ch7_LiquidityModel_Estimation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import warnings\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "plt.rcParams['figure.figsize'] = (10, 6) \n",
    "pd.set_option('use_inf_as_na', True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "liq_data=pd.read_csv('C:/Users/yim/Documents/카카오워크 받은 파일/mlfrm-master/bid_ask.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "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>Date</th>\n",
       "      <th>EXCHCD</th>\n",
       "      <th>TICKER</th>\n",
       "      <th>COMNAM</th>\n",
       "      <th>BIDLO</th>\n",
       "      <th>ASKHI</th>\n",
       "      <th>PRC</th>\n",
       "      <th>VOL</th>\n",
       "      <th>RET</th>\n",
       "      <th>SHROUT</th>\n",
       "      <th>OPENPRC</th>\n",
       "      <th>vwretx</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1031570</td>\n",
       "      <td>2019-01-02</td>\n",
       "      <td>3.0</td>\n",
       "      <td>INTC</td>\n",
       "      <td>INTEL CORP</td>\n",
       "      <td>45.77</td>\n",
       "      <td>47.470</td>\n",
       "      <td>47.08</td>\n",
       "      <td>18761673.0</td>\n",
       "      <td>0.003196</td>\n",
       "      <td>4564000.0</td>\n",
       "      <td>45.960</td>\n",
       "      <td>0.001783</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1031571</td>\n",
       "      <td>2019-01-03</td>\n",
       "      <td>3.0</td>\n",
       "      <td>INTC</td>\n",
       "      <td>INTEL CORP</td>\n",
       "      <td>44.39</td>\n",
       "      <td>46.280</td>\n",
       "      <td>44.49</td>\n",
       "      <td>32254097.0</td>\n",
       "      <td>-0.055013</td>\n",
       "      <td>4564000.0</td>\n",
       "      <td>46.150</td>\n",
       "      <td>-0.021219</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1031572</td>\n",
       "      <td>2019-01-04</td>\n",
       "      <td>3.0</td>\n",
       "      <td>INTC</td>\n",
       "      <td>INTEL CORP</td>\n",
       "      <td>45.54</td>\n",
       "      <td>47.570</td>\n",
       "      <td>47.22</td>\n",
       "      <td>35419836.0</td>\n",
       "      <td>0.061362</td>\n",
       "      <td>4564000.0</td>\n",
       "      <td>45.835</td>\n",
       "      <td>0.033399</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1031573</td>\n",
       "      <td>2019-01-07</td>\n",
       "      <td>3.0</td>\n",
       "      <td>INTC</td>\n",
       "      <td>INTEL CORP</td>\n",
       "      <td>46.75</td>\n",
       "      <td>47.995</td>\n",
       "      <td>47.44</td>\n",
       "      <td>22724997.0</td>\n",
       "      <td>0.004659</td>\n",
       "      <td>4564000.0</td>\n",
       "      <td>47.100</td>\n",
       "      <td>0.009191</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1031574</td>\n",
       "      <td>2019-01-08</td>\n",
       "      <td>3.0</td>\n",
       "      <td>INTC</td>\n",
       "      <td>INTEL CORP</td>\n",
       "      <td>46.78</td>\n",
       "      <td>48.030</td>\n",
       "      <td>47.74</td>\n",
       "      <td>22721240.0</td>\n",
       "      <td>0.006324</td>\n",
       "      <td>4564000.0</td>\n",
       "      <td>47.800</td>\n",
       "      <td>0.010240</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Unnamed: 0        Date  EXCHCD TICKER      COMNAM  BIDLO   ASKHI    PRC  \\\n",
       "0     1031570  2019-01-02     3.0   INTC  INTEL CORP  45.77  47.470  47.08   \n",
       "1     1031571  2019-01-03     3.0   INTC  INTEL CORP  44.39  46.280  44.49   \n",
       "2     1031572  2019-01-04     3.0   INTC  INTEL CORP  45.54  47.570  47.22   \n",
       "3     1031573  2019-01-07     3.0   INTC  INTEL CORP  46.75  47.995  47.44   \n",
       "4     1031574  2019-01-08     3.0   INTC  INTEL CORP  46.78  48.030  47.74   \n",
       "\n",
