Chapter 14 - Backtest Statistics.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as mpl\n",
    "\n",
    "%matplotlib inline\n",
    "mpl.style.use('ggplot')\n",
    "mpl.rcParams['figure.figsize'] = 16,6\n",
    "\n",
    "from load_data import load_contracts\n",
    "from sampling import dollar_bars, daily_bars"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 14.1\n",
    "\n",
    "A strategy exhibits a high turnover, high leverage, and high number of bets, with a short holding period, low return on execution costs, and a high Sharpe ratio. Is it likely to have large capacity? What kind of strategy do you think it is?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**A: Sounds like something akin to a statistical arbitrage strategy on a smaller number of instruments, such as futures. Strategies with short holding periods and high turnover generally tend to have low capacities.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 14.2\n",
    "\n",
    "On the dollar bars dataset for E-mini S&P 500 futures, compute\n",
    "\n",
    "* a) HHI index on positive returns\n",
    "* b) HHI index on negative returns\n",
    "* c) HHI index on time between bars\n",
    "* d) The 95-percentile DD\n",
    "* e) The 95-percentile TuW\n",
    "* f) Annualized average return\n",
    "* g) Average returns from hits (positive returns)\n",
    "* h) Average returns from misses (negative returns)\n",
    "* i) Annualized SR\n",
    "* j) Information ratio, where the benchmark is the risk-free rate\n",
    "* k) PSR\n",
    "* l) DSR, where we assume there were 100 trials, and the variance of the trials' SR was 0.5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "es_contracts = load_contracts('@ES').truncate(before='2017-1-1')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "dbars = dollar_bars(es_contracts, 1000000000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'ann_information_ratio': 0.8031419312703149,\n",
       " 'ann_return': 0.13022942020827788,\n",
       " 'ann_sharpe': 0.9187046149534217,\n",
       " 'avg_return_hits': 0.003013915357033664,\n",
       " 'avg_return_misses': -0.0032678142122765583,\n",
       " 'dd95th': 0.05700825382941044,\n",
       " 'rHHINeg': 0.0007544096935836607,\n",
       " 'rHHIPos': 0.0006160043025995248,\n",
       " 'tHHI': 0.0027732600439860644,\n",
       " 'tuw95th': 0.17785634049828378}"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def get_stats(series, rf_yearly_return=0.00):\n",
    "    yrs_in_ser = (series.index[-1] - series.index[0]).days / 365\n",
    "    obs_per_year = len(series) / yrs_in_ser\n",
    "    dd, tuw = computeDD_TuW(series)\n",
    "\n",
    "    returns = series.pct_change()[1:]\n",
    "    total_return = ((1 + returns).cumprod().iloc[-1] - 1)\n",
    "    \n",
    "    rf_period_returns = rf_yearly_return / obs_per_year\n",
    "\n",
    "    return {\n",
    "        'rHHIPos': getHHI(returns[returns >= 0]),\n",
    "        'rHHINeg': getHHI(returns[returns < 0]),\n",
    "        'tHHI': getHHI(returns.groupby(pd.Grouper(freq='M')).count()),\n",
    "        'dd95th': np.percentile(dd, 95),\n",
    "        'tuw95th': np.percentile(tuw, 95),\n",
    "        'ann_return': total_return / yrs_in_ser,\n",
    "        'avg_return_hits': returns[returns >= 0].mean(),\n",
    "        'avg_return_misses': returns[returns < 0].mean(),\n",
    "        'ann_sharpe': np.sqrt(obs_per_year) * returns.mean() / returns.std(),\n",
    "        'ann_information_ratio': np.sqrt(obs_per_year) * (returns.mean() - rf_period_returns) / returns.std()\n",
    "    }\n",
    "\n",
    "def getHHI(betRet):\n",
    "    if betRet.shape[0] <= 2:\n",
    "        return np.nan\n",
    "    wght = betRet / betRet.sum()\n",
    "    hhi = (wght**2).sum()\n",
    "    hhi = (hhi - betRet.shape[0] ** -1) / (1.0 - betRet.shape[0] ** -1)\n",
    "    return hhi\n",
    "\n",
    "\n",
    "def computeDD_TuW(series, dollars=False):\n",
    "    # compute series of drawdowns and the time under water associated with them\n",
    "    df0 = series.to_frame('pnl')\n",
    "    df0['hwm'] = series.expanding().max()\n",
    "    df1 = df0.groupby('hwm').min().reset_index()\n",
    "    df1.columns = ['hwm', 'min']\n",
    "    df1.index = df0['hwm'].drop_duplicates(keep='first').index\n",
    "    df1 = df1[df1['hwm'] > df1['min']]\n",
    "    if dollars:\n",
    "        dd = df1['hwm'] - df1['min']\n",
    "    else:\n",
    "        dd = 1 - df1['min'] / df1['hwm']\n",
    "    tuw = ((df1.index[1:] - df1.index[:-1]) / np.timedelta64(1, 'Y')).values # in years\n",
    "    tuw = pd.Series(tuw, index=df1.index[:-1])\n",
    "    return dd, tuw\n",
    "\n",
    "get_stats(dbars['Close'], 0.015)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 14.3.a\n",
    "\n",
    "Consider a strategy that is long one futures contract on even years, and is short one futures contract on odd years.\n",
    "\n",
    "Repeat the calculations from exercise 2."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "series = pd.Series(index=dbars.index)\n",
    "series[series.index.year % 2 != 0] = 1\n",
    "series[series.index.year % 2 == 0] = -1\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA6UAAAFcCAYAAAAwIxi+AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4wLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvqOYd8AAAIABJREFUeJzs3Xd4HNXZNvD7zFb1YjXLcpU7GBvjRgsmEcZAMKQZArzJ9wYIJpSYFgihpUAcDKHFQCgJ5A0BQ0gwEKpNLy5gExv3bsuqq7YrabVl5nx/zFbtqljaIq3u33VxaXZmdvbsSBZ69nnOc4SUUoKIiIiIiIgoCZRkD4CIiIiIiIiGLgalRERERERElDQMSomIiIiIiChpGJQSERERERFR0jAoJSIiIiIioqRhUEpERERERERJw6CUiIiIiIiIkoZBKRERERERESWNsacTHn30UWzcuBE5OTm4//77o56zdetWPPPMM1BVFVlZWfj1r38d84ESERERERFR6ukxKJ0/fz4WLlyIFStWRD3e1taGp556Cr/61a9QUFCAlpaWmA+SiIiIiIiIUlOP5btTp05FZmZml8c/+eQTzJ07FwUFBQCAnJyc2I2OiIiIiIiIUlqPmdKeVFdXw+v14q677oLT6cTZZ5+N0047rVfPraqq6u/Lx0VBQQFsNluyhzHk8L4nF+9/8vDeJwfve3Lx/icP731y8L4nF+9/cpSWlvbqvH4HpaqqYv/+/bj99tvhdrtx2223YcKECVEHsHr1aqxevRoAsGzZskB2daAxGo0DdmypjPc9uXj/k4f3Pjl435OL9z95eO+Tg/c9uXj/B7Z+B6XDhg1DVlYWrFYrrFYrpkyZgoMHD0YNSisqKlBRURF4PFA/reAnKcnB+55cvP/Jw3ufHLzvycX7nzy898nB+55cvP/J0dtMab+XhJk1axZ27NgBVVXhcrmwZ88ejBgxor+XJSIiIiIioiGgx0zpgw8+iG3btsHhcGDJkiVYvHgxvF4vAGDBggUoKyvDjBkzcOONN0JRFHzzm9/EqFGj4j5wIiIiIiIiGvx6DEqXLl3a40UWLVqERYsWxWRARERERERENHT0u3yXiIiIiIiIqK8YlBIREREREVHSMCglIiIiIiKipGFQSkREREREREnDoJSIiIiIiIiShkEpERERERHRIPLY+hp8ctCe7GHEDINSIiIiIiKiQWJdpQNv7W7GQ59XJ3soMcOglIiIiIiIaJDYVucEALhViZYOb5JHExvGZA+AiIiIiIiIopNSotLuhpTAvqYOuLwaAMBqVNDiUpFjHfwh3eB/B0RERERERCnqk4MO3PdpVdi+LLOCv/9gYpJGFHss3yUiIiIiIhqgmnwluj+ZWRTY53BryRpOXDAoJSIiIiIiGqD85bpnT8zDXd8cCQDIMKdWGMfyXSIiIiIiogHK5ZVQBGBUgOOK03HV3BIUZ5qSPayYYlBKREREREQ0QLlUDWaDAiEEDAJYMD432UOKudTK+xIREREREaWQXbYO5FgNyR5GXDEoJSIiIiIiGoB++c5B7LA5ceLIrGQPJa4YlBIREREREQ0wqiaxrd4JAMixMFNKREREREREcaZqEk6PBiklXGpw2Zc0U2qHbWx0RERERERENADcvuYQttY5YVRE2LIvqR6Upva7IyIiIiIiGiRqWj0oz7dg0eQ8tHSogf1pxtQO21L73REREREREQ0SHlVi4rA0/Pj4IswpywzsFyKJg0oABqVERERERJQQzR1e/HTVXvxrWwOaO7xwerSen5QA7R4Vh1pcyR4G7C4VbW79ntx4cimOK04HAGgymaOKPwalREREREQUd7sbnHjgs2rUtnrw7KZ6/PjlPbjwxV2oa/Uke2j446dVuOb1/fCoyQuSa1vdAICN1a0AAItRwbmT8wAA5fnWpI0rEdjoiIiIiIiI+kxKibs/rIRHA5ravTAbBe5bOCbivBvfOggAsBgEbj2tDHe+dxgAcPmqvfjnhZNgMiSvRnVfo54lPdzixrgkBYBb6/TlX351Wllg35yyLKy6eHJSxpNIzJQSEREREVGfrfy6ARuOtOGr6jYcbHFhd0MHDjR1dHm+KoEZwzNQmmUO7Ht/f0sihtqlEdn6WPaFjNujStz/aRW+qm6L2+tKKfHUl7V4a3cTdtQ7kWFWMLkwLW6vN1AxKCUiIiIioj57frMNAHDNvBLcfGopAGDNvvAgszJkvmaO1QAAqHK4A/tWrKtBS4c33kONSpMS2+r1LOURe3BM/97egI8O2PGBL2D++IAdexq6Drb7wuFS8dqOJjy2vhZv72lGUYYJSqp3NYqC5btERERERNQn+xqDQdqsEZnItRoxe4Qdaw+3wmJQYDQIXDitAF9W6dnGuytGYXiWCQBgNgi41WAHHzVJzXy21zvh9XUS8jcZqrK78eKWBgDAviY9oL7v0yoAwCsXTYLoY+Do8mp48otajM2z4pxJeajuNJ821RsadYVBKRERERER9ajVpSLDrAQCMreq4bo3DwSO51r10GLWiAxsONKKl7bqQd2F0wpQ1+aB1ajgmKK0wPP/eNYYrK9sxd++qgcAKElKEDpcwfVAbe0erKt04OWtjTAbBGYM199LaBbX7lKRY+1bGLV45S7fVgsqynOwra497PgxRUOvdBdg+S4REREREfWg1aXi0lf24s3dzYF9u0NKWc8ozwlsTxgWGVjVtrpRnGEKyzCOzLHg+OEZwZOSlCVsc+tBaY7FgC+r2nDPh0ewu8GJy2YVY2KB3vRoU8i8Uu9RpjNdXg1eTULK8Oe9t68Fnx5yBB5bjQI/nVXc17cxqDFTSkRERERE3dpe70SHV8OfN9RiR70T1500HC6vXup6xexinD0xL3Buuik879XS4cWm6jacUZ4bcd3Q7KgqkxOV2n2Z0p+fOBw1rR6My7dgTK4VaSYFm2v0YNTWHsyU/uTfe7FkdjHOCnnP3Vm8cheOK07Hzd8YEbb/8Q21AIDjitOxubYd88qy+lwWPNgxKCUiIiIiom5tqw+WmX54wI5r5pXA5dWDyKmdusWmdQpKdzd0wKsBp47JjriuN2RZ0ESFpKom8XVdO9o9GjyqhN2lwqgIzCzNiAgKjb6o+f98JcZ+j2+oxbTidOSnG7G5ph3zRmZ1+5qba9vR0B7ZyGlsngWzRmRic207MsxDt4iVQSkREREREUXl1SQ+PmDHLpsTWWYFDl8jIJdXwqXq22ZDeDCVZgx//Nj6GggAY3ItEdf3aMGoNBGJ0rpWDy5ftTdsX3GmCWXZ5qhZysIME4yKwNyyTIwfZsWzm4LB6aeHHNjf1IHPD7fiwbPHYGxe5PqmnpDuTbY2valRuknBJdML8cQXtbh23nDkpxnx4QE7zp8yLFZvc9BhUEpERERERFE9s7EOr+1sAgAcW5yOr2v1jOnP39gfKGlN75ThMxsEFk3Og63di01VbRida8GPjy9ChtnQ7WtpCYhKQ+eGXjhtGF7Y0oDaVg/OnRS9FLcww4SXfzgJALCu0hF2rKHdi88PtwIAbG1ejI1yiQPNwXm3/vv18DljUZhhwpkTcgOZ2D+eNabP7ykVMCglIiIiIqIIHx2wBwJSALAYBG47rQx/3VQHNaTZT1anYFMIgUtP0Bv2SCl7PU8yEcuhZFqCAfRpY3Lwgm/Zl+JMU4/PNXVqD/z2nmDTp7o2T+fTAQC7bHpQOjbPgkanfk5emh6CGZPVbngAGrqFy0RERERE1KV/ft2AcXkWlGQG1xWdXZaJR88dh4fOGQsAKM0ywdBNcNVTQCoQPJ6I8l3/OqRPnV+Owoxgfq6nLC4AKFHey6UnFMFiEKhpdUd