event_study_CAR_SURE.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "82a3147a",
   "metadata": {},
   "source": [
    "# Event-study"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "dcdb2d2a",
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "06b2a5cd",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import wrds\n",
    "import re\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import re\n",
    "from cleanco import basename\n",
    "from tqdm.notebook import tqdm\n",
    "import dask\n",
    "dask.config.set(scheduler=\"processes\")\n",
    "from tqdm.dask import TqdmCallback\n",
    "from linearmodels.system import SUR\n",
    "import seaborn as sns\n",
    "#sns.set(rc={'figure.figsize':(16,8)})\n",
    "import plotly.graph_objects as go\n",
    "from collections import OrderedDict\n",
    "import scipy.stats as stats\n",
    "import statsmodels.api as sm\n",
    "import matplotlib.ticker as mtick"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "6b940649",
   "metadata": {},
   "outputs": [],
   "source": [
    "#import functions from functions.py\n",
    "from function_definitions import *"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5274439b",
   "metadata": {},
   "source": [
    "# STEP 1 : estimate alpha and betas wrt CAPM and FF models\n",
    "\n",
    "## a) $R_i^e = \\alpha_i +\\beta_i^M R_M^e + \\beta_i^S R_{SMB}^e + \\beta_i^H R_{HML}^e+\\sum AR_{event_k}1_{event_k}$\n",
    "## b) $R_i^e = \\alpha_i +\\sum AR_{event_k}1_{event_k}$  (zero-benchmark)\n",
    "## matrix notation: $R_i^e = X \\beta_i $\n",
    "\n",
    "### $ \\begin{bmatrix} R_{i,1}^e \\\\ ... \\\\ R_{i,m}^e \\end{bmatrix} = \\begin{bmatrix} 1 & R_{M,1}^e & R_{SMB,1}^e & R_{HML,1}^e & 0 & 0 & 0 & ... & 0\\\\ 1 & R_{M,1}^e & R_{SMB,1}^e & R_{HML,1}^e & 1 & 0 & 0 & ...& 0 \\\\  1 & R_{M,3}^e &  R_{SMB,3}^e & R_{HML,3}^e & 0 & 1 & 0 & ... & 0 \\\\  1 & R_{M,4}^e &  R_{SMB,4}^e & R_{HML,4}^e & 0 & 0 & 1 & ... & 0 \\\\... & ... & ... & ... & ... & ... & ... & ... \\\\ 1 & R_{M,m}^e & R_{SMB,m}^e & R_{HML,m}^e & 0 & 0 & 0 &... & 0 \\end{bmatrix} \\begin{bmatrix} \\alpha_i \\\\ \\beta_i^M \\\\ \\beta_i^S \\\\ \\beta_i^H \\\\ AR_{event_0} \\\\ ... \\\\ AR_{event_k} \\end{bmatrix}$  \n",
    "\n",
    "###  m = number of days in the whole period, k = number of events pertaining to firm i  \n",
    "\n",
    "### use seemingly unrelated regressions: \n",
    "\n",
    "### $ \\begin{bmatrix} R_{1}^e \\\\ R_2^e \\\\ ... \\\\ R_{n}^e \\end{bmatrix} = \\begin{bmatrix} X & 0 & ... & 0 \\\\ 0 & X & ... & 0\\\\ ... & & ... & ... \\\\  0 & 0 & ... &  X \\end{bmatrix} \\begin{bmatrix} \\beta_1 \\\\ \\beta_2 \\\\ ... \\\\ \\beta_n \\\\ AR_{event_0} \\\\ ... \\\\ AR_{event_k} \\end{bmatrix} + \\begin{bmatrix} \\epsilon_1 \\\\ \\epsilon_2 \\\\ ... \\\\ \\epsilon_n \\end{bmatrix}$  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "1e082eb3",
   "metadata": {},
   "outputs": [],
   "source": [
    "#load returns\n",
    "returns = pd.read_csv('../data/event_study/excess_returns.csv', index_col = 'date')\n",
    "returns.index = pd.to_datetime(returns.index)\n",
    "\n",
    "#load refinitiv data\n",
    "refinitiv_data = pd.read_csv('../data/event_study/refinitiv_data.csv')\n",
    "refinitiv_data.Date_day = pd.to_datetime(refinitiv_data.Date_day)\n",
    "#load market caps\n",
    "mkt_caps = pd.read_csv('../data/event_study/market_caps.csv', index_col = 'date')\n",
    "mkt_caps.index = pd.to_datetime(mkt_caps.index)\n",
    "#load the event dataframe\n",
    "event_df = pd.read_csv('../data/event_study/event_df.csv', index_col = 'date')\n",