       "          VOL       RET     SHROUT  OPENPRC    vwretx  \n",
       "0  18761673.0  0.003196  4564000.0   45.960  0.001783  \n",
       "1  32254097.0 -0.055013  4564000.0   46.150 -0.021219  \n",
       "2  35419836.0  0.061362  4564000.0   45.835  0.033399  \n",
       "3  22724997.0  0.004659  4564000.0   47.100  0.009191  \n",
       "4  22721240.0  0.006324  4564000.0   47.800  0.010240  "
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "liq_data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "rolling_five=[]\n",
    "\n",
    "for j in liq_data.TICKER.unique():\n",
    "    for i in range(len(liq_data[liq_data.TICKER==j])):\n",
    "        rolling_five.append(liq_data[i:i+5].agg({'BIDLO':'min',\n",
    "        'ASKHI':'max',\n",
    "        'VOL':'sum',\n",
    "        'SHROUT':'mean',\n",
    "        'PRC':'mean'}))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "rolling_five_df=pd.DataFrame(rolling_five)\n",
    "rolling_five_df.columns=['bidlo_min', 'askhi_max', 'vol_sum', 'shrout_mean', 'prc_mean']\n",
    "liq_vol_all=pd.concat([liq_data, rolling_five_df], axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "liq_ratio=[]\n",
    "for j in liq_vol_all.TICKER.unique():\n",
    "    for i in range(len(liq_vol_all[liq_vol_all.TICKER==j])):\n",
    "        liq_ratio.append((liq_vol_all['PRC'][i+1:i+6]*\n",
    "        liq_vol_all['VOL'][i+1:i+6]).sum()/\n",
    "        (np.abs(liq_vol_all['PRC'][i+1:i+6].mean()-\n",
    "        liq_vol_all['PRC'][i:i+5].mean())))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "Lhh=[]\n",
    "for j in liq_vol_all.TICKER.unique():\n",
    "    for i in range(len(liq_vol_all[liq_vol_all.TICKER==j])):\n",
    "        Lhh.append((liq_vol_all['PRC'][i:i+5].max()-\n",
    "        liq_vol_all['PRC'][i:i+5].min())/\n",
    "        liq_vol_all['PRC'][i:i+5].min()/\n",
    "        (liq_vol_all['VOL'][i:i+5].sum()/\n",
    "        liq_vol_all['SHROUT'][i:i+5].mean()*\n",
    "        liq_vol_all['PRC'][i:i+5].mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "turnover_ratio=[]\n",
    "\n",
    "for j in liq_vol_all.TICKER.unique():\n",
    "    for i in range(len(liq_vol_all[liq_vol_all.TICKER==j])):\n",
    "        turnover_ratio.append((1/liq_vol_all['VOL'].count())*\n",
    "        (np.sum(liq_vol_all['VOL'][i:i+1])/\n",
    "        np.sum(liq_vol_all['SHROUT'][i:i+1])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "liq_vol_all['liq_ratio']=pd.DataFrame(liq_ratio)\n",
    "liq_vol_all['Lhh']=pd.DataFrame(Lhh)\n",
    "liq_vol_all['turnover_ratio']=pd.DataFrame(turnover_ratio)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "liq_vol_all['mid_price'] = (liq_vol_all.ASKHI + liq_vol_all.BIDLO) / 2\n",
    "liq_vol_all['percent_quoted_ba'] = (liq_vol_all.ASKHI -\n",
    "liq_vol_all.BIDLO)/\\\n",
    "    liq_vol_all.mid_price\n",
    "liq_vol_all['percent_effective_ba'] = 2 * abs((liq_vol_all.PRC -\n",
    "liq_vol_all.mid_price))/\\\n",
    "    liq_vol_all.mid_price"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "liq_vol_all['price_diff']=liq_vol_all.groupby('TICKER')['PRC']\\\n",
    "    .apply(lambda x:x.diff())\n",
    "liq_vol_all.dropna(inplace=True)\n",
    "roll=[]\n",
    "\n",
    "for j in liq_vol_all.TICKER.unique():\n",
    "    for i in range(len(liq_vol_all[liq_vol_all.TICKER==j])):\n",
    "        roll_cov=np.cov(liq_vol_all['price_diff'][i:i+5],\n",
    "        liq_vol_all['price_diff'][i+1:i+6])\n",
    "        if roll_cov[0,1]<0:\n",
    "            roll.append(2*np.sqrt(-roll_cov[0,1]))\n",
    "        else:\n",
    "            roll.append(2*np.sqrt(np.abs(roll_cov[0,1])))\n",
    "\n",
    "            \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "gamma=[]\n",
    "for j in liq_vol_all.TICKER.unique():\n",
    "    for i in range(len(liq_vol_all[liq_vol_all.TICKER==j])):\n",
