股指期货案例-葛新杰.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# GARCH 广义自回归条件异方差模型"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import arch #GARCH所用到的库\n",
    "import statsmodels as sm\n",
    "import warnings\n",
    "import datetime #用于时间转换\n",
    "warnings.filterwarnings(\"ignore\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "data=pd.read_excel('/Users/singal/Desktop/工作/第二次工作/案例一：股指期货案例/数据/表2数据.xlsx')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Unnamed: 0</th>\n",
       "      <th>沪深300指数现货收盘价</th>\n",
       "      <th>沪深300指数期货收盘价</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2009-04-16</td>\n",
       "      <td>2687.11</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2009-04-17</td>\n",
       "      <td>2650.69</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2009-04-20</td>\n",
       "      <td>2707.67</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2009-04-21</td>\n",
       "      <td>2675.44</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2009-04-22</td>\n",
       "      <td>2576.28</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Unnamed: 0  沪深300指数现货收盘价  沪深300指数期货收盘价\n",
       "0 2009-04-16       2687.11           NaN\n",
       "1 2009-04-17       2650.69           NaN\n",
       "2 2009-04-20       2707.67           NaN\n",
       "3 2009-04-21       2675.44           NaN\n",
       "4 2009-04-22       2576.28           NaN"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "data=data.rename(columns={'Unnamed: 0':'日期'})#更改列名"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "data['对数价格']=np.log(data['沪深300指数现货收盘价'])\n",
    "data['对数收益率']=data['对数价格']-data['对数价格'].shift(1)\n",
    "data=data.drop('沪深300指数期货收盘价',axis=1)#本案例不需要期货数据"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "y=data['对数收益率'].dropna()#第一行没有数据，故删去。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>日期</th>\n",
       "      <th>沪深300指数现货收盘价</th>\n",
       "      <th>对数价格</th>\n",
       "      <th>对数收益率</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2009-04-16</td>\n",
       "      <td>2687.11</td>\n",
       "      <td>7.896222</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2009-04-17</td>\n",
       "      <td>2650.69</td>\n",
       "      <td>7.882575</td>\n",
       "      <td>-0.013646</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2009-04-20</td>\n",
       "      <td>2707.67</td>\n",
       "      <td>7.903844</td>\n",
       "      <td>0.021269</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2009-04-21</td>\n",
       "      <td>2675.44</td>\n",
       "      <td>7.891869</td>\n",
       "      <td>-0.011975</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2009-04-22</td>\n",
       "      <td>2576.28</td>\n",
       "      <td>7.854102</td>\n",
       "      <td>-0.037767</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          日期  沪深300指数现货收盘价      对数价格     对数收益率\n",
       "0 2009-04-16       2687.11  7.896222       NaN\n",
       "1 2009-04-17       2650.69  7.882575 -0.013646\n",
       "2 2009-04-20       2707.67  7.903844  0.021269\n",
       "3 2009-04-21       2675.44  7.891869 -0.011975\n",
       "4 2009-04-22       2576.28  7.854102 -0.037767"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "在介绍GARCH之前，先引入ARCH自回归条件异方差模型，ARCH的具体构建步骤如下：\n",
    "1. 检验时间序列的平稳性，可用ADF检验。\n",
    "2. 绘制偏相关系数图，以得到均值方程，如m阶的AR模型。\n",
    "3. 构建AR模型，得到残差，并生成残差平方。\n",
    "4. 检验残差平方的相关性，检测方法可以为混成检验（Ljung-Box test）等，若存在相关性，则GARCH或ARCH效应存在；并确定残差平方的自相关阶数。\n",
    "5. 以方差作为被解释变量，残差作为解释变量构建AR模型。\n",
    "\n",
    "经过上述步骤即可得到ARCH(m)模型：\n",
    "$$ a_t=sigma_t \\epsilon_t$$\n",
    "$$\\sigma_t^2=\\alpha_0+\\sum_{i=1}^m \\alpha_i a_{t-i}^2$$\n",
    "\n",
    "对原案例作补充：混成检验，又称LB检验，是根据一系列滞后阶数\b来判断序列是否存在相关性。\n",
    "$$\\begin{aligned}\n",
    "&H_0:\\qquad\\hat\\rho_1^2=\\hat\\rho_2^2=\\cdots=\\hat\\rho_k^2=0\\\\\n",
    "&\\\\\n",
    "&H_1:\\qquad \\hat\\rho_k^2 \\not=0,\\quad 1\\leq k\\leq h\\\\\n",
