Empirical Analysis - Replication.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### *Hong Sraer (2016),  \"Speculative Betas\"*\n",
    "\n",
    "*Data retrieved from Prof. Sraer's website at http://faculty.haas.berkeley.edu/dsraer/data.html*"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Loading .dta files (from stata) into python/pandas and creating date index that is useful for plotting."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import os\n",
    "import matplotlib.pyplot as plt\n",
    "import statsmodels.formula.api as sm\n",
    "import numpy as np\n",
    "data20 = pd.io.stata.read_stata('portfolios/20_beta_sorted_portfolios.dta')\n",
    "data20['date'] = data20.apply(lambda x: str(int(x['year'])) + '/' + str(int(x['month'])), axis=1)\n",
    "data20.set_index('date', inplace=True)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# Defining lagged variables from data\n",
    "lags = [3,6,9,12,18]\n",
    "for lag in lags:\n",
    "     data20['mktrf_lag' + str(lag)] = data20['mktrf'].shift(lag)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We couldnt get BETA_1 variable, the post-ranking full sample beta of each portfolio. Attempting to reconstruct it from the data below. However, data uses daily returns to compute the beta, whereas we have monthly results. Spread is still extremely wide, and results are similar for all but the bottom two portfolios (compare with descriptive statistics on Table 1 of paper)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_beta(df):\n",
    "    '''\n",
    "    df: dataframe for each group, ranked by betas. \n",
    "    '''\n",
    "    # Regressing portfolio returns on the current and lagged market returns\n",
    "    # Ran first controlling for serial correlation and heteroskedasticity using newey-west std errors\n",
    "    # Results were similar than just the usual OLS, so left it as is (also no specification in paper).\n",
    "\n",
    "#### RESULTS USING NEWEY WEST    \n",
    "#     results = sm.ols(formula=\"ret_rf ~ mktrf + mktrf_lag3 + mktrf_lag6 + mktrf_lag9 + mktrf_lag12 + mktrf_lag18\", \n",
    "#                      data=df).fit(cov_type='HAC',cov_kwds={'maxlags':17})\n",
    "    \n",
    "    results = sm.ols(formula=\"ret_rf ~ mktrf + mktrf_lag3 + mktrf_lag6 + mktrf_lag9 + mktrf_lag12 + mktrf_lag18\", \n",
    "                 data=df).fit()\n",
    "\n",
    "    params = results.params\n",
    "    # Beta computed as sum of the coefficients of all lags of market returns, per the paper\n",
    "    out = params[1:].sum()\n",
    "    return(out)\n",
    "    \n",
    "# Full sample postranking betas computed, so we just need to groupby portfolios and do regression\n",
    "betas = data20.groupby('group').apply(lambda x: get_beta(x))\n",
    "betas_df = pd.DataFrame(betas).rename(columns={0: \"BETA_1\"})\n",
    "betas_df\n",
    "data20 = data20.merge(betas_df, how='left', left_on=['group'], right_index=True)\n",
    "data20['BETA_2'] = data20['BETA_1'].pow(2)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Displaying betas for each group"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "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>Betas</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.159782</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.251669</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.400479</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.550107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.557424</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.624215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.677083</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.672397</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.809993</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.808146</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.904018</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.965724</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.991483</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>1.076860</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>1.133268</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1.173566</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>1.273700</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>1.396761</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>1.413388</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>1.881448</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Betas\n",
       "1   0.159782\n",
       "2   0.251669\n",
       "3   0.400479\n",
       "4   0.550107\n",
       "5   0.557424\n",
       "6   0.624215\n",
       "7   0.677083\n",
       "8   0.672397\n",
       "9   0.809993\n",
       "10  0.808146\n",
       "11  0.904018\n",