    "        gamma.append((max(liq_vol_all['ASKHI'].iloc[i+1],\n",
    "        liq_vol_all['ASKHI'].iloc[i])-\n",
    "        min(liq_vol_all['BIDLO'].iloc[i+1],\n",
    "        liq_vol_all['BIDLO'].iloc[i]))**2)\n",
    "        gamma_array=np.array(gamma)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "beta=[]\n",
    "for j in liq_vol_all.TICKER.unique():\n",
    "    for i in range(len(liq_vol_all[liq_vol_all.TICKER==j])):\n",
    "        beta.append((liq_vol_all['ASKHI'].iloc[i+1]-\n",
    "        liq_vol_all['BIDLO'].iloc[i+1])**2+\n",
    "        (liq_vol_all['ASKHI'].iloc[i]-\n",
    "        liq_vol_all['BIDLO'].iloc[i])**2)\n",
    "        beta_array=np.array(beta)\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "alpha=((np.sqrt(2*beta_array)-np.sqrt(beta_array))/\n",
    "(3-(2*np.sqrt(2))))-np.sqrt(gamma_array/\n",
    "(3-(2*np.sqrt(2))))\n",
    "CS_spread=(2*np.exp(alpha-1))/(1+np.exp(alpha))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "liq_vol_all=liq_vol_all.reset_index()\n",
    "liq_vol_all['roll']=pd.DataFrame(roll)\n",
    "liq_vol_all['CS_spread']=pd.DataFrame(CS_spread)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "dvol=[]\n",
    "for j in liq_vol_all.TICKER.unique():\n",
    "    for i in range(len(liq_vol_all[liq_vol_all.TICKER==j])):\n",
    "        dvol.append((liq_vol_all['PRC'][i:i+5]*\n",
    "        liq_vol_all['VOL'][i:i+5]).sum())\n",
    "liq_vol_all['dvol']=pd.DataFrame(dvol)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "amihud=[]\n",
    "\n",
    "for j in liq_vol_all.TICKER.unique():\n",
    "    for i in range(len(liq_vol_all[liq_vol_all.TICKER==j])):\n",
    "        amihud.append((1/liq_vol_all['RET'].count())*\n",
    "        (np.sum(np.abs(liq_vol_all['RET'][i:i+1]))/\n",
    "        np.sum(liq_vol_all['dvol'][i:i+1])))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "florackis=[]\n",
    "\n",
    "for j in liq_vol_all.TICKER.unique():\n",
    "    for i in range(len(liq_vol_all[liq_vol_all.TICKER==j])):\n",
    "        florackis.append((1/liq_vol_all['RET'].count())*\n",
    "        (np.sum(np.abs(liq_vol_all['RET'][i:i+1])/\n",
    "        liq_vol_all[\"turnover_ratio\"][i:i+1])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "liq_vol_all['vol_diff_pct']=liq_vol_all.groupby('TICKER')['VOL']\\\n",
    "    .apply(lambda x: x.diff()).pct_change()\n",
    "liq_vol_all['price_diff_pct']=liq_vol_all.groupby('TICKER')['PRC']\\\n",
    "    .apply(lambda x: x.diff()).pct_change()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "cet=[]\n",
    "for j in liq_vol_all.TICKER.unique():\n",
    "    for i in range(len(liq_vol_all[liq_vol_all.TICKER==j])):\n",
    "        cet.append(np.sum(liq_vol_all['vol_diff_pct'][i:i+1])/\n",
    "    np.sum(liq_vol_all['price_diff_pct'][i:i+1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "liq_vol_all['amihud']=pd.DataFrame(amihud)\n",
    "liq_vol_all['florackis']=pd.DataFrame(florackis)\n",
    "liq_vol_all['cet']=pd.DataFrame(cet)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "import statsmodels.api as sm "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [],
   "source": [
    "liq_vol_all['VOL_pct_change']=liq_vol_all.groupby('TICKER')['VOL']\\\n",
    "    .apply(lambda x: x.pct_change())\n",
    "liq_vol_all.dropna(subset=['VOL_pct_change'], inplace=True)\n",
    "liq_vol_all=liq_vol_all.reset_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [],
   "source": [
    "unsys_resid=[]\n",
    "\n",
    "for i in liq_vol_all.TICKER.unique():\n",
    "    X1=liq_vol_all[liq_vol_all['TICKER']==i]['vwretx']\n",
    "    y=liq_vol_all[liq_vol_all['TICKER']==i]['RET']\n",
    "    ols=sm.OLS(y, X1).fit()\n",
    "    unsys_resid.append(ols.resid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "******************************************************************************************\n",