9TrVD39/Y7sULWxqQYVIYjEbBoJSIiIiIaAhTNRnoQBvKpWoYlWNBTaue4XOHzI+0GhXceXoZlp85JmbjSET5rr1Df5/ZFgNMBiXQ/dfQizgxdK1Sv3Mn5SHbYkCrO/L+AcHlY0b65tOW5UTOqyWW7xIRERERDVmalFj+yRFsrXPib98bH5bZVDUJJSSFdbjFFfbcmaWZsR1LTK8WXbPLC6tRgcXXjOlnc0rwp3U1SDf1nCmNtj6pEAJGg4A3ekwKTQI5VgPurhgFh0uFuTfR7xDUY1D66KOPYuPGjcjJycH9998fcXzr1q249957UVRUBACYO3cuvv/978d+pEREREREFFMvbmkINOs5bHfDpAiUZJoghIAm9ZLVgnQjbO1eXHdSacxff3RIR14tAZNK7R0qcqzBALSiPAdj8iwYnx/ZObezU0ZnY3u9Ez+aUYj/eXlPYL9JEWFdhENpUgbKfrMsPQe+Q1WPQen8+fOxcOFCrFixostzpkyZgltuuSWmAyMiIiIiSgX/2FyPE8YomBS5TGdSSSmxakcjsiwGOFwqrnl9PwDgztPLMLM0E6qUMAiBR749Fm5VItca+yLLNJOCW74xAss+OpKQdUpbXCqyQ4JDIQQmDEvr5hlBFqOCq+cNBwBcO68ERb65tkZFwKtJuLwaNla1YU5ZZmCerR7Yx/hNpKAef7KmTp2Kurq6RIyFiIiIiCglqJrEmn0teHlrA2paPVi5pQE3nVKKdJMS87LXvmrzaGj3aPjhcQVIMypwqxr+/l9bYOkSVQIGBUg3GZDec3PaPvNXCCem0ZGKzF40NerJt8pzA9smg4BHldhU3YZlHx/BiSOz8P1jhmFcvgWalL2arzrUxaTR0a5du3DTTTfhnnvuweHDh2NxSSIiIiKiQUnVJH757kGsWFcTaBIEAMs/qcKv369M4sjC2XzLmJRlm3HelHyc41urs9nphapJOFwqDrdE7yobS/5prH/ZGNtEWH2bB8/9tz5s+RqvJmGKcZToz5R2ePUS3s8PO3DDWwewZm8L7C4VTm8icsCDW79z8GPHjsWjjz4Kq9WKjRs3Yvny5Xj44Yejnrt69WqsXr0aALBs2TIUFBT09+Xjwmg0DtixpTLe9+Ti/U8e3vvk4H1PLt7/5OG9jx+vquFwcwfSrAp22jrw49kjsXLTkUCw4jdQ7v8uRyMAYHxpAQoK9NriY0pqsPZIO2aXlwAAttS2x328RpvW42tJKfGXdYew6NgSFGb2roPtYxt34a3tDThhbBFGlOo/91IcRIbVEtP3lG6tQbtbhTktAwBw9alj8KePD8CuGfFlVRuAgfM9H6j6HZSmp6cHtmfOnImnn34adrsd2dmRRfMVFRWoqKgIPLbZbP19+bgoKCgYsGNLZbzvycX7nzy898nB+55cvP/Jw3sfP2/sasKfN9TCatQzccVWFSOyzdjb2BF2Xm1dPSrt7rAmP8mwt7oJAGDytMFm0zOiU4aZ8e9tDXjraz2j+9DZY+L+89La2hrYDn2tn/xrD6YPT8fPTyzFvsYO/GXdYXy+z4ZlC0aHPb/do0ad85ou9DLk9Xtrccq4YbDZbOhwe6F5PTF9T9LrQYfbi6YWBwBgdpERVqNAi6Mt6vsaSkpLe9ccq9/lu83NzZC+AvA9e/ZA0zRkZWX197JERERERIPK/iY9+OzwlWtmmg1YfuboQFms3/99VY9r/7Mf1Y74l8Z2x9buhUEgLJgrzjRBlcCnB+0Ym2fBmLyeu9L21ymjo8cODU4v3ttnBxAs8a1r80Scd9NbB/Hjl/fgzxtqwvb7y3T9z/lgfwtqWj0xbzzkn1PqXzLGpAhkmA1odWtIMyo4t9P3nyL1mCl98MEHsW3bNjgcDixZsgSLFy+G16t/6rBgwQKsXbsW77zzDgwGA8xmM5YuXRq2vhERERER0VBwoCl8Hc9MswEGRaBiXA7+s1PPSlqNAjttTgBAo9OL4VnmhI8T0Bv+rD3sQH6aMdApFgguW+JwaxidG5P2Mz1SehE7+APMBl8TJkCfM/rspjpU2vXg/o1dzbhidknguMv34YBL1b/+bVM9AKDJGbxGLPjnlPqvazYIZJkNqG/3wOnVkMmlYHrUY1C6dOnSbo8vXLgQCxcujNmAiIiIiIgGG1WTONjswryRmVjrW/cz0xwZ1KWbDFB8QaCagHU5u/LnDbWotLuRaw0PmEJHnB6DLrWxUGV3w2LQR2ZUgIZ2D9ZXtuL9/fZAgO+3vtKBOWV65tXlm8vb4dG/Fmea0OD0ItZ9h4yKnind3diB8nwrTAYFmRYDNte0AwCyBsh9HMgS8/EHEREREVGKeOqLWpz33A5sqg7OGaxp9cClSswekYnTx+q9VfxlsaXZwWyoUUFgiRA1STGpqkmsq9QD58XHhjfgmTUiEyeOzMK4PAt+Oqs4YWOaUtj1WqG3vnswsG01KnhtRxMe31AbEZACwUyqR9Xw5u5mAAg0mvLP4b1yTmzfV47ViOYOFc1OL4oy9O95aCCaEeXDCQrHO0RERERE1ANNSvx1Yx122px4zVeKe9d7waUQbe36vMXiTBOWnlSKFy+YCItR/1Pbagz+ya0IEZh7mKz1K7fWtaPDq+GWU0dEzHc1KAK3fGMEHjh7LAoz4rg4aSfTitMRejtCs8hNHSo8qv+eCbhVDRkmBb8/Y1TEdfxPa3MHOx77g1JNArlWA8qyY9tgqiTTBI8mUWl3I9viC0otwe+5xciQqyf97r5LRERERJTqVqyrweq9LXhle2PY/pe+tqHVreFwiz6fNMcXlHQXiPgbISWqelfzNSX1z9387JADZoPA8aUZiRlAL4XeDk/IzZk9IgMuVQ8sTQYBVeolsxMLIrOr/gDU/3xFAPVtXjQ7PfBoGkyx7nIEhM0LzvGVQ5sNwe9/sj58GEwYthMRERER9WD13pao+//+Xxve2NWEXQ0dKM+3oDiz++yiRHCuo7PT+qXx8qOX92DZR0cA6AHq2spWnFCaEZbBTbbOvY7cIbXNBkXg4wN6F16zQUDVJBRFwBgSYF4zT29wNNkXqPqz0edNzodH0/Dgh/vw3j476ttj2+QIQNj3PNvX1OgtX+kwoGd3qXvMlBIRERERdUPVJPLTjBibZ8H1J5fi2U11aGj34ssqfU7pygsm9thBdmZZDjZWtkDTJLLT9D/B/1vdhhNHxn8pRYdLxbrKVnhUDXsbXWhyejEvAa/bH06PGtj2N44CAJOiQJPB7KMi/GW5+j31h7L+oLQ834ofHFOA57fUx22sWSHddf0Zcm9IpleJQ3Y21Qycj0eIiIiIiAYQVZP4qroNf/uqHo1OLzyqRKbZgKvmDscVs/VmOWNyLb1a0uS+847B7BEZ8GgSdpcecEVr1BNroWuh1rZ68PlhB4yK3tBoIPGX1db71hS1tekZzTtPL8Oz3x0fPM8goEoZWMYm3xfg+x/7S5X9c1KNBoHvHTMMU4r19xuP8DA05vQHyzefWhrYZ2H9bo+YKSUiIiIi8nG4VCz7qBJf10UGjJtr2wPbxZlm3D6/DMcWp/fquhajglx/l9YOPSjd1+TCinXVuGru8NgMvpNDzS7cEdKM6eG11Whs9+LYonRkDrBlSk4bk4OXvm7Ak1/U4tbTylDrC05LMs3ITQuGLEZFQNOCgeA9Z4zCltp2mH07JIBDLS4sfeMAAD3YNRkEHv/Bcaipt8UlKA0tz/UHxyeNysbt8xUcsbu77SxMOmZKiYiIiIigz/W8Y80hfF3n7FVzmlkjMo9qXubkwjRMLkjDb781MrDvnT2Rc1UPt7jwwmZboHlSXzQ7vbh19SFIGSwj3WnrQH27F/npAy8vVZRpwgXTCrCushUbKlvx5i69w3FhRvhYmzq8cKnBhkXFmWZUlOcG5qRKCfxtU32gjNcfJBoNCqxGJS6dcEMzpaFZ81kjMnHelHwIzintEYNSIiIiIiIAK7fYsK9JDwQfPmcs/vKd8rDjv6sYGe1pvVZRnos/nDkax5WEd71dsa4aDb4lZQDg6tf34/ktNlz9+v4+v9b+ZhccLhVLTyqNOObvEDzQLJqcj5E5ZjyzqQ67GjoAACZDeLhS2+pBS4eKHGv4ewgEpQhvINXqUhFvIixTGveXS0m8bUREREREAL6qaQtsW4wKhqWb8L8zC5FpVvDnReMwrTg+S6i8s6cFT35RG9Nr7m/Sg7oMkxIWXC+anIfF04bF9LVixWQQmDMiEzWtHpwyOgsjss1Rz9vf1BHocuvnz1BKKdHhCQalY/JiuyZpNOFzSpkV7QsGpUREREREAPY2Bstl/WW5508Zhme+OwElWdEDpFiJdYnns5v0brONTi+GpZswf2w2AOB/ZxYh3TSw5pOGshgVeDWJTw46wuZ/hjYLcqkSudbw9+A/qkmElT0nYtmb0JJdBqV9w6CUiIiIiIa8Zmf4+pXppuCfyaYEdE+N9Sv4Y7GyHD2YvnbecDy/eEKvOgUnk8UYHF+bO1h6+4/FE3HpCUWBx12V7zY6vXCFrHGaYU5suMPy3b7hbSMiIiKiIe83H+hdar85LhurLp4caJATL6f7Mpd+hpDy086e+qIWv3j74FFdf/7YHABAWbZevmpQxIDOkPpZQqK60LU+jYrAsJAGTcM6NWsSvrD+g/3hjaPSEpApBYAZw/XSbi3y20e9wKCUiIiIiFJatcONm98+GNZMKJSqSVTZ9WOXzixOyJiWnlSK+WOCgak/Bu7whkc1zU4vXtvZhJ02J3Y39H5d01a3ilE58S05jofQ7rjtIXNDAcAYkuWdUhi+FI///h1sDu9YnKjOt/45rq3u+DdWSkUMSomIiIgopa3Z24IdNifW7ItcfmWnzYnr3zwAp1fDD48rQKYlcdnE0NdSfH+Vdw6q3t3bHNi+6a3eZ0sb273ITxuYXXa7E1q+q3bKOhpDstfmTiXV/tiz1a1hblkmvjE6G7NK49OYKpqSTBMA/QMOOnqD7yeViIiIiKgXpJT4sqoNHl+gEG15kOUfH4EqgVtOHYF5IzMTOr6sKAHwinXVYY8/PugIbBdmmHp97fp2L05IYFAWK8ZuyqZD5/Z2Pi304TFF6ThvSn6MR9a9xccOQ6bZECibpqPDoJSIiIiIUtJbu5vx+IbgUisOtxZxTqtbwxnjc3DiqKxEDg0AkGUOBqVOjwYppd7Ap8WN+xaOxhMbagPrdQLA5MK0Xl3Xo0o0O70oTO99EDtQKN20fNrfFMwid+5yG9rAaWpR7+5TLJkMSsID4VTC8l0iIiIiSkmdS2EdUTKlErHvfNtboZnSNreGfU0ubKltBwDkpRnDAmWrUel1aajTo0ICyLQMvj/1u5sCGla+aww/MTSLOi7PGvNxUXwNvp9UIiIiIqJe6PCGZ0ajBqVSJqwZTmehQakmJTwhkygNQuCkkcGgNNOshHWj7Y5/SRTLIFyfJLQsN6/TnFh/h1sgcv3R0Mfx7pxMsTf4flKJiIiIiHqhcydbtxoepEopocrI+YmJEtYdV4iwLKFBACVZweM5VgM2VrXhvzVtPV63oV1fc7VzM6DBwL8izrg8C1Z8e2zYseLMYDly5/VW/W/1mCSU7lL/MSglIiIiopShahJL39iPR9ZWo9IeXr7bOUjd1+SCV5NRGw4lwrB0E07zLQvzdW07Hv482OTIn+174KwxuOeMUVCEgEeTuGPN4R6ve/M7epdec4LW6Iwl/3coL82IDHP496W7JkjZViNuO60Mt88fGcfRUbyw0RERERERpYz39rVgf5MrrCmOn8uXKb3hzQOYXJiGA016E6Gio+hqG2vXnjgcB5tdECK8kU+aSQ8ox+Xr8yNDA7TlnxzBEbsbLq+Ge88c02VQbRpiZayzyxLbPZliZ/B9fEJEREREFIWUEq/tbAo8fvo75bj1GyNCTtC/7GnswOs7m3Cw2QWjInDiyMR33vUzKgIPnTMW9y0cE7a/c3lqaKnv1jonqh1uVDk8ONDcga4kowttvBVnmvDD4wqSPQyKMWZKiYiIiCglbKpuC3TcXTK7GAXpJmRbDDh5VBbsLhWVdndYsyCHW0OGSRkQjXF6mv5Z7XAHtp/57njsqHfi5ncOhjVHCnXWhFykm5JTlhxPT5xXnuwhUBwwU0pEREREA1JoY6LaVjeaO7xdntvS4cXDn1ejLNuMFy+YiLMm5gEAzAYFvzh1BA41u9Dk9KK+zRP2vDZP5NqlydBTB+Azx+vvx58x9TfW7dS7CR/ubwGgB+iDkb/RUfI/JqBEYlBKRERERAPK9rp2/O6Dw/jBC7uw1bdu5x1rDuOKVXvxwmYbNBmZHfzkoANNHSquP7kUligNfs6amAsAeHZTXXwHHwMPnT0mYl+GWfF91bOf/qY/3k73Ynu9E0CwAy/RYMDyXSIiIiIaMLbXt+OWdw/BatSDrto2D46BvsaoQQg8v8WG8nxrRFObSrsL6SYF4/IsUa/7w+MKYWv3YvXelni/hX4bk2eN2Odfh9PflMngy6yqndYu9Wd+rabBmXuS6N1arJRaBudPKxERERGlpCannuG76RS9QZHLq2F9pQNtHg1zR+qBaE2rO+J5lXY3RmSbuy2DvWR6YcS+y04oisWwY6YoI3rOaFy+FTeeXIor55QACC4Z4+0UlPrf/U9mDqz3dbR6qGamFMNMKRERERENGP7GPf5lTp7ZVBeyvqgeqXx80IFzJ+eHPe9IixvTh6d3e+28NCMmFaRhp80Z2Nf5Osn0+KJxyLF23ZzoVN+apgDgr1DuFJNCkxKlWSZ8c1xOPIYYd8yTDk3MlBIRERHRgOHxRVkZvvLTYEAKWHwtanfanGgJaXrU7lHR4PRiRHb00t1QxxXrgev3jxmGVRdPjtm4Y2F4lrnXHXP9mdLKlvD1WD2ahElJhT/xmSodSlLhJ5aIiIiIUkSbW58TmW6ODM5Cy2//37/2BLY3VumdZsvzI+didjaxQD8nP21wFwwaffWtL29rxGeH7IH9HlXC2NP6MkQDDINSIiIiIko6KSW21LbhtR2NGJljRp7VgAumDQs7J9MSDFT9ZatSSryyvRHDs0yYXtJ9+S4AzCnLwj1njAp04x2sQjsMv7mrObDt1SRMA2Dd1b7yN3TqroyZUg+DUiIiIiJKur//14bbVh9GfbsXp4/NgRAC508Jzvcc6VufM5RH1bDD5sTuhg4smpwPpZfdcY4pSu/1uQNVWkh33b1NHYFtjyphGsSZ0uOK0/GzOSW47ITiZA+FEmhw1y0QERER0aD11u4mHGhyYfrwDPwvaET+AAAgAElEQVRza0Ng/95GPcgKnV/5p2+Pi3j+1jon3trdjEyzMmgb+/THsDQjGpxedPiWgQH0OaXR1mkdLIQQOHPC4M5i09FjUEpERERESfHajiZU2t14c3dz2P5DIc17rplXguFZkVlSANhY1Yp1lQ58d+qwQNnnUOJvAaWGtKx1ejQU+tYyJRosGJQSERERUVJU2t04eVQWThubDU0DxuVbcMWqfbi7YlTgnIry8KzZsgWjcLDZhcfW12L9kVZoEkMySwoApdlmNDq9YfuaOryYnpaRpBER9c3Q+0iJiIiIiJKurtUDADAqAnPLsnDiqCwUZ5rxysWTkWPtOm8ypTAdZ47PhSIAe4cKILh8zFBzy6kjMMo31/aV7Q1weTW0uTXksUkQDTI9/gt+9NFHcdlll+GGG27o9rw9e/bgggsuwNq1a2M2OCIiIiJKPaom0exbZ3TeyMyjfr4QAjlWI9p8cykNg7jbbH9kWQyBTPJfN9bj2a/qAQB5g3y5Gxp6egxK58+fj1tvvbXbczRNw3PPPYcZM2bEbGBERERElHpUTeK7z+/EA59VAwByu8mKdqcgPfg8w9BMlAIASjKD80f/s7MJAJDXx3tKlCw9/hOeOnUqMjO7/wTrzTffxNy5c5GdnR2zgRERERFR6vFnSKscbgBAhrlvpaahwexgXpezv2aN0P9OD70FzJTSYNPvz5UaGxuxfv16LFiwIBbjISIiIqIU1uybB+qXYe7bn6MbjrQGtodq+S6gv/fyfEtYYJ5l4ZxSGlz6/THKM888g4svvhiK0vMvlNWrV2P16tUAgGXLlqGgoKC/Lx8XRqNxwI4tlfG+Jxfvf/Lw3icH73ty8f4nT7Lv/eGOlrDHo0qKkN6HbOmvzpiAu9/dDQAoKiyMydjiKZ73Pd1yBC41uIxOYcEwFGREX0ZnqEr2zz11r99B6d69e/HQQw8BAOx2OzZt2gRFUTBnzpyIcysqKlBRURF4bLPZ+vvycVFQUDBgx5bKeN+Ti/c/eXjvk4P3Pbl4/5Mn2fe+tqE17HFbSyPaxdFnOovNwaVQBsPPUjzvu9BUKALQfOuVNjY2Ak6W8IZK9s/9UFVaWtqr8/r907pixYqw7RNOOCFqQEpEREREQ0tDuwdNThXl+RYIX+Dpn1PqJ/oQkALhcyiHuoUTcnHK6Gw8v8WGJqcXUspkD4noqPQYlD744IPYtm0bHA4HlixZgsWLF8Pr1X+ZcB4pEREREUVzqNmFa/6zHwDw01nFOGdSHgBgl60DGSYFf/3ueLi8Wp+vr/QxmE1FJ4/Wm41uONKKDUdah3TjJxqcegxKly5d2uuLXXXVVf0aDBERERGlhh02Z2Db32kXALbVt2NyYRosRgUWY997blqNeuDlD3YJuOHkUhxo6kA2l4ShQYY/sUREREQUc241mAX1qMFy0kanF8eVZPT7+sPSTfjTt8diRDYb+vilmRRMKUpP9jCIjtoQXmqYiIiIiOKlw6MHorlWA4443Gj36EvBeFQJc4zKS0fmWFjGS5QCmCklIiIiopjxahIGAbR5VBiEvi5pc0c7fvjibswbmQm3KmEyMJAkoiAGpUREREQUE1JKfO/5nVAEYFQERuZY0OHVUNPqAQCsPawvB+N/TEQEsHyXiIiIiGJkxboaAPp6mVOL0vHrb47E9ScH1ym8YnYxALA7LBGFYaaUiIiIiGLi3b0tge07Ty+DIgRq24JZ0SmFaXjo7DEYnsXmRET9ITUVQjEkexgxw0wpEREREcXUH88aE2hAVBdSqmtSBMbkWfu1FAzRUCfdLmhXfAfay88meygxw98IRERERBQT4/OtyDIrKM+3Bva1uLyBbTY4Iuo/ueZ1/etbL0N6U2N+NoNSIiIiIoqJNo+KY4vD18nMtQZnixk5l5SoX6THA7Q0BnfUVSdvMDHEOaVERERE1G/1bR5UOzw4a0Je2P6TR2VhuW87zcR8CNHRknu2Q3vqfj0Y9QYrD8RPfwHkDkviyGKHQSkRERER9dvBZhcAYFJBWth+IQT+8YMJONTsQropdRqzECWKPLAbaKgDxk8F9mwL7Fdmn5LEUcUWP64iIiIion5zeTUA0bOhGWYDphSlR+wnol5wdQAAlOt/C3H5jfq+9IwkDij2mCklIiIiom49srYaTo+GX5w6ostzXKoEAFjYzIgottwuQFEAoxFi5knAxVdCFJUke1QxxaCUiIiIiLq1OmT9Ub92j4r1la04bUw2hBCBTKmZy70QxVZ7K2CxQgihB6bzz0r2iGKOQSkRERERHbWHP6/B54cdGJdnxahcC9zMlBLFnNy9DfKjt4HyyckeSlwxKCUiIiKio/b5YQcAwOnLkLpUX6bUwEwpUSxIKaHdewsAQEw9PsmjiS/+1iAiIiKiXlE1GbHP7QtGbW1eKAJg9S5RjLjdwe201Gps1BkzpURERETUK7/7oBKXzCjEv7c1BPbZO1S0e1Ss2deCCcN8896IqE+0x/8AuWOz3l3XErK8Ulpqd69mUEpEREREXfriSCsAYGSOGZtr23D9m21hx9/f34I0kwKvJnFGeW4yhkiUMuS+nUBWNsSo8ZDrPwzsFwxKiYiIiGioqrS7AAB/WDAalXY3fvH2wcAxq1HBF0fasOGIHqhaWLtLvSBrq4Ci4cyqR+NxQ0yfA+XiJdAEINf5AtMUv1f8zUFEREREXfJPIzUoApMK0sKOLT1peNjfyuy8Sz2RlQeg3bYE2m1XQvvLg5A7tyR7SAAAeXAvtDdeSu4YVBVotUPu3wUAED+5Dph5YlLHlCjMlBIRERFRl/xBabRw88SRWZg3MgufHdI78TrcauIGRgOK3LZJ/2HJzgEcdohjIrvFak8/ALn2ff1BfQ1kXRXk5+9BnH8JlHMWJ3jEncb26D1AYz3kKRUQ2XnJGcS+nfrXgiIAgFAUKBctgczJA46dmZwxJQgzpURERETUJadHg0EApi6yoGeU5wS2p5ekdodQik7u2QbtgTuhPXQXtN9eB+3BOyFb7ZHn+QPSrBwoj/8ruP+Vv0M21CdquNGlZ+pf9+0K2y2//BSy0RbXl5bbvoKsOqQH9ooC5UdXB46JnDwoFy2BMJriOoZkY1BKRERERF2qcriRazVC6WJO28gcCwAg12pAYUZq/+FM0Wl/0NfSxPQ5wKhyAID8+N2wc6S9OfjA0QKhKIASDEW0Wy6F7GiP+1ijkVIC7g59++De4Jg2fKx3w33rn/qx/btiHqDKjnZoD9wB7c6rIT9/HygeAeEPkIcQlu8SERERUVQvbLbhs0MOFKQH/2Q0GwTcqkSGWQ8o8tOMmFKYhh8cMyxZw6QkklWHAtvK4p9AFJVCXf5LyM/WQFMEYDBAqTgP8stPAQDi9HMgZp2sP0HTwi/W4QSsSegye2APUFetbzv04FnaaiH/b4W+7cueavfcCABQnljV5yZNssMJ+drzwKhyKHNPA2qrgwcb6oCysX18E4Mbg1IiIiKiFKNqEltq2zG9JL3vfzxLiee36Fmh0Czp44vG4dNDDpw4MguA3gBp2YLR/R80DXiycj9QXAZh0jPissMJ7c6QUtOiUv3rrFMh//E45D+f0Q9UnAfUVgGWNIgf/jTwM6nceh/kF59CvvNv/TzFkLD3EqY5uO6urDoEbc3rkJ++C0ipz+X8eiNkU/ActDqArOw+vZR2zQXB1zpuNuS2r8KOi/GT+3TdwY7lu0REREQpZm2lA3e+dxj/3t7Y52u0uIJNi7IswWBhWLoJiybns1R3iJHNDdB+ex3kGy9Cej36zoN7AsfF+ZcEt8dPiXz+4X1AUUnYhyRi7ESIefNDTpIxH3dvBOa/DisCdm+DfOEJoMkG5X9/DjFpmn5OaJfgw3ujXKWb6+/bCdlo08uEQ/e/+DTkx28Hd6RnQPzwij69h8GOmVIiIiKiFPXCZhu+O7VvZbW1rZ7AdpaZeYyhTm77CtA0yNdXQr6+EsqfXwE69HmYylW/gpgxN3hyWngJrqyuBHZtDQtcA0Iz+VKLPJ4IvvmuyrV3AM2NQOkoICdPD6B3bIYEIDd8HDhde+BOIK8AyrInIXqR3dV+fxNgMkP5/ZNh++Un+rxbMe90yLXvQxw/T59rOwQxKCUiIiJKMf6pei5VYqfNGbG+aG8oIbFCpiVJZZU0IEivF9i+OXynxw3paw6EouHhx9LDuzDL914DAIiZJ0Ve3NUR3O48xzQOpKZBPvcY5M6v9UZLUkKMHAtkZEGUjtID0lAms/5184bw/U02yA2fAFk50J68D8rdf4ZI76b7tMcN2GqDj61pQIcTYv7ZEJOn6Z2JfU2ihiIGpUREREQpRg0pE9xQ2dqnoNSrBa+RZWZQOlTJr9ZCe+I+QPXqQdMhvXRVvvp8cC6otdPPV6fH8oM3gdHjgeLSKC8QUtKagKAUX62D/MhXMpudC9ibIWsqgWgBMwCMHg9x3sUQU2cAzY3QHvt98FjlAWhvvaxvH9gFTI1cm1U21AW3fUGpuPCnEDPm6tnRM7+rN4P6+V1AlLVdhwoGpUREREQpJjSgDN3uSV2rBzlWAyxGJex550/Jj+n4aHCQe7bpAanHre/IyoaYdQrkF58EA1IAyMoNe55QDFAeeh6QGuS2ryDGTIAoLOn5BRORKQ3JzCrX3gntd9cBAMS4SVHPF0YjxLf15kRyyxfh1/IHpABkow3RWorJXVt9F1L07roAxClnQFgsEOcsDp547MyjfSspZWgWLRMRERGlMDXkb3u1l81jdtmcuHzVXixeuQvtHhXVDn1O6eOLxqEkyxyPYdIApz3/JJCTp/8HAGYLlCt+AeWJVVDuf1bfl54Z6MYbSqRnQGRkQZl9au8CUiAxjY5ULwBAuesRoGREcH9GL9YGNURWDIhvnAkYjUB1ZfTnHNoXuL5c96HezMhiOdpRpzwGpUREREQpJjTLqfbi73xVk7jp7YOBx+/vs2PFuhoAQHEmu+ymOtneCnkwSkdZZ5veSbelyfe4HQAghIDIzoO49Hoov36kfy8e2ugoEeW7/vdSWAJhsQb3G3ouIJX2loh94uIrgew8wBF5DADg71Q8YjRQVxU5Z5UAsHyXiIiIKOX4s6NGBdB6Ub7735q2sMev7gguJaP0cZ1TGhxkmwPa8luB+mooj6wM7yarquHZQX8Zr48SupxLX4V2m01UUJqWAWHWs5XK1bdB+9PvIDo3a4pCmMzo/K9JKApgNAWDz840DcjKgeHGuyHdLr2MlyLwrhARERGlmCanFwKA1aj0qnx3e70z7HFNqwcV5Tn486JxcRrh0CJb7ZC+stGBRntiOXDkIOB2Q774F2grn4Z0+NbtVFVAMUB861wAgHLp9bEfwOjxwPCR+nYCloSR9qZgOTIAMX0OlIeehyif3POTj58H8aOroTzqm0tq9FURmEyQnq6C0mBgL8yWqKXOxKCUiIiIKKW4VQ1r9rZgdlmmHpT24u/8F79uCHtcUZ6Dq+eWcC5pP8lWO7TVq6Bddwnqf7JIL5P1lcAOBNLrBXZsBiZNAxQF8r3/QK5eBbn1S/0EX0ClXHg5DE++2vu5oUdBKAoU//qlCcmUNutdd0PH0N1SLqHnKQqUUxdAmExQ7nkCyr1/0Q/4MqXy8H6od1wFueXL4JM0LTwbTFHxDhEREdGgIO3N0F58GrLyQO/OT0TTlAHoowN2tLhUnDspDwZF9JgprWsNz/A8dPYY/GxOCQTLdvtFe30ltFsug1z5NADfz+/PL4J27YVJHlmIJhugaRAnng5lxUvB5kWtDv2rowVoT0AQ7Qva5NaNMb2sdLkgd24J3+nuAELnkvaRKCyByMrRH5hM+rqtVYeA6sPQHv411CXfhdy9TQ/sFS6p1JMeg9JHH30Ul112GW644Yaoxzds2IAbb7wRN910E2655Rbs2LEj5oMkIiIikl9+BvnuKmi/+Tm0j96Kfo6mQu7YDO25x6Hd8CPIruZ5pSgpJV7d0YSxeRZMK06HQfScfKq0uwLbx5WkY0yeFQaFAWlfSE2DtDfp/616DiifEt7IZ6Cp15tZiYJiCKMJyMgC8gshN30O6V+PdP2H8R+Hr7RZ/utv3Z6mff4+ZHtrry8r/7MS2n2/gtwbEp94PIApxhUAJrM+p9Q/53bKdED1Qm7bBLn2A8C3Pil1rcegdP78+bj11lu7PD5t2jQsX74cy5cvx5VXXonHH388pgMkIiIiAqB3rjRbgMISyK/WB3bL5kY9I7LlS2i/WQrt/tsgP3hDz/IMoFLJRNhc246DzS6cOykPQggooudMqcsbPP7bb7EzaH/IT1dDu+HH0G6/CgCgnHoGUFwaeZ7XA7l/d6KHFzkOf7BUoJflCkWBmHUKsG8n5KdrAADKtXfEfxydGij5qb+7HtrKp/Rzqg5B/uUBaE8/EPn8NgdkQ31kdYTvunJzyPqiXo8egMeS0aQHu755pcplN+jZ2JA1Ual7PQalU6dORWZm1+v2WK3WQHmHy+ViqQcRERHFhaytAopK9T+gW+2QUkI21EG76f9Bu/oH0B7+dUR3ULhd0S+WojbXtEMRwDfGZANAr8p3Xb2ZdEq948/I+bN5GVlQbnsAYvapYafJl5+Fds8NkDVdrG2ZKLYafSmUvPzgvpIRgNcL+d7rwKhxENNmxX0YYvY3oh84uAdy9av6tqrqX3d9HXGa9pul0G65FPKJ5eEHrOn61/pq/bw1r+vZ4VjP8TSZwjOlJjOQngm0tQKWtECjKOpaTL4j69evx9KlS/H73/8eV155ZSwuSURERBSutgooHg6RkQnUHoF85mFot1wWdorynf/R1wP0c0fPwKQqTUoYFQGTQf8TzyDQY6Oj0Ewp9Y880Cn7mZUNYbFCLPxecF96RqA0FvbmxA2uE3lwD+SbLwPZuWHLwIj0kGSUuf9zL3tDGLqfcyml1KskAKAj2Cla7tsJ9fJFQGO9/viLT8JL9l36udL3HLl6lf61sS5WQwcAPfPq9QA1lXqnXZMZyMqG3LNdH4N/7il1KSbrlM6ZMwdz5szBtm3bsHLlStx+++1Rz1u9ejVWr14NAFi2bBkKCgpi8fIxZzQaB+zYUhnve3Lx/icP731y8L4n19Hef+n1oq6hFumnVsA87QQ4jhyE+tmaiPMypIq0h5+Da91HaLn3VmR3tMEyhL7PFqsDimgO3FuLuRIGoynsXne+94ZDejb5BzNK+W+il7wH98JbUwnLjHkQFj1Y0pztqK8+jPRFF6L91RcAAPkjx8AwrABeZyv8/Y2VtAwYLVa4AWRnZibt57Nx+S+hATDmF2BYyBg84yfCv0pt+rTjkZWg8flnXYb+DPr3pX++BtaTTocNgMjIQuberWh74WmoB/ZEXCdj83qkLzgPAFC/4WNIACbVC6PRCPPYCXDbapE2fgqyY/i+WrJz4Ha7YKg9AkyZjvySEjTlF8Dtm2aQObwU6fy31a2YBKV+U6dOxYoVK2C325GdnR1xvKKiAhUVFYHHNpstli8fMwUFBQN2bKmM9z25eP+Th/c+OXjfE09u+wpwd0DMmHfU91/WVgGqCmdWHlwjxkLe9QiUHZsh9+2EfOXvgfNanR1ob2yENOhzxlrWvA5l7KSYv5eBqt3XKdV/bzVVxaFGN2rq6mH0NS/qfO8b7Xqn1QsmZ/HfRA9kzREgvwDazZfr8wXLxsJw50P6sX07AU1Dx8hyKDf/AXL9h2hUJYTNBmkNLjmiKQrcvrmH9uYmiCTcc9lQD233NuDYmdAuuDz8+55TAOXOhwFXBzrGTYIrQeMTpy2E/PCtwFhC54e2/vVhtE+cpu8HYN/wGeTh/UBhSaBZU+Bcux3tNhtkWyu0Zj289jha4PV64cnIAgC4zrkgpj/rWmYOZHMjNK8H4thZsNls0CxpwTGpQPsQ/bdVWho5pzqafpfv1tTUBH5o9u3bB6/Xi6ysrP5eloiIiFKM9sAd0FbcA9nm6NX50t4M6W8UUlcFABDFw/WvQkBMmQ7lnMVhzxGn6B9+i3F6ICrXfTiklobRpF6y69fS4UWVw43vPb8z4lxVk1A1CZdXwiAAk4F9QbqjrX0f2u1XQrvhR8EGNpX7IXdshtz4GeTGz/R9eQUQ46dAuWgJhG/uYkTPFf/zRXJWZ5Rr3wek1MdYMiLiuCgbA1E+ObG9YjKzwzsVh5bh5uQDLt/8cJNZ79abngnljocir+P//RI6n9zeDOlvRJRXAOEvBY4RMbxM32h1ADm+NVAzgwk6YYpxY6UU1GOm9MEHH8S2bdvgcDiwZMkSLF68GF6v3rZ5wYIFWLt2LT766CMYDAaYzWZcd911bHZEREREYeTBYJmdtvRi2M/5AXD+/3T7HO0XPwFy86H8/knI6sP6zuKyiPPE4ksh//lXKD+/K2xuXOC1X3seYtFF/XsDg4SG8L/ra1q7XhJnyat70dyh4swJubAYuXR9T6S/62vInEYA0O6/LfjAaAIKinq4kAxeI0GNuLQnlgMjx0I56/t6g7C17wMTpkIUliTk9XtFCP3e+IXMBxeTpkF+/aX+wGTSmx4pBghrMBspvn0B5OsrIUaV6zv8Qe2MecBXa9H63OOQ6z+KbIYWCyUhv5ey9aBUbvo8uK+HObPUi6B06dKl3R4///zzcf7558dsQERERJRapK0W2u+uD9vn/M9LUE44Rc8m5OQFMkqB51Tu17MhDXX62oXNjfp5WZHTg5QzzgPOOK/rATQOnbI5KWWvkwN1bXqS4Z3dzUg3MSjtjnS59AxdWjqUm+6B9o8/A+1tgO/DFuX2B/RGPFk5EBnRKwaN46fAu2e7vnBsmt4VVm7dBDF9TvzHv+FjYMPHkBXnAZX7gZojEAu+E/fXPTqdfm47gss5yfUfAv71Us0WQFMBXzMvCAWQGsSYiZD+4wDgS6KJ2acAufloX/V8/IaeG9K92N/xN/T3TpQPyygcfwMRERFRfPmznJ1of/ottF/8L7SrF0cck5+s1pdtmHkS5FsvQ677ILyr7tHQND07NATKeDUJKEdZsDZjeAYqynPjM6BBSnq9kB4PZEc7tHdegXzpaT3Dll8IUVIGw/W/hXLlL/WTx0+FGFUOUVLWZUAKAHm/eQSYNE3PkrY06a+zdaMe8MbzvezcEnxgq4H87D3AZIY44eS4vu5R8zWMkof364/r9GVclKt+BeWuR4LnGU16S2l/oDfCt7auyZdrk7520x79vgqjCWLxT2CarM9JRWYcphmGfqhm0MehXB2SQU/venlN0sW00RERERFRKOlsh/afFwEA4v9dC3HCSZArn0bmscfD8fi9+kkeN2SrHcI3B0tu/y/kmteA8slQltwM+cGbkCufDMwT7S3l+t9C++PtkJ+/B/n5e/q++56FyMmL3RscYKSMnnGwRJkvOrUwDQZF4NbTIkuihzJZcwTaXVcH18UMFbLkixhWCOXW+4CRY3t1XSUtHaJouB4k+ktL66qh3XkVDMueisXQI2gfvAn5whPBx3dcpW/MmAuRntHFs5JDnHwG5OpXoT11P5Tb/ghZe0Q/MHIcxLDC4ImKoldR+IJS5brfAAf3AkazflxKyC8+gfZn3+8XoxHCZEbe3Y/BVl8fkZCNzeBDMqG+DK6YPkfPoNdVQ4ydEIcXTS3MlBIREVFcyFa7Pt/uwG6Iy2+EcnIFhDUdyo+vQfqZ50PMOiVwrnbXNdA2fALpckH7Y3BpOSEElNPPhrLsaYizIzOq3RFTpkfs0278cd/f0CCgdVG+W1EeXCdxf0M7fvzybjQ6vWAbkHBy/25ot1+pB6RTpkdm5zs9FmMn6mtU9pI45QyIcy+Eck8wUERDHWR9TVgmX3v1eaiXL4L2779HuUrvyEP7IJ97TH8fnccR40Y/sSCysqH8v58DVYcg//V/kM89rh/IGxZ+Yu0RSGc74JtPKrJzIaadECwRkBq0T1eHXFjfLxQFwmCIOu+83wyRmVIAegY95PccdY1BKREREcWFfPvfwOF9UK78JZQ534g4Li65MvigpQnyiXuh/eJ/g/tCmsCI3Hx2sOyFrsp3tZDK5Re/OoLmDhU1rR4ojErDaG+8FNhWLvkZlNv+COTp60uKhd+DcuPd/bq+GDcJyqKLIrL12q0/hXx3VeCxfE2f/yjfeLHvL+ZbKkU594fAzBPDx3HiN/t+3TgSx86EmHsa5Mfv6Mu9ABHzzdHeBtRUhs/j1E/Uv2oScAbnoyakyVDIGOMS9A4BDEqJiFKI9HohE9TNkahH9mYgd1iXjVxERhbE+ZeE7wzpBqpcdkO/h6BcfRuUm//Q7+sMFhKyxzmlXjUYofIPwU62bgxuW9MgjCYov1wOccnPIL5zSaDEPB7k3u0xvZ721H36RnsrDP75rwCUP70EcezMmL5WTBUNB1xOPagePjL6ObbayDJ8/wcsUgINdcH9xZFL3sRc6NI+7LTbJ/xdRESUQrT7b4N21Q+SPQwawmR7G9Tbfwb51TrIPduBHuatibO+H/74W+fqGzPmQZSO6vd4xPQ5EOOnhI/Rv0ZkCtJklDUxAYS2eFJDykSPtilSKpNtreHLhfh+dkXeMCinLYx7Bizm1/d1n0WWXrqt3P8slGvvhLAMvNLdMBZrcNvZFthU7v2rvryLX070TKlstevduv3MVsRb2L85A1v29AWDUiKiVLJnG4DwNSGJEkl++CZQUwltxd1AXRXEnNO6PV8oCsQFlwYfHzMDGDEayncu6eZZfTDzpOB2aBYlxXRVvhvaeFgNqeXNMDOr46c99nt9Y9osKH/8+1HNFe2TznM9fSWg0bpEay88CfUPtxzV5cVJ39K/jh6vf83O0+deDnShQaknuM6uyBsGZV7I75P8gvDn+X7wZeh8UiA+3XajmTpD/+rv/ktHhUEpEVEq8ZU6ab+7Hup1F0PGY5FwGnKkqkJ76S+Qu77Ws0ndnbtvZ9hjUT65x+uLb34byu0PwvDkqxDZeTDc9UhMsloDHNYAACAASURBVKShlMtvCC7R0FAf02sPJFJKiCjtRWVIrjQ0KM2yDI2gVB7aC/XyRZBHDkU/3lAP+JZOUX6yNOp6uLGmXPcbiLkhQZZ/XmLofEgAsvKA3o16zzbITWt7fX3Z5uj7MkrJZEkLbrd3+n0T2kRowtTwY/5sZeffQQmaNy18GWnZak/I66UaBqVERKnEmhack9fqACoPJHU4lBrkmlch33kF2vJboS29KPo5TQ3QXn8B2PV1+IGSnudzCcUAMWpcLIba9WsYTcDocgCAbKiN62slU68ypXIIBqVffu77+knEMe3tf0O77QrAaIK49Lq4zhsNJYQAQudF+sp35Rcfh50nP3o7ONZH7+n9CzTUA/mFPZ83wISVF3fOGodmrzt3EPYH9R43xImnQ/zPz8KqMOJukr4Oqsjmmr99waJnIqIUII8cgvxsNVB1GOK4WcDp50C++HTk/9CJ+kDu2hr+uNOyI1JTod19PdDSBEw+Dsqii4DxUwBnG8RAWjQ+O0//w7WpsedzB6kug9KQ7bBMaYqX78q6KmjPPhJsROOIzGLJ/7wIjJkA5bIbw9fDTITM4FI9UPU5oHJ9p6D043eCD/wloj2QUgK2GohJx/Z7iAnX3dxalzO43XnuZmizoanHQ5k3P6bD6ok45QyI0eUQo8oT+rqpgkEpEVEK0O66OrAtTl0AGAz6H6Edzi6fQwToASWE0mWJmzxySM9+zjwRonwy5Et/1csLfU1gpJTQ7v0l0NIEsegiKOdeGHzyQApI4VtawmAEVE/PJw9SEtHXKQ2fUxrcTvVMqXz/TSDkQxXpaI48SdMgxkxMfEAKACFlwrLNAeloCZYRL3sa8rV/QH66Rj/BaILI6OX8SGe7/vu/87zLwaDzEjAhZKMt+KBzl9vQZVl6GbzHkhACYEDaZyzfJSIa5KSmhu8YPT44J8edul1Gqf+kqkK74jvQfnoeZFND5PHmBmgP3wWYLVAWXwpk+crS7E2Qh/dDe+0FyA/fAvbu0Pf3olQ36RQF0FKzEYnDpWJ7vTMsU5pl9jXPCTnP6Qn+zkj1oBTOTnMSo2RKgS7Sywngn4cIQF+Xt+ZI8LHBEGhWBACwWiF92dQe+X/3W9O6P28g6iYoFWVjgg+s6eEH04KPWUI7+DAoJSIa7Gyd5seZzIBvTo7c8ElKL39B/RSypq224u6wQ9LZDu3uG4G2NijX3gExrChQlqbd/jNoK+6GfPUfkM89FnhOoubi9UsKB6V/+PgIWjpUtHuC7+/x88phVAB/WLpqeyP+WxUMzMypviZMR6fffx3tkedoWrBJTqKVlAW3hRKe/TMYgPEhzXzyC4FdX0P6PwTqjr9EvfO8y8HA/wmKwQDljofCj4U0NxLGTgWf/rLfotL4jY3ihkEpEdEgpz3/RPgOozGwLptc/xG0O34GaW9KwshowAvtztypy6X8/D2guQHix9cE50iVhixk32lZFfGtc4GxE+M10tgRqRuUbqnVA67a1mB5cqbZgPw0E97bZ4e9w4uN1W1hz0nFmFR6PVCv/SHUh+6CPLwv/KBL/yBGetyQqgq562v930GSPlARxaX6/GsA2LkF2p9+FzxoMEAoCpTrfwtlyS166XmrA9qyXwAApKZBhnywFEq75wb9+glYozP2fFHp1OMhRo4NP9TNGqAiOxfioiVQbv59HMdG8cKglIhosAsp9xKnn6PPawldl63RBu3pB5MwMBrw/HOOjcbIZSi+WqevFzr7lMA+IQTENxb6nhO+hqNy4eUQg6FUUFFSch1BVZNdBph1bXqQev9n1VA1idKckEAlFYPSd1cBzjbg641AXXX4QV8Qp/3s+9DuuQHaqn8AAMTwkZ0vkzDKDXfrH+hMmAo4WoIHfOWpYsp0iBNOCguc1V8tgbbku9B+flH4PMvOLIMxKO1aT8u7KKefDZGd1+05NDCx0RER0SAmayoBe7P+x8yRgxALvwcAEJ1LtmoqI5772o5GGBWBsybyf+BDkfbJu5DPPqI/mHAMcGAPAEBWHdIbqzQ3Bta9DSVOOQPweoCRYyBXPp3IIcdGipbvVtrd8DfVzTBFzznsrHdibJ4FJVkWVLWkZlm/dLkg3/m3/iA9E8r9fwMO7Ib2h5t9J2h6Z1oAOOTLohYNB46bnfjB+gijEYZb79O7WF/xneD+TnMrRUFRcG7w8DIgIxPYv0v//d5VQ6OJg7D7bg9d48W3L4xco5QGPQalRESDlDxyCP+fvfMMjKO62vBzZ1e76l2yiovcC8YNsE0xGDBgCIHQCQQIIQEn9I8SQgmdGEICBEJogRBKCMVgIFRjMMY24IaNu1xkWbIlq3dtm/v9mN2dberFknWfP5qdtndni+6Zc8776n+9E+zRaJdc3eqdfjFyHACFNQ4sQpCbaOOF1Ub5pQpKBwZy9w5kSRHajOOMFYVmWaOIT0Q2NRhKuo/e4c/WiBHh5bhi+GjE8OuRzY3ILz8Kz0T1dUKCUql7kG/9CzKz0I7/SadPK5ubDHXU0D63XqKk3ijFvv/EIYxKi5wda3LreKTEejDW7GIoQcs3njM8mqcdhXbWL4z3Y9R4xJW3GNYqm9dBaJ/9/n1tZuB6A9GaFQoEBWuWa+5E7tyK/qdbwBMsdid939/pxyKigisa+gVtOJlpZ0b2Slb0b1T5rkKhUPRDpMuF/thdIDS0W/4UMSDVnnrLfJA+CIBrP9zF7z7YGbav4uBGejzoD9yIfOEvSG9PXaBdkNywBgD9gRuDywdj4lo8p4iODRch6Q+EZkr3FiIXLUS+/ixy59ZOn1a/9gL0x+/uhgF2jiavuFFabBSxUS0HN/VOHasmsFsOfBDWGtLhQF+2CM8z85EVZQDo/3wMz9MPIXflRz5o3ffIbz5HnHY+lt/ehggQEdKOmGX46BLi+9lHERE8NsV07w0lXw+q7wZIiMWR/Oht4+/Kb3pqeApFt6OCUoVCoeiPlJcavpBnXYLIHhxxF2G3oz34jPEgUNBGMeCQH77hX9avOQ/5w3fICkO1WXvwWbSHnkUcf5pRshtIqOVCCKI/9quFCh3pAWmZSMqsHWHrj8hAS49OIJ0Os7y0A/iC0pgWSnd9FNc6mZGXQl5K31VllfuK0G+9HPmvv8Hq5cj8Dcb69Sth7bfoD90UMTCV2zeDxYoI9MoNQLvxXrBakW/2/bJz7Yr/C1/pu4/gE/vx/Q3JlFLvVVeObfmmkkLR11BBqUKhUPQT9BVfon/xAfprz6D/8XcAiEHZrR4jMnMgISm8XE1x0CM3rkVf8LIR4IQIoeh/fxC2bUSceREiMxsRn4h20Ty0+54OFslKSWv7idIyISOrm0ffg4RmSgN9H+MSwvdvB0FewcUFnRuXF/3+G9D/cieyob7tnQNochuvKdraegb02plZnDvZtMwQfVHpaF8hNNYjzr/CeOx0IDf9YKw77GgAZEGEoLS4ALKHtFhCLZLT0O74a/j6mcd329C7hZYUZkeMRcw5A+1XNwTtJ90h3qXez6OYfWpPjbCH8d6U6QMl1YreQ/WUKhQKRT9A/3xh+N19ixVGjm/74OwhyML+WbLr8hilhn2h36s/IRvqzVLSmmrk8i+MUtwm0w5EXHgl2omnBx0n4uKxPPYanjt/C6XF7VIk1R58tn8puGqWYPXdgAm9LNyJGDaq4+d0mLYcUpdduxw1VVBSjD7/VrTbHkHExbfrMF+mNNracr4hKz6KOSOTAYjxlvha+mB6QrqMclQxchwSkK88bbYZensk5acL4PjTgg8sKkCMm9zqucXgPEPpdtc2/zrtihu7Z+DdgPb7hyE9M+I2oVkQF/zaXOHzNPWEBKVSQlom4syLe2iUPYxUQelApA/+FCkUCoUiEH3lN0ZAethRCJ9IDaDd+qd2BWti7EQo3NnhzMuBpqrJzblvbOPDrcpjtb1IXUc2NcLeQnPdqqUAiFPPRbvpAcThx6Dd8lBYQBqEz7O0hdLwQITF0rZAS19C04Izm26zH0/++6nOld/6yiVDztcpdAkjxkJJEXLpp+0+rMmtE23V0CL8JvzzLMNn9qajzQzpjUdmc8nkDEal9sESbF+7QaTMdab3NVTsRwYElrKuxig/H5LX5unFxMOMv6echeX597s62m5FjBqPSG5HhQL4bZnk4v8FZUulywXRMepmnqJfoYJShUKh6OtsWQdxCWi/vtkv+MHQEYgRY9t1uBgywsgMlZf24CBNpNuF7OrEHNNbcUlBbRt79h+ky4Xsat9ia+f/9kv06y5Ef+Q2sNnQrrsb7fHX0Z5biHbqOYhxk9CuuhXRhk2Edu1diNPOQwT4Ih40hJbvhpQ+6k/e3/FzVpaZy5vXdXJgXqSOGDUBMrORP65u92FNLr3FftL02CgWXjyOMemmj2xyjJVzJ6b1zcClpNjIkkX4/PlsrwD0h272L8uvPja2t5EpBRBTZxp/J0zp6kgPLL5M6e7tyI8ChO3cToiyHZgxdQsd76lW9H9UUKpQKBR9HFlfBwlJRp/ULq86aEoLnnSR8Cs0ulvfr5vQn3kY/U+3IlUfaxjy7ZfQb7oUfeFrhoUIhtes57G7u+d67dnlXxRHzUEcehgiytbhwEMMH4N21iVdH09fpKYK1n7rv8Eja0LEnbSOT42kr2d38HDkisVdG5+uGwHZ/n2wbUO7D2ty6cS0UrrbH5D79+L5zRnITxcgph2FiItHu+2RoH0iWZzIhnrk5wthygzE0BFtPo8YMhztuYWICVO7bewHhFiztFt+95W53uXylzn3S3LzABCHH31gx6HoVfr3r5dCoVAMBIp2gc/aQBg/29olV7f/eF9pZUDJosOtt7Bz68j8TcaksbXJcmkxFO5Af/FxpN655wkkv6LZ3y/Xn5GOZuSenYBAfvhf9DvnoS/9DP21Z2DTWti8zphcOxydKiGVG9ciV3xprmio677BH0x4S211X492SXGQsIwYMrzj5/RlSouMmwKyvgvZfSk7FRg3u1vOlPZlZOEO5NpvkQ116HfMMzfkePuZh4/2r9JunR9+/L49yM/fg6aGDvlX9skMcQcRVqtp21RVYW5wu/p1plRkZKE98y5aXxOgUvQo/e/XS6FQKAYQsq7GMHYf6S3V9amhdkTq3y+GYQZ2r68vb2Hn1tGfNbIW+jsvt7xTfS0kpcCa5UFWJB0lcMr42rqyFvfrD8h9RejXnA/5mxBHzDKyP+mDkP9+CrasN/ZprEe/4SL0a85Dv+u3yMb29wDrX3+C/sQ9kJxqruyMYM9AYs1yAGRpMWSaKtZy5dKOn6uq3FAt9glD1VR3flxS9998ApCBwUYrtFa+25fRX34K/emH0G8IFuWR3ywCDHEfxhyC+OmFiNETwo6Xa1YgF32AOOxoxOBO3FDo79i91j6Btl/NTWDru5Y/7UFY+lGfuqJb6H+/XgqFQjGQKNgO4O8f1W68D3H5DYiO3AXXwhUa39tc2cLOrSOmHWkstGAVInUPNDQgjjkJMf1Y5Idv+stUu8KOyn5eClxeYi6Pm4QYOQ7t9w+jXX0H4kgjGyD/9bfgY/YVtfv0cvH/YOhItD/82Z/1Eyed2eVhH+zIhnrjpkDITR5Z0v5rDyCrKyE5De2Ge8BmQ1/4ahcGJUETaPc+ZTxcv7Idh0ia3P20fLdwB4yZaPQwH3k8wuvPqV13l38Xyy1/QjvDzIKKX16HOPIEAOR3S8DRhDj1HAYkkUSRqisRgTeoFIp+QD/89VIoFIqBg6zwihN5MzkiIwvtqBM6dhJLePluR9GXfoZn/q1Ib3YJpzPyjo0NRqYnPtHwE5S6oSK6ZgX6C38J2/3p70r4cGvkANmimbnSPbUtPF8/QTaZ4kZivCHEIoRATJmB9qsb0f7vfsQxJwUdo8+/FVm0i3bRUI8YnIewR6P99RW0x19HdKIEdKCh33CRkVXasSV4/d8f7JDIEC4n2KMRqRmI0y80elbbEUyGIqU0glIhArKurd9AklJy++eF7Kpy9LtMqfSKr4nUDLSzLjG+CzNnY3n+/VazntrRcxCXX2/8tlV7r097PHUPQrRr7vQHpvo7Lxu96Q11HdMdUCj6AP3r10uhUCgGCFL3oL//OnLdSiPzlZjc+ZP5y3c7H5TKVd/A3j2QO8xY4WohSKz39jHGJ/p7wmT+JvR//An53ZIwMZ9Pt1fz/Kr9kZ8zQICxzuFhY2kj1c29I9bU7QSU4oqklLDNYvxktEuvCVuv33t9m6eWug71NX6lUhEb125vS4WJ+MXvEEediDj7MsMn9G/3triv3LkV3av2ChgKvoEZ6kG56IFqqO3F55+qaUbPowhRCo7AjkoHm8qMaoT+lCmV+Zv81QDisCM7fLwQwmgT8HnvWvqxsE8XEEkpiLlnAyA/eQd9/u+NDSooVfQz+s+vl0KhUAwg5LIvkB+8ARtWw5TpXfOB1LoelOJ2w5A8LDfeB3mjYct69LdfCt/PK/Ai4hMgI9vwhPQJygC0ojDr1iV/WbaXPTUOAHQZbAtw+6JCfv/p7s6/hgOE1D3Izw0vRO3q2zt28PAxbe9TU2W8P+mDOjG6gYd2z5P+ZRFQEqodNxft8usRc37a5jn0v/4R+do/kI3egMjt8ntGCmsUYswhhnpuR/H1fft6SjURfHcmAl/sNPtX+0umVDod6I/cZgb+ieE3atpFepa5bLW2vN9BTpA9mLe6QgzQzLGi/9I/fr0UCoVigCHXr/IviykzunYybwZHRlBjTYtp50TOY2aCKMg3zvfpu+H7+VRH4xMNoYrUDONxQpLx1+mIePoVe+o45z9b+bqglie/NfovPREm4yX1Xfc/DeR/W6vYXR15TN1GwXbYv9fIxE2Z2bFj25MhLzOul1BBafuIifUvipyhxkKS2X8nomyI4+YCIFu6kROfYPzducUQI9u5lSBvxdR0qKtB/89zHVOg9ng/3z47jzYypU6PztcBPr79JlNaESJcFpfQufMUB9yksgzgoNR38yrgsx2x11Sh6MP0k18vhUKhGGBUmeq4oiuluwAW46devvJ3/6rxGTEAxNmC/w1IKSOrfbrd/kxQa/gDX28pqfabmw0xnwt+baxvoex3/temBYrdavSS6r3gn/7cqlJu/KidfZsdQDY34nnyfjzPzEd//lGIjUdMnt7u47WHnjOud0gQLxvrkbt3BK/z9uWpTGkn0DS0B55BuydEZCrD6OGWLz0e+TivMJJsajJVYgMDqxTjZoxc/CH6/Tca2fI2Mp4A+n03GAu+75omzJLeCHxfVE+909w+OKmf2ICE3iDrbLl5XoDC9EBXax0+BgJ614mOOXBjUSg6gQpKFQqFoo8g3S48Tz2A5+mHoKjA3JCR3eIx7SJCBsHtjfhCAz/59afot16O9D6/lNLwXAzIlIoZxxk7RwqC/JlSY4IuRoxFTJnhD6xl/qY2h9vgNLJTJXWRA9iSOidr9rbfLqUlfOXBnk4Ev9LlQrpbydoWF8L6lYaATmM94txftl8NMzYekZGFmDoTCncimxqRDXV4fnMG+vUXoT9wY3CAU15qCOOkZXb8hQxIAsyGhEAMykF4b6L4sRnBnfxuSQun8E6fXE5wGiXp4sLfmJtTA/r5inahX3UW0mun1BJy01p/1ttfiiosrWZKv9hRQ1qsld9OH8Spo5M5emgnM469TXNj8OOOWFwFoJ33K//yweA72iVCLWAG+vVQ9DsGbq2DQqFQ9DHk2m9h3fcAiJmzDcGV2HiEvYt+c8lpMHUm2uzTYK2xKr8icm+nXLHYWKipQgoN/R6v+E5SCmJQLgDar2/Cs3UDYuyh4ce//S9jwR5yl37soTB6AvKdl5FTpoMtOigbHMiOSgcOt85/1peTEWulrDFY3Oiq93cCsPDica2/bi8uj45bD++3K6s3gl67JXzypn/xIXLxh4gTT0fMOtlvwSPXLEc21CNffRowMpoiUjDozVhoV/0eMWp8u8YJoD3wjD9rJE7+GXLlUuTnCxETpwXv6HSgr/oGmhoNhd7UDETUwBR66TCBmceWFIoDqgKklEEBj2xsMGxMwOgldTrAZg+uaPCVrQc+7eplrQ5Lf+xu/7KYeJh3fC33lLp1yQ8lDZwxLpW5ozvZk9kLyPJSI9gee6hfEVrWBvu4drpnXilM+xGTp8OQEciVXxt95h0pG1co+gAqKFUoFIo+gnzzRQCj3LWrfaQBiKgoLL/zCuysDba+aHbrrCyqJ96uMS7V7rfG0Jd8DNs3mzvWVAULiVgsQcJJnifu9WeXIDxrITQN7ZJr0O+7DvnGC4YP5J5dMDty9ujlH8ooa3Rz/4lDWLuvgQWbOuerCnDnoj1sKW8KC2ILKo3AMbTkUUqJfOM5Y/k/zyE/XYB20wNIb6AaiP6f52Dd92hPvRV888CXCYrpWAZIDMoxl/NGw9SZyM/fC88sNTUE+ZqKE07v0PMMaALLbLNyI+8TmHX64TuYavYC6//4k7nN5YI6U/nYT0LI4/aQkWUEb8lpCF9Q20pPqcsj0SUkR/fdslVZkI/+4E0AiHMvR5xylrFh2waIjTfK1Bu7UPWgsoF+NK8vsV5TiVy5NOj3WKHoD6hbTAqFQtEHkE4HVFcY3oQd6D3sKuWNbh5YUsRtnxXS3BSQPd24xphsBxIUlFqDfU83rIY1KwAQV9wY8blE9mDETy4wJkx7Wu/j/N/WKqYPjmdSVhxRETKZHWFLuWGX4fIET+53VRiBXm5iSCa6zFRNFRddBZXl6HfM8wek4vwr0K6+w9jBm9lmg+lpqX/7JfqST4wHCV0rp9TO/AU0NyE/Xxi8IdD3dObxiPOv6NLzDCREdAzacwvRnngdkZkTeZ+ArLNfYdfHrm3msttpCB35hLx8x8dG6JG0R7c+Lp9/7ZiJ5sqQntJ9dU6u/XAn1c1ufwm+Veu7gZks3Gk+qDTFjeTWDTBuEiLOKFXvND4bpJ9f2flzHGSIy69Hu/uJ4B5nhaIfoIJShUKh6APINcsBEKec1au9UbceY07K39nsLamzR6M99DyEZmsDbRssFqTHKKsNVRcVE6a0+Hzi8KP9y221cv5yqlEWa+umEr29dcE9oMU1RhD+Y2lwFlLu2OpfDsxc+telZcDI4JJcudPMQMt/PgZb1kNWLqKzVhe+58odGnlDQKAkph9rKB0r2o0QInLg6CPwM5e/IbiH12n2OstVywzrlwhiZOL404JXOJqRG9a0NiiwWBC/vC5gXXCmdOHmSgprnCzbXdcvglK/BVSoaFd9bXDfbScR8Ylo/1iApioF/IgoG2Lw8AM9DIWiw7T5n/7pp5/m17/+NTfddFPE7UuXLuXmm2/m5ptv5s4776SgoKC7x6hQKBQHP/mbIS4BceQJvfaUaTbB0cMSyU00yrze2lZHgzUacc5liKQUtFPODtpfTD/WfBBYvtsUUlramkpvlJmVdAsjkJo9PDFsYj05K9Y/rsBM6SWTzV6974uCFTxldSWeP92CvmxRxKeuagruTXW4PRHX41Wz1R7+J1givJYoe7gwS3V4ebEYfUjEcXQLgdfc0702OQoga7B/US77ArnkY3NbYE9qQT7s34c49PCwU4iQ7w85Q9Ffejysn9KPxwMJScG9wZrWak8p0OVKgh6lyfRxld98jufxu5HrVxoCUVHdU14qBrA/qUJxMNFmUDp79mxuv71ls+/MzEzuueceHn30Uc455xyee+65bh2gQqFQDARkQy0kJvuFQHoDDWNyHTilnT/xMjMLF1q+mzvMXI5LgM3rkLURBDVaDUrNba+NMLwghyXbmTs6ONM0Os0USrIFTLqTAvrnHlxSHHSM3PQD7Nwa1GsZiMNtjFNKyQNfFfG/Tfv924JKe/ftMYSDUjMi20zExhmZyYDrIb2qw4EZtUi+sN2F3PqjsTB0JEw6oseeZ6AiAoJSALZuAEBWh9slafc+hRaaFYVgAazRE9B+czM0NiDfejHyk3o8ECr4I0SLPaV9OVOqv/YP9A//a2SS4wPKSDeuRX/yfuO1tmAPpVAoBiZtzn4mTJhAfHzLJS5jx471bx89ejQVFRH87RQKhULROvV1nffq6yS+WE8PCKR2x2WZti5DgkvAAsuKtTln4HY6qdu+Pbi3FIJ7T0MJ2Pb+EMNaprjWSWvJnsBMkDvEw6a0PmBi6wuiWyjLbPYGpfsbXKwsDhZXqXUYr0FKiczfaCrmBvQBav94B+3Ge2HEWAAs9zyJ9uy7hrKwoxlZuAP9j78zTxpQBtwlxoRnXOUn70D2ELRbHuq8cqmi3fhK1S/6tIy3h54AQ4Yjfn4l2h+fQOS0UGIdiGZBDM6DcYciS4oj76N7wm+CaFqLPqW+z3Nfy5TqSz5BfvUxcuFr0FiPdvkNiF+ZfebiiFkAyNK9B2qICoWiD9Ktt+QXL17M1KlTu/OUCoVCcdAjHQ4oyEcEZiJ7Ac2blXUFGHUmuRqMABkQ6YMQZ1xkbJgQ8tuemsHfxl3IJZsT0AN7xWjD3iEmDvGT8xGzT2VcjSF2dHhOPJaQbM+wZLPM12Yx/1XtqQ3Ortz2WaH5oLjA+8IiT9Id3tfZ5Aqf5PuCUqoqjFLckV6l3oD3RFijEBOmBgXnQrNATCzs2YX+0M0QEHCIMy+KOI6Ool17F9rdTxjP4yN3GNrNDyKiY1o+UNF9eL8rjW54fcRctNPOQzvhdMSQdvbu+T4zFqvh+RsJXQ8PSgN6SnUp/b7CmjBurgBkxPYtKyCfXRKAdu/fEZOOCFaVPvVctHm/R7v8+gMxPIVC0UfptkL8DRs28OWXX3Lfffe1uM+iRYtYtMjo9Zk/fz7p6V1vcu8JrFZrnx3bwYy67gcWdf27B4/Hw7rv17OjsISfnT0Hj9NNVLQNSytCNK41K8DRTNLsudh7+D1Ic9VREWWU09msGunp6dz/Exvz3loPQJKznpTLLifKO456u40GIG7SYcQHjM3dWMs3gwxBI/nDKn8JQ6WA5AAAIABJREFUsCV3WNufo1/fQNPSz0j8bCcj4gQ/nTacZ5cXAJVMH5rMFTOHMjHbtNRIqwYwsionTchlSFoizyzfDUBlk5v09HSklJRvWW+IJ9XXkVRTQdTIsUFP6xA20tPT2e+OUFZrjyc9PRlXfTWVQOKQYUR7X4f+6mfoNVVYW3hdNWkZNDsdRM+eS/NXhupu5ttLu1d8aPBQmq64gdqnHiJ69lwSLr8OLYK4Tn+ir//myDe+ZP+FxwNgj4khOWCsCUnJ/s9Ha5R6/0ZpGqnp6VTHxOKGsNet11RRtnIpIi4haFu51UqUzcaKbft5ckMtvmL7nPRkKhqMGzTjh2aR0sHAtCevve81pz7yAlHD8gBo2rgab/0FKZmZWKcOzJLzvv6ZP9hR179v0y1B6e7du3n22Wf5wx/+QEIr8vdz5sxhzpw5/sfl5ZGN0w806enpfXZsBzPquh9Y1PXvOtvWb+OhNTVURSUACeQ/vYD/kctPKObKi0/07+d2u7FarcjqCuT6VVhWLAagNiMX0cPvgVWapbYxeCgvLyfbZmRedAlJCTHUxCeDdxx6rTGVbHQ4aA4YW0OVGdg1fPIu8ZOno511KTI7t12fIznyENzWYrT6asrLyynxnu/7wmrumJUVdA5nk6k0OzzWjS3NDPZm5yXy7dYiRlqb0KsqELNORi79jMqbL8fy/PuA+do2762ivDyWsooQYSagoqqa8lg3cr8xpa5rdlAf+DpsMf5rEvZaTjkHbfpxuEaMBW9QWlFV1eY16DCTZ6I9+x4uTaPS6W5xPP2F/vSb43S6Kdvwg/9xXX198OejDVy1xudc93iQDkfY65ZrvzX+NtQFbfNIieeH79le4KJ6+Fxzf0cjO0vribYK3A3VlDd2rIS3R6/9sFHQ1EhNSqb5O1Ji2ixV1dX3+O9cX6U/feYPRtT1PzDk5ES23gqly0FpeXk5jz76KNdcc027n1ShUCgORjYXlFAVlck1aZV8v6+Z/9lyAfgfuax+8VticbHfEke91SzDnFlWyxEim8N/ejjJdntLp+4Rjk83y3ajrRqNLp0EQkoLff2iluB/F7UecxJc7rGQeNp5LduXREDY7Lij7Fi950+yt/zvaFBccBbIp8oL8FVBLV8V1PLmCYnGP7RRE2DpZ3gQaFJS2eT2lzwW1hhlxg5PePmuf51PfKUDyqAiJQ1S0tq9f1foTSEshcHGpOF8ZJ/GTU//CcZea6xs7/uQMxT2FvpL4n3luzd/UsCYtGiuPMLr0en9Hoif/SL4eE2DxgY8KcaNmCS7hRqHB+lsprSmiUFxtl61kGoL6XZDWQli6syg9WL0eNMCqjUhNIVCMWBpMyh9/PHH2bRpE3V1dcybN4/zzz8ft9uYtJx88sm8/fbb1NfX88ILLwBgsViYP39+z45aoVAo+iBOb6/isbOnMWFfGd8vNzJ80R4HUXhIFi7GWKvY76xmjc24ibc151C+dR1KisvCS1L2+AQz8OxZUWZw9vDJw7j2f7vM3jff/ocfg/x8IWJSsOWF1Mx/H9sShzFyRHCpbHtwa1as0vh/cv6haUzKimV4SnTYfhlx4ZPYp386gv/7uMAv9uJxOLACwh5NsxbFz499kItW7eH1bWZW1Gf9sqW8CYAXfz6FbUVlzF9azL/XlhFt0ViwQeeShMGM7Sa7CkX/566pvwWgqawMfB/zdn5PtavvQL/jKvCp9losPJE7l/yKZvIrmjl5VDJ5KdFIr/epOOKY4BMIjb0Jg3gz7ySElNwxezC3frqbHa/8m9rUcSRFEME6oOzaBo31iBBFaDF0JOLSa5DffgWJSQdmbAqFok/TZlB6ww03tLp93rx5zJs3r9sGpFAoFP2Vd+uTwApRNhu5w4fwD1GExWph0ODssH3PfG0LAM/8NI/FJZJnl++mtN5FVkLPBkOB2rUW3cyKDtHrSHXUoNuCJ9t63hi+vetVjsoKbs3QA/olN+YdwamdGItLWInzZohsFo1JWXER9/Opi07LNrfnJtqIs2n+oNS57AvsADExNP7uj7ABPv6xBOxGb6omoN6p88CrS1mjZXDssETGZsZTV2N4RqZX7uGDj3eyIWYIa1PHMTahcxNn7ZaHkPuKOnWsom+zJnWc+aC9mdIMbybU1/9rsbIkdaJ/847KZvJSosHlFQuLCqmW0DQeHnUuAFIIv4bXK8ONb9xYT2QP0wNGndeD1fe6A9BmnQyzTu7lASkUiv6CqgNSKBSKLrJ7WwFLF39Pg9VQQvWp2ubkDY4YkAKkuIxyPk1ozB5lCC+E2pT0BDIgV2rZtcW/rD96O5rU0UP+LXywtZJHlu7l64LaoPW6MPdbam9/2W4gbs0S1OPaGm8dbeWOycET9id/kke2zQhKb26ewCdTz4GRE3AnGaW0VfYAsSSL8TwrRQZJOLnqiEEADIo2rsfUiq1+Vd7m7GGIzMjvW1uIMRPRjpvb9o6KfsHuuEH+ZWdAdQCifdMnIQTa3X9Du+OvxooQ8audVQ5kxX7kK17F2qio0BMQ4NiEFpKh3erN+vcG+hcfoH/8dqv7SIc3uLb1biuCQqHo/6igVKFQKNqBy+miqSFcIOe7r1dz43cNPLovkWxHFf85Y3C7zvfISUO4OrUCW7SNoSkxDE60sXZfQ9sHdpHAbkrLrgAfzfJSLFLH4w7uKa1qMoK5isbg9XqI7cvOyuYOj8UtLFi92VrpdiF35SMbI18Dy4P/h7j3uqB1Ma/8jac+u43bfvwXtqRknkuaQX6dTr0MLwL685L7/ctXr3uFuOLtxnm9QfGHOUexM97oAS6MHRR2vGJgcuMRN/mX3SPNUtl1Ve27mQIgBuchUo0bT9VasIXPrqpm9Ef+QEl0qrEiQqa0yWKu01r4fvQG8o3nkQv+Hbyuvhb9y4+QvsjZ4f0d6OX+eIVC0f9RQalCoVC0QdHOPVz92moe+u93QetXLlvLI7vtDHdVcGlsKXccm0tsQny7zpmZm8XJpx7tf5wVH+XveexJpG6mXXzlu3LPLnzpmCVRQ6h3mBNuq7de0C2DywSlV/joqP3riLEK3t5YgcMdLiDUGi5hIco3ho/eQn/oJvTrfx4+5mbvzYAm86aAdDqQP65GANMrNnHT0UaPbkWTm+c2GxP3ayq+BmB6+QYSXY38+5u7eXC0k6lUor/4ONLpIGp/MRfs+oyx9XuY4C4DYI2ezN4QP9T2UtHo6lSAruj7uJJNMavdJZ1TVy7QzOy9DQ/byhp5J34iv5t5GyvTxoMtpHw/NYOKaNP6R3MFf7ayw1uwexX9+UeRrz9jegQ7veOzHeCBKRSKfocKShUKhaIV9u0u5ravyyi1JbPenk19jVF2u2bFOh7eaWWYq4p7zp7COWcdx5BRnStjBUiwW6hztD/70llkQPmfUxr/AvSnHgCgNMaYdL+6rsy/j8X7X8KjhwSl3vLFWft/4Ccj4llWWMf5/93GM9+XtGschTUOSqwJ5Dq8k/uqipZ33h9+Tv3q86DBtKWxRxmZW6dbZ2uVC4DjLzuPN3N28fsNRnYnPj6GidMnIc66BEqLaVr8EXLzei7YvYhbSz/jnqPS+b8G48ZDnbNz78Vv39/JjR8XdOpYRd+mUTMDxsao2Fb2bJmCmAz/8imihBhnI6+NMPpDPx97SpjQmQxR49Vk8I2f6fW7OjWObmOT1ybHV7br+2tXQalCoegYKihVKBSKFnC73Pzlix3URcUy3GEEao8uWElZcSl/yhcMdlVzz1mTSEhObONMbZMVb6Os0U1Fo6vL52oNPaCn1OPLfoZYNPjmxS6P5L8/GsHiGz8GB42+GFVIyYljTTPyj/Or2xzD3lonP3hLlWdvW4RsbgKfsND4ycimkDLpavO5pa4jAwPYaUcizr6M6FhjElwWUGZsTUrBfvypCF//X1qmMWav6Ezds39GLngZAO30CxFjDyX1TENUpqNZXx+OviY8o+g2ljaZYl9j88KFfNpDTWyqf9lVVMh1+WaP5ipbjlkG62VRY/BviyVku8MWXA7cG8gt65G6B1m4M2AgzcZ3c+0KsEcjQnpnFQqFoi1UUKpQKBQt8NHHy8m3G4HML8YaE9K1thwWLtmAW7Nw20kjSUzpHnuD44Ybk883fuxZY+/ATKnmU9/1CjMlOwwxowS7MaFcvbdl4aXtlYbAiv2a24mJCp6A6rLlwMzh1vntBzv55+r9xvEeJ/Kjt0w/1N070K+7ELllvXmQyyyllV9/ilyz3Ny2ZgXaqeeQZLdgswj+s964eXDeIQG+od4SZDE4z3gcWiIJfgEauzc17OxAcOlw69R3MrOq6D8Uu82bNyKzc0GpJ94MMmeVrmFaqoVhyWb/ZejnbtGOmqDHoZnS0bWFnRpHR5FlZrWC/pc70a86C/1+051BblkH1ZWwZxfimJN6ZUwKheLgQgWlCoVC0QK1zWbWLS3NzJKsbo5lnLOMrCE53fZc2V4rmM+211DW0HPZUolgdG0h46t3MbzeO9H0BpF3r38egJRoo180zhb8LyIwe7hoRw1psVYOyYrHEmLZ6NZbD0oD0aQ0gmKPd32jEQhLX1kg+D0cAajYj3zjefOxN8CMsmgcOigWtw6xURq/mGKWSeLxBozZQ4y/kTxIvWqhNu+LcXjMcTY4Peyra7nH9JV1ZVy+YDsLNgVkdFsJzDuLR5dsLW9qNehX9A56K5/xUB5bvpfPtxsVBB7NQoKrgQVf3cr42t2IoSO4a7YpjhaaaQ/9LsVowd+f2ataV8PtLuQPRlm7dudjiONPC9/+5Ufov/+V8SBvdK+MSaFQHFyooFShUChaoNltTgjj4k2PzHJrHIla92fGThxhZF1//d4OVhTWhfVxdgXfuXQhGFW7hwd/+Ae26jL0z96F0mIA0mbPAWDFHuO5t5QF2018V2RmTp1uyei0aOxWLcymwtVKltEV8poE0ug/00OuZ+A5AzKlhFi1aNfc5V8el2GUMoYGyX5ivH2A1ghBaYJR0pvoDch9asO6lFz0Vj7z3t8ZfoyXsgYXLo/k5bVmL+7tnxd2u+DRk9/u49ZPd/Ord3dQUKXElHqT40pW+z1CIdjvtzWa3Tpf7arlqe+MG0BuXWIJ/KwPGUFGXBRXzzAyr05PcNCpS4JskxItkqu2vuN/3NJHvdvxjXlQDtpFwd704oJfQ7P5WyFUP6lCoegEKihVKBSKCDQ1NLLCYSrpJqYk8cu4UgCcFht58d3/83nZVDO7N39pMb/7YCf3fbmHhi6WhjY4PVzyTj6vrStDFxpawJRavvWSfznqpDMAWFfSyDUf7uLVdcGlxD+WmnYUbimxeAPH2JCMamjgGUhoeaJAIhf8G7n4w6D1ct+egBO2nDkW4yf7l30Z3jpn5H5Q/2Q5JS18Y4JRVpkaYyUzLopN+42+1oIqhzkMj+SHfQ1BWVApJc0unTHp0dxyjJk531TWxP1fFbU47o6wrqSBx5btZVmhIexU1eTmzkWF5Ff0nkflQEdOP45zA0rCPe3MVn/hLb+NtnpVrHWwxpoiSWLEWACivBGv0x18Xl1KrCJgna5zyr4AFfD0XrIv0r3fKS38d0+bcwZMnu5/LJsOnG2NQqHov6igVKFQKCLwzDvLKbeZ/aKa1UJ2qpktnTV1ZLc/Z1K0lYUXj2PBz8eSm2ijpN7F6r0N7OxiVqzW4aHBqfPmhgoarTF8l276LZKRhXb339Aeeo6oaDPDYYuQbgzMgHp0icU7kQ7NlG4oDfdz9ZFfEfxatJYm92u/RVZ4M48uMzAkf6N/UVx6TdAhY9KNTOlpY5KJiNemQsTGof1+fvC2eLM8e0JmjD9LHFjGe+4bW7l78R62eV/DzZ8U8LPXt/JDSSN2q8YxwxJ576Kx/v0HJ0XIyHaCP36xh68KaoMCeodHcteiPVT2go3QQKWwxvzc6UJw0aR0rt7yJuCveG+T51YZN7JSY4x+VI8uscbEIGYej/jldQhv5t/mDVq/2BncQ6pLiPKWqOc0lgUpTgOQ0bne1g7TUGcIokUqfccbmHoRE6f1zpgUCsVBhQpKFQqFIgI/6MGBjRACjzcosOpuBo/svP1LW1g0waA4U1SltrlrmVJfX9rwFKNvsjw6BYaPMTZaoxCD8xAZWURZBNkJxvNePi2Te08Y4j9HXJQW1N/m0SXWgP8gxw9P5JRRxjX78zd7I46jttnNC6tLGZlqCrvYPS33asq3vVncgEypXPGlf1mbdXLQ/sOS7Sy8eBxXHdHCRD0mIEM1agIxJwVMpDVTrGlwoo2qZg+byxoj9sfWOzy4dRkUYCd5xaECLT1yE7onKI3Er6Zl0uTWuXzB9i5n0hWRufZD025Fl8Z7O7KuyPu4Y6X18d5qArcusWoC7Yob0Y6e49/u8wN+e2OwyrWU0p/5z6vfi/7EPQA8f0gzT1d8AI09n5WUDXXIT9+F5NQwyxo/uXnG32GjEIkpPT4mhUJx8KGCUoVCoYhAih5cGik0weRp4zhJL+LFU3N7/PnHZ5pWD+tKWs48tgdfP+l5EwPKVn2WDSGTzLmjjcAy3mZhSnYcb14whvtPHEJ6bJQ/QCuoaqas0U2Ty5yY33BUDr+bYQaDkYR+nl1VSoPTw3Uzs3ny9OFcPSMrqJQ4EDH7NOSqb5BbfwzuKe0KWcHvm/ApoU6ZEbQ+J9EIJm/7rJBIzjAOjx4WCKbFRoXtt7W858prcxPNgHftPlUu2dP4glDh/bx2VGvKbjWDUosWHtgdkmncMEmKDlWyhqnZcYy1NPCLnR/712dYPUblRnlpxwbSCfR7rzcWAr6H4ti53r+nGH8TEhGXX4/22z/0+HgUCsXBifVAD0ChUCj6Gi6ni33WROZSzCcYgYymacQnJXDNJXPaOLp7OHNcKgVVDnZXO1hSUMO86YPCymTbi6/q0+o9/vDyTVBTFXHfM8alMjU73m9TYbdqTMqKw2oxM67Xf1QAGIJIofxyagb/WltGk1snNsAqZktZE9/sruPiyenkpRhltEOT7LSU4xPnXY78cRX6q09DSXFHX3Lkc8YHez5ahw43FpzBQe8hGWZGdUVh+GvcXtFMWUNw2ey4dPMmwhWHZfLZ9mp2VjmobXb7xZM6Q6Mr8hUKvLYxVnV/ubsJFRnz9U9bL70aNsMj3+xldlE9J45MYlJWXKRTtHjeSG9XvM3CrGEJ7Kh0BK3XJSRHW7h7ZiKyaSqyqhw2r0PW1Rqluw11yMZ6RGx8+Em7Ael2QZW3tzzOLHHXLvkdXPK7oH21o07skTEoFIqBgfpPplAoFCHs3LyDZoudQ3MSONtqlKJqEQQ+ehK7VePWWbmcNiaFZrekpgslvL5g0qIJXlp2L7dsfAV8voN7g30ONSGCfBN9WDWBK7J+UBBJ3gDMN95t5U18V1THfq/NzcwhCS0e62foSITNjjjz4sgBqc9vtJ1o9z6FdtMDYeutQ0cYCyV7gtYnx5hB5Kfbq5kxOJ7XzhvNvCMGIYB3NlXy4hrDZ3Xu6GQy46I4LNcMTM4Yl8oFE9MBqO5i6fV7mysjro8PEJd6dlUpxbXdlE1WAPBjSF/0oYOMGxXphx7qX/dVQS13fbGHpQW1bSplS4zvYWmDC5sl8m+JRYiwsmBdSoQQiLzRaJdfjxjkFdNqbkSkeysTejJbuqfAXE7oHk9mhUKhiITKlCoUCoWX6vJKli7fyAt1hgruIRNHcExGOpcdwDENijfKQvfWOUmJsVLW4GJbeRNHD0ts40gTPSAoTXJ1rtTTqgl/cGuzCJweSVZ8eMlqore3stbhITsB/rJsLyX1Lr/nqb0lvxYhzJrIOCPrI9IyIxb3ar9/uEO9dCJnKOSE9wBbfX1wleVh2x6aM5RlhbWcMjrFH6SfOiaFMekxNLo8fvuXiyal89vp4T2svtfb6NKR3sCiPTQ4PdgsgiiLRkWji/c2VXLMsAROH5vCbZ+ZNxDibGamtNGl8+zKEu47sef6nAcagQJXZ09I9Ze1B2aofTy6bC/nVqdxSaA3LgSJUG0obeSVH8rYU+MMUmkORNNEWHArJUSo9jXwKe+W74eh3S+8BgR5zoRWGigUCkV3ooJShUIxoMnfmM89q+oZ5qlha1Q6bi2DaI+Dce5KUjLGHejhMSbNKHXdXNbEIZmx3LN4D0W1Tt4aHN9ixiUUtzfYa9G/sx1YNUFRjZOX1+7H6ZHEWDX+fMqwsP38QWmzB5dHUlpvZEgbvGIt9hZKTbW/voL88L/ILz5AeD1D/X2vIYjoGIiOibitIwh7eEbYxyGDYjlkUGzY+pGpxvvx6NzWSzZ9wct9X+3B4da5fFomp49NbXNMF72Vz6jUaP5yah5vb6zAIyW/mJwR1sMYG6Xxws9G4tYlCzdXsqSgtkPB70Cm0eXh52/mA/DPs0aSHqEfONCaJTXGGnRdA2/QzBmZxM7KZrZHsOd588fgmx3vba7k1NHJHNPCDSWLgBBHGHQpWw5Kvd8B6XT0nF+pJyDTn6CCUoVC0XOooFShUBx0bFi9kbT0ZCxWC5m5rVsmvPH9HuptOWy22Jkr9nHyYXkMH3fgg1EfidFWBifaDN/MQ9Io8QZ55Q1uvyBPW/iSPtYWZ7dtMyo1mk37m/zlpBMyYyL2SvqEWmocboprHUjg6hlZ/P07o1y4pf5HEZ+IuPA3yDGHwPgpxkotclDarVgsiDN/0e2nHZ5i59i8RKI0wbaKJl75oYzDc+LJaoci7/bKZqqb3CwvrGNcegzZ3mPuPn4wTq9XqkUTZHgVmnMSbTS6dGodHn/5tKJlGgJ8bAuqHJGD0oBM6YzBwSXnNosZlA5NslPr8LC/PthLV0rJx/nVQeuirRpXHJbZ4rismqCqyU1Zg8v/3uoy1HIpYNnqHbe7ZR/fziBXL4O8MYi0DHB4FaYTkhA//Xm3Po9CoVAEonpKFQrFQUNtVQ3/eftL7thiYd43dfzmq2rqa4KFapqbmnjxP4tprKsHoE5ayHZU8fSsJK66+ESGj+uhMrguMCzZzr46o2fQp2r77Kr295EF9pT6EHPP6dAYLp2ayds/H8tPxxp2Dy1laRMCyncLqg3RlnHpMcw/eSgXHpoWnim1RsGhh5vjmnYUwmfdEtrHe8hUxJW3dmjcbWF55l20Uzt2LdqD3apx09E5XHdkNjcfnUOzW7I8gjBUS1y2YDvVzR5iosxrMC0nnplDEpgXUi7sK6Muqe/e4GQg4GlBRjfQEzYzpEw9sFw7yiJIjrZQ1RwsfFXnDG/AbnbrRLVS3dDkbdp+eKnZR60TUr7r7SkVqRlg9d6AcHefV63+7ivozzyMftsVyKIC9M8XAqDd8yQisQX/X4VCoegG1C1VhULRrynaUcjuwlIG52Zw24pqGq3ZQdsrSiuIT0pg8Wffkl/WwNBkOwv1HJoWfsex4waRb0vnFFFC9rCet3npLEnRFioajQzKlOw4Vu9t6JA3pT8oDci4iLMvRX7yTofH8m2REczXNEeeCMdYNaI0wd5aJ4t31pBg08hJtGHVBOMzwsthLf9obQwBAUNcApYb7u3wePsCw5LtTMiI4eW1ZTS5dC6enNH2QV7aI5Tky77uq3MyNr3rZc0DCSmhsNpBYrSF5IAsc2BQGsqxeYm8t7mSHZXN2CyCeJslKPsKUNXU8UDR990KbCvVQ0qyxQmnI4aMQIydiGzw3uTwdE9QKivLkB+9ZT7+7D3YsNp4XhWQKhSKHkZlShUKRb9jxVer+HHVBgAeXbKbR/YmcN3KZhqt0Tw0XufV00whketWNnPbi0t4oiyZj8hlV5VRjvaZyOXOrVZsupvjJkQWHukrnDQyGasm+P1nu/2qtvkVzXy5s6Zdx++qakYTpmgS0Onew+tmGkH/zyelR9wuhCDRbuHzHTXsq3Nx66zczpcN6wET/Q4q7vYlhBCcMMJQLn1zQ0WL+4UqrwL+z2trZMVHIVCZ0s6gS8m1/9vF1R/sDFpf773p8+wZIyIe56tYsFk0fznvV7vM72NZQ/h7MSSp9dLttFgjKN5R2cwGr/pvqNCR0DTE2InGA4svU9o977tc+U3w4xWLjec87fxuOb9CoVC0hgpKFQpFn8XZ7OTd95ZQvCvYsmN+cTx3brXSWN/ALruZdUp0NTBu8lgSUhL5+0wzK7fZPsi//KkYDEC8u5Gr0yr51zmjGT91fA+/kq4xIjWaB08aikeX/pJYgMdX7KPe0XYmbe2+RkalRvtLa7vCxEGxLLx4XKvejInevtIbjszukIdja4jxk7vlPAeK2AALl4KQQNPh1nluZQm/WrA97LgnfjK8zXPbLBqJ0RYqG7uvjPNg5rV1Zf7lpbuNbGN9QKbTo0v/zYNBERSmwaw+iLIIRnh9d7/cVevfXhpwg+CCQ9N44WcjeSSCMFggD51kqifvqTG+57qUQRUOQUTZjMC0Yn+r520v8sM3glfkDkN77FW0s7q/51qhUChCUeW7CoWiz/LBx8v5d2MW/1rewMgvlzJnkGD0UFMo5OcLg4PVa0aAxavYOnjkUPh2CwBvnTuC1SvWM7/YsBoZ6yjlkV8d10uvonsYnhLNk6ePACnZXtnMP74vYX+Dm41ljWFCLKHsq3NyZHv8QbuJ8yam4dFhVl4X1ToDMoeiB/o+e5O4ACuRNXsbyEuJZmt5E//9sZzVew17m2HJdqoCynX/emoegxNbVggOJMFmYV+dE48ueXhpMY0unQfmKIuYSAQGj8sLzT7fT/Kr2FbezPAU85q3VFFQWGP0eOcl28lOsDE6LTroLr8vU/rgnKFMjKDiHInkaCvXH5nNEyv24fIGvbqkRWVdYbEgps5Efvc1emIycscWtKvvRFg7PrWT1RXQbCoIi59fiThilrKBUSgUvYYKShUKRZ+kuamJD6vjyKCGkxKbeJ0sdlQD1UZGY15yOZ+XSnbYM3jltGw2rstn2szgbNrtQxqoqGvCZh/HjGOnwX+2AbA1IHPan/De8RhzAAAgAElEQVTZrUzLiefRuXlc+s52lhbUcnhOfJCIUSCfba+m1hEsmNPTHD20myeyeaMRvaHE24OMTY/hxBFJfLGzhmWFdWwqa2RlcQNJAdnrCw9NY2VxAzaLYF1JA0OT2heQAkwfHM+CTZXcs3gP672ln8t213bIz3YgMiEjhk1lRjD2j+9N8bDkaAu/mtayUu41M7JYtKPGr4wcpQl/IAlQVOskJyGq3QGpj6OHJhhBqccMSlurfhfHnoJc9Q1y4euAUXIrZp3coecEYE+BuTxmItoJp3f8HAqFQtEFVFCqUCj6HG8t+IpXm7LAlsAv40o562ez2f6vxXwfZfZ+nvqTYzg14JgjZx8edp4Zxx7mX9Y0jRP1Ir7QBjPFsQ/oO7YvnSEp2soFh6bx3x8raHIVccusXKJDlG0X7aj2W7F0wQ3mwOH1YRQZrdv69AdiojSuOzKbOqeH74vqibNpXDI5g5+MTeHDrZW8uq6c7AQb1x/ZuSDy0ikZ5CbaeG6lGVgtVUFpmzx00lD+8HkhFk34+zgB5p88zB9wRuKkUcmcNMoU/7FahD+QXF/SwMrien42vm1f2lB8/dduf6ZUhljChDBuEmRmw/59xuPK8pb3bQ2nWVJuueWhzp1DoVAouoAKShUKxQGlqaGRdas3sbG4hh1NFtKtHpZYTCXc+Gijp+uOX57Ama8Z5bg/EcV0Jqi88tyjyf3kO045PTyA7Y9cNCmDlGgrz64s5eW1+4m2alw0Kd1vO/HktyX+fb8rqueyqZmIi+aBpX9kHUXOULTf3Q79vJ80kLMnpDI+I4aTRyUTbzPeh3MOSWPGkIQOZUZDEUIwZ2Qyu6sdvL+lCkB5lrYDIQRZ8VGsKzED0gS7pdWANBI2TdDg9NDk0nnquxKyE6K4qAUxsNbHY/z1Va5Laa6LvL8wsqVv/8tY0dzY8s6tIJvbFtRSKBSKnkT9x1IoFL3Kd0vXMHx4DpmDs9iwegN3bwK3Fg/EQ4Q5eXy0OTlMdDVQGxXH3CM65yUaHRPDOWfN7tzA+yinjklhwaYKPtpWDRh9aWeOT8WjByu5RluNma12/Gm9PsauIKbOPNBD6FbGZ8SGWeNoQnQpIA3EsIQxgtJP8qvJSbAxfXA80VaNeJvWqk9mIAVVzazZ28DZh6R1y7j6ElYN3Do8cVoeAEOS7EF9pmPSojt8zpQYKzsqm3l+VSml9S7uPWFIuCdvO/BlRXUkDU4PEkNMqTXEUSci33vV8Ctt6lxQGpgpVSgUigOBCkoVCkWvsW93MQ8VxkJhNYc7N7HKlhOkAf7OBaOoq6qmqaGZJxbvYIt9EDExZlAaI13UYpTiKkwy4qLY32Aor764Zj9njEuhJkSV99IpLfTH2bonGFL0DY4ZlsjRQxPIr2jmlk938+Ka/by4xlBnPWVUMhMyY3jm+1Ium5rB3NHJLQr5PPBVEWWNbuaMSvb3Mh8s5CTYyE20kedVzc1NDM6K5ld0PEDLjI+iqtnDF16bpryUzn+vNGG4Ic3/uhiAyW0oWIuEJMRl1yL/+ViQWFGH8GZKtfue7tzxCoVC0UXUzE6hUPQaS1Zu8y+vspn9oY8fZuP+MS6sVispGenk5A3mj+cexuXx+5kw2SzTPSfLCLSS0pSReyBXz8gOeuzWJU0uQxBqUHwU/zxrJFOywye22lNvoT32aq+MUdF7CCEYlmzHl2DzCfas2FPHY8v30eTWeWZlKQ8v3UujK9xSqKbZjc2b5SsMsCBqCZdH+nsg+wOGeJAZjIf2Yvs8ZTtCZpzRZpCTYPxN7kLptCZAgl+wanQ7MrfazONh+BhkJ8t3cTQZdcJZuW3vq1AoFD2AypQqFIpeY1WthdHs5/pj88jIzmDLj/kMHzOUpNTwIDMuMZ6fnXls0LpTTj2aU3prsP2I7IRgL0WHR9LsNoLSK6Zlkh4b2WtR2FWW9GDFbtV47mcjKapxMiU7jp1VzXwVUKIqgO+K6nhsueSO4wb716/ZW8+9Xxb5H++udrSpIHvlwh1YBLxw1qhufx09Qaj3Z1SACtiMwfH8cmpGpMNaZZA3KN1b52LOyI4HtYEIBHqAHVKrQkeBRMd0OFPqeeD/YP8+xNEngi26xcy5QqFQ9DQqKFUoFL3CS298Qb49l9MoZsgowz9xysxJB3hUBweaEOQkRLG3zvBGXF5YR45XqKU3rWAUfYv02Cj/DYlTRiX7g9L0WCvzTx7GS2v2s6ywjrc3VjAoLopHl+0NO0dVk7vN56lsxz59CU+IzYo1oGfz/InpnQrMMuLNGz+HdtAGJhRNGNncOJtGakwHpmkxsVBT1bEn270dALnofUhUFSgKheLAoYJShULR42xau4n3PLnMdBVz2XlHHejhHJT89dTh7Kxq5s9Li1lVXE+tt6c0tDRRMTCZkBnLnJFJbNrfxD/OGAHgV/995YcyYr03L341LdPfgwpQ7wwv7w1kT03b5b19DV2XBLalB2ZKI5Uzt4cEm9l3G9VF/yUhBGv3NdDg1Dl/YvuzriI6ttPlu+L40yB7SKeOVSgUiu5ABaUKhaLHKMwvYOG3O9joigV7ClecdAjRcTEHelgHJTFRGodkxpKdYOO7onr/+jjbwSVSo+g818zICsoCBgoYzRgczze76zhzfCpVTW7e3VyJABq8vcmBONw6e+ucuHXJzZ/s7o2hdyuhPaXWgCCyM4q5ALaAbKuli0GpRRhl0xMyYjhlVEr7D4yJhab2le9KXUdu3QCjJoDVinbRvE6OVqFQKLoHFZQqFIoe490VO1hsyWUS+zgheh/p2WMO9JAOejaVGZPS44cncsKIpDBlUcXAJbQs9czxqdisgtfWlfN9Ub3fNujSqRmce0gady/eE7F8d+HmSl5bX86UrK6VqR4oPCE9pUOSbFw2JYOkaIvXUqfjBF7bw3JaV8ttiynZcQyKj+LiyRlBAXObeHtKpZQRS5Cl04Hwqm03fvgm+kt/MzYcRD7ACoWi/6KCUoVC0SPsLShiqcjkGHcxt/zqxAM9nAGDrx/t+iOzlWiJolUS7BZOHpnMa+vKaXDpZMYZUwJNCOLtFiZlxbJgUyVby5uCgrXX1pcD8ENJcKnoxv2NTMiI6fOfOz2kp1QI0S1+rDcelc2QJHu7vWBb4tZZnVTAjYkFqYPTAfZgxV59+WLkS4+jXXUr4vBjcO/ZZW5UFlsKhaIPoIJShULRIXZu3sFnawoYNyieXeX17GrSuPK4UQweYfYjrfxmDQ/sjgUtiktmjz+Aox14vPCzkdQ0e/p8YKDoG9is5uckwR48JTh9bAofbq3iwSVFPDBnKEOT7JTWO/3bY6waTW6zvPf2zwv58ynDGNPJbGNvoUtJFytsIzJ7eNdUd7tMtDdz3dQYFpTKlx4HQH/2EbTDjgZ3QAZcqKBUoVAceNQvkUKh6BBL1hXwMbk8VprEe55c1tmyuXpFA19+/i0ANeVVRkDqJWtITkunUvQAabFRjEht29dQoQCwB2T17pw9OGhbWmwUfzk1Dw24a1Ehe2ocbPaWh//tJ8O5dVYOU7LjODYv0X9MeaOrV8bdFTw6aD0RlR5oor03A9oQO9L//iB6TaX/sYiN78lRKRQKRbtQmVKFQtEu3C43RTv38GNjFElaPXdOS2T33nJ+KG3iG2suj+9PZstrX/AJZunZ7UM7aeSuUCh6BYsmWHjxuBa3D02y88Ccofzh80JeWrPfbzGTEWdlWLKdaTnxeHTJ2n0N1Dk8bC1v5sghCX06Ux/qU3qwIGJikYDclQ9SIlpS092wBqcnIFOaeIAzvAqFQkE7gtKnn36aNWvWkJSUxF/+8pew7cXFxTz99NPs2rWLCy+8kDPOOKNHBqpQKHofXddZ++163thSyzZ7JgDR1iQuTa1jzKTDGTNpDCcBeQu+YnG1ncXWTPAKempSZ/I0VbqrUPR3BifZOSQzlhV76vzrYgJUai2a4OWzRzHv/R28t7mSFXvquGRyBjOHJLC9sonxGX1LECm0p/SgIcfwf5YvPoYELM+/H3m/IcOhIN98nKCCUoVCceBpMyidPXs2c+fO5e9//3vE7fHx8Vx++eWsXLmy2wenUChapqa8ijVrt3L8STN77Dn+9PJXfG/LIUM4OC9qL8mxUcyaOZGk9GCbgvPOns15QGN9A1e9vYU5MTVcdsEJPTYuhULRu5w2JtkflJ44IiksE2rRBI+dOpxvi+p4fV05jy7b69/2/JkjyYyP6tXxtkSD04NLl36P1oOKtEyjP1SG2/gEYQ2Z+qmgVKFQ9AHaDEonTJjA/v37W9yelJREUlISa9as6daBKRQDEafDie72EBVtw2JpfdL0h/c3U2xPZVplNUmpyV1+3rrq2qB1juZmvrflcIJexG8vOhabvW1rkdj4OF755WFdGotCoeh7TMqK49kzRpAeF9WiTUm83cKckcmsL2lkSYH5e1JY42gzKK11eHhxdSlnT0hjaLKdfXVOshO6187IrUu2VTQDkNPN5+4LCCHAYgG3bvyNxKjxYAme+gkVlCoUij6AEjpSKPoQV72+jgve283Zb+Sz7Euz+sDtclNfU8eaFet45rUvWLXsB4rtqQA01jWEncfpcIatawlHczPnvb2TK5//Kmh9fbWRFRmdYm9XQKpQKA5ushJs7fLNnDMyOMj5eFtVm8c8vnwvX+6q5dr/7WJLWRPz3t/J59urOz3WSPx12V7uWbwHgOyEvpG57W60a+8yFgL6SWVDnV9hV/vNLeDxBB8U07fKqxUKxcCkV4WOFi1axKJFiwCYP38+6enpvfn07cZqtfbZsR3MDMTrvnDB53yxs4Y7LjyapNRkKm0J/m2P7E1g7Etf04CVIluqd60dyOXjAvMcVktU0HX78tNvuHMLPDcrmUOmTWxzDHc/8TaQRaEtlfT0dHRdR9M0qksNdcaM1KQB9770NgPxs98XUNe9ZzghPZ2kpCTGZMbz4Gfb2FHRGPE6+67/xpI6Vu81b659V2LcVPt4ey3nTR/ZrkC4PSwr3OJfnpiXTXTUQVjCe+wcqpd/gWffHtK813z/dReC1Ik+4TSSxoyjYdYcGivL0CvLAEhKScWmvge9gvrNObCo69+36dWgdM6cOcyZM8f/uLy8vDefvt2kp6f32bEdzAy0697c1MQje2IgKoaz39lBtqMK7Clck1bJUxVGELrVlhl23O