       "12  0.965724\n",
       "13  0.991483\n",
       "14  1.076860\n",
       "15  1.133268\n",
       "16  1.173566\n",
       "17  1.273700\n",
       "18  1.396761\n",
       "19  1.413388\n",
       "20  1.881448"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "betas_display = pd.DataFrame({'Betas': data20['BETA_1'].unique()})\n",
    "betas_display.index = range(1,21)\n",
    "betas_display"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Replicating figure 6\n",
    "\n",
    "0 = Low\n",
    "1 = Mid\n",
    "2 = High"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "quartiles = data20['aggregate_disp_daily'].quantile([.25,.75])\n",
    "low_q = quartiles.loc[.25]\n",
    "high_q = quartiles.loc[.75]\n",
    "data20['boolean_disagreement'] = 1\n",
    "data20['boolean_disagreement'] = np.where(data20['aggregate_disp_daily'] <= low_q, 0, data20['boolean_disagreement'])\n",
    "data20['boolean_disagreement'] = np.where(data20['aggregate_disp_daily'] >= high_q, 2, data20['boolean_disagreement'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "def figure6(df, var):\n",
    "    avg = df[var].mean()\n",
    "    return(avg)\n",
    "\n",
    "# lags = ['3', '6', '12', '18']\n",
    "lags = ['12']\n",
    "for lag in lags:\n",
    "    var = 'ret_rf' + lag\n",
    "    fig6_totalDF = data20.groupby(['BETA_1', 'boolean_disagreement']).apply(lambda x: figure6(x, var)).reset_index()\n",
    "    fig6_totalDF = fig6_totalDF.loc[fig6_totalDF['boolean_disagreement'] != 1]\n",
    "    fig6_totalDF = fig6_totalDF.rename(columns = {0: 'AvgRet_'+ lag})\n",
    "    \n",
    "    # Fit quadratic vs linear\n",
    "    x_L = fig6_totalDF.loc[fig6_totalDF['boolean_disagreement']==0,'BETA_1']\n",
    "    y_L = fig6_totalDF.loc[fig6_totalDF['boolean_disagreement']==0,'AvgRet_'+ lag]\n",
    "    linear_LowD = np.polyfit(x=x_L, y=y_L, deg = 1)\n",
    "    x_H = fig6_totalDF.loc[fig6_totalDF['boolean_disagreement']==2,'BETA_1']\n",
    "    y_H = fig6_totalDF.loc[fig6_totalDF['boolean_disagreement']==2,'AvgRet_'+ lag]\n",
    "    quadratic_HighD = np.polyfit(x=x_H, y=y_H, deg = 2)\n",
    "    # Plotting\n",
    "    groups = fig6_totalDF.groupby('boolean_disagreement')\n",
    "    fig, ax = plt.subplots()\n",
    "    ax.margins(0.05) # Optional, just adds 5% padding to the autoscaling\n",
    "    labelDict = {0: 'Low Agg Disagreement', 2: 'High Agg Disagreement'}\n",
    "    colorDict = {0: 'red', 2: 'blue'}\n",
    "    markerDict = {0: 'o', 2: '^'}\n",
    "    \n",
    "    for name, group in groups:\n",
    "        ax.plot(group['BETA_1'], group['AvgRet_'+ lag], marker=markerDict[name], linestyle='', \n",
    "                color=colorDict[name] ,label=labelDict[name])\n",
    "        if name == 0:\n",
    "            ax.plot(x_L, linear_LowD[1] + linear_LowD[0]*x_L, color=colorDict[name])\n",
    "        else:\n",
    "            ax.plot(x_H, quadratic_HighD[2] + quadratic_HighD[1]*x_H + quadratic_HighD[0]*x_H**2, color=colorDict[name])\n",
    "    ax.set_ylabel(lag + 'month forward excess returns (%)')\n",
    "    ax.set_xlabel('Portfolio Postranking Betas')\n",
    "    ax.set_title('SML varying with aggregate disagreement \\n (Using 12 Month forward returns)')\n",
    "    ax.legend()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "While the relationship between excess forward returns and β is quite noisy\n",
    "at the 3- and 6-month horizons, two striking facts emerge at the 12- and\n",
    "18-month horizons. First, the average excess returns-to-β relationship is mostly\n",
    "upward-sloping for months with low aggregate disagreement. This result is generally consistent with our theory, which holds that low aggregate disagreement\n",
    "means low or even no mispricing and hence a strictly upward-sloping Security\n",
    "Market Line. Second, in months of high aggregate disagreement, the excess\n",
    "returns-to-β relationship appears to exhibit the inverted-U shape predicted by\n",
    "the theory."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Regression results\n",
    "\n",
    "First, we estimate first stage cross-sectional regression over 20 $\\beta$ sorted portfolios:\n",
    "$$r^{(12)}_{P,t} = \\kappa_t + \\pi_t \\cdot \\beta_P + \\phi_t \\cdot (\\beta_P)^2 + \\epsilon_{P,t}$$\n",
    "$$For:  P = {1,2,...,20}$$\n",
    "\n",
    "$r^{(12)}_{P,t}$ is the 12-month excess return of the $P^{th}$ beta-sorted portfolio, and $\\beta_P$ is the full-sample post-ranking beta of that portfolio.\n",
    "\n",
    "You will get a time series of $\\kappa_t$, $\\pi_t$, and $\\phi_t$, for each month's regression. We are mostly interested in time series of $\\phi_t$ which represents concavity of SML. \n",
    "\n",