      "OLS Result for INTC\n",
      "                                 OLS Regression Results                                \n",
      "=======================================================================================\n",
      "Dep. Variable:                      y   R-squared (uncentered):                   0.157\n",
      "Model:                            OLS   Adj. R-squared (uncentered):              0.154\n",
      "Method:                 Least Squares   F-statistic:                              46.31\n",
      "Date:                Sun, 02 Apr 2023   Prob (F-statistic):                    7.53e-11\n",
      "Time:                        21:28:56   Log-Likelihood:                          1444.9\n",
      "No. Observations:                 249   AIC:                                     -2888.\n",
      "Df Residuals:                     248   BIC:                                     -2884.\n",
      "Df Model:                           1                                                  \n",
      "Covariance Type:            nonrobust                                                  \n",
      "==================================================================================\n",
      "                     coef    std err          t      P>|t|      [0.025      0.975]\n",
      "----------------------------------------------------------------------------------\n",
      "VOL_pct_change     0.0008      0.000      6.805      0.000       0.001       0.001\n",
      "==============================================================================\n",
      "Omnibus:                      373.849   Durbin-Watson:                   1.908\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            53506.774\n",
      "Skew:                           7.228   Prob(JB):                         0.00\n",
      "Kurtosis:                      73.344   Cond. No.                         1.00\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] R² is computed without centering (uncentered) since the model does not contain a constant.\n",
      "[2] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "******************************************************************************************\n",
      "OLS Result for MSFT\n",
      "                                 OLS Regression Results                                \n",
      "=======================================================================================\n",
      "Dep. Variable:                      y   R-squared (uncentered):                   0.044\n",
      "Model:                            OLS   Adj. R-squared (uncentered):              0.040\n",
      "Method:                 Least Squares   F-statistic:                              11.45\n",
      "Date:                Sun, 02 Apr 2023   Prob (F-statistic):                    0.000833\n",
      "Time:                        21:28:56   Log-Likelihood:                          1851.0\n",
      "No. Observations:                 249   AIC:                                     -3700.\n",
      "Df Residuals:                     248   BIC:                                     -3696.\n",
      "Df Model:                           1                                                  \n",
      "Covariance Type:            nonrobust                                                  \n",
      "==================================================================================\n",
      "                     coef    std err          t      P>|t|      [0.025      0.975]\n",
      "----------------------------------------------------------------------------------\n",
      "VOL_pct_change  9.641e-05   2.85e-05      3.383      0.001    4.03e-05       0.000\n",
      "==============================================================================\n",