    "event_df.index = pd.to_datetime(event_df.index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "0ee78686",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Factor returns\n",
    "FF5 = pd.read_csv('../data/FF_factors/FF5_daily.csv',skiprows = 3, index_col = 0).astype(float)\n",
    "FF5 = FF5/100 # in percent?????\n",
    "FF5.index.rename('date', inplace = True)\n",
    "FF5.index = pd.to_datetime(FF5.index, format = '%Y%m%d')\n",
    "momentum = pd.read_csv('../data/FF_factors/Momentum_daily.csv',skiprows = 13, index_col = 0)[:-1].astype(float)\n",
    "momentum = momentum/100 # in percent?????\n",
    "momentum.index.rename('date', inplace = True)\n",
    "momentum.index = pd.to_datetime(momentum.index, format = '%Y%m%d')\n",
    "momentum.columns = ['UMD']\n",
    "FF5 = FF5.merge(momentum,left_index = True,right_index = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "ba005bc7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0ba1d61af5da4be19a875cb6da1dac64",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Building event indicator vectors:   0%|          | 0/116 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#build the indicator vectors for each event of each firm\n",
    "event_indicator = {}\n",
    "indexing_date = {}\n",
    "indexing_index = {}\n",
    "nb_events_dropped_mkt_cap = 0\n",
    "nb_events_dropped_large_day_delta = 0\n",
    "nb_events_dropped_exception  = 0\n",
    "total_events = 0\n",
    "not_dropped = 0\n",
    "\n",
    "for firm in tqdm(returns.columns, desc = 'Building event indicator vectors'):\n",
    "    events = refinitiv_data[refinitiv_data.CUSIP == firm].Date_day.values\n",
    "    precise_events = refinitiv_data[refinitiv_data.CUSIP == firm].Date.values\n",
    "    nb_events = len(events)\n",
    "\n",
    "    df = pd.DataFrame(np.zeros([returns.shape[0],nb_events*11]), index = returns.index,\n",
    "                      columns = [f\"event{i//11}_day{(i%11)-5:+}\" for i in range(nb_events*11)])\n",
    "    \n",
    "    retained = []\n",
    "    closest_indexes = []\n",
    "    for i,event in enumerate(events):\n",
    "        event = pd.to_datetime(event)\n",
    "        \n",
    "        if firm == '30303M10' and i == 9:\n",
    "            closest_indexes.append(None)\n",
    "            continue #overlaps with another cyber event\n",
    "        \n",
    "        total_events += 1\n",
    "        \n",
    "        #the day of the event could be a holiday or weekend,\n",
    "        #hence find the first trading day after the event and use\n",
    "        #a 3 day window around that day\n",
    "        timedelta = (event-df.index)\n",
    "        try:\n",
    "            closest_index = min((j for j, td in enumerate(timedelta) if td.total_seconds() <= 0))\n",
    "                                 #key=lambda j: abs(timedelta[j].total_seconds()))\n",
    "            \n",
    "                \n",
    "            #if after trading hours, take the next day\n",
    "            if pd.to_datetime(precise_events[i]).hour > 16:\n",
    "                closest_index +=1\n",
    "            \n",
    "            #do not consider events pertaining to low mkt cap firms\n",
    "            if mkt_caps[firm].iloc[closest_index] < 300e6:\n",
    "                nb_events_dropped_mkt_cap += 1 \n",
    "                closest_indexes.append(None)\n",
    "                continue\n",
    "            \n",
    "            #drop if the window is larger than 5 days\n",
    "            nb_actual_days = (df.index[closest_index+1]-df.index[closest_index-1]).days+1\n",