    "Essentially, we want to perform a time-series regression of $\\phi_t$ on aggregate disagreement and other control covariates (as suggested by literature), and see how this determines it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "def regTable3(df):\n",
    "    try:\n",
    "        results = sm.ols(formula=\"ret_rf12 ~ BETA_1 + BETA_2\", data=df).fit()\n",
    "        params = results.params\n",
    "        params.index = ['slope0', 'slope1', 'slope2']\n",
    "    except:\n",
    "        params = pd.Series([np.nan]*3)\n",
    "        params.index = ['slope0', 'slope1', 'slope2']\n",
    "    return(params)\n",
    "\n",
    "coeffs = data20.reset_index().groupby(['month', 'year']).apply(lambda x: regTable3(x)).reset_index()\n",
    "coeffs['date'] = coeffs.apply(lambda x: str(int(x['year'])) + '/' + str(int(x['month'])), axis=1)\n",
    "coeffs = coeffs.set_index('date').drop(['month', 'year'], axis=1)\n",
    "data20 = data20.merge(coeffs, how='left', left_index=True, right_index=True)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, we perform second stage regression as explained above. We use Newey West estimators for the standard errors while still using OLS -- no reason to think regressor endogeneity is present, so OLS estimators are still unbiased/consistent. However, we need to use NW std errors, which are heteroskedasticity and auto-correlation robust, to properly perform inference.\n",
    "\n",
    "#### ADD DERIVTION OF NEWEY WEST COVARIANCE MATRIX vs. USUAL OLS W/ HOMOSKEDASTICITY COVARIANCE MATRIX"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['group', 'year', 'month', 'ret_rf', 'ret_rf1', 'ret_rf3', 'ret_rf6',\n",
       "       'ret_rf9', 'ret_rf12', 'ret_rf18', 'inflation', 'TED', 'dp', 'pe',\n",
       "       'mktrf', 'smb', 'hml', 'umd', 'aggregate_disp_daily', 'mktrf_1',\n",
       "       'mktrf_3', 'mktrf_6', 'mktrf_9', 'mktrf_12', 'mktrf_18', 'hml_1',\n",
       "       'hml_3', 'hml_6', 'hml_9', 'hml_12', 'hml_18', 'smb_1', 'smb_3',\n",
       "       'smb_6', 'smb_9', 'smb_12', 'smb_18', 'umd_1', 'umd_3', 'umd_6',\n",
       "       'umd_9', 'umd_12', 'umd_18', 'mktrf_lag3', 'mktrf_lag6', 'mktrf_lag9',\n",
       "       'mktrf_lag12', 'mktrf_lag18', 'BETA_1', 'BETA_2',\n",
       "       'boolean_disagreement', 'slope0', 'slope1', 'slope2',\n",
       "       'aggregate_disp_lag1', 'aggregate_disp_lag1_scaled', 'dp_scaled_lag1',\n",
       "       'TED_scaled_lag1', 'inflation_scaled_lag1'],\n",
       "      dtype='object')"
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Creating lagged variable of agg disp\n",
    "data20['aggregate_disp_lag1'] = data20.groupby('group')['aggregate_disp_daily'].shift(1)\n",
    "\n",
    "# Scaling and lagging necessary variables\n",
    "scaleVars = ['dp', 'TED','inflation']\n",
    "for var in scaleVars:\n",
    "    newVar = var+'_scaled_lag1'\n",
    "    data20[newVar] = data20.groupby('group')[var].shift(1)\n",
    "    data20[newVar] = data20[newVar].sub(data20[newVar].mean()).div(data20[newVar].std())\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Regressing spefication: Spec1\n",
      "Regressors: aggregate_disp_lag1\n",
      "Regressing spefication: Spec2\n",
      "Regressors: ['aggregate_disp_lag1', 'mktrf_12', 'hml_12', 'smb_12', 'umd_12']\n",
      "Regressing spefication: Spec3\n",
      "Regressors: ['aggregate_disp_lag1', 'mktrf_12', 'hml_12', 'smb_12', 'umd_12', 'dp_scaled_lag1', 'inflation_scaled_lag1']\n",
      "Regressing spefication: Spec4\n",
      "Regressors: ['aggregate_disp_lag1', 'mktrf_12', 'hml_12', 'smb_12', 'umd_12', 'dp_scaled_lag1', 'inflation_scaled_lag1', 'TED_scaled_lag1']\n"
     ]
    }
   ],
   "source": [
    "def regFunc(big_df, varList):\n",
    "    '''\n",
    "    Params\n",
    "    1) big_df: Dataframe with all variables, regardless of specification\n",
    "    2) varList: list with variables, and last element is dependent variable (y)\n",
    "    '''\n",
    "    df = big_df[varList]\n",
    "    \n",
    "    # Automating parameter to be passed to sm.ols()\n",
    "    formula_str = varList[-1] + '~' + varList[0]\n",
    "    if len(varList) > 1:\n",
    "        for regressor in varList[1:-1]:\n",
    "            formula_str = formula_str + '+' + regressor\n",
    "    \n",
    "    # Computing regression object\n",
    "    results = sm.ols(formula=formula_str, data=df).fit(cov_type='HAC',cov_kwds={'maxlags':11})\n",
    "    return(results.summary())\n",
    "    \n",
    "\n",
    "specMax = ['aggregate_disp_lag1', 'mktrf_12', 'hml_12', 'smb_12', 'umd_12', 'dp_scaled_lag1', 'inflation_scaled_lag1',\n",
    "          'TED_scaled_lag1']\n",
    "regDict = {'Spec1': specMax[0],\n",
    "           'Spec2': specMax[:5],\n",
    "           'Spec3': specMax[:-1],\n",
    "           'Spec4': specMax}\n",
