Fin_Quant_2_Pairs_Trading_vf.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Modelo de Pairs Trading Baseado em Cointegração"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Autores: Alexandre Fernandes Theoharidis, Débora Darin, Nelly Pires Colnaghi e Rafael Couto**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Este trabalho contém:**\n",
    "\n",
    "**1)** Importação da série de preços das ações PETR3 e PETR4 e tratamento de dados;\n",
    "\n",
    "**2)** Cálculo da regressão linear dos preços das ações & computação dos resíduos da regressão (spreads) para verificação de desvios da média;\n",
    "\n",
    "**3)** Backtest da estratégia de pairs trading;\n",
    "\n",
    "**4)** Cálculo do retorno para o par PETR3 x PETR4."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Introdução"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Neste notebook, foi desenvolvido um modelo de pairs trading de ações baseado em cointegração, inspirado no artigo de Caldeira e Moura (2013). A estratégia tem como objetivo identificar, no curto prazo, desvios entre preços de pares de ativos que estejam relacionados no longo prazo. Assim, A cointegração se verifica quando os resíduos de um modelo de regressão entre duas séries são estacionários - dessa forma, dizemos que as séries são cointegradas.\n",
    "\n",
    "Assim, a primeira etapa da estratégia resume-se a encontrar um par de ações com spread estacionário. Matematicamente, sejam $A_t$ e $B_t$, $t = 1, 2,…, T$, a série de preços das ações A e B, respectivamente. Uma abordagem simples para verificar se essas ações são elegíveis consiste no seguinte modelo de regressão (1):\n",
    "\n",
    "\\begin{align}\n",
    "ln A_t = \\alpha + \\beta ln B_t + \\varepsilon_t\n",
    "\\end{align}\n",
    "\n",
    "onde $\\varepsilon_t$ é o spread do par, $\\beta$ é a proporção do ativo A em relação ao ativo B e $\\alpha$ é uma constante. Teoricamente, o spread deve ter média zero. Os log preços dos ativos são usados em vez dos seus preços de mercado, de modo a impor uma relação entre os retornos desses ativos, semelhante aos modelos de fatores aplicado aos retornos das ações, como os introduzidos por Carhart (1997) e Fama e French (1993), relacionando o retorno esperado da ação A ao retorno esperado da ação B, sendo E o operador de valor esperado (2):\n",
    "\n",
    "\\begin{align}\n",
    "E[ln(A_t/A_{t-1})] = \\beta E[ln(B_t/B_{t-1})]\n",
    "\\end{align}\n",
    "\n",
    "Esse resultado é uma condição necessária para a implementação de estratégias de negociação de pares usando um único par de ações, pois garante que os retornos contínuos esperados das ações A e B sejam proporcionais. Portanto, se o spread seguir um processo estocástico de reversão média, é possível explorar desvios de curto prazo em relação às equações (1) e (2) negociando essas ações por meio de posições compradas e vendidas.\n",
    "\n",
    "Com o objetivo de ilustrar o conceito por trás dessa estratégia, suponhamos que, em um determinado momento, o $\\beta$ e o spread sejam positivos e distantes de sua média. Por hipótese, o spread é estacionário e a equação (1) sugere que a reversão à média só pode ocorrer se a diferença $[ln A_t - \\beta(ln B_t)]$ diminuir, pois $\\alpha$ é uma constante. Assim, existem apenas cinco cenários nos quais isso ocorre: (i) $ln A_t$ declina e $ln B_t$ aumenta; (ii) $ln A_t$ declina, enquanto $ln B_t$ permanece inalterado; (iii) $ln B_t$ aumenta, enquanto  $ln A_t$ permanece inalterado; (iv) ambos os preços logarítmicos aumentam, mas $ln A_t$ cresce menos que $\\beta (ln B_t)$; (v) ambos os log preços caem, mas $\\beta (ln B_t)$ cai mais lentamente que em $ln A_t$. Assim, a estratégia de vender ativo A e comprar ativo B em uma proporção adequada pode produzir retornos positivos em qualquer desses cenários. Assim, para cada real apostado na depreciação da ação A, deve-se apostar $\\beta$ reais na valorização do ação B.\n",
    "\n",
    "Uma vez modelado o spread, a estratégia de negociação de pares pode ser implementada. O modelo mais usado com essa intenção é o de limite fixo (Fixed Threshold), uma abordagem não paramétrica que estabelece como orientação de negociação dois limites externos fixos equivalentes em módulo ao desvio padrão histórico do spread multiplicado por um fator predeterminado, geralmente definido igual a 2, mas com sinais opostos (Gatev et al. 2006). Devido à sua essência e simplicidade não paramétricas, a implementação do modelo é direta. De fato, analogamente ao exemplo discutido anteriormente, para $\\beta > 0$, sempre que o spread ultrapassa o limite superior (positivo), uma posição comprada na ação B e uma posição vendida na ação A são abertas, com pesos variando de acordo com o beta encontrado via equação (1), esperando a convergência do spread para seu valor médio. Segue-se naturalmente que essas posições devem ser invertidas quando o spread atingir o limiar oposto (negativo). Para $\\beta < 0$, as posições na ação B devem ser invertidas em cada caso, ou seja, a posição da ação deve ser reduzida quando o spread ultrapassar o limite superior, enquanto uma posição comprada deve ser aberta quando o spread atingir o limite inferior.\n",
    "\n",