    "            if nb_actual_days > 5:\n",
    "                nb_events_dropped_large_day_delta += 1\n",
    "                closest_indexes.append(None)\n",
    "                continue\n",
    "\n",
    "            #3 vectors per event with a single one each\n",
    "            df.iloc[closest_index,   i*11+5] = 1 #event day\n",
    "            df.iloc[closest_index-1, i*11+4] = 1\n",
    "            df.iloc[closest_index+1, i*11+6] = 1\n",
    "            \n",
    "            #extra window for analysis\n",
    "            df.iloc[closest_index-2, i*11+3] = 1\n",
    "            df.iloc[closest_index-3, i*11+2] = 1\n",
    "            df.iloc[closest_index-4, i*11+1] = 1\n",
    "            df.iloc[closest_index-5, i*11] = 1\n",
    "            \n",
    "            df.iloc[closest_index+2, i*11+7] = 1\n",
    "            df.iloc[closest_index+3, i*11+8] = 1\n",
    "            df.iloc[closest_index+4, i*11+9] = 1\n",
    "            df.iloc[closest_index+5, i*11+10] = 1\n",
    "            \n",
    "            closest_indexes.append(closest_index)\n",
    "            retained.append(i)\n",
    "            \n",
    "            not_dropped +=1\n",
    "            \n",
    "        except:\n",
    "            closest_indexes.append(None)\n",
    "            nb_events_dropped_exception +=1\n",
    "            if event.year != 2023:\n",
    "                firm_ = firm\n",
    "                event_ = event\n",
    "            \n",
    "            \n",
    "    # Drop the columns with all zeros\n",
    "    zero_columns = df.columns[(df == 0).all()]\n",
    "    df = df.drop(columns=zero_columns)\n",
    "    \n",
    "    if df.shape[1] == 0:\n",
    "        event_indicator[firm] = []\n",
    "    else:\n",
    "        event_indicator[firm] = df\n",
    "    indexing_date[firm] = [date if index in retained else None for index, date in enumerate(events)]\n",
    "    indexing_index[firm] = closest_indexes\n",
    "\n",
    "nb_events_step3 = not_dropped"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b63a83db",
   "metadata": {},
   "source": [
    "### Seemingly Unrelated Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "47024c5f",
   "metadata": {},
   "outputs": [],
   "source": [
    "#load the optimal sample parameters\n",
    "optimal_sample = pd.read_csv('../data/event_study/optimal_sample.csv', index_col = 0)\n",
    "x_ = optimal_sample.loc['sample_start','idx']\n",
    "y_ = optimal_sample.loc['sample_end','idx']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "212b9aa7",
   "metadata": {},
   "outputs": [],
   "source": [
    "SUR_data = OrderedDict()\n",
    "FF3_factors = ['Mkt-RF','HML','SMB']\n",
    "\n",
    "exog_ = {}\n",
    "nb_events_end = 0\n",
    "shapes = []\n",
    "\n",
    "for firm in returns.columns:\n",
    "    dependent = returns[firm].iloc[x_:y_].dropna()\n",
    "    shapes.append(dependent.shape[0])\n",
    "    \n",
    "    #drop firms that don't have all observations \n",
    "    #(exog and dependent have to be the same length for all regressions)\n",
    "    if dependent.shape[0] != y_-x_:\n",
    "        continue\n",
    "    \n",
    "    exog = sm.add_constant(FF5.loc[dependent.index, FF3_factors])\n",
    "    if len(event_indicator[firm]):\n",
    "        event_vectors = event_indicator[firm].loc[dependent.index,:]\n",
    "        zero_columns = event_vectors.columns[(event_vectors == 0).all()]\n",
    "        event_vectors = event_vectors.drop(columns=zero_columns)\n",
    "        nb_events_end += event_vectors.shape[1]/11\n",