    "\n",
    "# Reducing data matrix to relevant variables of interest, \n",
    "# and dropping duplicates as these variables are common across portfolios\n",
    "\n",
    "regDF = data20[['slope2'] + specMax].dropna()\n",
    "regDF = regDF.drop_duplicates()\n",
    "\n",
    "regResults = []\n",
    "for key, value in regDict.items():\n",
    "#     print('Regressing spefication: ' + key)\n",
    "#     print('Regressors:', value)\n",
    "    allVars = [value] + ['slope2'] if key == 'Spec1' else value + ['slope2']\n",
    "    regResults.append(regFunc(regDF, allVars))\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<class 'statsmodels.iolib.summary.Summary'>\n",
       " \"\"\"\n",
       "                             OLS Regression Results                            \n",
       " ==============================================================================\n",
       " Dep. Variable:                 slope2   R-squared:                       0.098\n",
       " Model:                            OLS   Adj. R-squared:                  0.095\n",
       " Method:                 Least Squares   F-statistic:                     4.350\n",
       " Date:                Thu, 28 Nov 2019   Prob (F-statistic):             0.0377\n",
       " Time:                        14:23:40   Log-Likelihood:                -1515.3\n",
       " No. Observations:                 384   AIC:                             3035.\n",
       " Df Residuals:                     382   BIC:                             3043.\n",
       " Df Model:                           1                                         \n",
       " Covariance Type:                  HAC                                         \n",
       " =======================================================================================\n",
       "                           coef    std err          z      P>|z|      [0.025      0.975]\n",
       " ---------------------------------------------------------------------------------------\n",
       " Intercept              -4.1048      1.450     -2.830      0.005      -6.947      -1.262\n",
       " aggregate_disp_lag1    -4.0734      1.953     -2.086      0.037      -7.901      -0.246\n",
       " ==============================================================================\n",
       " Omnibus:                       11.561   Durbin-Watson:                   0.638\n",
       " Prob(Omnibus):                  0.003   Jarque-Bera (JB):               12.716\n",
       " Skew:                           0.343   Prob(JB):                      0.00173\n",
       " Kurtosis:                       3.569   Cond. No.                         1.02\n",
       " ==============================================================================\n",
       " \n",
       " Warnings:\n",
       " [1] Standard Errors are heteroscedasticity and autocorrelation robust (HAC) using 11 lags and without small sample correction\n",
       " \"\"\", <class 'statsmodels.iolib.summary.Summary'>\n",
       " \"\"\"\n",
       "                             OLS Regression Results                            \n",
       " ==============================================================================\n",
       " Dep. Variable:                 slope2   R-squared:                       0.259\n",
       " Model:                            OLS   Adj. R-squared:                  0.250\n",
       " Method:                 Least Squares   F-statistic:                     5.077\n",
       " Date:                Thu, 28 Nov 2019   Prob (F-statistic):           0.000162\n",
       " Time:                        14:23:40   Log-Likelihood:                -1477.5\n",
       " No. Observations:                 384   AIC:                             2967.\n",
       " Df Residuals:                     378   BIC:                             2991.\n",
       " Df Model:                           5                                         \n",
       " Covariance Type:                  HAC                                         \n",
       " =======================================================================================\n",
       "                           coef    std err          z      P>|z|      [0.025      0.975]\n",
       " ---------------------------------------------------------------------------------------\n",
       " Intercept              -2.3650      1.493     -1.584      0.113      -5.291       0.561\n",
       " aggregate_disp_lag1    -3.3267      1.606     -2.071      0.038      -6.474      -0.179\n",
       " mktrf_12               -0.0529      0.075     -0.704      0.481      -0.200       0.094\n",
       " hml_12                 -0.3856      0.133     -2.895      0.004      -0.647      -0.125\n",
       " smb_12                  0.2176      0.151      1.438      0.150      -0.079       0.514\n",