    "&\\\\\n",
    "&Q=n(n+2)\\sum_{k=1}^h \\frac{\\hat\\rho_k^2}{n-k}\\sim X(h)\\\\\n",
    "\\end{aligned}$$"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "adf=sm.api.tsa.stattools.adfuller(data['对数价格'])#检验平稳性"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "sm.tsa.stattools.adfuller函数返回值依次为：\n",
    "1. 统计量的值\n",
    "2. 伴随概率\n",
    "3. 阶数\n",
    "4. 样本个数\n",
    "5. 3种显著性水平及对应的置信区间"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-2.3739455291571887,\n",
       " 0.14922550984098537,\n",
       " 0,\n",
       " 485,\n",
       " {'1%': -3.443905150512834,\n",
       "  '5%': -2.867517732199813,\n",
       "  '10%': -2.569953900520778},\n",
       " -2480.089452553145)"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "adf#无法拒绝原假设，对数价格序列是非平稳的"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "绘制偏相关系数图，确定自回归阶数。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sm.api.graphics.tsa.plot_pacf(y)#十一阶自相关系数显著不为0，故建立AR(11)模型"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "构建AR(11)模型，作为均值方程。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = sm.tsa.arima.model.ARIMA(y,order=(11,0,0)).fit()#均值方程"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Residuals Squares')"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1440x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "a=y-model.fittedvalues#得到残差\n",
    "a2=a**2#生成残差平方项\n",
    "\n",
    "fig=plt.figure(figsize=(20,5))\n",
    "ax1=fig.add_subplot(121)\n",
    "ax1.plot(a)\n",
    "ax1.set_title('Residuals')\n",
    "\n",
    "ax2=fig.add_subplot(122)\n",
    "ax2.plot(a2)\n",
    "ax2.set_title('Residuals Squares')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "观察可得，对数收益率的残差大致不存在相关性，但是残差的平方存在相关性，即存在GARCH效应，下面对残差平方进行混成检验，以得到显著结论。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([ 1.00000000e+00,  1.33204103e-02,  9.99207620e-02,  1.65006872e-01,\n",
       "         5.82346498e-02,  6.50969938e-04,  9.26627944e-02,  1.53437064e-01,\n",
       "         2.80107863e-02,  3.56551078e-02,  1.54913017e-01,  4.02175085e-02,\n",
       "         4.51039438e-02,  1.49925222e-01,  2.66898584e-02, -1.33290922e-02,\n",
       "         6.72842566e-02,  2.44933449e-02, -4.31556427e-03,  2.77584801e-02,\n",
       "         9.96190695e-02]),\n",
       " array([ 0.08658856,  4.96900753, 18.31121607, 19.97650103, 19.97670955,\n",
       "        24.21066196, 35.84397357, 36.23248421, 36.86330682, 48.79639677,\n",
       "        49.60237383, 50.61824491, 61.86631731, 62.22354271, 62.31282679,\n",
       "        64.59277391, 64.89554963, 64.90496915, 65.29551871, 70.33635634]),\n",
       " array([7.68559733e-01, 8.33669138e-02, 3.79395549e-04, 5.04761787e-04,\n",
       "        1.26236971e-03, 4.77648986e-04, 7.75760752e-06, 1.59201233e-05,\n",
       "        2.78381757e-05, 4.43851493e-07, 7.37836752e-07, 1.08762561e-06,\n",
       "        2.43032767e-08, 4.76243393e-08, 1.00524139e-07, 8.65531248e-08,\n",
       "        1.60035159e-07, 3.22262415e-07, 5.46614126e-07, 1.60487335e-07]))"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test=sm.tsa.stattools.acf(a2,nlags=20,qstat=True)#返回值依次为自相关系数，混成检验统计量，伴随概率\n",
    "test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>AC</th>\n",
       "      <th>Q-statistic</th>\n",
       "      <th>p-value</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>lag</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.013320</td>\n",
       "      <td>0.086589</td>\n",
       "      <td>7.685597e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.099921</td>\n",
       "      <td>4.969008</td>\n",
       "      <td>8.336691e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.165007</td>\n",
       "      <td>18.311216</td>\n",
       "      <td>3.793955e-04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.058235</td>\n",
       "      <td>19.976501</td>\n",
       "      <td>5.047618e-04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.000651</td>\n",