    "        \n",
    "        exog = pd.concat([exog, event_vectors], axis = 1)\n",
    "    \n",
    "    if exog.shape[1] > 4:\n",
    "        SUR_data[firm] = {\"dependent\" : dependent, \"exog\" : exog}\n",
    "        exog_[firm] = exog"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "48409782",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "126 events relating to 47 firms.\n"
     ]
    }
   ],
   "source": [
    "#stats\n",
    "nb_firms_end = len(SUR_data)\n",
    "print(f'{nb_events_end:.0f} events relating to {nb_firms_end} firms.')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "5a2e931d",
   "metadata": {},
   "outputs": [],
   "source": [
    "#fit SUR model\n",
    "model = SUR(SUR_data)\n",
    "res = model.fit()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "b63364a1",
   "metadata": {},
   "outputs": [],
   "source": [
    "#correlation matrix of residuals\n",
    "resid_corr = res.resids.corr()\n",
    "\n",
    "#cross_correlations (upper triangle)\n",
    "cross_corrs = resid_corr.where(np.triu(np.ones(res.resids.corr().shape),k = 1).astype(bool))\n",
    "\n",
    "#mean cross-correlation of residuals\n",
    "r = np.nanmean(cross_corrs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "65b317e8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1600x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The average cross-correlation of residuals is 0.012\n"
     ]
    }
   ],
   "source": [
    "plt.figure(figsize = (16,8))\n",
    "ax = sns.heatmap(resid_corr,vmax=0.8,  square=True)\n",
    "ax.set(xticklabels=[], yticklabels = [])\n",
    "plt.show()\n",
    "\n",
    "print(f'The average cross-correlation of residuals is {round(r,3)}')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2464b002",
   "metadata": {},
   "source": [
    "# STEP 2 : test the significance of the mean CAR\n",
    "\n",
    "## scaled abnormal returns: $A_{i,t} = \\frac{AR_{i,t}}{s_i \\sqrt{1+ x_t'(X'X)^{-1}x_t}}$  \n",
    "### $s_i$ is residual standard deviation (from step 1?), X is the matrix of explanatory variable values (from step 1?) and $x_t$ is vector of explanatory variable values on event day t (row t of X?)\n",
    "\n",
    "## feasible estimator of the variance of A: $s_A^2 = \\frac{s^2}{1-\\bar{r}}$\n",
    "### $s^2$ is the sample cross-sectional variance of event-day scaled abnormal returns and $\\bar{r}$ the average sample cross-correlation of the residuals (from step 1?)\n",
    "\n",
    "# ADJ BMP t-statistic: $t_{AB} =  \\frac{\\bar{A} \\sqrt{n}}{s_A \\sqrt{1+(n-1)\\bar{r}}} \\sim N(0,1)?$   \n",
    "### n is the number of firms\n",
    "## same formula for CARs but replace $\\bar{A}$ with the scaled cumulative abnormal returns and $s_A$ with the cross-sectional standard deviation of SCARs\n",
    "\n",
    "# ADJ PATELL t-statistic: $t_{AB} =  \\frac{\\bar{A} \\sqrt{n}}{\\sqrt{(m-p-1)/(m-p-3)}\\sqrt{1+(n-1)\\bar{r}}} \\sim N(0,1)?$   \n",
    "### p is the number of explanatory variables in the expected return regression\n",
    "### m = number of days in the whole period"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "9b115b55",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean CAR = -0.78%\n",
      "t-stat = -2.434 and p value = 0.016\n"
     ]
    }
   ],
   "source": [
    "#abnormal return\n",
    "AR = res.params.T.filter(like = 'event')\n",
    "AR = AR.reset_index()\n",
    "AR[['CUSIP', 'event', 'day']] = AR['index'].str.extract(r'(.+)_event(\\d+)_day(.+)')\n",
    "\n",
    "AR['t'] = np.nan\n",
    "for firm in AR.CUSIP.unique():\n",