      "Omnibus:                      285.769   Durbin-Watson:                   1.533\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):            11207.666\n",
      "Skew:                           4.937   Prob(JB):                         0.00\n",
      "Kurtosis:                      34.349   Cond. No.                         1.00\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] R² is computed without centering (uncentered) since the model does not contain a constant.\n",
      "[2] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "******************************************************************************************\n",
      "OLS Result for IBM\n",
      "                                 OLS Regression Results                                \n",
      "=======================================================================================\n",
      "Dep. Variable:                      y   R-squared (uncentered):                   0.134\n",
      "Model:                            OLS   Adj. R-squared (uncentered):              0.130\n",
      "Method:                 Least Squares   F-statistic:                              38.36\n",
      "Date:                Sun, 02 Apr 2023   Prob (F-statistic):                    2.43e-09\n",
      "Time:                        21:28:57   Log-Likelihood:                          1547.1\n",
      "No. Observations:                 249   AIC:                                     -3092.\n",
      "Df Residuals:                     248   BIC:                                     -3089.\n",
      "Df Model:                           1                                                  \n",
      "Covariance Type:            nonrobust                                                  \n",
      "==================================================================================\n",
      "                     coef    std err          t      P>|t|      [0.025      0.975]\n",
      "----------------------------------------------------------------------------------\n",
      "VOL_pct_change     0.0005   7.43e-05      6.193      0.000       0.000       0.001\n",
      "==============================================================================\n",
      "Omnibus:                      446.818   Durbin-Watson:                   2.034\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           156387.719\n",
      "Skew:                           9.835   Prob(JB):                         0.00\n",
      "Kurtosis:                     124.188   Cond. No.                         1.00\n",
      "==============================================================================\n",
      "\n",
      "Notes:\n",
      "[1] R² is computed without centering (uncentered) since the model does not contain a constant.\n",
      "[2] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n"
     ]
    }
   ],
   "source": [
    "market_impact={}\n",
    "for i, j in zip(liq_vol_all.TICKER.unique(),\\\n",
    "    range(len(liq_vol_all['TICKER'].unique()))):\n",
    "    X2=liq_vol_all[liq_vol_all['TICKER']==i]['VOL_pct_change']\n",
    "    ols=sm.OLS(unsys_resid[j]**2,X2).fit()\n",
    "    print('***'*30)\n",
    "    print(f'OLS Result for {i}')\n",
    "    print(ols.summary())\n",
    "    market_impact[j]=ols.resid \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "append1=market_impact[0].append(market_impact[1])\n",
    "liq_vol_all['market_impact']=append1.append(market_impact[2])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "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>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>liq_ratio</th>\n",
       "      <td>738.0</td>\n",
       "      <td>7.368514e+10</td>\n",
       "      <td>2.569030e+11</td>\n",
       "      <td>8.065402e+08</td>\n",
       "      <td>1.378496e+10</td>\n",