       "      <td>19.976710</td>\n",
       "      <td>1.262370e-03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.092663</td>\n",
       "      <td>24.210662</td>\n",
       "      <td>4.776490e-04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.153437</td>\n",
       "      <td>35.843974</td>\n",
       "      <td>7.757608e-06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.028011</td>\n",
       "      <td>36.232484</td>\n",
       "      <td>1.592012e-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.035655</td>\n",
       "      <td>36.863307</td>\n",
       "      <td>2.783818e-05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.154913</td>\n",
       "      <td>48.796397</td>\n",
       "      <td>4.438515e-07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.040218</td>\n",
       "      <td>49.602374</td>\n",
       "      <td>7.378368e-07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.045104</td>\n",
       "      <td>50.618245</td>\n",
       "      <td>1.087626e-06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.149925</td>\n",
       "      <td>61.866317</td>\n",
       "      <td>2.430328e-08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.026690</td>\n",
       "      <td>62.223543</td>\n",
       "      <td>4.762434e-08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>-0.013329</td>\n",
       "      <td>62.312827</td>\n",
       "      <td>1.005241e-07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0.067284</td>\n",
       "      <td>64.592774</td>\n",
       "      <td>8.655312e-08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.024493</td>\n",
       "      <td>64.895550</td>\n",
       "      <td>1.600352e-07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>-0.004316</td>\n",
       "      <td>64.904969</td>\n",
       "      <td>3.222624e-07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.027758</td>\n",
       "      <td>65.295519</td>\n",
       "      <td>5.466141e-07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.099619</td>\n",
       "      <td>70.336356</td>\n",
       "      <td>1.604873e-07</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           AC  Q-statistic       p-value\n",
       "lag                                     \n",
       "1    0.013320     0.086589  7.685597e-01\n",
       "2    0.099921     4.969008  8.336691e-02\n",
       "3    0.165007    18.311216  3.793955e-04\n",
       "4    0.058235    19.976501  5.047618e-04\n",
       "5    0.000651    19.976710  1.262370e-03\n",
       "6    0.092663    24.210662  4.776490e-04\n",
       "7    0.153437    35.843974  7.757608e-06\n",
       "8    0.028011    36.232484  1.592012e-05\n",
       "9    0.035655    36.863307  2.783818e-05\n",
       "10   0.154913    48.796397  4.438515e-07\n",
       "11   0.040218    49.602374  7.378368e-07\n",
       "12   0.045104    50.618245  1.087626e-06\n",
       "13   0.149925    61.866317  2.430328e-08\n",
       "14   0.026690    62.223543  4.762434e-08\n",
       "15  -0.013329    62.312827  1.005241e-07\n",
       "16   0.067284    64.592774  8.655312e-08\n",
       "17   0.024493    64.895550  1.600352e-07\n",
       "18  -0.004316    64.904969  3.222624e-07\n",
       "19   0.027758    65.295519  5.466141e-07\n",
       "20   0.099619    70.336356  1.604873e-07"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "acf,q,p=test\n",
    "df=pd.DataFrame()\n",
    "df['AC']=acf[1:]#自相关系数多出一项，即lag=0，故需删除\n",
    "df['Q-statistic']=q\n",
    "df['p-value']=p\n",
    "df['lag']=range(1,21)\n",
    "df=df.set_index('lag')\n",
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "一阶自相关系数所对应的Q统计量的伴随概率小于0.1，其他阶的自相关系数所对应的Q统计量的伴随概率均小于0.01，故拒绝原假设，认为残差平方存在相关性，即存在GARCH效应。下面建立ARCH模型。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "sm.api.graphics.tsa.plot_pacf(a2)#大致定位10阶，建立AR(10)模型"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "接下来即可根据残差的10阶滞后项建立关于方差的AR(10)模型，所得模型即为ARCH模型。实际中可以直接使用arch库，方便快捷。\n",