    "    \n",
    "    ts = [value-x_ for i,value in enumerate(indexing_index[firm]) if (value is not None)\n",
    "          and (indexing_date[firm][i] >= returns.index[x_]) \n",
    "          and (indexing_date[firm][i] <= returns.index[y_])]\n",
    "    for day in ['-5','-4','-3','-2','-1','+0','+1','+2','+3','+4','+5']:\n",
    "        AR.loc[(AR.CUSIP == firm) & (AR.day == day),'t'] = ts\n",
    "\n",
    "#cumulative abnormal return\n",
    "CAR = AR[AR.day.isin(['-1','+0','+1'])].groupby(['CUSIP','event']).params.apply(cumulate_returns)\n",
    "plt.hist(CAR, bins = 20)\n",
    "plt.title('Distribution of CARs')\n",
    "plt.show()\n",
    "print(f'Mean CAR = {CAR.mean():.2%}')\n",
    "stat = sm.OLS(CAR.values,np.ones(len(CAR.values))).fit()\n",
    "print(f't-stat = {round(stat.tvalues[0],3)} and p value = {round(stat.pvalues[0],3)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "c042b6f4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mean_daily_AR = AR.groupby('day').mean().params.loc[['-5','-4','-3','-2','-1','+0','+1','+2','+3','+4','+5']]\n",
    "mean_daily_AR *= 100\n",
    "ax = mean_daily_AR.plot(figsize = (12,5))\n",
    "plt.ylim([-0.6,0.7])\n",
    "plt.xticks(range(0,11),['-5','-4','-3','-2','-1','0','+1','+2','+3','+4','+5'])\n",
    "plt.axhline(0, color='gray', linestyle='--')\n",
    "ax.yaxis.set_major_formatter(mtick.PercentFormatter())\n",
    "plt.title('Average abnormal returns')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "7bdbcbd0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cum_mean_daily_AR = (AR.groupby('day').mean().params+1)\n",
    "cum_mean_daily_AR = cum_mean_daily_AR.loc[['-5','-4','-3','-2','-1','+0','+1','+2','+3','+4','+5']]\n",
    "cum_mean_daily_AR = cum_mean_daily_AR.cumprod()-1\n",
    "cum_mean_daily_AR *= 100\n",
    "ax = cum_mean_daily_AR.plot(figsize = (12,5))\n",
    "plt.ylim([-0.4,1.1])\n",
    "plt.xticks(range(0,11),['-5','-4','-3','-2','-1','0','+1','+2','+3','+4','+5'])\n",
    "plt.axhline(0, color='gray', linestyle='--')\n",
    "ax.yaxis.set_major_formatter(mtick.PercentFormatter())\n",
    "plt.title('Cumulative average abnormal returns')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "86434c71",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fef8612409094c6b90ddebab6ea2b6f3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Computing SARs:   0%|          | 0/47 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#regression residual standard deviation\n",
    "s = res.resids.std()\n",
    "SAR = {}\n",
    "\n",
    "for firm in tqdm(SUR_data.keys(), desc = 'Computing SARs'):\n",
    "    si = s[firm]\n",
    "    \n",
    "    X = exog_[firm]\n",
    "    \n",
    "    SAR[firm] = pd.DataFrame(np.ones(X.shape[0])*np.nan, index = X.index)\n",
    "    \n",
    "    ts = [value-x_ for i,value in enumerate(indexing_index[firm]) if (value is not None)\n",
    "          and (indexing_date[firm][i] >= returns.index[x_]) \n",
    "          and (indexing_date[firm][i] <= returns.index[y_])]\n",
    "    \n",
    "    for t in ts:\n",
    "        for day in ['-1','+0','+1']:\n",
    "            x = X.iloc[t+int(day)]\n",
    "    \n",
    "            #correction term\n",
    "            d = x.T @ np.linalg.inv(X.T @ X) @ x\n",
    "        \n",
    "            ARit = AR[(AR.CUSIP == firm)  & (AR.day == day) & (AR.t == t)].params.values[0]\n",
    "        \n",
    "            SAR[firm].loc[X.index[t+int(day)], :] = ARit/(si*np.sqrt(1+d))\n",
    "\n",
    "SAR = pd.concat(SAR, axis = 1)\n",