       "      <td>2.261858e+10</td>\n",
       "      <td>4.505784e+10</td>\n",
       "      <td>3.095986e+12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lhh</th>\n",
       "      <td>738.0</td>\n",
       "      <td>3.340167e-05</td>\n",
       "      <td>5.371681e-05</td>\n",
       "      <td>3.966368e-06</td>\n",
       "      <td>1.694354e-05</td>\n",
       "      <td>2.368095e-05</td>\n",
       "      <td>3.558960e-05</td>\n",
       "      <td>5.824148e-04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>turnover_ratio</th>\n",
       "      <td>738.0</td>\n",
       "      <td>6.491127e-03</td>\n",
       "      <td>2.842668e-03</td>\n",
       "      <td>1.916371e-03</td>\n",
       "      <td>4.897990e-03</td>\n",
       "      <td>5.764112e-03</td>\n",
       "      <td>7.423111e-03</td>\n",
       "      <td>2.542853e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>roll</th>\n",
       "      <td>738.0</td>\n",
       "      <td>8.190794e-01</td>\n",
       "      <td>6.066821e-01</td>\n",
       "      <td>7.615773e-02</td>\n",
       "      <td>4.574986e-01</td>\n",
       "      <td>6.975982e-01</td>\n",
       "      <td>1.011879e+00</td>\n",
       "      <td>4.178873e+00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CS_spread</th>\n",
       "      <td>738.0</td>\n",
       "      <td>3.305464e-01</td>\n",
       "      <td>1.267434e-01</td>\n",
       "      <td>1.773438e-40</td>\n",
       "      <td>2.444225e-01</td>\n",
       "      <td>3.609800e-01</td>\n",
       "      <td>4.188028e-01</td>\n",
       "      <td>5.877726e-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>amihud</th>\n",
       "      <td>738.0</td>\n",
       "      <td>2.777021e-15</td>\n",
       "      <td>2.319450e-15</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.117729e-15</td>\n",
       "      <td>2.220439e-15</td>\n",
       "      <td>3.766086e-15</td>\n",
       "      <td>1.320828e-14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>florackis</th>\n",
       "      <td>738.0</td>\n",
       "      <td>2.284291e-03</td>\n",
       "      <td>1.546181e-03</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>1.059446e-03</td>\n",
       "      <td>2.013517e-03</td>\n",
       "      <td>3.324181e-03</td>\n",
       "      <td>7.869841e-03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>cet</th>\n",
       "      <td>738.0</td>\n",
       "      <td>-1.113583e+00</td>\n",
       "      <td>3.333932e+01</td>\n",
       "      <td>-4.575246e+02</td>\n",
       "      <td>-1.687807e-01</td>\n",
       "      <td>5.654237e-01</td>\n",
       "      <td>1.660166e+00</td>\n",
       "      <td>1.845917e+02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>market_impact</th>\n",
       "      <td>738.0</td>\n",
       "      <td>8.614680e-05</td>\n",
       "      <td>5.087547e-04</td>\n",
       "      <td>-1.596135e-03</td>\n",
       "      <td>-3.010645e-05</td>\n",
       "      <td>3.383862e-05</td>\n",
       "      <td>1.309451e-04</td>\n",
       "      <td>8.165527e-03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>percent_quoted_ba</th>\n",
       "      <td>738.0</td>\n",
       "      <td>1.565276e-02</td>\n",
       "      <td>7.562850e-03</td>\n",
       "      <td>3.779877e-03</td>\n",
       "      <td>1.041887e-02</td>\n",
       "      <td>1.379992e-02</td>\n",
       "      <td>1.878123e-02</td>\n",
       "      <td>5.545110e-02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>percent_effective_ba</th>\n",
       "      <td>738.0</td>\n",
       "      <td>8.334177e-03</td>\n",
       "      <td>7.100304e-03</td>\n",
       "      <td>0.000000e+00</td>\n",
       "      <td>3.032785e-03</td>\n",
       "      <td>6.851479e-03</td>\n",
       "      <td>1.152485e-02</td>\n",
       "      <td>4.656669e-02</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      count          mean           std           min  \\\n",
       "liq_ratio             738.0  7.368514e+10  2.569030e+11  8.065402e+08   \n",
       "Lhh                   738.0  3.340167e-05  5.371681e-05  3.966368e-06   \n",
       "turnover_ratio        738.0  6.491127e-03  2.842668e-03  1.916371e-03   \n",
       "roll                  738.0  8.190794e-01  6.066821e-01  7.615773e-02   \n",
       "CS_spread             738.0  3.305464e-01  1.267434e-01  1.773438e-40   \n",
       "amihud                738.0  2.777021e-15  2.319450e-15  0.000000e+00   \n",
       "florackis             738.0  2.284291e-03  1.546181e-03  0.000000e+00   \n",
       "cet                   738.0 -1.113583e+00  3.333932e+01 -4.575246e+02   \n",
       "market_impact         738.0  8.614680e-05  5.087547e-04 -1.596135e-03   \n",
       "percent_quoted_ba     738.0  1.565276e-02  7.562850e-03  3.779877e-03   \n",
       "percent_effective_ba  738.0  8.334177e-03  7.100304e-03  0.000000e+00   \n",
       "\n",
       "                               25%           50%           75%           max  \n",
       "liq_ratio             1.378496e+10  2.261858e+10  4.505784e+10  3.095986e+12  \n",
       "Lhh                   1.694354e-05  2.368095e-05  3.558960e-05  5.824148e-04  \n",
       "turnover_ratio        4.897990e-03  5.764112e-03  7.423111e-03  2.542853e-02  \n",
       "roll                  4.574986e-01  6.975982e-01  1.011879e+00  4.178873e+00  \n",
       "CS_spread             2.444225e-01  3.609800e-01  4.188028e-01  5.877726e-01  \n",
       "amihud                1.117729e-15  2.220439e-15  3.766086e-15  1.320828e-14  \n",
       "florackis             1.059446e-03  2.013517e-03  3.324181e-03  7.869841e-03  \n",
       "cet                  -1.687807e-01  5.654237e-01  1.660166e+00  1.845917e+02  \n",
       "market_impact        -3.010645e-05  3.383862e-05  1.309451e-04  8.165527e-03  \n",
       "percent_quoted_ba     1.041887e-02  1.379992e-02  1.878123e-02  5.545110e-02  \n",
       "percent_effective_ba  3.032785e-03  6.851479e-03  1.152485e-02  4.656669e-02  "
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cols=['vol_diff_pct', 'price_diff_pct', 'price_diff', 'VOL_pct_change', 'dvol', 'mid_price']\n",
    "liq_measures_all=liq_vol_all.drop(liq_vol_all[cols], axis=1)\\\n",
    "    .iloc[:, -11:]\n",
    "liq_measures_all.dropna(inplace=True)\n",
    "liq_measures_all.describe().T "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.mixture import GaussianMixture\n",
    "from sklearn.preprocessing import StandardScaler \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "liq_measures_all2=liq_measures_all.dropna()\n",
    "scaled_liq=StandardScaler().fit_transform(liq_measures_all2)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "kwargs=dict(alpha=0.5, bins=50, stacked=True)\n",
    "plt.hist(liq_measures_all.loc[:, 'percent_quoted_ba'],\n",
    "**kwargs, label='TC-based')\n",
    "plt.hist(liq_measures_all.loc[:, 'turnover_ratio'],\n",
    "**kwargs, label='Volume-based')\n",
    "plt.hist(liq_measures_all.loc[:, 'market_impact'],\n",
    "**kwargs,label='Market-based')\n",
    "plt.title('Multi Modality of the Liquidity Measures')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "n_components=np.arange(1,10)\n",
    "clusters=[GaussianMixture(n, covariance_type='spherical',\n",
    "random_state=0).fit(scaled_liq)\n",
    "for n in n_components]\n",
    "plt.plot(n_components, [m.bic(scaled_liq) for m in clusters])\n",
    "plt.title('Optimum Number of Components')\n",
    "plt.xlabel('n_components')\n",
    "plt.ylabel('BIC values')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "def cluster_state(data, nstates):\n",
    "    gmm=GaussianMixture(n_components=nstates,\n",
    "    covariance_type='spherical',\n",
    "    init_params='kmeans')\n",