    "\n",
    "arch.arch_model函数的主要参数：\n",
    "1. y，表示时间序列。\n",
    "2. x，传入外生变量。\n",
    "3. mean，指定均值函数的类型，如“Constant”、“AR”等。\n",
    "4. lags，指定均值函数的滞后项。\n",
    "5. vol，指定波动率模型的类型，如“ARCH”、”GARCH“等。\n",
    "6. p，波动率模型的滞后阶数。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                           AR - ARCH Model Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                  对数收益率   R-squared:                       0.016\n",
      "Mean Model:                        AR   Adj. R-squared:                 -0.007\n",
      "Vol Model:                       ARCH   Log-Likelihood:                1280.90\n",
      "Distribution:                  Normal   AIC:                          -2515.80\n",
      "Method:            Maximum Likelihood   BIC:                          -2420.10\n",
      "                                        No. Observations:                  474\n",
      "Date:                Tue, Oct 26 2021   Df Residuals:                      451\n",
      "Time:                        11:25:30   Df Model:                           23\n",
      "                                  Mean Model                                 \n",
      "=============================================================================\n",
      "                 coef    std err          t      P>|t|       95.0% Conf. Int.\n",
      "-----------------------------------------------------------------------------\n",
      "Const      8.7200e-04  1.234e-03      0.706      0.480 [-1.548e-03,3.292e-03]\n",
      "对数收益率[1]       0.0192  8.874e-02      0.216      0.829      [ -0.155,  0.193]\n",
      "对数收益率[2]       0.0123      0.104      0.118      0.906      [ -0.192,  0.217]\n",
      "对数收益率[3]       0.0689  9.419e-02      0.731      0.465      [ -0.116,  0.254]\n",
      "对数收益率[4]      -0.0373  6.062e-02     -0.615      0.538    [ -0.156,8.152e-02]\n",
      "对数收益率[5]   9.7660e-03  7.772e-02      0.126      0.900      [ -0.143,  0.162]\n",
      "对数收益率[6]      -0.0379      0.107     -0.354      0.724      [ -0.248,  0.172]\n",
      "对数收益率[7]       0.0393      0.108      0.363      0.717      [ -0.173,  0.252]\n",
      "对数收益率[8]      -0.0288  6.583e-02     -0.438      0.662      [ -0.158,  0.100]\n",
      "对数收益率[9]      -0.0612  6.875e-02     -0.891      0.373    [ -0.196,7.350e-02]\n",
      "对数收益率[10]      0.0402  8.342e-02      0.482      0.629      [ -0.123,  0.204]\n",
      "对数收益率[11]     -0.0574      0.138     -0.415      0.678      [ -0.328,  0.214]\n",
      "                               Volatility Model                              \n",
      "=============================================================================\n",
      "                 coef    std err          t      P>|t|       95.0% Conf. Int.\n",
      "-----------------------------------------------------------------------------\n",
      "omega      1.6201e-04  1.560e-04      1.038      0.299 [-1.438e-04,4.678e-04]\n",
      "alpha[1]   4.8132e-18      0.191  2.514e-17      1.000      [ -0.375,  0.375]\n",
      "alpha[2]       0.0778  6.122e-02      1.271      0.204   [-4.216e-02,  0.198]\n",
      "alpha[3]       0.0330      0.124      0.265      0.791      [ -0.211,  0.277]\n",
      "alpha[4]   6.2050e-18      0.193  3.217e-17      1.000      [ -0.378,  0.378]\n",
      "alpha[5]       0.0208  6.407e-02      0.324      0.746      [ -0.105,  0.146]\n",
      "alpha[6]       0.1169      0.150      0.778      0.437      [ -0.178,  0.411]\n",
      "alpha[7]       0.0962      0.109      0.882      0.378      [ -0.118,  0.310]\n",
      "alpha[8]   6.4011e-18      0.326  1.966e-17      1.000      [ -0.638,  0.638]\n",
      "alpha[9]   1.8616e-20      0.454  4.101e-20      1.000      [ -0.890,  0.890]\n",
      "alpha[10]      0.0669      0.106      0.634      0.526      [ -0.140,  0.274]\n",
      "=============================================================================\n",
      "\n",