    "SAR.columns = SAR.columns.get_level_values(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "a2f9a9e1",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ADJ-BMP = -1.037\n",
      "p value = 0.3\n",
      "\n",
      "\n",
      "ADJ-PATELL = -1.811\n",
      "p value = 0.07\n"
     ]
    }
   ],
   "source": [
    "#compute SCARs\n",
    "SCAR = SAR.apply(product_of_consecutive_non_nulls)\n",
    "\n",
    "#mean \n",
    "SCAR_mean = SCAR.explode().mean()\n",
    "\n",
    "#cross-sectional standard deviation of SCARs\n",
    "sa = np.sqrt(SCAR.explode().var()/(1-r))\n",
    "\n",
    "#number of firms\n",
    "n = len(SUR_data.keys())\n",
    "\n",
    "#ADJ-BMP test statistic\n",
    "\n",
    "ADJ_BMP = (SCAR_mean*np.sqrt(n))/(sa*np.sqrt(1+(n-1)*r))\n",
    "print(f'ADJ-BMP = {round(ADJ_BMP,3)}')\n",
    "\n",
    "p_value = 2 * (1 - stats.norm.cdf(abs(ADJ_BMP)))\n",
    "print(f'p value = {round(p_value,3)}')\n",
    "print('\\n')\n",
    "\n",
    "m = y_-x_\n",
    "p = 3 #nb of explanatory variables in the expected return regression\n",
    "\n",
    "#ADJ-PATELL test statistic\n",
    "ADJ_PATELL = (SCAR_mean*np.sqrt(n))/(np.sqrt((m-p-1)/(m-p-3))*np.sqrt(1+(n-1)*r))\n",
    "print(f'ADJ-PATELL = {round(ADJ_PATELL,3)}')\n",
    "\n",
    "p_value = 2 * (1 - stats.norm.cdf(abs(ADJ_PATELL)))\n",
    "print(f'p value = {round(p_value,3)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "016daeaf",
   "metadata": {},
   "outputs": [],
   "source": [
    "#save data\n",
    "CAR.to_csv('../data/event_study/CAR_SURE.csv')\n",
    "AR.to_csv('../data/event_study/AR_SURE.csv', index = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4d3c06a7",
   "metadata": {},
   "source": [
    "## Zero benchmark regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "4c0dac58",
   "metadata": {},
   "outputs": [],
   "source": [
    "SUR_data = OrderedDict()\n",
    "\n",
    "exog_ = {}\n",
    "nb_events_end = 0\n",
    "shapes = []\n",
    "\n",
    "for firm in returns.columns:\n",
    "    dependent = returns[firm].iloc[x_:y_].dropna()\n",
    "    shapes.append(dependent.shape[0])\n",
    "    \n",
    "    #drop firms that don't have all observations \n",
    "    #(exog and dependent have to be the same length for all regressions)\n",
    "    if dependent.shape[0] != y_-x_:\n",
    "        continue\n",
    "    \n",
    "    exog = pd.DataFrame()\n",
    "    if len(event_indicator[firm]):\n",
    "        event_vectors = event_indicator[firm].loc[dependent.index,:]\n",
    "        zero_columns = event_vectors.columns[(event_vectors == 0).all()]\n",
    "        event_vectors = event_vectors.drop(columns=zero_columns)\n",
    "        nb_events_end += event_vectors.shape[1]/11\n",
    "        \n",
    "        exog = sm.add_constant(event_vectors)\n",
    "    \n",
    "    if exog.shape[1] > 1:\n",
    "        SUR_data[firm] = {\"dependent\" : dependent, \"exog\" : exog}\n",
    "        exog_[firm] = exog"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "45421597",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "126 events relating to 47 firms.\n"
     ]
    }
   ],
   "source": [
    "#stats\n",
    "nb_firms_end = len(SUR_data)\n",
    "print(f'{nb_events_end:.0f} events relating to {nb_firms_end} firms.')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "acde13bc",
   "metadata": {},
   "outputs": [],
   "source": [
    "#fit SUR model\n",
    "model = SUR(SUR_data)\n",
    "res = model.fit()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "23cd2e43",
   "metadata": {},
   "outputs": [],
   "source": [
