u8bc9\u4f60\u7684\u662f\u672a\u676530\u5929\u5e02\u573a\u5927\u6982\u4f1a\u6709\u591a\u201c\u98a0\u7c38\u201d\u3002\u800cVIX\u671f\u9650\u7ed3\u6784\uff0c\u662f\u628a\u672a\u6765\u4e0d\u540c\u65f6\u95f4\u70b9\uff08\u6bd4\u59821\u4e2a\u6708\u540e\u30012\u4e2a\u6708\u540e\u3001\u534a\u5e74\u540e\uff09\u7684\u201c\u98a0\u7c38\u9884\u671f\u201d\u4e32\u8d77\u6765\uff0c\u5f62\u6210\u7684\u4e00\u6761\u66f2\u7ebf\u3002\u8fd9\u6761\u7ebf\u957f\u4ec0\u4e48\u6837\uff0c\u76f4\u63a5\u66b4\u9732\u4e86\u5e02\u573a\u6700\u771f\u5b9e\u7684\u201c\u5185\u5fc3\u620f\u201d\u3002<\/p>\n\n\n\n
    2. \u6570\u636e\u83b7\u53d6\uff1aCBOE\u5b98\u65b9\u514d\u8d39CSV\u5168\u81ea\u52a8\u5316<\/strong><\/h2>\n\n\n\n
    CBOE\u6bcf\u5929\u5728\u5b98\u7f51\u516c\u5f00\u6240\u6709VIX\u671f\u8d27\u7ed3\u7b97\u6570\u636e\uff0c\u5730\u5740\u56fa\u5b9a\uff0c\u683c\u5f0f\u7a33\u5b9a\u3002\u6211\u4eec\u4f7f\u7528\u793e\u533a\u7ef4\u62a4\u7684 vix-utils<\/code> \u5e93\u5f02\u6b65\u4e0b\u8f7d\uff0c\u901f\u5ea6\u6781\u5feb\u3002<\/p>\n\n\n\n
    pip install vix-utils nest-asyncio pandas\n\nimport nest_asyncio, asyncio\nfrom vix_utils import async_load_vix_term_structure\nimport pandas as pd\n\nnest_asyncio.apply()\n\nasync def fetch_vix():\n    df = await async_load_vix_term_structure()\n    # \u8fc7\u6ee4\u5468\u5408\u7ea6\u4e0e\u5df2\u5230\u671f\u5408\u7ea6\n    df = df[(df['Weekly'] == False) & (df['Expired'] == False)].copy()\n    df['Trade Date'] = pd.to_datetime(df['Trade Date'])\n    return df\n\ndf = asyncio.get_event_loop().run_until_complete(fetch_vix())\ndf.head()<\/code><\/pre>\n\n\n\n
    \u8fd4\u56de\u5b57\u6bb5\u5173\u952e\u5217\uff1aTrade Date\u3001Futures Contract\uff08\u6708\u4efd\uff09\u3001Tenor_Days\uff08\u5269\u4f59\u5929\u6570\uff09\u3001Settle\uff08\u7ed3\u7b97\u4ef7\uff09\u3002<\/p>\n\n\n\n
    3. \u57fa\u7840\u53ef\u89c6\u5316\uff1a\u4ea4\u4e92\u5f0f\u671f\u9650\u7ed3\u6784\u6ed1\u5757\u56fe<\/strong><\/h2>\n\n\n\n
    \u7528Plotly\u505a\u51fa\u5e26\u65e5\u671f\u6ed1\u5757\u7684\u4ea4\u4e92\u56fe\uff0c\u4e00\u76ee\u4e86\u7136\u770b\u51fa\u66f2\u7ebf\u5f62\u6001\u53d8\u5316\u3002<\/p>\n\n\n\n
    import plotly.express as px\n\n# \u6bcf\u6708\u5408\u7ea6\uff08\u53bb\u5468\u5408\u7ea6\uff09\nmonthly = df[df['Weekly'] == False].copy()\n\nfig = px.line(monthly, x='Tenor_Days', y='Settle', color='Trade Date',\n              title='VIX\u671f\u9650\u7ed3\u6784\u65f6\u95f4\u6ed1\u5757\uff08\u62d6\u52a8\u67e5\u770b\u5386\u53f2\uff09',\n              labels={'Tenor_Days':'\u5230\u671f\u5929\u6570', 'Settle':'\u671f\u8d27\u4ef7\u683c'})\nfig.update_layout(height=600, template='plotly_dark)\nfig.show()<\/code><\/pre>\n\n\n
    \n
    \"\"\/
    \u56fe1\uff1a\u4ea4\u4e92\u6ed1\u5757\u56fe\uff0c2024\u5e748\u6708\u660e\u663e\u5012\u6302\uff0c2025\u5e7412\u6708\u6062\u590d\u5178\u578b\u6b63\u5411<\/figcaption><\/figure>\n<\/div>\n\n\n
    4. \u91cf\u5316\u6307\u6807\u4e00\uff1a\u6bcf\u65e5\u66f2\u7ebf\u659c\u7387\uff08Slope\uff09<\/strong><\/h2>\n\n\n\n
    \u6700\u7b80\u5355\u6709\u6548\u7684\u5355\u65e5\u6307\u6807\uff1a\u5bf9\u5f53\u5929\u7684Tenor_Days\u548cSettle\u505a\u4e00\u5143\u7ebf\u6027\u56de\u5f52\uff0c\u5f97\u5230\u659c\u7387\u3002<\/p>\n\n\n\n
    import numpy as np\n\nslopes = []\nfor date, group in df.groupby('Trade Date'):\n    group = group.sort_values('Tenor_Days')\n    if len(group) >= 3:\n        slope = np.polyfit(group['Tenor_Days'], group['Settle'], 1)[0]\n        slopes.append({'date': date, 'slope': slope})\n\nslope_df = pd.DataFrame(slopes).set_index('date')<\/code><\/pre>\n\n\n\n
    5. \u6838\u5fc3\u6a21\u578b\uff1aHMM\u8bc6\u522b\u4e09\u79cd\u9690\u542b\u72b6\u6001<\/strong><\/h2>\n\n\n\n
    \u659c\u7387\u5e8f\u5217\u662f\u975e\u5e73\u7a33\u7684\uff0c\u76f4\u63a5\u9608\u503c\u5224\u65ad\u5bb9\u6613\u8bef\u5224\u3002\u6211\u4eec\u4f7f\u7528\u9ad8\u65afHMM\u81ea\u52a8\u5b66\u4e60\u4e09\u79cd\u9690\u85cf\u72b6\u6001\u3002<\/p>\n\n\n\n
    from hmmlearn.hmm import GaussianHMM\nfrom sklearn.preprocessing import StandardScaler\n\nX = StandardScaler().fit_transform(slope_df[['slope']])\n\nmodel = GaussianHMM(n_components=3, covariance_type=\"full\", n_iter=1000, random_state=42)\nmodel.fit(X)\nhidden_states = model.predict(X)\n\n# \u6839\u636e\u5747\u503c\u6392\u5e8f\u547d\u540d\u72b6\u6001\nmeans = model.means_.flatten()\nstate_names = ['\u5012\u6302\uff08Backwardation\uff09', '\u8fc7\u6e21\uff08Flat\uff09', '\u6b63\u5411\uff08Contango\uff09']\norder = np.argsort(means)\nname_map = dict(zip(order, state_names))\n\nslope_df['regime'] = [name_map[s] for s in hidden_states]<\/code><\/pre>\n\n\n\n
    \u5f97\u5230\u7684\u8f6c\u79fb\u77e9\u9635\uff082025\u5e741\u6708\u6700\u65b0\u62df\u5408\uff09\uff1a<\/p>\n\n\n\n
    \u4ece\u2192\u5230<\/th>\u6b63\u5411<\/th>\u8fc7\u6e21<\/th>\u5012\u6302<\/th><\/tr>
    \u6b63\u5411<\/td>0.942<\/td>0.057<\/td>0.001<\/td><\/tr>
    \u8fc7\u6e21<\/td>0.623<\/td>0.351<\/td>0.026<\/td><\/tr>
    \u5012\u6302<\/td>0.012<\/td>0.089<\/td>0.899<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n
    \u5012\u6302\u72b6\u6001\u81ea\u4fdd\u6301\u6982\u7387\u9ad8\u8fbe90%\uff0c\u4e00\u65e6\u8fdb\u5165\u77ed\u671f\u5f88\u96be\u5feb\u901f\u9000\u51fa\uff0c\u8fd9\u4e5f\u662f\u4e3a\u4ec0\u4e482020\u5e74\u75ab\u60c5\u5012\u6302\u6301\u7eed\u6570\u6708\u3002<\/p>\n\n\n
    \n
    \"\"\/
    \u56fe2\uff1aHMM\u81ea\u52a8\u6807\u6ce8\u7684\u6b63\u5411\/\u8fc7\u6e21\/\u5012\u6302\u72b6\u6001<\/figcaption><\/figure>\n<\/div>\n\n\n
    6. \u91cf\u5316\u6307\u6807\u4e8c\uff1aCarry\u4ef7\u5dee\uff08VX2-VX1\u3001VX6-VX1\uff09<\/strong><\/h2>\n\n\n\n
    # \u6784\u5efa\u6708\u5408\u7ea6pivot\u8868\npivot = df.pivot(index='Trade Date', columns='Tenor_Monthly', values='Settle')\n\ncarry = pd.DataFrame({\n    'Near_Mid': pivot[2] - pivot[1],\n    'Near_Long': pivot[6] - pivot[1]\n})\ncarry = carry.dropna()<\/code><\/pre>\n\n\n\n
    \"\"
    \u56fe3\uff1a\u7ed3\u8f6c\u98ce\u9669\u3002VX2-VX1 \u7684\u8870\u51cf\u901f\u5ea6\u5feb\u4e8e VX6-VX1\u3002\u5f53\u8fd1\u6708\u4ef7\u5dee\u9996\u5148\u6536\u7a84\u65f6\uff0c\u7ed3\u8f6c\u4e3a\u6b63\uff1b\u5f53 VX2-VX1 \u8f6c\u4e3a\u8d1f\u503c\u800c VX6-VX1 \u4fdd\u6301\u9ad8\u4f4d\u65f6\uff0c\u7ed3\u8f6c\u53d1\u751f\u9006\u8f6c<\/figcaption><\/figure>\n\n\n\n
    \u5f53Near_Mid\uff1c0\u65f6\uff0c\u5356\u51fa\u8fd1\u6708\u3001\u4e70\u5165\u4e2d\u6708\u53ef\u9501\u5b9a\u6b63carry\uff1b\u53cd\u4e4b\u5219\u9700\u8b66\u60d5\u3002<\/p>\n\n\n\n
    7. \u91cf\u5316\u6307\u6807\u4e09\uff1aPCA\u4e3b\u6210\u5206\u5206\u89e3<\/strong><\/h2>\n\n\n\n
    \u5bf9\u524d8\u4e2a\u6708\u4efd\u5408\u7ea6\u505aPCA\uff0c\u901a\u5e38\u524d\u4e09\u4e2a\u4e3b\u6210\u5206\u89e3\u91ca\u7387\uff1e95%\uff1a<\/p>\n\n\n\n