       " umd_12                 -0.0010      0.082     -0.012      0.990      -0.163       0.160\n",
       " ==============================================================================\n",
       " Omnibus:                        7.845   Durbin-Watson:                   0.599\n",
       " Prob(Omnibus):                  0.020   Jarque-Bera (JB):                7.796\n",
       " Skew:                           0.345   Prob(JB):                       0.0203\n",
       " Kurtosis:                       3.104   Cond. No.                         27.1\n",
       " ==============================================================================\n",
       " \n",
       " Warnings:\n",
       " [1] Standard Errors are heteroscedasticity and autocorrelation robust (HAC) using 11 lags and without small sample correction\n",
       " \"\"\", <class 'statsmodels.iolib.summary.Summary'>\n",
       " \"\"\"\n",
       "                             OLS Regression Results                            \n",
       " ==============================================================================\n",
       " Dep. Variable:                 slope2   R-squared:                       0.361\n",
       " Model:                            OLS   Adj. R-squared:                  0.349\n",
       " Method:                 Least Squares   F-statistic:                     8.130\n",
       " Date:                Thu, 28 Nov 2019   Prob (F-statistic):           3.17e-09\n",
       " Time:                        14:23:40   Log-Likelihood:                -1449.2\n",
       " No. Observations:                 384   AIC:                             2914.\n",
       " Df Residuals:                     376   BIC:                             2946.\n",
       " Df Model:                           7                                         \n",
       " Covariance Type:                  HAC                                         \n",
       " =========================================================================================\n",
       "                             coef    std err          z      P>|z|      [0.025      0.975]\n",
       " -----------------------------------------------------------------------------------------\n",
       " Intercept                -2.6558      1.303     -2.039      0.041      -5.209      -0.103\n",
       " aggregate_disp_lag1      -6.2047      1.716     -3.616      0.000      -9.568      -2.842\n",
       " mktrf_12                 -0.0760      0.058     -1.303      0.193      -0.190       0.038\n",
       " hml_12                   -0.2928      0.119     -2.452      0.014      -0.527      -0.059\n",
       " smb_12                    0.3242      0.117      2.764      0.006       0.094       0.554\n",
       " umd_12                    0.0143      0.069      0.208      0.835      -0.121       0.149\n",
       " dp_scaled_lag1           -2.1027      1.311     -1.603      0.109      -4.673       0.468\n",
       " inflation_scaled_lag1    -3.4697      0.964     -3.599      0.000      -5.359      -1.580\n",
       " ==============================================================================\n",
       " Omnibus:                       13.488   Durbin-Watson:                   0.747\n",
       " Prob(Omnibus):                  0.001   Jarque-Bera (JB):               14.329\n",
       " Skew:                           0.473   Prob(JB):                     0.000774\n",
       " Kurtosis:                       2.952   Cond. No.                         33.5\n",
       " ==============================================================================\n",
       " \n",
       " Warnings:\n",
       " [1] Standard Errors are heteroscedasticity and autocorrelation robust (HAC) using 11 lags and without small sample correction\n",
       " \"\"\", <class 'statsmodels.iolib.summary.Summary'>\n",
       " \"\"\"\n",
       "                             OLS Regression Results                            \n",
       " ==============================================================================\n",
       " Dep. Variable:                 slope2   R-squared:                       0.361\n",
       " Model:                            OLS   Adj. R-squared:                  0.348\n",
       " Method:                 Least Squares   F-statistic:                     7.271\n",
       " Date:                Thu, 28 Nov 2019   Prob (F-statistic):           5.51e-09\n",
       " Time:                        14:23:40   Log-Likelihood:                -1449.1\n",
       " No. Observations:                 384   AIC:                             2916.\n",
       " Df Residuals:                     375   BIC:                             2952.\n",
       " Df Model:                           8                                         \n",