    "#correlation matrix of residuals\n",
    "resid_corr = res.resids.corr()\n",
    "\n",
    "#cross_correlations (upper triangle)\n",
    "cross_corrs = resid_corr.where(np.triu(np.ones(res.resids.corr().shape),k = 1).astype(bool))\n",
    "\n",
    "#mean cross-correlation of residuals\n",
    "r = np.nanmean(cross_corrs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "37cea9d7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1600x800 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The average cross-correlation of residuals is 0.315\n"
     ]
    }
   ],
   "source": [
    "plt.figure(figsize = (16,8))\n",
    "ax = sns.heatmap(resid_corr,vmax=0.8,  square=True)\n",
    "ax.set(xticklabels=[], yticklabels = [])\n",
    "plt.show()\n",
    "\n",
    "print(f'The average cross-correlation of residuals is {round(r,3)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "cf4b8beb",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean CAR = -0.79%\n",
      "t-stat = -2.325 and p value = 0.022\n"
     ]
    }
   ],
   "source": [
    "#abnormal return\n",
    "AR = res.params.T.filter(like = 'event')\n",
    "AR = AR.reset_index()\n",
    "AR[['CUSIP', 'event', 'day']] = AR['index'].str.extract(r'(.+)_event(\\d+)_day(.+)')\n",
    "\n",
    "AR['t'] = np.nan\n",
    "for firm in AR.CUSIP.unique():\n",
    "    \n",
    "    ts = [value-x_ for i,value in enumerate(indexing_index[firm]) if (value is not None)\n",
    "          and (indexing_date[firm][i] >= returns.index[x_]) \n",
    "          and (indexing_date[firm][i] <= returns.index[y_])]\n",
    "    for day in ['-5','-4','-3','-2','-1','+0','+1','+2','+3','+4','+5']:\n",
    "        AR.loc[(AR.CUSIP == firm) & (AR.day == day),'t'] = ts\n",
    "\n",
    "#cumulative abnormal return\n",
    "CAR = AR[AR.day.isin(['-1','+0','+1'])].groupby(['CUSIP','event']).params.apply(cumulate_returns)\n",
    "plt.hist(CAR, bins = 20)\n",
    "plt.title('Distribution of CARs')\n",
    "plt.show()\n",
    "print(f'Mean CAR = {CAR.mean():.2%}')\n",
    "stat = sm.OLS(CAR.values,np.ones(len(CAR.values))).fit()\n",
    "print(f't-stat = {round(stat.tvalues[0],3)} and p value = {round(stat.pvalues[0],3)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "94bf5d43",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mean_daily_AR = AR.groupby('day').mean().params.loc[['-5','-4','-3','-2','-1','+0','+1','+2','+3','+4','+5']]\n",
    "mean_daily_AR *= 100\n",
    "ax = mean_daily_AR.plot(figsize = (12,5))\n",
    "plt.ylim([-0.8,0.6])\n",
    "plt.xticks(range(0,11),['-5','-4','-3','-2','-1','0','+1','+2','+3','+4','+5'])\n",
    "plt.axhline(0, color='gray', linestyle='--')\n",
    "ax.yaxis.set_major_formatter(mtick.PercentFormatter())\n",
    "plt.title('Average abnormal returns')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "44a10bee",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "cum_mean_daily_AR = (AR.groupby('day').mean().params+1)\n",
    "cum_mean_daily_AR = cum_mean_daily_AR.loc[['-5','-4','-3','-2','-1','+0','+1','+2','+3','+4','+5']]\n",
    "cum_mean_daily_AR = cum_mean_daily_AR.cumprod()-1\n",
    "cum_mean_daily_AR *= 100\n",
    "ax = cum_mean_daily_AR.plot(figsize = (12,5))\n",
    "plt.ylim([-0.4,1.1])\n",
    "plt.xticks(range(0,11),['-5','-4','-3','-2','-1','0','+1','+2','+3','+4','+5'])\n",
    "plt.axhline(0, color='gray', linestyle='--')\n",
    "ax.yaxis.set_major_formatter(mtick.PercentFormatter())\n",