1DG0lOjGVt/j7+48jm10sqeStjLza7jYKtu7jTO+9at3EHg4ZmtToGXddZhLnPvEfeoljEc15GM3UOD5CFRcgB9b4cCAbaZ7+voK57zzEsBhx11SRHScrqHCzesJtJWXFB+6Snp1NUsp8r/7staP2C9fsA2FHRyMOfbuSqI1r/HWuLGz/axeAke9C6+poq6rt01r6Lx+0Cl8v/2ZZNhgq6w+Ew1h01h9Sjjqf8N2cBUFNbg1Dfg15B/eYcWNT1PzDk5LTPGlCV7yoUvYjb5aaxzpgKLV68GoCRDuNu9SDRzBUJ+5l13FT//hfa9/HUjFh+GVfKbE8xz89OZsasaYydPI5JIwf59yvft5+SPXu5fpXDv67RGVKiFQGXM7jMd7eWSKUtgWdrMni92ZgIJsT/f3t3Hl91fed7/HX2k33fN7IQAiQQdhBRFBQR3BdQ1Fq1HWtnWm/vo9Y6d8bemWnHsWrVaWc67lq9sglYKhYFZJEqi0EDhAABAiEhkBVC9pNz7h9HDoRFINsvy/v5T3LO7yzf8+bwy/fz+35/359fBz+tiAx0N2eFExNo4+k1JXy0u/00Xo/Hwy9WHmx334yMUJ6+5vT1UvPOGEGtbXLR2ua57Dbsr2lmffGJiz+wnzCZLeA+z2JHptNdPnPEmQc8++NSxCLS11x0pPTFF1+koKCAuro6Hn30Ue6++25cLu/1ra6//npqa2t58sknaWxsxGQysWLFCl544QX8/XWOgsjZ7lhYBEBCczWljkiSmqt47sHJmM1m4PS1Av9tiIvF2yuYPnUYUQkxJGUkn/NaGUMzSN+2iX2OKDZ9vY+V1XZwhHF1WynrLAkcPnHx80r3FxYDMKT5KDOS/Jk0cSj/sLiASnuw7zFxyXGd+9AiMmBFB9r47Q0p/Mf6Ul7ZepQ9lY3Mzgrj1+tKuSq9luLa0wfS3rgtnYhvr4P6y6sS+Pf1pQQ7LLjcHhbvqGLBjkruHB7BvJFRl/z+xTVN59yXHd3PD7SZzOA+z0HJM1ZM7s3XkRWRgemiRenjjz/+ndtDQ0P54x//2GUNEulNGk7Wc2DPQYaPHtbp19pXUHTOfSmWpm8L0vZyxmaTM/a7X8/utPObe8by4MLdvFUfg8PawjPDYWjuNEzvrGIDUTxUdxL/oECKdhbx5qbDTIq2cLzBxby7rwEgf185EMucrBBm3DydyspKXN8eTX88upa0lFgCQ4K+oxUiIt/N32bhzuwIvi5vYG3xCdZ+O2r54Y5ycmL82X60geszQnwFKcDEpCDuHB7B4p1VPPjBXupa3NgtJnYcbbis9/7pimIArkkN5v7cKLYfbWBqaj9fbdZsvsBI6VmFaFwSHCnpmTaJiFxEj55TKtLX/GbhZrY74ng1upzoxM6d1/TZ1wext0Xzxs0pHCuzMn/zIR6Zldup13T6+RHfdoJ9VidTTBUMzZ0GwJjEYNYecVBxpJKUoEBe3VRKoSOWHce9z5tSdJDkjBS+OQ4pVDLmiit9r3lqclxubgZhUVoQQEQ6LzPCjzCnhZqm9iN4T12dgP8FFhyaNSSMxTurqGtx88SV8eysaGT1vlra3B4sZy1+1ORy0+JyE2C3nLMNvNdRjfC39f+CFLxF6XmvVXpWLlp1V0R6ERWlIudRV3uCt5ZvYbsjAYD31uzkodlOfvzn/QyjhvuvHHzeKbUXUlF6lOWeBHJcRwgKzSEoNJh/HJbRJW114u18eDh9rpXD6u3ktbZ6p9pHWlztnvPi+oOUbayh0RHL9Z7SdtueHh/KX/MOEhw+uEvaJyLisJp59dYM7py/G4DrM0L44ZTB2FouvNxQuJ+VO4aFExdkZ3JKMM1tHj7a7aGsroWksxYumrdoDy43TE4O4okpCee81tCofj5l90wXGintolWMRUS6g4pSGTBWr/yCpaVu4s1NZAZbuenGCez6eg8Oh40hI4fwzsK1mE1w311Tef+jLawxxXKrpZStDX6sdSSw9uNysPmzCX82bWpg0PrPseLm6dtGEhx24aPvLU0tPLLWu8DHfaO7/vzMarydM5fndIfDdlZRWtLmaPe//aAtHJfZe0eIvX1HJX1YBj/uooJZROQUm8XE342LobHVzR3DI4gMdlJZ+d1r4D4w6vSCPINCvfu6Q7XNvqK0ptHF0ZOtuL6twTYeqqO4polBYc5vVxCH0XEBTE8P7YZP1Etd4vRd89wf4n7/fyA5rYcaJiJyYSpKZUA4dric/zkaQLPDTnPLcTY1hvCnD4oBOwCzdn3GRx7v0fXov/6NihYTSZ4avn/fNPLe2NjutX6RUMd/lAZR7PBObV3+aR7DUyJ4fWcdsaZmnnzgaiyW09PRqiu8y49HtJwgK3d8l38267cjpfddO9x3X2NzK+DkzbyjPJ2WyEFHJPc6jnDr7EmUFZcRGhHJz/5ygGp7EPGhmsIlIj3jxsywDj83IdiO2QTFtc2E+zfw7jeV5z3HdEvpSVJCHeyq8G4bMpBGSeHCRamr/YwZU+pgLE8910ONEhH5brokjAwI+TsP0Gyx8w8RNfzHjennbP/Ik8DUtlJyW47walUoRQThj/coe67f6dUbp7aVcsXUcTwYcNR338LWOJ4usnPIEcEeUzC3z9/L2wvW4Ha7WfPJl/xoXS0ADw3qns/21LUpPDPM0+6c18hw7+JEux0xfPHFdgBiw/xxOJ2kZqURFhXJNf51AEy+qnPntYqI9ASH1UxsoJ1Dx5v5y+4aiqqazruSbkygnX9cdYhfr/OemjA5eYAt1mYygefcS+d4NnxiQGNERC6NRkplQDjZ1ArAhAnDCQoL5oVRtfznl0e4K9XJs2VB+Lma+PHcKTTWN/Cvy/LZ64hmUIu3aHvgjqu440QdweGhQCYAY7NTeWvT6SP0qc0VjPBr5UOb9wLBS1zxLHl/DxDqW1vCYT+9smRXih+USPyg9vdlZA+GHXsB+M/KMJxtzaQPaj91+L67pnKPqw1bN7VLRKSrpYQ6KK5tItLfRmqYg19fl0J1o4vvLzm9uvnvvzxC8xnXM3VaB9jx9wtdp1REpBdTUSr9Tkuz9/qcdofdd19lfStmjxv/4ADAe97ki9+eN/kvm/MZPDQDu8OO3WHn3+4ZxxsfbCQjzjut1Wa3ERoZ3u49zlxUCGB6NIwdmcX2T/ex33H6Gnpz7EdY0hRJq9mG1dJzHSOLxcKrV4ey4vMChieGkjNqKM6A9iMKZrMZs32AddZEpE9LCbXzZUkdlfUuJiYFAt4FkZbcMwSL2cTB2mb+XFjNZ/uPc6oudQy0otR0gem7IiK9mIpS6Ve+2ZzPP+/1FqMzPIcB+Lo1kKP2BNJaKtqd63nKyPEj2t12+vnx2H3Tv/N9EtOSmLNtHQtavKOPN944GbPZzHPfi+H2+Xt9j7vtxglkF+zj+Z0nSR6U0qnPdrmiE2N5cG7nLmMjItKbjIgJYGlBNZOSg7hv5OkDgKcuA5MS6uAfJsbx9xNiKa5tZvPhkwQNtINvZ18SJiIaqo4Z1x4RkUugolT6lY17joEpEYCVpkQCXI3U270jhHendd1iF2azmXvvuoYF7xX6boN3hPJ/xRznd0e9q/H6BfgzYlwOb4/rsrcWERmwhsf4s3DukIs+zmQykRrmJDXM2QOt6mVMJjhzNk9QCLhaMf/mFcOaJCJyMSpKpc/ZtCGPw5V13HHb1edsO9ZqJt1zjKdmZmGxWgiJCGPfziJe3XKEIUOHn+fVut648cNh+WF+Gl3bI+8nIiLiYzaB+4yitM0FKRmY7I4LP0dExGAqSvuJupoTzP94Cw/cPhmHs28fGXa1umhubMLh72T+kvWMz0okIjqU+au2c7jFQoEjFvAnZOUXTJ8xyfe8g3uK2W0OZ6Snisi409e2G5yTybM5md3S1hfH2Ck7UtXuvoDgQD6cl9Ut7yciIvKd/ALA1Yqn/DCm2ETv+aXnOXVFRKQ3UVHaT7y3Ygsfk0Dq2m1Mv2HSxZ9gsIJtBVgtVjJHnFss/vf8dawyJxDfXEOZI55F291ANZgSiDQd//b+MNaWNnHqzM+62hP8amMFmG08NL1nRkQBUrPSSM3ShcdFRKR3ME2ejuejRbiXvovlR09CmwuTRd09EendtJfqJxraAAvsqmjgu5boOby/hMDggHNWk+1JrS2t/LLADLhZmu0m74t83t1dh8tk4c4kC6vMCQAE0HrOc2eHNXHbrVdz79vbSHa4KSk6xPyNe9lsiqLFHsSTifXtrtcpIiIykJiCQzFdfyue5e/j2b8b2to0UioivZ6K0n6iyQ1YYKM7gnuOHGs3ffWUwq8L+af8FiJch/nt7cOxOWxYrTastp79GixYtgHwXs/ztvf3AE5weKcc/+6o9zFpzRU899BVACxZuo63G2IAGJbq/elwu2hye/jVhiNU2hO4orWU7HAHE6Zc0aOfRUREpLcxXX8LnrUrcC95x3tOqUZKRaSXG2DrpPdf5W7vAgaNVicPr6k+Z3vpgRJ+/XU9wW1NVNiC+OWS7dy/aC/zl66nuamJHV/t6LG2bjlpb3d7bEsZb0w7PXL7WHgVv3toiu/2DdPHAHAjpQwZ6T1X08/TynGXmUq7d5XbXzw4jVmzr/StgisiIjJQmZz+mK6eCbu3Q0O9RkpFpNfTobN+ot7UvtD7zVurKXc7+OnEWPz8nfzLZ6VgtvGv18STX1DM/1SHY/a4yT9pYemiIlxmKy8HFZOSOeii79XS1MKiP2/kputGExwWcsltdLvdnDxeR7Ejknud5dx121V89ukmRo8eQVhUJINbdnMCGzNmTm73PP+gQP50Yxx+Aem++0Y7G1nuSbjk9xYRERlQAoO8P5ubNFIqIr2e9lL9wLIP11Npj2ZMSxlf2b3TYjfZvAXbz7a5gJPYbAH8W46N+EGJxA9K5Or6Bn6zYBP5jjjf65SWVVxSUZq3OZ+FrXHUfrSVH9837ZLauOOrHfzfnW4GuWrBEU1sqB9ms5lpZ6ye+9rjN1FRUXHe559d/I5Nj2J5kff3O61lgFa7FRER8bHavD89Ho2Uikivp6K0j9nyeR7v7mnAA4TRigsTOxzehX0GB5l46tYMTlTVYjKbePCT0wXedeZjZOWeLiD9Avx5/MZsjh2ppNXl4p/22DhcdfK875n3xTe8seskdw+ycdW14zlW2wAEYjrz4twXcaTiBC2WcCrcfgAkxEac8xirzYrNbruk1ztzmu6U3NRLboeIiMiAYD2ji6eRUhHp5bSX6iMa6xv407Iv+IgEcPgDcBBIba7A2dZMk8XBqIw4rFYr4TGRACy+K5g7F+0DYHB0wDmvGREbRURsFACROzZR4nK32+52u1mybAN/aowBh4MlxRXUfLiew3WtYIJmN/zundXcPDaFnUVlTJk0jLCoyHPex9Xq4m9HmsAOv70uGbvDTkhkWKfysJ+xOFOsVtsVERFp78xC1P/cPoCISG+iorQPOFC4n3//4hhH7eeeQ/n0TcMIizp31BHAZrfx/i1JVJRVkJQx/jvfI8LTxHF3+0WC1q/Z4i1IgciW45RaQ3jjpB1M3u1rLYnen9tcQDRblm/nXx+65pzX/vAvG8n7dlqx1W7pdEEKEBoWBNQB4Azw6/TriYiI9Ccmm803n8l05XWGtkVE5GK0VGkv5nK5+MvyDSzZdIATZie/znLx4bwsPpiTwV22Mp4bYb5gQXqKf2AAKZmDLroqrQM339jj+MO7q9lXUPTt+3tHTse0lPHq98bx/r3Dzvvc7OZyAPIdcezdufec7akoRyEAABdSSURBVLVNLt/v5i6aQhSbksBVrlLuc5Z3yeuJiIj0Jx63tyQ1TboGU0jnDwaLiHQnjZT2Yp+t2syrJ6LACsnNVWSP8a5Ka7Vaue/ua7v0vewm7x+vT0wJfLLNxeKMVvZVNWBvC+CJuZMwm83tCtvrPaUkBNmYOWM8DmcWt7xXCMCTec08Z9lHalY6+Vu2s67wGKvMp0d4LV242ML//t6lLbIkIiIy0JhGjIW5P8B01QyjmyIiclEqSnuxmvpm3+9W3N/xyK63M28XK0ggvK0Op9+502PPXnU3obmaUkc4LrOVx79qZfqW1d5i9IyC1NnWjH+QzmsRERHpbiaHE9O0m4xuhojIJdH03V7so7og3+9zB3dvMZfgbL+S7oHyGgBOWpzt7n/l6hCezT73+bfHm9rdXmOKa3c7obma9+4dftFpxCIiIiIiMrCoQuil6mpPUGsLZExLGR/Oy2LClNHd+n733DyJ6z2l/CisEoDQAAcAD0eeaPe4mMQ4how895qg02+YxHMjTn+d/j3bzPu3JPHMMG+xe0eCCatVA/MiIiIiItKeqoReyNXq4pdLtoMjggmxjh55T78Af3583zS+2ZwPNfBGqQ1s4O+8tOuGAgzOyYR877mlWbnewnXoqKF8OKpbmiwiIiIiIv2AitJeZumydfy5xp9qh3dV3bGjBvfo+8fGRxGxs4RkTjI2pIEJV4y7rOc/ndbMifqmbmqdiIiIiIj0NypKe5HSAyW8VR/DUM9R7o9qY8LEbAKCA3u0DTGJcbzx/biLP/ACRk8a2YWtERERERGR/k5FaS/y0Rd7gAT+7opEUrPSjW6OiIiIiIhIt9NCR72E2+1mY0sIk1tLVZCKiIiIiMiAoZHSXuJoyRFqbYFkhzQa3RQREREREZEeo5HSbnSo6CBvzV9DTUXlOdtqK6upqagC4Hh1Lbt2lwAwNCO+R9soIiIiIiJiJI2UdqOX1x9kryOepZ9U8s4MCyGRYb5t31t5DH9XEz8ZVMwzhwOAUPxdTSRlZBjXYBERERERkR6mkdJu4mp1UWIN8d0u2Lnf9/u+giIAGqzObwtSr4y2GqxWHScQEREREZGBQ0VpNzm4t5gmi4P7/Y4CsP1wLQBbN37NhvxDAIxvKeMHwRUkN1cxqLmSe0d3/FIsIiIiIiIifZGG5S5i+9Yd7Ciu4J47r7ms5+0sKgOimTI+k4p1BXxEAmVvrmGbPR7wnjf697eMJiQ8lNld32wREREREZE+QUXpRfyf3VYgjjluNx+v+BsLKp3cGlKPn93CzFlX+h7ndrvZ8vk2ckYNwT8okC2VbSRSTUxiFj+YE8Xxd9fxhT2h3WsHBAX28KcRERERERHpXTR99wJaW1r5/Z9W+W7/41vreeV4JMdtgbzdEMMfayN57f+tpq72BCVFhzhQuJ/flATws0XbqT9xkp22aMb5NwFgtVr52dwp7V7/XscRrDYdExARERERkYHtolXRf/3Xf5GXl0dISAjPP//8Ods9Hg9vvvkm27Ztw+Fw8Nhjj5GWltYtje1Jry9az6fmRN/tGhznPOYv7jg+/fMBmiyntx1xhFF+uJw2s4XBMUG+++1OO2NayvjKHs+LY2ykZl3edGAREREREZH+6KIjpVOnTuWpp5664PZt27ZRXl7Oyy+/zA9/+ENee+21Lm2gUYqabb7f/3tyIH98aBI/jz/hu29E8xGeH21nAudeg/TJrY0AhAT7tbv/n79/LR/OyyI1K72bWi0iIiIiItK3XHSkdNiwYRw7duyC27du3cpVV12FyWQiMzOT+vp6ampqCAsLu+Bz+oK9jmgAfpFQR/ygLACuvGY8lnVf8czhAHbZIkkflsHPhmXwM+DQ3mIW/W0fOTF+7K5opKrFxKCMcQZ+AhERERERkd6v0yc1VldXExkZ6bsdERFBdXX1eYvSVatWsWqV9zzNZ555pt3zepMzrxV6850z22278dbpPPP7L7jZ2f5zR0ZGMnrS2B5rY39ktVp77XdiIFD+xlH2xlDuxlL+xlH2xlDuxlL+vVuni1KPx3POfSaT6byPnT59OtOnT/fdrqw8d+prb3DmF/Z8bVwydzAWS1avbX9fFRkZqUwNpPyNo+yNodyNpfyNo+yNodyNpfyNER8ff0mP63RRGhER0e4fuKqqqs9P3QV4wL+cIcnnP5pisVh6uDUiIiIiIiL9U6cvCTN27FjWr1+Px+Nhz549+Pv794ui9I7bppI9JtvoZoiIiIiIiPRrFx0pffHFFykoKKCuro5HH32Uu+++G5fLBcD111/PqFGjyMvL4yc/+Ql2u53HHnus2xstIiIiIiIi/cNFi9LHH3/8O7ebTCYeeeSRLmuQiIiIiIiIDBydnr4rIiIiIiIi0lEqSkVERERERMQwKkpFRERERETEMCpKRURERERExDAqSkVERERERMQwKkpFRERERETEMCpKRURERERExDAqSkVERERERMQwKkpFRERERETEMCaPx+MxuhEiIiIiIiIyMGmk9DyefPJJo5swICl3Yyl/4yh7Yyh3Yyl/4yh7Yyh3Yyn/3k1FqYiIiIiIiBhGRamIiIiIiIgYxvKrX/3qV0Y3ojdKS0szugkDknI3lvI3jrI3hnI3lvI3jrI3hnI3lvLvvbTQkYiIiIiIiBhG03dFRERERETEMCpKRQYITYoQkZ6m/Y6I9CTtc/ouFaXSY8rKynC73UY3Y8DSjtp4+v73vOrqaqObICIDiPo6xlJfp+8acAsdbd68mQ0bNpCdnW10UwaM/Px8XnrpJerq6sjJycFs1rGQnpSXl8fbb79NSUkJNpuNyMhIo5s0oGzdupW8vDwyMzPxeDyYTCajmzQg5Ofn88ILL9DQ0KD9vQG2bt3KwoULKS8vJzIykoCAAKObNGCon2MM9XWMpb5O32c1ugE9wePx4PF4+Oyzz1i2bBmVlZWMHDmSoUOHGt20fsvj8dDW1sYHH3zAxo0bmTdvHhMmTGi3XZ3z7rd//34WLVrEXXfdRWNjI+vWraO8vJypU6fidrv1R7MbtbW1sXz5cj799FMqKyvJzs5m0KBByr0bndrvvPXWW+zevZu77rqL8ePHt9uu/U73y8/P54MPPmDOnDkUFRXx8ccfk52dzejRo/X97ybq5xhDfZ3eQX2d/mFA/CuZTCbMZjOxsbE8++yzPPzwwyxYsMDoZvVrJpMJq9WKyWRi4sSJvp30rl27cLlcBrdu4MjPz2fo0KGMHj2acePGERoayscff0xDQwNms1nTXLqRxWIhPj6e3/3udzzwwAO88sorAPrj2I1O7Xeam5sZN24c48ePx+12U1xc7Nsu3W/79u2MGTOG3NxcrrvuOpqamlizZg1NTU36/ncT9XOMob5O76C+Tv/Qr6fvrlixgtWrV3P8+HHS0tKIiorCZrORlpbGypUrMZvNpKam4na71VnpIqcyr62tJS0tjYSEBDZv3szWrVuZP38+JSUl5OfnU19fT2pqqo4idrGz87darSxbtowpU6bg5+dHYWEhtbW1VFZWkpOTo+y72IoVK9i8eTMnT54kKSmJmJgYbDYbmZmZLFmyhJCQEJKTk3G5XOqcd6Gzc09OTmblypUcPHiQ9957j/3797Nz506sViuxsbFGN7ffOTv/lpYWNm3axMSJEwkICPDt8xsbG8nIyDC6uf2K+jnGUF/HWOrr9E/9tihdu3Yta9euZcaMGaxcuZLa2lrCwsIIDAwEICIigsWLFzN58mQcDofBre0fzs68urqazMxM7HY7hw4d4uGHH2bmzJmYzWY2bNjA0KFDdZ5RFzo7/5qaGrKzs6muruazzz7jk08+wWQycd1111FcXMzw4cOxWgfEDP5u5/F4+Oijj/jyyy+ZMGECCxcuxOl0EhERgdPpBCAqKoq33nqLm266SQVpFzlf7na7nZycHBoaGigqKuLRRx9l6tSpVFdXs2vXLgYPHqx9fhe50Pd+0KBBHDx4kHXr1rF69WrsdjsjR46krq6OIUOGqIPYRdTPMYb6OsZSX6f/6rdF6fLly5k2bRpjxowhOTmZw4cPU1JSwrBhwwCIi4tjx44d1NbWkpKSwtatW0lKSjK41X3b2ZmXlJRw6NAhpk2bxsiRIwkPD8dkMmGz2dizZw+5ubn4+fkZ3ex+43z5l5WVMXfuXLKyskhKSmL27NlUV1ezf/9+Jk2aZHST+w2TycTSpUu57bbbGD16NLGxsRQWFtLS0kJKSgoACQkJ5OXlceLECYYMGUJ+fj4xMTEGt7xvO1/uu3btorGxkWuvvZaxY8cSFhaGzWajtbWV3bt3M3HiRHVQusjZ+cfExLB7926sViu33noriYmJJCcnc/PNN3P48GGKi4sZN26c0c3uN9TPMYb6OsZSX6f/6neH608tw52amspXX30FQHp6OpmZmVRXV1NYWOh77Lx583j//ff5yU9+Qm1trSHt7Q8ulHlWVhbHjh2jsLDQN1oE3qNcLS0tOnLYRb4r//LycgoLC4mMjGTEiBGAd4U6FUMdd/a5KafyT09P9+1fcnNziY+P9/2xPOWRRx7hvffe4wc/+IEuVXKZLjX3xMREDhw4QFlZWbv9Tn5+Pg6HA5vN1nON7kcuJf9Ro0YRHx/Pvn37KC8vJzU1ldGjRwPehUgGDx7cs43up9TPMYb6OsZSX6f/6/MjpYWFhbhcLt90lVPTgkwmE/v27SMgIMB3jkV5eTl2u52kpCTKy8t5/fXXSU5O5oknnvB9ieXiOpr5l19+yUsvvYTNZuPBBx/UjrqDOpp/QUEBv//972lpaeHuu+/WdK4OamlpaTfSdir/6upqKioqCAsLIyQkBIfDwa5du0hKSiI8PJzi4mJee+01MjIy+PnPf65VMS9TR3PfuHEjL730Eg6Hg3nz5mnEooM6mn9eXh4vv/wyHo+HW2+9FbvdbtRH6LNOnQ966rxE9XN6RkdzV1+na3Q0f/V1+q4+W5Tu37+fP/zhDyxevJgpU6YQFhYGnP4SOxwOqqqqKCwsJDc3l4CAAP72t79hMpkYMmQIbreb7OxsZsyYoU7KJepo5gBZWVm43W5GjhzJzJkz2x1NlEvT2fxNJhMjRoxg5syZ2kl3wJ49e3jzzTfZsWMHdrud6OhoTCYTbW1tmM1m7HY7xcXFHD9+nPT0dEJDQ1m1ahUOh4P09HRcLhc5OTnccMMN2udchs7m3tTUxKhRo5g1a5b2Ox3Q2fwBRowYwaxZs1SQXqY9e/awaNEijhw5QlJSki+/U9mrn9M9Opo7qK/TFTqbv/o6fVefm77rcrl45ZVXeOWVV5g5cyYjR46koKAAoN21iJxOJ0OHDsXlcvHOO+/gcrmor6/3jS4FBwcTFxdn2OfoSzqbeVBQEADJyclkZmYa9jn6qq7KPzIyUucTddDOnTt5/fXXmTBhAvHx8WzYsIGTJ0/idruxWCwAxMbGkp6eTnl5OcuWLQPAZrMRFRUFePNPTk427DP0RV2Re2ZmpkalO6gr8o+LiyMtLc2wz9BXFRQU8Prrr5OdnU1NTQ3Lli3jm2++AfBlr35O1+tM7urrdF5X5K++Tt/V50ZKW1paaGlp4f777ycpKQmXy8X+/fsZNWqU7wu7cOFCPv30U0aNGsWoUaPIy8tj+fLlREVFcfvtt2vlv8ukzI2l/I23detWrFYrt99+O1FRUXz99ddMmjTJN51x/vz5bNiwgcmTJ5OWlsbatWtZvnw5sbGx3Hjjjcq/g5S7sZS/cTZt2oSfnx+33HILGRkZbNq0iZKSEgYPHozT6WT+/PmsWrVK+/wuptyNpfwHtj6xBOGePXsIDAwkPj4eh8PBlClTfNtOjRRZLBY8Hg+HDh3iyJEj3Hvvvb7r0f3oRz+iublZ01cugzI3lvI31pn5AwwbNoxFixYRFhbGJ598QkJCAq+99hq5ublERERw9OhR5syZQ3R0NAA//elPaWtr07lEl0m5G0v5G+fs7OPj4ykuLqa6uprw8HCcTid1dXVs3bqVYcOGcfToUe3zu4ByN5bylzP16pHS+vp6nn/+eZYsWUJAQABpaWnYbDY8Ho/vxGc/Pz/effddrr32WhwOByEhIUycOJHAwEDfuXanluaWi1PmxlL+xjpf/larldDQULKzs9m+fTuzZs3ivvvuo66ujoKCArKysrjhhhva5W+1WnX+3GVQ7sZS/sa5UPYej4eDBw+yZs0aNm/eTENDAxkZGbS2tjJ27Fjt8ztJuRtL+cv59Oqi9OTJk7hcLsaPH+9byjwuLs73RXS73fj7+1NSUoLdbichIcE3dH/muXZy6ZS5sZS/sS6UP0B4eDirV69m3LhxhIeHExAQwFdffcW4cePw9/dX/p2g3I2l/I1zdvYej4e4uDiCgoLIyckhMjKSqKgo5s6dS0NDA1988YVv5oyy7zjlbizlL+fT6/5V161bR0FBAQ0NDYSHhzN9+nQmTZqEzWZj7969vmv7eTwezGYzra2tAL6js6eupaYv7KVT5sZS/sa61PxbW1sZMmQIK1euBGD79u2cPHnSd5RW+V8e5W4s5W+c78q+qKjIl73VaiU7O5vJkycD+NYSOEXZXx7lbizlLxfTK0ZKPR4PtbW1PPvssxw8eJCqqiq2bNnC0KFD8ff3x2q1YrFY2L9/Py6Xi5SUFN+okc1mY9OmTTQ3NzN8+HCd5HyJlLmxlL+xOpK/xWIhMDCQ/Px8/vrXv3L48GEeeugh3/l0cnHK3VjK3zgdyf6UwsJCfvvb31JXV8fs2bN1zu5lUO7GUv5yOQwvSk8Nw9fU1HDgwAGeeOIJRo0axa5du/j888+54oorAO8Sz6WlpVRVVZGWlubrnAOMGTOGnJwcIz9Gn6LMjaX8jdWR/FNTUwHvVMZT2c+aNYvg4GAjP0qfotyNpfyN09F9fltbG1arFbPZTGJiInPmzFHH/DIod2Mpf7lchhWlbW1tzJ8/n/z8fKxWK9XV1ZSWljJ+/HjMZjO5ubm8/fbbpKen+653lpyczNdff82SJUtYvHgxV155JX5+fhrKv0TK3FjK31idyX/p0qW+/AMDA33XAZSLU+7GUv7G6ew+/4MPPmDy5MlEREToeqOXQbkbS/lLRxnSsy0oKODJJ5+kvr6e2NhYFixYgNVqZefOnRQVFQFgMpm48847WbRoke95eXl5rFy5kpSUFJ577jnCw8ONaH6fpMyNpfyNpfyNodyNpfyN01XZR0REGPUR+iTlbizlL51hyEhpZWUliYmJ3H777aSlpbFv3z6sVisjR45kwYIFXHfddbjdbiIjIykoKCA9PZ2AgADKy8uZPn06M2fOxOl09nSz+zRlbizlbyzlbwzlbizlbxxlbwzlbizlL51hyEhpWloakyZNwu12AzBkyBAqKyuZOnUqbrebjz/+GLPZTFVVFWaz2begwrhx4xg2bJgRTe7zlLmxlL+xlL8xlLuxlL9xlL0xlLuxlL90hiFFqcPhwGaz+c6Ly8/P9y2c8Nhjj1FaWsozzzzDSy+9RFpaGnD6shfSMcrcWMrfWMrfGMrdWMrfOMreGMrdWMpfOsNq5JufOpJy/Phxxo4dC4Cfnx/33HMPJSUlREdH+85l0WUvuoYyN5byN5byN4ZyN5byN46yN4ZyN5byl44w/JIwbW1tbN++ncDAQN59912KiooYPXo0cXFx+Pn5Gdm0fkuZG0v5G0v5G0O5G0v5G0fZG0O5G0v5y+UydKTUZDJx4MABPv/8c44dO8Y111zDtddea2ST+j1lbizlbyzlbwzlbizlbxxlbwzlbizlLx1h8hg8mbuqqor169cze/ZsbDabkU0ZMJS5sZS/sZS/MZS7sZS/cZS9MZS7sZS/XC7Di1IREREREREZuAxZfVdEREREREQEVJSKiIiIiIiIgVSUioiIiIiIiGFUlIqIiIiIiIhhVJSKiIiIiIiIYVSUioiIiIiIiGFUlIqIiIiIiIhhVJSKiIiIiIiIYf4/lS6CgJ1FOmwAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 1152x432 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "returns = dbars['Close'].pct_change()\n",
    "long_short_returns = returns * series\n",
    "\n",
    "reg_pnl = (1 + returns).cumprod()[1:]\n",
    "strat_pnl = (1 + long_short_returns).cumprod()[1:]\n",
    "reg_pnl.plot()\n",
    "strat_pnl.plot();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'ann_information_ratio': 1.1711684339746886,\n",
       " 'ann_return': 0.20028988683961113,\n",
       " 'ann_sharpe': 1.2867282233743575,\n",
       " 'avg_return_hits': 0.0031285352891812397,\n",
       " 'avg_return_misses': -0.003134265651482499,\n",
       " 'dd95th': 0.06370056133354067,\n",
       " 'rHHINeg': 0.0007395923828325928,\n",
       " 'rHHIPos': 0.0006321610656659775,\n",
       " 'tHHI': 0.0027786614941559943,\n",
       " 'tuw95th': 0.12288905468436859}"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "get_stats(strat_pnl, 0.015)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 14.3.b\n",
    "\n",
    "Consider a strategy that is long one futures contract on even years, and is short one futures contract on odd years.\n",
    "\n",
    "What is the correlation to the underlying?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "For this short backtest the correlation is -0.15.\n"
     ]
    }
   ],
   "source": [
    "print(\"For this short backtest the correlation is {:.2f}.\".format(long_short_returns.corr(returns)))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 14.4\n",
    "\n",
    "The results from a 2-year backtest are that monthly returns have a mean of 3.6%, and a standard deviation of 0.079%.\n",
    "\n",
    "What is the SR? What is the annuallized SR?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The SR is 0.46\n",
      "The annualized SR is 1.58\n"
     ]
    }
   ],
   "source": [
    "monthly_returns = 0.036\n",
    "std = 0.079\n",
    "sharpe = monthly_returns / std\n",
    "ann_sharpe = np.sqrt(12) * sharpe\n",
    "print(\"The SR is {:.2f}\".format(sharpe))\n",
    "print(\"The annualized SR is {:.2f}\".format(ann_sharpe))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 14.5.a\n",
    "\n",
    "Following on exercise 1:\n",
    "\n",
    "The returns have a skewness of 0 and a kurtosis of 3. What is the PSR?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The PSR is 0.92.\n"
     ]
    }
   ],
   "source": [
    "from stats import psr\n",
    "print(\"The PSR is {:.2f}.\".format(psr(sharpe, 12, 0, 3)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 14.5.b\n",
    "\n",
    "Following on exercise 1:\n",
    "\n",
    "The returns have a skewness of -2.448 and a kurtosis of 10.164. What is the PSR?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The PSR is 0.83.\n"
     ]
    }
   ],
   "source": [
    "print(\"The PSR is {:.2f}.\".format(psr(sharpe, 12, -2.448, 10.164)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 14.6\n",
    "\n",
    "What would be the PSR from 2.b, if the backtest had been for a length of 3 years?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The PSR is 0.95.\n"
     ]
    }
   ],
   "source": [
    "print(\"The PSR is {:.2f}.\".format(psr(sharpe, 12*3, -2.448, 10.164)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 14.7.a\n",
    "\n",
    "A 5-year backtest has an annualized SR of 2.5, computed on daily returns. The skewness is -3 and the kurtosis is 10.\n",
    "\n",
    "What is the PSR?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The PSR is 1.00.\n"
     ]
    }
   ],
   "source": [
    "print(\"The PSR is {:.2f}.\".format(psr(2.5, 252*5, -3, 10)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 14.7.b\n",
    "\n",
    "A 5-year backtest has an annualized SR of 2.5, computed on daily returns. The skewness is -3 and the kurtosis is 10.\n",
    "\n",
    "In order to find that best result, 100 trials were conducted. The variance of the Sharpe ratios on those trials is 0.5. What is the DSR?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The DSR is 1.00.\n"
     ]
    }
   ],
   "source": [
    "from stats import dsr\n",
    "\n",
    "print(\"The DSR is {:.2f}.\".format(dsr(2.5, 0.5, 100, 252*5, -3, 10)))"
   ]
  }
 ],
 "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": "3.5.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}