       " Covariance Type:                  HAC                                         \n",
       " =========================================================================================\n",
       "                             coef    std err          z      P>|z|      [0.025      0.975]\n",
       " -----------------------------------------------------------------------------------------\n",
       " Intercept                -2.6115      1.320     -1.978      0.048      -5.199      -0.024\n",
       " aggregate_disp_lag1      -6.1426      1.669     -3.681      0.000      -9.413      -2.872\n",
       " mktrf_12                 -0.0776      0.060     -1.286      0.199      -0.196       0.041\n",
       " hml_12                   -0.2962      0.116     -2.560      0.010      -0.523      -0.069\n",
       " smb_12                    0.3250      0.117      2.770      0.006       0.095       0.555\n",
       " umd_12                    0.0123      0.067      0.182      0.856      -0.120       0.144\n",
       " dp_scaled_lag1           -1.9086      1.604     -1.190      0.234      -5.052       1.235\n",
       " inflation_scaled_lag1    -3.3624      0.870     -3.866      0.000      -5.067      -1.658\n",
       " TED_scaled_lag1          -0.3419      1.111     -0.308      0.758      -2.520       1.836\n",
       " ==============================================================================\n",
       " Omnibus:                       13.799   Durbin-Watson:                   0.746\n",
       " Prob(Omnibus):                  0.001   Jarque-Bera (JB):               14.694\n",
       " Skew:                           0.478   Prob(JB):                     0.000645\n",
       " Kurtosis:                       2.950   Cond. No.                         39.9\n",
       " ==============================================================================\n",
       " \n",
       " Warnings:\n",
       " [1] Standard Errors are heteroscedasticity and autocorrelation robust (HAC) using 11 lags and without small sample correction\n",
       " \"\"\"]"
      ]
     },
     "execution_count": 147,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "regResults"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Results presented above from the second-stage regressions.**\n",
    "\n",
    "A higher Agg.Disp.t−1 is associated with a\n",
    "smaller φt, that is, a more concave Security Market Line, or equivalently, lower\n",
    "average returns of the square portfolio. The t-statistics are between −2.1 and\n",
    "−3.7 depending on the specification. Importantly, the estimate is significant by\n",
    "itself even without any controls, although the inclusion of the D/P ratio and\n",
    "the year-over-year past inflation rate increases the effect of aggregate disagreement on the concavity of the Security Market Line"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Additional Plots"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1) Showing time series of SML concavity coeff and agg Disagreement\n",
    "\n",
    "RELEVANT: Note period during dotcom bubble"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, 'Time Series of Concavity vs Aggregate Disagreement')"
      ]
     },
     "execution_count": 161,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot1_df = data20[['slope2', 'aggregate_disp_daily']].drop_duplicates()\n",
    "\n",
    "fig, ax = plt.subplots()\n",
    "ax2 = ax.twinx()\n",
    "rspine = ax2.spines['right']\n",
    "rspine.set_position(('axes', 1.15))\n",
    "\n",
    "plot1_df.slope2.plot(ax=ax, style='b-')\n",
    "plot1_df.aggregate_disp_daily.plot(ax=ax2, style='r-')\n",
    "ax2.legend([ax.get_lines()[0], ax2.get_lines()[0]], ['Concavity','Disagreement'])\n",
    "\n",
    "ax.set_xlabel('Date')\n",
    "ax.set_ylabel('Concavity Coefficient')\n",
    "ax2.set_ylabel('Aggregate Disagreement')\n",
    "ax.set_title('Time Series of Concavity vs Aggregate Disagreement')\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2) Show plot of diasgreement vs concavity, and add fitted values from regression (Figure 3 in draft paper)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1a1d14e3c8>"
      ]
     },
     "execution_count": 174,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot2_df = data20[['slope2', 'aggregate_disp_daily']].drop_duplicates()\n",
    "fig,ax = plt.subplots()\n",
    "plot2_df.plot.scatter(x='aggregate_disp_daily', y='slope2', ax=ax)\n",
    "f = lambda x: -4.1048 + -4.0734*x\n",
    "ax.plot([-1, 3], [f(-1), f(3)], c='r')    \n",
    "ax.set_title('Aggregate Disagreement and Failure of Capm')\n",
    "ax.legend(['Fitted Values','Concavity Coeff'])\n"
   ]
  },
  {
   "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}