    "plt.title('Cumulative average abnormal returns')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "bbdb5f3f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "900d392995684901a840c30a66429d0c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Computing SARs:   0%|          | 0/47 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#regression residual standard deviation\n",
    "s = res.resids.std()\n",
    "SAR = {}\n",
    "\n",
    "for firm in tqdm(SUR_data.keys(), desc = 'Computing SARs'):\n",
    "    si = s[firm]\n",
    "    \n",
    "    X = exog_[firm]\n",
    "    \n",
    "    SAR[firm] = pd.DataFrame(np.ones(X.shape[0])*np.nan, index = X.index)\n",
    "    \n",
    "    ts = [value-x_ for i,value in enumerate(indexing_index[firm]) if (value is not None)\n",
    "          and (indexing_date[firm][i] >= returns.index[x_]) \n",
    "          and (indexing_date[firm][i] <= returns.index[y_])]\n",
    "    \n",
    "    for t in ts:\n",
    "        for day in ['-1','+0','+1']:\n",
    "            x = X.iloc[t+int(day)]\n",
    "    \n",
    "            #correction term\n",
    "            d = x.T @ np.linalg.inv(X.T @ X) @ x\n",
    "        \n",
    "            ARit = AR[(AR.CUSIP == firm)  & (AR.day == day) & (AR.t == t)].params.values[0]\n",
    "        \n",
    "            SAR[firm].loc[X.index[t+int(day)], :] = ARit/(si*np.sqrt(1+d))\n",
    "\n",
    "SAR = pd.concat(SAR, axis = 1)\n",
    "SAR.columns = SAR.columns.get_level_values(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "e17aa94a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ADJ-BMP = -0.273\n",
      "p value = 0.785\n",
      "\n",
      "\n",
      "ADJ-PATELL = -0.455\n",
      "p value = 0.649\n"
     ]
    }
   ],
   "source": [
    "#compute SCARs\n",
    "SCAR = SAR.apply(product_of_consecutive_non_nulls)\n",
    "\n",
    "#mean \n",
    "SCAR_mean = SCAR.explode().mean()\n",
    "\n",
    "#cross-sectional standard deviation of SCARs\n",
    "sa = np.sqrt(SCAR.explode().var()/(1-r))\n",
    "\n",
    "#number of firms\n",
    "n = len(SUR_data.keys())\n",
    "\n",
    "#ADJ-BMP test statistic\n",
    "\n",
    "ADJ_BMP = (SCAR_mean*np.sqrt(n))/(sa*np.sqrt(1+(n-1)*r))\n",
    "print(f'ADJ-BMP = {round(ADJ_BMP,3)}')\n",
    "\n",
    "p_value = 2 * (1 - stats.norm.cdf(abs(ADJ_BMP)))\n",
    "print(f'p value = {round(p_value,3)}')\n",
    "print('\\n')\n",
    "\n",
    "m = y_-x_\n",
    "p = 3 #nb of explanatory variables in the expected return regression\n",
    "\n",
    "#ADJ-PATELL test statistic\n",
    "ADJ_PATELL = (SCAR_mean*np.sqrt(n))/(np.sqrt((m-p-1)/(m-p-3))*np.sqrt(1+(n-1)*r))\n",
    "print(f'ADJ-PATELL = {round(ADJ_PATELL,3)}')\n",
    "\n",
    "p_value = 2 * (1 - stats.norm.cdf(abs(ADJ_PATELL)))\n",
    "print(f'p value = {round(p_value,3)}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "id": "43e0fa15",
   "metadata": {},
   "outputs": [],
   "source": [
    "#save data\n",
    "CAR.to_csv('../data/event_study/CAR_SURE_0B.csv')\n",
    "AR.to_csv('../data/event_study/AR_SURE_0B.csv', index = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6a2cf0c7",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "1e4339f7",
   "metadata": {},
   "outputs": [],
   "source": [
    "refi_copy = refinitiv_data.copy()\n",
    "indexes = []\n",
    "for firm in CAR.index.get_level_values(0).unique():\n",
    "    for event in CAR[[firm]].index.get_level_values(1).astype(int).unique():\n",
    "        indexes.append(refi_copy[refi_copy.CUSIP == firm].iloc[[event]].index[0])\n",
    "#refinitiv_data.loc[indexes,:].to_csv('end_data_refinitiv.csv')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.9.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}