    "    gmm_fit=gmm.fit(scaled_liq)\n",
    "    labels=gmm_fit.predict(scaled_liq)\n",
    "    state_probs=gmm.predict_proba(scaled_liq)\n",
    "    state_probs_df=pd.DataFrame(state_probs, \n",
    "    columns=['state-1', 'state-2', 'state-3'])\n",
    "    state_prob_means=[state_probs_df.iloc[:,i].mean()\n",
    "    for i in range(len(state_probs_df.columns))]\n",
    "    if np.max(state_prob_means)==state_prob_means[0]:\n",
    "        print('State-1 is likely to occur with a probability of {:4f}'.format(state_prob_means[0]))\n",
    "    elif np.max(state_prob_means)==state_prob_means[1]:\n",
    "        print('State-2 is likely to occur with a probability of {:4f}'.format(state_prob_means[1]))\n",
    "    else:\n",
    "        print('State-3 is likely to occur with a probability of {:4f}'.format(state_prob_means[2]))\n",
    "    return state_probs "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "State-3 is likely to occur with a probability of 0.549953\n",
      "State probabilities are [0.38713538 0.06291121 0.54995341]\n"
     ]
    }
   ],
   "source": [
    "state_probs=cluster_state(scaled_liq, 3)\n",
    "print(f'State probabilities are {state_probs.mean(axis=0)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.decomposition import PCA \n",
    "\n",
    "pca=PCA(n_components=11)\n",
    "components=pca.fit_transform(scaled_liq)\n",
    "plt.plot(pca.explained_variance_ratio_)\n",
    "plt.title('Scree Plot')\n",
    "plt.xlabel('Number of Components')\n",
    "plt.ylabel('% of Explained Variance')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "def gmm_pca(data, nstate):\n",
    "    pca=PCA(n_components=3)\n",
    "    components=pca.fit_transform(data)\n",
    "    mxtd=GaussianMixture(n_components=nstate, covariance_type='spherical')\n",
    "    gmm=mxtd.fit(components)\n",
    "    labels=gmm.predict(components)\n",
    "    state_probs=gmm.predict_proba(components)\n",
    "    return state_probs, pca \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "State probabilities are [0.01626015 0.24913847 0.73460138]\n"
     ]
    }
   ],
   "source": [
    "state_probs, pca=gmm_pca(scaled_liq, 3)\n",
    "print(f'State probabilities are {state_probs.mean(axis=0)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[9.31402915e-30, 3.93192273e-02, 9.60680773e-01],\n",
       "       [4.69720066e-27, 2.91446997e-01, 7.08553003e-01],\n",
       "       [1.55509285e-33, 8.43360623e-02, 9.15663938e-01],\n",
       "       ...,\n",
       "       [1.00000000e+00, 2.10084170e-10, 3.88534013e-32],\n",
       "       [9.99993949e-01, 6.05079723e-06, 2.19210255e-22],\n",
       "       [9.99999435e-01, 5.65415554e-07, 4.32335123e-23]])"
      ]
     },
     "execution_count": 81,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "compo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "State-3 is likely to occur with a probability of 0.734601\n"
     ]
    }
   ],
   "source": [
    "def wpc():\n",
    "    state_probs_df=pd.DataFrame(state_probs, columns=['state-1', 'state-2', 'state-3'])\n",
    "    state_prob_means=[state_probs_df.iloc[:,i].mean()\n",
    "    for i in range(len(state_probs_df.columns))]\n",
    "    if np.max(state_prob_means)==state_prob_means[0]:\n",
    "        print('State-1 is likely to occur with a probability of {:4f}'.format(state_prob_means[0]))\n",
    "    elif np.max(state_prob_means)==state_prob_means[1]:\n",
    "        print('State-2 is likely to occur with a probability of {:4f}'.format(state_prob_means[1]))\n",
    "    else:\n",
    "        print('State-3 is likely to occur with a probability of {:4f}'.format(state_prob_means[2]))\n",
    "\n",
    "wpc()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "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": "3.10.10"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}