      "Covariance estimator: robust\n"
     ]
    }
   ],
   "source": [
    "import arch\n",
    "\n",
    "model=arch.arch_model(y,mean='AR',lags=11,vol='ARCH',p=10)\n",
    "result=model.fit(disp='off')#display=off，表示输出一个较为简洁的结果\n",
    "print(result.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                           AR - GARCH Model Results                           \n",
      "==============================================================================\n",
      "Dep. Variable:                  对数收益率   R-squared:                       0.022\n",
      "Mean Model:                        AR   Adj. R-squared:                 -0.001\n",
      "Vol Model:                      GARCH   Log-Likelihood:                1274.50\n",
      "Distribution:                  Normal   AIC:                          -2519.01\n",
      "Method:            Maximum Likelihood   BIC:                          -2456.59\n",
      "                                        No. Observations:                  474\n",
      "Date:                Tue, Oct 26 2021   Df Residuals:                      459\n",
      "Time:                        11:25:32   Df Model:                           15\n",
      "                                  Mean Model                                 \n",
      "=============================================================================\n",
      "                 coef    std err          t      P>|t|       95.0% Conf. Int.\n",
      "-----------------------------------------------------------------------------\n",
      "Const      4.6466e-04  7.619e-04      0.610      0.542 [-1.029e-03,1.958e-03]\n",
      "对数收益率[1]       0.0361  4.486e-02      0.805      0.421   [-5.183e-02,  0.124]\n",
      "对数收益率[2]   4.1538e-03  4.989e-02  8.326e-02      0.934   [-9.363e-02,  0.102]\n",
      "对数收益率[3]       0.0621  4.809e-02      1.292      0.196   [-3.213e-02,  0.156]\n",
      "对数收益率[4]      -0.0228  4.496e-02     -0.508      0.612    [ -0.111,6.530e-02]\n",
      "对数收益率[5]   3.1536e-04  4.618e-02  6.829e-03      0.995 [-9.020e-02,9.083e-02]\n",
      "对数收益率[6]      -0.0126  4.960e-02     -0.253      0.800    [ -0.110,8.465e-02]\n",
      "对数收益率[7]       0.0338  4.975e-02      0.679      0.497   [-6.375e-02,  0.131]\n",
      "对数收益率[8]   1.0088e-03  4.259e-02  2.369e-02      0.981 [-8.247e-02,8.449e-02]\n",
      "对数收益率[9]      -0.0498  4.805e-02     -1.037      0.300    [ -0.144,4.435e-02]\n",
      "对数收益率[10]      0.0647  4.931e-02      1.311      0.190   [-3.200e-02,  0.161]\n",
      "对数收益率[11]     -0.0888  4.472e-02     -1.986  4.704e-02   [ -0.176,-1.162e-03]\n",
      "                              Volatility Model                              \n",
      "============================================================================\n",
      "                 coef    std err          t      P>|t|      95.0% Conf. Int.\n",
      "----------------------------------------------------------------------------\n",
      "omega      2.8782e-05  1.401e-06     20.538  9.892e-94 [2.604e-05,3.153e-05]\n",
      "alpha[1]       0.0501  2.188e-02      2.291  2.195e-02 [7.248e-03,9.303e-02]\n",
      "beta[1]        0.8498  1.818e-02     46.748      0.000     [  0.814,  0.885]\n",
      "============================================================================\n",
      "\n",
      "Covariance estimator: robust\n"
     ]
    }
   ],
   "source": [
    "model=arch.arch_model(y,mean='AR',lags=11,vol='GARCH').fit(disp='off')\n",
    "print(model.summary())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "表格最下方汇报了波动率模型的两个参数 $\\alpha$ 和 $\\beta$ 且均在1%的置信水平下显著。故得到波动率模型：$$\\sigma_t^2=0.00003+0.05 a_{t-1}^2+0.85 \\sigma_{t-1}^2$$\n",
    "最后参照教学案例，在GARCH模型中引入哑变量 $D_t$，股指期货推出(2010-4-16)前后分别为0和1。则原模型改进为：\n",
    "$$\\sigma_t^2=\\omega+\\alpha a_{t-1}^2+\\beta \\sigma_{t-1}^2+\\gamma D_t$$ \n",
    "若哑变量的系数$\\gamma$显著为正，则股指期货的推出加剧了股票市场的波动，若为显著为负，则减缓了股票市场的波动，若显著为0，则股指期货的推出对现货市场的影响甚微。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    246\n",
       "1    240\n",
       "Name: 股指期货推出前后, dtype: int64"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data=data.set_index('日期')\n",
    "data['股指期货推出前后']=1\n",
    "data.loc[:'2010-4-16','股指期货推出前后']=0\n",
    "data['股指期货推出前后'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                           AR - GARCH Model Results                           \n",
      "==============================================================================\n",
      "Dep. Variable:                  对数收益率   R-squared:                       0.022\n",
      "Mean Model:                        AR   Adj. R-squared:                 -0.001\n",
      "Vol Model:                      GARCH   Log-Likelihood:                1274.50\n",
      "Distribution:                  Normal   AIC:                          -2519.01\n",
      "Method:            Maximum Likelihood   BIC:                          -2456.59\n",
      "                                        No. Observations:                  474\n",
      "Date:                Tue, Oct 26 2021   Df Residuals:                      459\n",
      "Time:                        11:27:33   Df Model:                           15\n",
      "                                  Mean Model                                 \n",
      "=============================================================================\n",
      "                 coef    std err          t      P>|t|       95.0% Conf. Int.\n",
      "-----------------------------------------------------------------------------\n",
      "Const      4.6466e-04  7.619e-04      0.610      0.542 [-1.029e-03,1.958e-03]\n",
      "对数收益率[1]       0.0361  4.486e-02      0.805      0.421   [-5.183e-02,  0.124]\n",
      "对数收益率[2]   4.1538e-03  4.989e-02  8.326e-02      0.934   [-9.363e-02,  0.102]\n",
      "对数收益率[3]       0.0621  4.809e-02      1.292      0.196   [-3.213e-02,  0.156]\n",
      "对数收益率[4]      -0.0228  4.496e-02     -0.508      0.612    [ -0.111,6.530e-02]\n",
      "对数收益率[5]   3.1536e-04  4.618e-02  6.829e-03      0.995 [-9.020e-02,9.083e-02]\n",
      "对数收益率[6]      -0.0126  4.960e-02     -0.253      0.800    [ -0.110,8.465e-02]\n",
      "对数收益率[7]       0.0338  4.975e-02      0.679      0.497   [-6.375e-02,  0.131]\n",
      "对数收益率[8]   1.0088e-03  4.259e-02  2.369e-02      0.981 [-8.247e-02,8.449e-02]\n",
      "对数收益率[9]      -0.0498  4.805e-02     -1.037      0.300    [ -0.144,4.435e-02]\n",
      "对数收益率[10]      0.0647  4.931e-02      1.311      0.190   [-3.200e-02,  0.161]\n",
      "对数收益率[11]     -0.0888  4.472e-02     -1.986  4.704e-02   [ -0.176,-1.162e-03]\n",
      "                              Volatility Model                              \n",
      "============================================================================\n",
      "                 coef    std err          t      P>|t|      95.0% Conf. Int.\n",
      "----------------------------------------------------------------------------\n",
      "omega      2.8782e-05  1.401e-06     20.538  9.892e-94 [2.604e-05,3.153e-05]\n",
      "alpha[1]       0.0501  2.188e-02      2.291  2.195e-02 [7.248e-03,9.303e-02]\n",
      "beta[1]        0.8498  1.818e-02     46.748      0.000     [  0.814,  0.885]\n",
      "============================================================================\n",
      "\n",
      "Covariance estimator: robust\n"
     ]
    }
   ],
   "source": [
    "model=arch.arch_model(y,x=np.array(data['股指期货推出前后']),mean='AR',lags=11,vol='GARCH',p=1,q=1).fit(disp='off')\n",
    "print(model.summary())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "引入虚拟变量后的估计结果和未引入虚拟变量的估计结果一致。"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "da5d3c13d5aa829161cf843792884de182f91ab2e540693bb706e0e23cabc9e9"
  },
  "kernelspec": {
   "display_name": "Python 3.8.10 64-bit ('py38': conda)",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.10"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}