    "Assim, nesse projeto, o cálculo dos spreads entre as ações do par foi realizado utilizando o método de filtro de Kalman e a estratégia implementada baseou-se em modelo com thresholds para identificação dos pontos de abertura e fechamento do trade, já citado anteriormente. No caso do filtro de Kalman, este foi configurado de tal forma que os coeficientes da regressão linear fossem variáveis no tempo e assumissem o comportamento de um passeio aleatório, conforme verificado abaixo em (3) e (4):\n",
    "\n",
    "\\begin{align}\n",
    "\\alpha_t = \\delta_1 + \\alpha_{t-1} + w_t \\\\\n",
    "\\beta_t = \\delta_2 + \\beta_{t-1} + z_t\n",
    "\\end{align}\n",
    "\n",
    "Onde $\\delta_1$ e $\\delta_2$ são parâmetros a serem estimados e $w_t$ e $z_t$ são ruídos brancos não correlacionados.\n",
    "\n",
    "Os dados utilizados no projeto compreendem as séries históricas de preços das ações ordinárias e preferenciais da Petrobras (PETR3 e PETR4). Foram coletados preços diários de janeiro de 2008 a março de 2020. A escolha dessas ações se justifica por serem ações da mesma empresa e, portanto, deveriam apresentar cointegração. \n",
    "\n",
    "Os parâmetros utilizados nesse trabalho consistem em uma definição das janelas de estimação dos parâmetros da regressão e de negociação do par. O tamanho da janela de estimação, definida no trabalho como janela_dados, foi calibrado de modo que houvesse um volume de dados razoável para ajuste do filtro de Kalman. Foi escolhida uma janela de 1 ano em dias úteis para essa análise, equivalente a 252 dias. \n",
    "\n",
    "Além disso, foi definida uma janela para determinar se o spread era estacionário ou não, chamadada de janela de cointegração (janela_coint). Nessa janela, a conclusão de cointegração para o dia em análise é encontrada somente se, em todos os dias da janela, o teste ADF indicar estacionariedade do spread. Para isso, foi escolhida empiricamente uma janela de 5 dias.\n",
    "\n",
    "Ademais, como premissa do trabalho, assumimos um custo de execução igual a zero e um nível de significância alfa igual a 1% (a ser utilizado no teste ADF para verificação da cointegração). Para a regra de trading, também foi definido um threshold de distância de 0.25 desvio-padrão da média para abertura de posição. Por fim, também foi definida uma variável para reduzir a frequência de ajustes de posição (delta_beta_lim), que determina um intervalo de variação para o beta dentro do qual não há ajuste das posições em relação àquelas quando o par foi aberto - para ilustração: um delta_beta_lim igual a 0.10 significaria que a posição não sofreria alteração se o beta abrisse para menos desse valor. Com esta medida, impedimos que haja variação para qualquer movimento do beta dentro deste intervalo, limitando-o.\n",
    "\n",
    "A figura abaixo ilustra a estratégia de pairs trading com thresholds. Ela mostra que as posições são abertas quando o spread se situa nas regiões inferior ou superior de fora da banda delimitada pelos threholds, definidos com um múltiplo do desvio-padrão do spread. Ademais, as posições são fechadas quando o spread cruza o eixo 0. \n",
    "\n",
    "Quando um par é aberto, o tamanho da posição não sofre ajuste até que haja convergência do spread para  a média, mesmo que o spread continue se desviando da média, tornando a estratégia teoricamente mais atrativa. O único ajuste implementado é um ajuste decorrente da variação do beta ao longo do tempo, que determina a proporção relativa das ações do par que devem ser compradas."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABjYAAAMQCAYAAACE7VwoAAEAAElEQVR4nOzde3iU9Z3//1cIpxAO4RBORkRqEAwiWtwWqIJrwWpxScVCt7ZWSbFke1UhtVsp7qpd2Wh/bcC2WxCb1NqqJYoEe1CJtkQFasEDaQFNKiIgpwSYBMgBCPP7I98JTO7PnO853s/HdXG1+czMPR8Yk8zcr/v9fqe53W63kFSKi4slSYsXL47zTgAAAAAAAAAAiK0u8d4AAAAAAACAnaqrq1VQUKB+/fqpsLBQf//73+O9JQAAYCOCDQAAAAAAkFIee+wxPffcc3K73Xr66ae1atWqeG8JAADYiGADAAAAAACkjKamJj333HNea88880ycdgMAAKKBYAMAAAAAAKSMAwcOWNZOnz6tEydOxGE3AAAgGgg2AAAAAABAyti/f79x3eVyxXYjAAAgagg2AAAAAABAyjBVbEhSQ0NDjHcCAACihWADAAAAAACkDF/BBhUbAACkDoINAAAAAACQMmhFBQBA6iPYAAAAAAAAKYNWVAAApD6CDQAAAAAAkDKo2AAAIPURbAAAAAAAgJTBjA0AAFIfwQYAAAAAAEgZtKICACD1EWwAAAAAAICUcOjQIZ05c8Z4GxUbAACkDoINAAAAAACQEnzN15AINgAASCUEGwAAAAAAICX4akMl0YoKAIBUQrABAAAAAABSAhUbAAA4A8EGAAAAAABICf4qNgg2AABIHQQbAAAAAAAgJdCKCgAAZyDYAAAAAAAAKcFfK6qTJ0/q1KlTMdwNAACIFoINAAAAAACQEvxVbEi0owIAIFUQbAAAAAAAgJQQKNigHRUAAKmBYAMAAAAAACS9EydOBKzIoGIDAIDUQLABAAAAAACSXqBqDYmKDQAAUgXBBgAAAAAASHrBBBtUbAAAkBoINgAAAAAAQNLbv39/wPsQbAAAkBoINgAAAAAAQNKjFRUAAM5BsAEAAAAAAJIeFRsAADgHwQYAAAAAAEh6zNgAAMA5CDYAAAAAAEDSoxUVAADOQbABAAAAAACSHq2oAABwDoINAAAAAACQ9GhFBQCAcxBsAAAAAACApHbgwAGdPXvWa61fv36W+9GKCgCA1ECwAQAAAAAAkpqpDdWwYcMsa1RsAACQGgg2AAAAAABAUjO1ocrJybGsEWwAAJAaCDYAAAAAAEBSMwUbw4cPV+/evb3W3G437agAAEgBBBsAAAAAACCpmVpRDR8+XFlZWZZ1qjYAAEh+BBsAAAAAACCpmSo2hg0bRrABAECKItgAAAAAAABJzVcrqn79+lnWaUUFAEDyI9gAAAAAAABJzdSKiooNAABSF8EGAAAAAABIarSiAgDAWQg2AAAAAABA0jp+/LgaGxu91nr06KHs7GxaUQEAkKIINgAAAAAAQNLy1YZKEhUbAACkKIINAAAAAACQtHy1oZLMwQYVGwAAJD+CDQAAAAAAkLRMwcbw4cMlydiKiooNAACSH8EGAAAAAABIWrSiAgDAeQg2AAAAAABA0qIVFQAAzkOwAQAAAAAAkpapYsPTioqKDQAAUhPBBgAAAAAASFr+KjaYsQEAQGoi2AAAAAAAAEnL3/BwWlEBAJCaCDYAAAAAAEBScrvdfoeH9+rVS927d/e6raWlRc3NzTHZHwAAiA6CDQAAAAAAkJRMocagQYPUo0ePjq+ZswEAQOoh2AAAAAAAAEnJ33wND9OcDdpRAQCQ3Ag2AAAAAABAUjJVbHjma3hQsQEAQOoh2AAAAAAAAEkpmIoNgg0AAFIPwQYAAAAAAEhKpmCjc8UGragAAEg9BBsAAAAAACApmVpRUbEBAEDqI9gAAAAAAABJiVZUAAA4E8EGAAAAAABISuEOD6cVFQAAyY1gAwAAAAAAJKVgKjZMMzao2AAAILkRbAAAAAAAgKTT0NCgEydOeK317NlTAwcO9FqjYgMAgNRDsAEAAAAAAJKOqVqjcxsqiRkbAACkIoINAAAAAACQdEzzNTq3oZJoRQUAQCoi2AAAAAAAAEknmPkaEq2oAABIRQQbAAAAAAAg6ZgqNmhFBQCAMxBsAAAAAACApBNsxYapFVVjY6Pa2tqisi8AABB9BBsAAAAAACDpBDs8PC0tzRhu0I4KAIDkRbABAAAAAACSTrDDwyXaUQEAkGoINgAAAAAAQNIJthWVRLABAECqIdgAAAAAAABJpa2tTQcPHrSsm1pRSeY5G7SiAgAgeRFsAAAAAACApGJqQzV48GB169bNeH8qNgAASC0EGwAAAAAAIKmE0oZKItgAACDVEGwAAAAAAICkYgo2fLWhkszBBq2oAABIXgQbAAAAAAAgqZhaUfmr2DDN2KBiAwCA5EWwAQAAAAAAkgqtqAAAcDaCDQAAAAAAkFRoRQUAgLMRbAAAAAAAgKRCKyoAAJyNYAMAAAAAACQVWlEBAOBsBBsAAAAAACCpmCo2aEUFAIBzEGwAAAAAAICkcezYMTU1NXmt9erVS/379/f5GFpRAQCQWgg2AAAAAABA0gi1DZVExQYAAKmGYAMAAAAAACQNU7Dhrw2VJPXo0UMZGRlea6dPn9aJEyds3RsAAIgNgg0AAAAAAJA0TPM1AlVsSAwQBwAglRBsSKqurta0adOUlpamtLQ0TZs2TZs3bw7pGC6XS2VlZV7HGTdunMrKynijBAAAAACATcJpRSWZ52zQjgoAgOTk+GCjsrJSV1xxhXbs2KGSkhKVlJRox44dmjx5siorK4M6hsvlUn5+vgoKCiRJpaWlKi0tlSQVFBQoPz8/WtsHAAAAAMBRTBUbgVpRSVRsAACQSrrGewPx5HK5dNtttyk7O1s1NTUdb3LuvPNOjR49WrfddpvXui+vvfaaqqqqtGTJEj388MMd6/PmzVNhYaFWrlypNWvWaPbs2VH82wAAAAAAkPrCrdgg2AAAIHU4umJjy5Ytqqur0+LFi73e4GRlZWnFihWqq6vTa6+9FvA47777riRpzpw5ltuKiookSeXl5fZsGgAAAAAABwtneLhkDjZoRQUAQHJydLBRVVUlSbr++ustt40fP16S9OqrrwY8zsMPPyy3293xmPNlZGREuEsAAAAAAOAR7vBw04wNKjYAAEhOjg42amtrJUkjRoyw3JabmytJOnr0aETPsX79eknSZz/72YiOAwAAAACA050+fVqHDx+2rNOKCgAAZ3F0sOHhb4ZGJC2kamtrdd9990mSvvzlL4d9HAAAAAAAYG5DNWTIEHXtGniEKK2oAABIHQQbUVJbW6spU6aorq5O69evV05OTry3BAAAAABAUjO1oQpmvoZEKyoAAFJJ4EsaELLOocb06dPDPlZxcbGNOwMAAAAAIHmZKjaCaUMl0YoKAIBUQsWG/L+RmTNnTkjHqqystC3UAAAAAOAsc+fOVVpamvHPtGnTVFZWFvUTsWlpaZo7d67f+5SVlSktLU3V1dVR3QvQmSnYCLZig1ZUAACkDkdXbHgGhO/Zs8fyBsfzBn3AgAFBH6+yslIzZsxQdna2tm3bpvHjx0e8x8WLF1vWqOIAAAAAUltpaall7amnnlJBQYFeeeUVrV69OmZ72bdvn1auXKk5c+bY8hkHiISpFVWwFRu0ogIAIHU4OtiYOnWqli5dqq1bt1reoNfW1kqSPv/5zwd1rPNDjY0bN3aEJgAAAAAQqnnz5hnX5s6dq/Lyci1cuFCTJk2KynO73W6vr9evX6+lS5d6VbPPmzfPuEcg2mhFBQAAJIe3orr66quVnZ2t++67ryPIkNrf2BQWFio7O1vXX399wONUV1d3hBo1NTWEGgAAAACiYuHChZKknTt3xncjQJxEMjycVlQAAKQORwcbWVlZevrpp1VXV6cpU6Zo2bJlWrZsmUaPHq26ujo9/fTTljc+np635/eSvfvuuyVJgwcP1re+9S3NnTvX8mfZsmWx/KsBAAAASEGZmZnG9bKyMo0bN65jHkdhYaHXxVu+7jd37lxt3rzZ6z7nz9iYO3euCgoKJElXXHGF0tLSOo5jmrERzD48j3W5XLr//vu9Zoh0Pp7nPoMHD1ZaWpoGDx7s8+9mUllZ6TW3xPT46upqpaWlafPmzR33nTZtWsfttbW1Kiws7DjGuHHjtGbNmqCeH/aLpGKjT58+Sk9P91o7efKkTp06ZcveAABA7Di6FZUkTZ8+XZs2bdLixYtVVFQkqb1FVXFxcdCl3VVVVZKk7du3a/v27VHbKwAAAABn27p1qyTvWQHTpk1TVVWV5syZo6KiIjU0NKi4uFgrV670mv1XVlamgoICy/3Ky8t9zgj85je/KUkqLy/XkiVLNGrUKJ97C3YfHvn5+ZLa54ns2rVLS5cu1RVXXKG9e/cqJydHkvStb33L67k991uzZo1qamqMV+B7eP6+U6dO7ZhZ8sorr2jlypV644039I9//MPr/rNmzdJ1113nNd+ktrZWU6ZMUV1dXcceXnnlFd16660qKSnRokWLfD4/oiOS4eFS+/fO0aNHvdYaGhqUnZ0d8d4AAEDsOD7YkKRJkyZpw4YNQd139erVlkF9nXvQAgAAAICdXC6XtmzZovvuu8+rZW5lZaWqqqq0ZMkSPfzwwx33nzlzpqZMmaKvfvWrHSfwS0pKNHXqVK/PMzNnztTo0aN16NAh4/NOnz5de/fuVXl5ud/h4aHsw2Ps2LFasWJFx9dXXnmlbr31Vr311lvKycnRvn37VF5ebgkQrrzyShUWFqqurs5vsPHKK68oLy/P67PevHnzlJubq6VLl6q6utrr73PZZZdZPuvdf//9qqur82o5PG/ePA0YMEBFRUWaOXMmrYhj6OjRo2pubvZa6927t3EouC9ZWVmWYMPlchFsAACQZBzdigoAAAAAEpGn7ZHnT//+/TVjxgxJ0saNGztO6P/yl7+UJN17771ej8/NzdXixYu1ffv2jvZOeXl52rFjhyorK7Vv376O+7ndbk2fPj2i/YayD4/CwkLLfaVzMw88VRvr1q3T5s2bO4Y8z549W4cPHw4YKKxevdoSpkjyWXVy++23e33tCVYWLFhgea4lS5ZIkv7whz/43QPsZZqvEWwbKg8GiAMAkBqo2AAAAACABHN+OySPsWPHGtvlTp061Xiydty4cV5fP/zww5oyZUpHQJKXl6eCggLbqg6C3Uco1q9frxkzZmjy5Mkdz/Gd73xHn/nMZzqCj0Bqa2tVX1+vnTt3asuWLUHPx/Bc1X/06FGVlZUZ7/PXv/41qGM52bZt2/TjH/9Yr776qj73uc9p1apV6t+/f1jHimS+hgcDxAEASA0EGwAAAACQYObNm2f7MXNzc1VTU6MtW7bohRde0Jo1a1RUVKSioiKVlpZG5Tkj5WmF9dZbb6m8vFzl5eUdMw59zQXxqKys1G233aa6ujpJUnZ2tq677jpdd911Ki8vD3oPnudFeH70ox/p97//vaT29mDf/va39cwzz4R1LFPFRijzNSQZ21ZRsQEAQPKhFRUAAAAAJLGqqirjiVlTG6asrCxNnz5dK1as0OHDh1VTU9NRuRHLfYQiJydHs2fP1urVq+V2u7Vp0yZJ0tKlS30+Zt++fZoxY4Yuu+wybdu2TW63W4cPH9bq1at1ww03BPW8GRkZktqrZ9xut/FP55kc8Hbs2LGOUMPjD3/4gz788MOwjhetig2CDQAAkg/BBgAAAAAkqW9+85uSpB//+Mde67W1tSouLlZeXp7Gjx8vl8uladOmGedaDBo0KOLBycHuIxTV1dUaN26cpXXU2LFjJUkDBgzw+VhPG6lZs2Z5Pa/L5VJJSUlQz5+bm6u8vDzdd999lhPflZWVSktL89miCu18hVoVFRVhHc8UbIRasUErKgAAUgOtqAAAAAAgSU2fPl1Tp07V0qVLVVtbqxtuuEENDQ0qLi5WXV2d1q1bJ6n9ZO7YsWO1cuVK7dy5s2NQ9lNPPaWqqirjTA8PT5CwYsUKXX311caWVcHuIxQjRoyQJN16661asGCBrr76ajU0NHTstXNIc74BAwYoOztbRUVFOnLkiEaNGqVdu3Zp1apVIe1h7dq1mjJlikaPHq277rqr4zhLly5VdnZ2x7wSmG3fvt24XlFRoe9+97shH8+O4eG0ogIAIDVQsQEAAAAASWzDhg0qLS3V9u3bVVBQoKKiIs2ePVs1NTVew8aLi4tVWlqq+vp6FRQUdLSfWr9+vd/5GmPHjtWcOXO0cuVKvy2rgt1HsLKysvTyyy9ryZIlWrNmTccxr7nmmoDzNXJycrRx48aOsKWgoEBvvvmmHnnkEW3cuFGS9NprrwXcQ25urjZu3KjZs2dr1apVKigo0KpVq7RkyRK98847QQ8wdypfFRvbtm1TdXV1yMejFRUAAPBIc7vd7nhvAqEpLi6WJC1evDjOOwEAAAAAwOy6667T22+/bbzt3nvv1X//93+HdLxPfepTHcPgPXbu3KkLLrgg6GO88MILuuOOO7zW8vPz9dRTT4W0F1+2bdum5cuX64UXXtCcOXP0n//5n8rNzbXl2AAA4BwqNgAAAAAAgO38DY4Pdc5Ga2urJdSQEq9iY/ny5VqzZk3HcPnHHnvMtmMDAIBzCDYAAAAAAICt3n//fbW2tvq8/Z///Kc2b94c9PF8taHq0iW00xrRnLFx8uRJy7D7p59+2pZjAwAAbwQbAAAAAADAVv6qNTxCqdowBRvDhw8PZUuSzBUbDQ0NIR/HZO/evZa1trY27du3z5bjAwCAcwg2AAAAAACArbZv3x7wPqEEG/v377eshdqGSopuKypfAQbBBgAA9iPYAAAAAAAAtgqmYuPAgQN67bXXgjqer1ZUoYpmKypTxYa/dQAAED6CDQAAAAAAYCtTxca0adMsa8FWbdjViqpr167q3bu315rb7balHRUVGwAAxA7BBgAAAAAAsE19fb3xZP7Xv/51y1qwwYZdraik6LWj8lWZQbABAID9CDYAAAAAADFXVlamtLQ0VVdXx3srsJmpDdW4ceN0yy23KDMz02u9oaFBL774YsBj2tWKSoreAHFfAQatqAAAsB/BBgAAAAAAsI2pDVVeXp7S09M1a9Ysy23BVG2YKjbCaUUlRW/OBhUbAADEDsEGAAAAAACwja+KDUnKz8+33FZRUaHm5ma/x4x2xYYdwQYzNgAAiB2CDQAAAAAAYBtTxYYn2PjCF76g7Oxsr9vOnDnjt2qjvr5era2tXmt9+vRR3759w9pfNFpR7du3T21tbcbbXC6XLa2uAADAOQQbAAAAAJBANm/erGnTpiktLU1paWmaNm2aNm/e7HUfz3wKl8ulwsLCjvsWFhYarw6vra31ut+4ceO0Zs0ay/3S0tJUVlamzZs3a9y4cUpLS9PgwYO1bNkyyxXtLpdL999/vwYPHuz3fp3/Tp77caI3dZkqNvLy8jr+v6+qDV/sbEMlRacVVaA5GlRtAABgL4INAAAAAEgQy5Yt0+TJk1VfX6/S0lKVlpaqvr5ekydPVllZmeX++fn5euONN1RaWqolS5Zo5cqVuuqqq7xOotbW1mrKlClauXKllixZotLSUuXl5enWW2/VsmXLLMd86qmnNGvWLOXn56u0tFSDBw9WUVGRfvzjH3fcx+VyKT8/X0uXLtV1112n0tJS3XXXXSoqKrKctN68ebMmT56sHTt2qKSkRIsXL1ZxcbGKiors+4dDwti+fbvOnDnjtTZ06FANHTq042tTsPHSSy/pyJEjxmPa2YZKik4rqkDBBcEGAAD26hrvDQAAAAAA2k98FhUVaerUqaqoqOg4+XrLLbcoPz9fBQUFmjFjhnJycjoeM2TIEK/7Tp06VTNmzNDKlSv18MMPS5Luv/9+1dXVqaamRrm5uZKkefPmacCAASoqKtLMmTM71iWpqqrK67633HKLPve5z2nVqlUdx3zhhRdUVVWl559/XrNnz+547JVXXqlbb71Va9as6VifP3++srOzVVNT07HPO++8U6NHj1ZdXZ39/5CIq0DVGpJ0zTXXaOTIkdq9e7fXekVFhQoKCiyPj0WwEWkFUaCKjUC3AwCA0FCxAQAAAAAJYP369ZKk4uJirxOvWVlZKi4u9rqPx8KFC73uO336dE2dOlWrVq2S1B6WlJeXa8GCBV7hhSQtWbJEkvSHP/zBa33OnDle983KylJeXp5XCFFSUqLs7GyvUEOSZs+erby8PP3sZz+T1F4tsn37di1evNjyd1q8eHHAfxMkH3/zNc4XSjsqu1tRUbEBAEDyo2IDAAAAABLI2LFjLWsXXnih8b6TJk2yrA0ZMkRVVVWSpKNHj3b8r6mVlST99a9/DXmP27dvV15ens9jep6/ublZkvnE9mc/+9mQnxeJL5iKDUmaNWuWli9f7rVWVVWlPXv2aMSIEV7rdldsMGMDAIDkR7ABAAAAACmuvLxc5eXlth5z+/btxrZBwcrMzLRxN0gUwVZsfPrTn1ZeXp7l/hUVFbr77ru91kzBht0VG5G2ogoUXNCKCgAAe9GKCgAAAAASyM6dOy1rvk6K1tbWWtY81RSSlJGRIUkqLS2V2+02/lm9enXIe8zOztacOXN8HtPtdns9v+kq/q1bt4b8vEhshw4dsoQQ6enpxooNqb1qozNTOypTK6pEGx5OxQYAALFFsAEAAAAACWDGjBmSpMWLF3udZHW5XB3zKDz38SgpKfH6urKyUtu3b1dRUZEkKTc3V3l5ebrvvvssJ24rKyuVlpbms52UP3fddZfKy8tVXV3ttV5bW6vBgwdr7ty5Xs9fXFzsFcK4XC7L3pH8fFVrpKWlGe9vmrOxdetWS7iX6K2ojh07psbGRr/32bdvn86ePRv2cwAAAG+0ogIAAACABJCTk6OSkhIVFRXpc5/7XEc4UVJSou3bt6ukpEQ5OTlej1m5cqWOHj2qG264Qbt27dLSpUs1depU3XLLLR33Wbt2raZMmaLRo0frrrvu0qhRozrum52dbQlLgnHvvffqzTff1BVXXKEFCxbo6quvVkNDg4qLi1VXV6eFCxd23PeZZ57RFVdcoSlTpnQENJ5h6Egtwc7X8BgzZoyuvvpqbdmyxWu9oqKiY9ZMS0uLjhw54nV7WlpaQrWiMlVjpKenq62treNrt9utffv2WeaHAACA8FCxAQAAAAAJYtGiRdq0aZMGDRqkgoICFRQUaNCgQdq0aZMWLVpkuf/69eslSQUFBVq1apWWLFmiiooKrxO3ubm52rhxo2bPnq1Vq1Z53fedd96xhCXByMrKUkVFhZYsWaI33nhDBQUFKioq0uzZs7Vt2zavoebjx49XTU2NZs+eraKiIhUVFemuu+7SunXrQv8HQkIzBRum+RrnM1VtnN+OytSGKpJQQ5J69eql7t27e621tLR0DLsPlakNlWmGDO2oAACwT5rb0/wUScNzdZPnaicAAAAAzlJWVqaCggJt27ZN48ePj/d2AEnSlClT9Pe//91rbd26dbruuut8Pmbv3r3Gqo6//OUv+vSnP62NGzfqxhtv9Lpt4sSJ+vOf/xzRXi+55BIdPnzYa62mpkZDhw4N+VirVq3Svffe67U2YsQI7dmzx2vtiSee6GjTBgAAIkPFBgAAAAAAiEhbW1tYFRsXXnihpk2bZln3VG2Y5mtEWrEh2Ttnw1SxYQpIqNgAAMA+BBsAAAAAACAi27dvV+eGEBdccIGys7MDPnbWrFmWNU+wYWpFFcngcA/TnI1wgw1TYGGapUGwAQCAfQg2AAAAAABAREIdHH4+05yNjz/+WK+//rqxYiNawUa4A8RNFRu5ubmWNYINAADsQ7ABAAAAAElm3rx5crvdzNdAwti+fbtlLVAbKo+BAwfqi1/8omW9oqIiKsPDpei3ojL93U33AwAA4SHYAAAAAAAAEYmkYkPy3Y4qlhUb4QQbp0+fNu7xqquusqxRsQEAgH0INgAAAAAAQEQiqdiQ2ttRde/e3Wutvr5eH374oeW+dlRs2BVsmKowhg8frgsuuEA9e/b0Wm9sbNSxY8dCfg4AAGBFsAEAAAAAAMK2f/9+HT582Gute/fuGjt2bNDH6Nmzp7Fq48iRI5a1RJqxYQo2cnJyJEkXXnih5TaqNgAAsAfBBgAAAAAACFuk1RoepiHiZ8+e9fq6X79+6t27d8jH7syuGRumoMITaHgCjkD3BwAAoSPYAAAAAAAAYYt0vobHzTffbKykOJ8d1RpSdFtR+avYYIA4AAD2INgAAAAAAABhs6tiQzJXbZzPjvkakn2tqKjYAAAgPgg2AAAAAABA2Oyq2JBknLNxPrsqNuxqReWvYoNgAwCA6CHYAAAAAAAAYTl16pR27NhhWQ+3YuP666/XBRdc4PP2ZK/YoBUVAAD2INgAAAAAAABhMbWhuvDCCzVgwICwj+mvaiPZZ2xQsQEAgD0INgAAAAAAQFhMbajCrdbw8DdnI5rBRmNjo86ePRv0MQ4dOqSWlhavtb59+6p///6SzBUbn3zyic6cORPaZgEAgAXBBgAAAAAACIupYiPc+Roen/3sZzV69GjjbXa1opIin7Phrw2VJPXo0UNDhw4N6nEAACA0BBsAAAAAgLjbsGGDKioq4r0NhCgaFRuS73ZUdlVsSJG3o/LXhsrX1xLBBgAAdiDYAAAAAADE3UMPPaQ777wz3ttAiEwVG3YEG6Z2VOnp6cYKiHBFOkA8UMWGRLABAEC0EGwAAAAAAOJq+fLl2rBhg1wulx588MF4bwdB2rt3r44cOeK1lpGR4bONVCguv/xyTZ8+3Wtt7ty5ER/3fJG2ogqmYsM0QNz0OAAAEBqCDQAAAABA3LhcLj300EMdXz/00EPavXt3/DaEoEWrWsNjxYoVmj59urKysvSVr3xFCxYssO3YUuStqKjYAAAgfgg2HGju3LlKS0sz/pk2bZrKyspCejMXjrS0tIBX21RXVystLU1lZWVR3QsAAACA+Fm+fLnl88eiRYvisxmExDRfI9LB4ecbPHiw1qxZoz179mjVqlWaMGGCbceWIm9FxYwNAADip2u8N4D4KS0ttaw99dRTKigo0CuvvKLVq1fHbC8ul0svvPCC+vXrp9mzZ8fseQEAAADEz+7du/XYY49Z1isqKrRhwwZNmzYt9ptC0KJdsRFtsajYoBUVAADRQbDhYPPmzTOuzZ07V+Xl5Vq4cKEmTZoUled2u91eX+/Zs0cFBQVeYcv48eMt9wMAAACQOhYtWiSXy6WFCxfqySef7Pj/y5cv15133qmPPvoo3luEH9Gu2Ii2SGZsnDx5UvX19V5r6enpVGwAABAjtKKCxcKFCyVJO3fujO9GAAAAAKSsiooKVVRUKCsrSw888EDH+gMPPKCRI0dq9+7devLJJ+O3QfjV3NysDz74wLLulIqNYNpQSdKgQYPUq1cvr7UTJ05Yhq4DAIDQEGzAIjMz07heVlamcePGdczjKCwsVG1tbcD7zZ07V5s3b/a6z/kzNsrKynTFFVdIkgoKCpSWlqbq6mqfMzY2b96sadOmec0F6Xx8z2Orq6tVVlamwYMHKy0tTePGjdOaNWu87utyuXT//fd33Gfw4ME+/24mlZWVXnNLfD3e83dZtmxZx14896mtrVVhYWHHMUz7BAAAAFLF+QPD77nnHssJ5mXLlklqr+hgkHhiMrWhGjlypLEKIlFFMmMjmDZU/tap2gAAIDIEG7DYunWrJO+y3GnTpqmgoEB5eXkqLS1VSUmJ1qxZo9GjR6u6urrjfmVlZZb7/eUvf9HkyZO97ne+iRMnasmSJZKkOXPmqLS0VAMGDDDed9myZZo8ebLq6+tVWlqq0tJS1dfXa/LkycYh43fffbdKSkq0ePFilZSU6PDhw7r11ltVWVnZcZ9vfetbWrp0qe666y6Vlpbqrrvu0sqVKzVlypSAV+uUlZVpxowZOnToUMd+rrvuOq1cuVJf+tKXLPe/7777Ou6Xn5+v3Nxc1dbWasqUKVq5cqWWLFmi0tJS5eXl6dZbb+34QAcAAACkkuXLl+u9997TyJEjOyrGz5efn6/8/HyvAASJxdSGKpmqNaTIWlEFW7Hha51gAwCAyDBjAx1cLpe2bNmi++67T9nZ2br++usltVckVFVVacmSJXr44Yc77j9z5kxNmTJFX/3qVzve1JaUlGjq1Kleg8dnzpyp0aNH69ChQ8bnHT9+vCRp6dKluuGGGzpmfxw9etTrfvv27VNRUZGmTp3aUbIuSbfccovy8/NVUFCgGTNmeL1prK+v15tvvtlxX89eXnjhBU2fPl379u1TeXm5SkpKtGjRoo7HXXnllSosLFRdXZ3xKh6PV155RXl5edqwYUPH2rx585Sbm6ulS5equrq64+8nSXV1daqpqfE65v3339+xnpub23GMAQMGqKioSDNnzuxYBwAAAJLd+QPDf/WrX/l8v71s2TJVVFToySef1D333KMJEybEbpMIyFSxkUzzNaTIWlGFUrFhCjYYIA4AQGSo2HAwT9sjz5/+/ftrxowZkqSNGzd2vMn75S9/KUm69957vR6fm5urxYsXa/v27R3VGHl5edqxY4cqKys73ujl5ubK7XZr+vTpEe13/fr1kqTi4mKvN6BZWVkqLi72uo9HUVGR1309AYEnNPG8wVy3bp02b97c8SZ29uzZOnz4cMBAYfXq1cYrlUaNGmW8/5w5c7z24wlWFixYYHkuTxXLH/7wB797AAAAAJLJQw89JJfLpfz8fE2bNs3n/UaOHKk77rhDkrwuQkJiSIWKjUhaUYVSsUErKgAA7EfFhoOVlpZa1saOHatJkyZZ1qdOnWp809f5jevDDz+sKVOmdAQkeXl5KigosLXqYOzYsZY1X1fGBGP9+vWaMWOGJk+eLKn97/qd73xHn/nMZ3y+Me2strZW9fX12rlzp7Zs2RL0fAxPwHL06FFjKy1J+utf/xrUsQAAAIBEt2HDBj355JOWgeG+LFu2TBs2bNCGDRtUUVGh/Pz86G8SQaFiI7KKDYINAAAiQ7DhYJ6WT3bKzc1VTU2NtmzZohdeeEFr1qxRUVGRioqKVFpaGpXnjNT06dO1d+9evfXWWyovL1d5ebmqqqokSdu2bfNqJdVZZWWlbrvtNtXV1UmSsrOzdd111+m6665TeXl50HvwPC8AAACQys4fGB5MaylPAHLnnXfqzjvvJNhIELt377YEAL1799Yll1wSnw2FiRkbAAAkL1pRIShVVVXGN3im8uOsrCxNnz5dK1as0OHDh1VTU9NRuWGHnTt3WtYi7U+ak5Oj2bNna/Xq1XK73dq0aZOk9rkfvuzbt08zZszQZZddpm3btsntduvw4cNavXq1brjhhqCeNyMjQ1J79Yzb7Tb+OX9eCQAAAJCsPJUXvgaG+3LHHXdowoQJcrlcWr58edT2h+CZqjWSrQ2VJPXo0aPjM5nHmTNndOLEiYCPDaViw7TOjA0AACJDsIGAvvnNb0qSfvzjH3ut19bWqri4WHl5eRo/frxcLpemTZumwsJCr/vl5uZq0KBBys7OjmgfnvZWixcv9gpZXC6XFi9e7HWfYFVXV2vcuHGW1lGedlcDBgzw+VhPG6lZs2Z5VXW4XC6VlJQE9fy5ubnKy8vTfffdZwmOKisrlZaW5rNFFQAAAJBMJkyYoGnTpmnZsmU+B4b7snbtWuXn5zNAPEGYLnBLtjZUHuHM2di3b5/a2tq81gYNGqTMzEzj/U0VGwcOHNDp06eD3ygAAPBCKyoENH36dE2dOlVLly5VbW2tbrjhBjU0NKi4uFh1dXVat26dpPY3hGPHjtXKlSu1c+dO3X777ZKkp556SlVVVcaZHh4jRozouK9kDihycnJUUlKioqIife5zn1NRUZEkqaSkRNu3b1dJSUnQMzE6P++tt96qBQsW6Oqrr1ZDQ0PHXjuHNOcbMGCAsrOzVVRUpCNHjmjUqFHatWuXVq1aFdIe1q5dqylTpmj06NG66667Oo6zdOlSZWdnhxzWAAAAAIkoKytLf/nLX8J67MiRI7V27Vqbd4RwpUrFhtTejurAgQNeay6XSxdccIHPx4TShkqSunXrpmHDhlmeZ+/evRo1alSIOwYAABIVGwjShg0bVFpaqu3bt6ugoEBFRUWaPXu2ampqvIaNFxcXq7S0VPX19SooKOhoP7V+/Xq/8zWysrJUUlKiqqoqFRQUdFRDdLZo0SJt2rRJgwYN6jj+oEGDtGnTJi1atCjkv1dWVpZefvllLVmyRGvWrOn4u11zzTUB52vk5ORo48aNHaFPQUGB3nzzTT3yyCPauHGjJOm1114LuIfc3Fxt3LhRs2fP1qpVq1RQUKBVq1ZpyZIleuedd0IOawAAAAAgmlK9YiPQnI1Q2lD5u505GwAAhC/N7Xa7470JhKa4uFiSOtovAQAAAECy69+/v1wul44dOxZyqyrEzsmTJzVs2DDL+v79+9W7d+847Cgyc+bM0csvv+y19rvf/U433XSTz8eUlJTowQcf9ForLCzUo48+6vMxd9xxh1544QWvtZUrV+qrX/1q6JsGAABUbAAAAAAAgOCYqjU+9alPJWWoIcW3YoMB4gAAhI9gAwAAAAAABCWV5mtI7TM2OgsUbIQ6Y8PX7bSiAgAgfAQbAAAAAAAgKKk0X0OKXcUGwQYAAPYi2AAAAAAAAEExBRvJXLFhCjYaGhr8Psauig1aUQEAED6CDQAAAAAAEBRTK6pkrtgItRXVsWPH1NjY6LXWs2dPDRkyxO/zmCo6qNgAACB8BBsAAAAAACCgDz/8UMePH/da69evny6++OI47ShyobaiCqcNlSQNGDDAMmC9qalJdXV1AR8LAACsCDYAAAAAAEBAqVatIYXeisrUPiqYYENizgYAAHYi2IBRdXW10tLSVFZWFu+t+JWWlqa5c+fGexsAAAAAkPJSbb6GZE/FRqD5Gh60owIAwD4EGwAAAAAAICBTxUayBxuhztigYgMAgMRAsAEAAAAAAAIyVWw4rRWV3RUbpqAEAAAERrABAAAAAAD8amxs1EcffWRZT/aKjT59+ig9Pd1r7eTJkzp9+rTx/lRsAACQGAg2HGrz5s2aNm2a0tLSlJaWpmnTpmnz5s22Pa6srEzjxo3ruN/cuXON96utrVVhYWHH/caNG6c1a9ZY7udyuXT//fdr8ODBHc9bXV0d3l8eAAAAABASU7XG6NGjlZGREYfd2CuUdlSmYCPYig2CDQAA7EOw4UBlZWWaPHmy6uvrVVJSotLSUtXX12vy5Ml+w41ly5Z1PK60tNTrcecPGS8rK1NBQYHy8vJUWlqqkpIS/eUvf9HkyZO9woja2lpNmTJFK1eu1JIlS1RaWqq8vDzdeuutWrZsmddz5+fna+nSpbruuutUWlqqIUOG6IorrrD/HwcAAAAAYJGK8zU8gh0gfvr0aR04cMCyHmzFBq2oAACwT9d4bwCx5XK5VFBQoKlTp6qioqLjDdwtt9yiz33uc/rjH/+oSZMmWR63b98+FRUVGR+Xn5+vgoICzZgxQzk5OSopKdHUqVO1evXqjsfPnDlTo0eP1qFDhzrW7r//ftXV1ammpka5ubmSpHnz5mnAgAEqKirSzJkzlZubq8rKSlVVVam0tFTz5s3ruN8NN9yggoKCKP1LAQAAAAA8UnG+hkewczZMIcTw4cPVrVu3oJ7HVLFx6NAhtba2qkePHkEdAwAAtKNiw2G2bNkiSVqyZInXm7esrCz94x//0MMPP2x83Pr16yVJxcXFlscVFxd73ScvL087duxQZWVlR1ltbm6u3G63pk+fLqk9KCkvL9eCBQs6Qg2PJUuWSJL+8Ic/SJJeeOEFSe0hyvk6fw0AAAAAiA4qNiJrQyVJ6enpuuCCC4I6LgAA8I+KDYfxvGEaMmRIWI8fO3asZa1zOe3DDz+sKVOmaMaMGZLag46CgoKOCgxJOnr0aMf/nt/G6nx//etfO+6Tl5dnebOZlZWlqVOnhvX3AAAAAAAEL5UrNoKdsWGahxFsG6rz7//JJ59YjnvJJZeEdBwAAJyOYAO2y83NVU1NjbZs2aIXXnhBa9asUVFRkYqKirzaSUlSeXm5ysvLw36ucAMaAAAAAEBwampq1NTU5LU2YMAAjRgxIk47slckrahCqdjwdX8GiAMAEDpaUTmM52qS82ddeBQWFgYsJd65c6dlzfTmLisrS9OnT9eKFSt0+PBh1dTUdFRuSFJGRoYkqbS0VG632/jHM6NjwIAB2r59u/GKmUhCEQAAAABAYKY2VKlSrSEF34rKjooNU7BBKyoAAEJHsOEwV199tSRp6dKlXm/UamtrtWbNGl1zzTXGx3naSi1evNjrcS6XS4sXL+64j8vl0rRp01RYWOj1+NzcXA0aNEjZ2dkdX+fl5em+++6zvGGsrKxUWlpaR4sqzyyNH//4x173W7NmTQh/cwAAAABAOExtqFJlvoYUfCsqOyo2TEEIFRsAAISOVlQOk5WVpfXr12vGjBn63Oc+p6KiIjU0NHQMAC8qKjI+LicnRyUlJSoqKup4nCSVlJRo+/btKikp6XhDN3bsWK1cuVI7d+7U7bffLkl66qmnVFVVpdLS0o5jrl27VlOmTNHo0aN11113adSoUdq1a5eWLl2q7OzsjjBl+vTpmjNnjpYuXara2lrdcMMNeuWVV1ReXt4RlAAAAAAAooOKjXbRqtgg2AAAIHRUbDjQ9OnTtWnTJg0aNEgFBQUqKirS7NmztXHjxo7h3iaLFi3yelxBQYEGDRqkTZs2adGiRR33Ky4uVmlpqerr6zvuJ0nr16/3mq+Rm5urjRs3avbs2Vq1apUKCgq0atUqLVmyRO+8847XG77Vq1ertLRU27dvV0FBgbZv365Nmzbpuuuui8K/EAAAAADAI9UrNpixER/vvvuu5s2bp759++pb3/qWsfU1AAC+pLndbne8N4HQeKorPC2gAAAAACDZ9e/fXy6XS8eOHTOeaEZ8NDc3a8iQIZb1uro69ejRIw47st+f//xn5efne61NmzZNL774YsfXhw4dslwI2Ldv35BDCZfLZRm63rNnTx0+fDi0TaeAefPm6fnnn+/4+hvf+IZ+9rOfxXFHAIBkQsUGAAAAAAAwqqurs6wNGzYsZUINKbgZG3a0oZLaq0P69u3rtdbS0uK4YKOpqckr1JCkp59+Ok67AQAkI4INAAAAAABgZDrhPnjw4DjsJHqCaUVlRxsqf48zHT+VHTp0yLJ25swZ42wTAABMCDYAAAAAAICRU4ONaFVs+Hqc0+ZsmIINSaqvr4/xTgAAyYpgAwAAAAAAGJlaUWVnZ8dhJ9ETTCuqaFdsEGy0I9gAAASLYAMAAAAAABg5oWKja9eu6t27t9ea2+32akdlZ8UGragINgAAkSPYAAAAAAAARk4INqTAczbsrNigFRXBBgAgcgQbAAAAAADAyAmtqKTAczaiXbFBsNGOYAMAECyCDQAAAAAAYOSUig1/czZOnjxpOeGenp5u64wNp7WiOnjwoHGdYAMAECyCDQAAAAAAYOSUYMNfKyo721BJ5kqPuro6NTc3h33MZEPFBgAgUgQbAAAAAADAiFZU9rahkqS0tDTj451UtUGwAQCIFMEGAAAAAACwOHXqlI4dO+a1lpaWlpIVG/5aUdldseHr8U6as+Er2Dhy5EiMdwIASFYEG0hYGzZsUEVFRby3AQAAAACOZKrWSMVQQ4ptxYavxzsl2Dh06JDcbrfxNio2AADBIthAwnrooYd05513xnsbAAAAAOBIpvkaqdiGSortjA1fj3dKsOFrcLhEsAEACB7BBhLS8uXLtWHDBrlcLj344IPx3g4AAAAAOI5TBodLiVGx4ZQZG77aUElSa2urGhsbY7gbAECyIthAwnG5XHrsscc6vn7ooYe0e/fu+G0IAAAAABzISa2omLERO/6CDYmqDQBAcAg2kHCWL19uCTIWLVoUn80AAAAAgEPRiqq9FVU0KjYINnwj2AAABINgAwll9+7dXtUaHhUVFdqwYUPsNwQAAAAADkUrKpf27duntrY2r/VBgwYpMzMzouejFZVvR44cidFOAADJjGADCWXRokVyuVxauHBhxxvLBx54QJIYJA4AAAAAMWRqRZWqFRu+WlFFow2VJPXt29cSppw6dcrvYO1UQcUGAMAOBBtIGBUVFaqoqFBWVlZHmCFJCxcu1MiRI7V79249+eST8dsgAAAAADiI0ys2GhoaotKGysMUkDihaoNgAwBgB4INJASXy6WHHnpIknTPPfdY3lQuW7ZMUntFB4PEAQAAACD6nBRs9OrVS927d/daa2lp0UcffWS5rx0VG76O44Q5G4GqUgg2AADBINhAQnjyySf13nvvaeTIkVq4cKHl9vz8fOXn53sFIAAAAACA6HFSKyrJXLWxa9cuy5pdFRum4zgh2KBiAwBgB4INxN35YcWvfvUr45tJ6VzVhicEAQAAAABER1tbm/EEc6pWbEjmORsff/yxZS2aFRup3oqqsbFRTU1Nfu9DsAEACAbBBuLuoYceksvlUn5+vqZNm+bzfiNHjtQdd9whqb0lFQAAAAAgOkzVGoMGDVJ6enocdhMbpovs9u/fb1mjYiN8gao1JIINAEBwCDYQVxs2bNDy5cstA8N9WbZsmUaOHKkNGzaooqIi+hsEAAAAAAcyzddI5TZUkjnYMJ2Ij+bw8FQPNgLN15AINgAAwSHYQFydPzB8woQJAe9/fgBy5513RnNrAAAAAOBYThoc7mFqRdW5bVLPnj1t+3dwYrBBxQYAwC4EG4ibDRs2aMOGDT4Hhvtyxx13aMKECXK5XFq+fHnU9gcAAAAATmVqRZXqwYaveY/ns6taQ2oPNrp08T4tU19fr5MnT9r2HInGFGx0bm/W3Nyc0v8GAAB7EGwgbiZMmKBp06Zp2bJlQb2BPN/atWuVn58fVJUHAAAAACA0tKIyszPY8HW8VK7aMAUbmZmZljWqNgAAgXSN9wbgXFlZWfrLX/4S1mNHjhyptWvX2rwjAAAAAIDkzFZUwQQbpvZRkcjJydHHH3/stbZ3715deumltj5PojAFG3379lVjY6PXWn19vS666KJYbQsAkISo2AAAAAAAAF6c2IrKNGOjM7srNpw2Z8MUbAwcONCyRsUGACAQgg0AAAAAAOCFVlRmdlds0IrKHJgRbAAAAiHYAAAAAAAAXmhFZRaLio29e/fa+hyJ5ODBg5a14cOHW9YINgAAgRBsAAAAAAAAL7SiMqNiI3ynT5/WkSNHLOsjRoywrJnuBwDA+Qg2AAAAAACAF1pRmTFjI3ymNlRDhgwx/ndFxQYAIBCCDQAAbPDJJ5/oK1/5ioYOHaqCggK988478d4SAABAWA4fPiy32+211r9/f3Xv3j1OO4qNQMHG8OHD1a1bN1uf00mtqHwFG4MGDbKsE2wAAAIh2AAAwAbf//739ac//UlNTU167rnntGLFinhvCQAAICymNlSpXq0hBQ427G5DJUl9+vRR//79vdbOnDmj/fv32/5c8UawAQCwE8EGkCJqa2t18803a8iQIZo9e3bKXuUDJKL9+/frxRdf9FpbvXq1zp49G6cdAQAAhM+Jg8M9/M3ZsLsNlb/jpuLnOdPgcIINAEC4CDaAFNDQ0KCvfe1rqqqqUnNzsyorKzV//vx4bwtwjNdff924bvrwBgAAkOicODjcw1/VRjQqNnwdNxXnbPiq2Bg4cKBlnWADABAIwQaQAubPn6+dO3d6rW3atEkvv/xynHYEOAvBBgAASCVOHBzu4S/YiGXFhpOCjf79+6tr165e6ydOnFBzc3OstgYASEIEG0CS+973vuczwPjFL34R490AzkSwAQAAUgmtqMyo2IiMKdgYOnSoJBnbUR05ciTqewIAJC+CDSCJ/fznP9fjjz/u8/YNGzboL3/5Swx3BDjPBx98oD179hhvO3DgQIx3AwAAEDlaUZlFq2LDFGyk4owNXxUbkjnYoB0VAMAfgg0gSf3+97/XD37wg4D3o2oDiC5f1RoSFRsAACA50YrKLFoVG05vRSURbAAAQkewASSh6urqoIeDv/LKK9q0aVOUdwQ41xtvvOHzNoINAACQjJzcispXsNG3b1/1798/Ks/p5FZUBBsAgHARbABJpqGhQfPnz1dTU5Pltm9/+9vKy8uzrFO1AUQPFRsAANjjgQce0LJly/xeMY/YcHIrKl8zNqLVhkqShg8fbhmeffToUR0/fjxqzxlrR44c0alTp7zW+vTpo8zMTEnSwIEDLY8h2AAA+EOwASSZ+fPna+fOnZb1mTNnqri4WIWFhZbbXnzxRb399tux2B7gKG+//baOHj3q83ZmbAAAELyFCxdq4cKF8d4GRCsqk2i1ofJ3/FSq2vA3OFyiYgMAEDqCDSCJfO9739PLL79sWb/88su1atUqSdLtt9+uSy65xHKfFStWRH1/gNP4q9aQqNgAAADJ58iRIzpz5ozXWt++fZWRkRGnHcWWr2AjmhUbvo6fSgPE/bWhkgg2AAChI9gAksTPf/5zPf7445b1vn37atWqVerdu3fHmqlqo7y8XP/4xz+iukfAaQIFG4cPH1ZbW1uMdgMAABA5J7ehkny3oqJiIzIEGwAAuxFsAEng97//vX7wgx8Yb1u1apVlrsb8+fONV/xQtQHY5+zZs34Hh3vQjgoAACQTJ7ehkhKrYsPpwcaRI0eiuicAQHIj2AASXHV1tb75zW8ab/vRj36km266yXibqWrjN7/5jWpra23dH+BUb7zxhmUAYuehjxLtqAAAgH3ef/993X777erbt68KCgpUU1Nj+3OYgg0nVWwk0oyNVGpFZXpPTMUGACASBBtAAmtoaND8+fPV3Nxsue3b3/62FixY4POxhYWFxg8gVG0A9jC1oTJ9ICPYAAAAdnnkkUdUUVEhSXruuee0bNky25/D6a2o4lWxYQo23njjDb3yyivGz4PJhlZUAAC7EWwACWz+/PnauXOnZX3mzJkqLi72+9j09HRj1cYvf/lL7dmzx7Y9Ak5lCjYuvvhiyxrBBgAAsIPL5dILL7zgtfbMM8/I7Xbb+jxOb0VlmrGRnp4e9YoNU3Cyf/9+ffnLX9aQIUM0c+ZM/ehHP9Jf//rXqO4jWgIFGwMHDlRaWprX7Q0NDZYKaQAAPAg2gAT1ve99Ty+//LJl/fLLL9eqVauCOkZhYaHxjTlVG0BkGhsb9dZbb1nWL7/8cssaMzYAAPDvnXfeUUFBgfr27atvfetb2rZtW7y3lJDef/99y5rb7bb9oiWnV2z06NFDX/7yl73Wvv71r0f9eQMFJ6+//roefvhhzZgxQxdeeKFuu+02Pf7448b/LhKRKdgYOnSo19dUbQAAQkGwASSgn//853r88cct63379tWqVavUu3fvoI7Tq1cvY9XGL37xC+MbSwDBMVVrXHLJJRozZoxlnYoNAAD8+8UvfqHnnntOkvTss8/qiSeeiPOOEpOvE9gff/yxrc/j9IoNSbrrrrv0la98RT179tS///u/a9GiRVF/zszMTN14441B3behoUG///3v9b3vfU//8i//ossvv1zf/va3tXr16oS9qCZQxYZEsAEACA3BBpBgfv/73+sHP/iB8bZVq1YpLy8vpOMVFhaqZ8+eXmtut5uqDSACpmDjmmuusVx1JhFsAADgT2trq8rLy73WnnnmmTjtJrHFM9hwUsWGJH3mM5/RqlWrdPjwYT3++OPGdqPR8MMf/lA33XRTyI/7+OOP9Zvf/Ebz58/XpZdeqq997Wv65z//GYUdhqepqUkNDQ1ea127drUEGQQbAIBQdI33BoBU1dbWJpfLpWPHjsnlcvn8c/7thw8f9llJ8aMf/SisN7n9+/dXYWGhZbDgL37xCy1cuNDncDwAvr3xxhuWtWuvvVbDhg2zrBNsAADgm+mk5ZkzZ9TU1KRevXrFYUeJK1bBhtNbUcXTpZdeqt/97nc6ePCgqqqq9Prrr+v1118P+TV+8cUX1a1bN/3qV7+K0k5DE0y1hmQONo4cORKVPQEAkh/BhqTq6mrdfffdqqqqkiRNnTpVxcXFmjRpUtjHXLNmjW699VbbB7kh8W3ZskUrV67sKKe3w7e//W0tWLAg7McXFhZq+fLlXv89trS0aMWKFVq8eLEdWwQcY//+/dq+fbtl/dprrzUONyTYAADAN18nLRsbGwk2OqEVlXMMHTpUc+fO1dy5cyW1v/avv/56R9jRufrBZO3atSkRbFCxAQDwxfGtqCorK3XFFVdox44dKikpUUlJiXbs2KHJkyersrIyrGPW1tYa5xrAGewONWbOnKni4uKIjjF06FCfszaam5sjOjbgNKY2VBMmTFB2draxFVVdXZ1Onz4di60BAJB0fJ20bGxsjPFOEtvRo0e1f/9+4212Dg93uVxqbW31WsvMzAx6xh+iY8yYMbrrrrv09NNPa+/evaqsrNT999+va6+91udjzp49q3379sVwl74FMzhcItgAAITG0cGGy+XSbbfdpuzsbNXU1GjRokVatGiRampqlJ2drdtuu00ulyukY1ZXV2vKlCnG8l2kvjNnztgaalx++eVatWqVLccyBRsNDQ3M2gBCZAo2PB8q09PTjVefUbUBAIAZwUZwfFVrSPZWbNCGKjl85jOf0X/+53/qD3/4gw4dOqTnnnvO+DrZGXpFgooNAEA0ODrY2LJli+rq6rR48WKvOQNZWVlasWKF6urq9NprrwV1LJfLpbKyMl1xxRWSFPKAZ6QGU9l2ONLS0jRv3jz97ne/s+3qqIsuukgFBQWW9RUrVujs2bO2PAfgBP6CDcl89dmBAweiuicAAJKVr5OWx48fj/FOEtvOnTt93rZ//35LlUW4aEOVfDIyMnTDDTfos5/9rOW2vXv3xmFHVgQbAIBocPSMDc9Mjeuvv95y2/jx4yVJr776qmbPnh3wWC+88IIKCgo0depU/fa3v9V3v/tdYw92pDbTFU7du3dXXl6esrKy/P7p37+/19fRUFhYqNLSUq+1Q4cOacWKFfr2t78dlecEUskHH3xgvPLtmmuu6fj/Q4cO1bZt27xup2IDAAAzKjaC469iQ2qv2hg9enTEz2MKNqjYSA4XXnihZS3Zgo2BAwda1gg2AAC+ODrYqK2tlSSNGDHCcltubq6k9l6mwejXr59KS0s1b948+zaIpGP6IDB58mS9+OKLcdiN1ejRo/W1r31Nv/3tb73WCTaA4JiqNT73uc8pIyOj42tTxQbBBgAAZv6Gh+OcQMHGnj17bAk2aEWVvBI52DC9F6ZiAwAQKUe3ovLwd3V8eXl5UMeYPXs2oQaS4gon06yNPXv26Je//GUcdgMklzfeeMOy1nlo47Bhwyz3IdgAAMCMVlTBCaZiww60okpeiRxs0IoKABANBBsJzu12W/4gcZmucEq0DwKXX365vvzlL1vWGSIOBBZovoZExQYAAKGgFVVgdXV1xhPD54tmsJFoF2rBLNmCDdN7ZlOwcezYMbW1tUVlXwCA5OboVlTJ4JFHHon3FhCCZPkgUFhYqOeee85rrba2Vr/5zW/09a9/PU67AhLb22+/bWlPmJmZqcmTJ3utMTwcAIDgEWwEFqhaQ7Iv2KAVVfJKtmDDVLHRpUsXDRw40NKirr6+3nh/AICzUbEhyeVy+bxtzpw5sduIQVpamuUPEleyBBsTJ07Uv/3bv1nWqdoAfAumWkOiYgMAgFAwYyOwYIKNPXv22PJctKJKXgMHDlSfPn281pqamoxhVSwdPHjQ0nliwIAB6t69u/H+tKMCAATL0RUbngHhe/bssczZqK6ultT+Czee7rvvPstacXFxHHaCYCRDKyqPwsJCy1Dzf/zjHyovL497oAckIoINAADs1dbWZqmG9GDGxjlUbCBYF154oXbs2OG1tnfv3rh+Jg22WsNj4MCBljWCDQCAiaMrNqZOnSpJ2rp1q+W22tpaSdLnP//5mO4JyS1ZKjYkacqUKbrhhhss61RtAFZnz541Dg6/5pprLGum4eFHjhxRa2trVPYGAECy8neykoqNc3bu3GlZS09P9/q6vr7eljAomT7PwCoR21GFGmxQsQEACJajg42rr75a2dnZuu+++zqCDKm9NVVhYaGys7N1/fXXx3GHSDbJ9kGgsLDQsvb2229r3bp1cdgNkLjeeOMNnTp1ymttyJAhGj9+vOW+aWlpVG0AABAEX22oJIKN85kqNoYPH25Zi7Rq4/jx42pqavJa69mzp/r27RvRcRE7yRJsmN4rexBsAACC5ehgIysrS08//bTq6uo0ZcoULVu2TMuWLdPo0aNVV1enp59+2tKiau7cuUpLS+toVQV4uN3upGpFJUn/+q//qmnTplnWqdoAvAXbhsrDVLVBsAEAgDd/JytpRdXu4MGDln+nHj166JJLLrHcN9JggzZUyS8nJ8eylojBBhUbAAA7ODrYkKTp06dr06ZNuuyyy1RUVKSioiJddtll2rRpk6ZPnx7v7SGJmKo1BgwYoG7dusVhN8EzVW1s2rTJMuANcLJQgw0qNgAACIxWVIGZqjXGjBmjiy66yLIe6QBxBocnv2Sp2Ag12PBX3QUAcC7HBxuSNGnSJG3YsEFut1tut1sbNmzQpEmTjPddvXq13G63sf2I6X5wjmS9wunGG29U//79Leumvw/gRI2NjXrrrbcs66EGGwcOHLB1XwAAJDuCjcBCCTYirdhItra6sErVYIOKDQCACcEGYJNkvsKJN4+Ab6ZqjUsuuUQXX3yxz8dQsQEAQGD+3m+eOHFCZ8+ejeFuElMsg41kvVAL5yRisGF6D0ywAQCwA8EGYJNkvsLJFMDw5hFoF2obKolgAwCAYARqL0PVhu9gY8SIEZb1aFRsJMuFWmh3wQUXWFohHz16NK4za6jYAABEC8EGYJNkGxx+PtObR1pRAe3eeOMNy9o111zj9zEMDwcAILBAJysJNszBxtixY2M2YyNZLtTCOYlWtZGIwcaOHTv0jW98Q1lZWfrWt76l6upqW48PAIgNgg3AJsn8QYCrYgCz/fv3a/v27ZZ1KjYAAIhcoPeb8bzKPBHs379fR48e9Vrr1auXRo0apSFDhqhXr15etzU2NkY0ZJlWVKkhkYKNxsZGNTU1ea1lZGQoKyvL52MGDhxoWauvr7d1hun/9//9f1q7dq3Onj2rZ599VqtWrbLt2ACA2Oka7w0gPDt27NBXv/rVgPe76KKLVFRUFFTlQF1dnUpKSoIuYf7617+uG2+8Maj7/uY3v1FVVZXlTY3JxIkTVVRUFNRxd+zYod/85jdB7Tk7O1tf//rXNXHixID3bWpq0sqVK7Vjx46g9xxssPHSSy9pzZo1QR03Vq9foFZUTnj9gt1zIr5+gfD6nRPq65ebm2tZnzBhguU1Nb1+M2fOtDzW83Ob1y+8PfP9dw6v3zm8fufw+nnj9TsnUV8/f8FGXl6efvzjH6u1tTXgcVP19fNUa1x88cUaPXp0R4shz/uJf/3Xf7U85n/+53+0ZMmSsF+/zu9fnn/+eT3//POS+P4Ld8+x/v7r06eP5XX89a9/rbNnz8b89etcrTFw4EBdeeWVAc9lzJw5U83Nzdq+fXvHxUFHjhwxXpAnhf76db7g6Nlnn9XPf/5z430T9eenP3z/ncPrdw6v3zm8fuck4+t3Pio2klRbW1tQ9/v444/18ssvB3XfrVu3htSXNdjjNjU16aWXXgrqP2TPPrZu3Rr0HoLdc11dXdB73r17t7Zu3RrSnk0fzEzfkMF+U0uxe/38taJyyusX7H0T8fXzh9fPW6ivn6kNlalag9fvnER6/fj+a8fr543X7xxeP+998Pq1i+Xr5yvY6Natmy6++OKgQg3PPlLx9du5c6ckeYUagRw+fJjvv/8n3q+fR6J+//li5+vXOUC4+OKL1bNnz6COm5GRoYsvvrjja39BaKiv39ChQ72+p06fPq0TJ04Y7+vk18+0B77/zu2B168dr985vH7eovX6nY9gI8VlZ2fr05/+dFD3nThxYtDJWK9evQK2YgnnvpJ02WWX6bLLLgvqvl/4whcsJdh27GPkyJFBJZue406cONHYO9RUsTF79uygjivF7vXz14rKKa9fsPtIxNfPrvtKvH7ny87O1o4dOyzrpr8Hr985ifT68f0X+j54/cK7r8Trdz5ev/D24cTXz1fbpNOnT4fUOidVXz9PxUZNTU1Qx5WkHj168P0Xxj5S/ftPCtxK1cPO16/zZ+SPPvoo6ON2/jngL9gI5fXr1q2bDh48qNOnT3utu1wu4/0T4fXj+8/7uMn2/cfr531cXr9zeP3OCeX1O1+a285GhYiJ4uJiSdLixYvjvJPo6t+/v1wul44dO+a3B2eiGD16tOWKlO3btxt7nCaa119/3VKuPHny5LDSUiBVfPDBB7r66qst64cOHVJGRkbAx0+bNk3vvPOO19orr7yiSZMm2bZHAACS1ZEjR7yuxjZZvny55s2bF6MdJZ7p06frrbfe8lorLy/XF77wBUnSz372My1ZssTr9m9+85sqKSkJ+bmamposM8K6desW0cwOxEdVVZVuvvlmr7XPfOYzqqysjPle/u///s9y3mL+/Pn6yU9+4vdxX/3qV/WHP/zBa+2pp55Sfn5+xHtatmyZHnjgAcv6pk2bNG7cuIiPDwCIHSo2AJuYhu2F2hsuXvy1ogKc6vXXX7esXXPNNUGFGhIDxAEA8CfQ4HCpffCwk3kqNs43ZsyYjv9/0UUXWW4PpbXG+RgcnjpMF9bt2bMnDjuxVmxI5vfInfnrKBCpXbt2Gdd9VWwAABIXwQZgg/r6esvck379+gXdPzTeovnGEUhWpvka11xzTdCPHzZsmGWNYAMAgHbBVAI4OdjYu3evGhoavNZ69+6tkSNHdnxtZ7Bx+PBhyxrBRnIyBRsHDhywtF6KBVOwMWTIkICPGzhwoGXNrs+nvtphEWwAQPIh2ABskOwfBEyVJceOHdOZM2fisBsgMZgqNkLpl0nFBoBE9c4776igoED9+vXT/PnzjVeFA9EWzEnK48ePx2AniSlQtYYkjRgxwnIfO4ONZKk+h7du3bpp+PDhlvVQ5tbYJdxgI5oX3hFsAEDqINgAbJDMbagkKS0tjXZUwHnefvttHT161GstMzNTkydPDvoYpmDjwIEDEe8NACL1i1/8Qs8995zcbrdWr16tn/3sZ/HeEhyIVlT+BRNs9O/f3zKLsLW1Naz3G7SiSi2mqg2CDenUqVM+/x0INgAg+RBsADZI9ooNiXZUwPkirdaQqNgAkJhaW1tVXl7utfbb3/42TruBkxFs+BdMsCHZ144qFT7P4JxECTZM733jGWz4mq8hEWwAyWL79u26/fbb1bdvX82fP18ffPBBvLeEOCLYAGyQCh8ECDaAc+wINpixASARma7Kdrvd/HxCzAUzY8PJrah27txpWRs7dqxlLZrBRjJVoMNbIgwQP336tPH7PNxgI5ifGYH4akMlEWwAyeLRRx9VRUWFJFF5DIINwA7J3opKMu+XVlRworNnz0Y8OFyiYgNAYvL1u51WeYg1Kjb8C7Ziw645G7SiSi2JULHhqw1Vly6BT0NRsQHA5NixYx2hhsczzzwTn80gIRBsADagYgNIHW+88YZOnTrltTZ06FCNHz8+pOMMHjxY6enpXmsul0tNTU0R7xEAwuUr2CB4RawRbPi2e/dunThxwmutX79+xpPVpoqNcK7Mp2IjtSRysBGMaH02pWIDSG61tbWWtTNnzvA+1sEINgAbEGwAqcPUhirUag0P2lEBSDSm9ywSFRuIPYIN34Kt1pCYsQGzZA82evTooT59+nittbW16ejRoxHtiWADSG6mYEOS9u/fH+OdIFEQbAA2oBUVkDrsmK/hQTsqAImGig0kimCCDafO2IhHsEErqtSSCMFGuIPDPaJx4R2tqIDkVlNTY1wn2HAugg3ABqlwhRMVG0D7laFvvfWWZd3OYIOrogHEE8EGEoVpEHD37t29vm5tbVVLS0ustpQwQgk2fM3YcLvdQT9fa2ur5aRuly5djJ8PkBz69OmjAQMGeK2dOXNGn3zyScz2EEnFhhSdz6dUbADJzVfFRix/tiGxEGwANkiFig2CDcBcrXHJJZfo4osvDut4VGwASDS0okIiaGhosMyzyszMtLSekZxZtRFKsJGZmWm5oMrtdoc0Z4NqjdQU76qNaAQbpkA0WB999JHOnj3r83aCDSDx0YoKnRFsABE6duyY5YNZ7969lZmZGacdhYdWVIC0du1ay1q41RoSwQaAxEPFBhKB6eKZQYMGqW/fvpZ1J87ZMAUbY8eO9Xn/SNtRpUL1OawSMdgwvTf2xe4L7/xVa0jSqVOndPLkybCPDyC6zp49SysqWBBsABFKlQ8CVGzA6davX681a9ZY1sMdHC4xPDwa6urqlJ+fr5ycHBUUFGjbtm3x3hKQVAg2kAhMV10TbLTbtWuXmpqavNb69++v4cOH+3xMNIKNZKs+h1UiBhvxbEXlb76GB1UbQOKqra312WaRVlTORbABRCgV2lBJ0sCBA9Wli/ePhMbGRrW2tsZpR0Ds/PrXv9att95qLE+nYiOxFBYW6s9//rMaGxv13HPP6ec//3m8twQkFV/BxqFDh9TW1hbj3cCpqNjwbefOnZY1X22oPCINNmhFlZoINrwFqtiQCDaAROarDZVExYaTEWwAEUqVig3JHMhQtYFU9+ijj+o73/mO8bZZs2ZFFFQyPNxeO3bs0Pr1673WysvL47QbIDn5mrEh8fMJseMr2GDGRuhtqCTfA8SDlUqfZ3AOwYY3KjaA5OarDZVEsOFkBBtAhFLpg4DpzSNzNpDKFi5cqKVLlxpvy8vL0w9/+MOIjk/Fhr1efPFFy5rb7eZDKBCkuro6nyX8Ej+fEDumk5MDBw6kYkOhDQ73MFVshDI8nFZUqSmewcaRI0cscyj79OkT0hzKgQMHWtao2ACcy1/FRktLCxflOhTBBhChVGlFJTFnA87R3Nysr3zlKyorKzPefv311+ull17SxRdfHNHzZGdnq1u3bl5rjY2NOnHiRETHdap169YZ148ePRrjnQDJKdDFCgQbiBVmbPhmV7BBKyrEM9iIdHC4RCsqAN78BRsSczacimADiFCqV2wQbCDV7Nq1S1/4whf0pz/9yXj71772Na1du1ZZWVm2PB9VG/bYtm2btm/fbrzt2LFjMd4NkJwCBRu0okKs0IrKN7uCjf379wc9Ky+VPs/gnOzsbPXq1ctr7cSJE8Zg0W6RtqGS7P1sun//fjU3N3utde/e3XI/gg0gcflrRSXRjsqpCDaACKXSBwFTpQmtqJBKNm3apBtvvFHvvvuu8fZ7771Xv/jFL2x9ToINe/iq1pAINoBg+ZuvIfGzCbHD8HCz2tpaSxgxaNCggFe6d+3aVTk5OZb1YKs2aEWVuuJVtZFowYZpvsaAAQMsawQbQGI6ePBgwO9Pgg1nItgAIkQrKiA5rFmzRjfeeKPPK5J/8pOf6L//+79tf95hw4ZZ1jh5GDrTfA0Pgg0gOFRsIFEwY8MsnGoNj0jmbNCKKnWZBssnS7DRq1cvy0yO06dPq6GhIeT9mNpQmf4bJ9gAElOgNlQSwYZTEWwAEUqlig2CDaSq//u//9Odd95pHJrbvXt3Pf3005o/f35UnpuKjcht2bLFb+kxMzaA4DBjA4mCig0zu4ONYCo2zpw5Y2xNlKyfZ+AtXhUbpt8noQYbkn2fT00VG8OHD7esEWwAiSlQGyqJGRtORbABRCiVgg1aUSEV/dd//ZcWL15svO2iiy7SSy+9pJtvvjlqz28KNrgqOjT+qjUkKjaAYNGKConC1/Bwp8/YiEew4asNVVpaWlDPi8SWzK2opPZKrs7CCTZMFRum7xmCDSAxUbEBX7rGewNAMmtsbFRLS4vXWkZGhvFDWTKgYgOp5pvf/KbKy8uNt33mM5/Rr371K2NPajtRsRE5gg3AHrSiQiI4efKkmpqavNZ69OihPn36OL5iY+fOnZa1sWPHBvVYU8uhYIIN2lCltmQPNuz6fGoKNj71qU9Z1gg2gMQUTMUGwYYzUbEBRCCVqjUkgg2kjrq6Os2cOdNnqJGfn6+XXnop6qGGxIyNSG3atMn4YfR8BBtAcAIFG0eOHLEMLgbs5qsNlSTHBxvxmLHB4PDUlkjBhulin0Ci2YrK9L1FsAEkpmAqNmhF5UwEG0AEUu0KJ1pRIRW8//77+sIXvqDXX3/dePuCBQv01FNPqWvX2BQtUrERmUDVGhIzNoBgBfM7nZ9PiDZfbagkOboV1fvvv68zZ854rQ0ZMiTokCHcVlSp9nkG3kwX8TitYqO+vt4ycLxnz54aPXq05b4EG0DiaW5uNv4+y8jI8Pr65MmTXPDmQAQbQARS7QqnrKwsdevWzWvN1C4ASGTf+973fF7R8eCDD+pHP/pRTPdDsBGZYIIN3sACwQk0Y0Pi5xOij4oNs0iqNaT2K/M7X7RRX18fMBhKtQp0eLvwwguVnp7utVZfX6+TJ09G7TmbmposQULXrl2NIUUgpseYwlF/TJW/F198sbKysizrBBtA4jF9th85cqSGDx9uWacdlfMQbAARSMUPArSjQjJ7++23VVVVZbzt8ccfV1FRUYx31D70sEePHl5rx48fd8wVqJGoqqrSvn37At6PYAMIzDQXrPPJLok5G4g+gg2zSIMNKbyqjVS7UAtWsW5HZVe1hmTPZ1NTG6pRo0apV69elvfora2tXNQHJBjTfI3c3FxdcMEFlnWCDech2AAiYCrdTvYPAqb9E2wgWaxevdqy1rVrV61bt07//u//HocdtaNqIzymao3p06db1mhFBQRmes/SuYRf4mcTos/0vnLgwIGS2sO2zMxMr9vcbrcjLgawI9gIZ4A4rahSn9ODDV8VG5Ko2gCSgKliIzc311ixwZwN5yHYACLglIoN5mwgWZiCjXvvvVfXXXddHHZzjinY4KrowEzBxi233GJZo2IDCMz0u9x0dTzBBqLNX8WG5NyqjWhVbAQaIJ6Kn2fgLRGCjXAGh0vRCzZGjRoliWADSAamio3Ro0dTsQFJBBtARFLxgwCtqJCsysvLLSe4u3XrpnvuuSdOOzqHio3Qvfrqq5YPxunp6frSl75kGS7rdrv5EAoEYHrPMmDAAMsaoau3t956S/PmzdPAgQN15513avfu3fHeUtLzNzxccmawcfbsWWOwMXbs2JCOQysqmCRCsJForaio2ACSRygVGwQbzkOwAUSAVlRA4jBVa8ydO9fS0iIeCDZCZ6rW+Ld/+zf16tVL/fv3t9xGOyrAv2DbzfCzydvjjz+u559/XqdPn9aaNWv06KOPxntLSS9QxUbn8FpSyreiev/993X27FmvtWHDhhnDR3/CCTZoRZX6kjnY8LSpOx+tqJCK3n33Xd1xxx3q27evvvWtb2nHjh3x3lLCoBUV/CHYACLglIoNWlEh0e3atUuVlZWW9blz58ZhN1bDhg2zrHHy0L9169ZZ1v7t3/5Nkvkqc9pRAf6Zfpfn5ORY1qjYOOfUqVN6/vnnvdaefvppnT59Ok47Sg20orLauXOnZS3UNlRS6DM23G53Sn6egTdTsBGoRVkkTO9xww02+vTpo549e3qttbS0BB12Hj9+3PLfeFpaGq2okHB++tOf6oUXXpAkPfvss1qxYkWcd5QYdu/erebmZq+1/v37a+jQobSigiSCDSAiqfhBgFZUSEamao0xY8Zo6tSpcdiNFRUbofnTn/5kCSp69uypWbNmSZKxYoNgA/DPFGyYru7mZ9M5virB/vnPf8Z4J6nF3/BwyZnBhh1tqKTQZ2z4alHXtWvXkJ8biSuZKzYk8+dTU0s7E3/zNSSCDSSGkydPas2aNV5rzzzzTJx2k1h8VWtIohUVJBFsAGE7efKkTp486bXWvXt345ujZEKwgWRUXl5uWUuUag2J4eGhMrWhmjVrVseJFoINIHSmE5gXX3yx0tPTvdZcLpeamppita2E5uvEmelDNoIXaMaGU1tRdRZOxcaQIUPUq1cvr7XGxkaf/y3ThsoZTMHGJ598ora2tqg8XyyCjWA/n/qbryERbCAxmD4Xnj59Wg0NDXHYTWLxF2wMHjxY3bt397qtoaEh5d8zwBvBBhCmVKzWkMwzNmhFhUT26quv6sMPP7SsJ3qwwVXRZm632+d8DQ9mbACh8zUXjFZ5vvn6uUKwEb7W1lZL9UXXrl29fq5TsdEunGBDCm3ORqp+noG3Hj16GN+LRqtqwxRsmJ4/WJEEG/7ma0gEG0gMvqoMgq1MSmU1NTWWtdGjR3f8f+ZsgGADCFOqfhCgYgPJxtSG6qabbjL2jo8XThwG78UXX9SJEye81vr27aubb76542tmbACh83VlNhVlvtGKyn6B2lBJzgs2Tp8+rQ8++MCyHq9gw3SRE5JfLNtRJXrFBq2okGh8ve8i2PBfsSHRjgoEG0DYfF35mOwINpBMXC6XMdj4yle+Eofd+Na/f3/L4MOTJ0/aUl784Ycf6mtf+5r69u2r22+/Xbt37474mPEUqFpDohUVEA5fJzAJXn2jFZX9ArWhkpwXbJiqNXJycsJub2saIO5rzgatqJwjVgPEDx48KLfb7bU2YMAAS7uYUFCxgVRHsOFboGCDAeIg2ADClKoVG3379rWcgG1ubrZcQQ0kAlOoMXjwYOXn58d+MwFEqx3Vgw8+2BEGVFRU6Ic//GHEx4yXU6dOEWwAUdDa2moJUrt06aJBgwbRKs8PKjbsZzoZ2fmkpdNmbNjZhkqiFRXMYlWxYXe1hmSt6pIiCzao2ECi8RVsOP0C02PHjlnek3bp0iVgKyqCDWch2ADClMofBExXxTBnA4nIFGwk0myN80Xjquj6+nqtW7fOa+35559P2itbX3zxRbW2tnqtDRw4UF/4whe81pixAYTGX5Uprah88/Vz5ejRo8b3gQgsmGCDio3YBRupWoEOq2QONsKt2Dh16pTx70jFBhINMzbMfFVrpKWldXxtqthgxoazEGwAYUrlDwK0o0IyePvtt7V161bLeqIGG9G4KnrLli3G9bfffjui48ZLMNUaEjM2gFD5azdDKyrf/J1QoGojPMzYsNq5c6dlbezYsWEfj4oNmMQz2IhkcLgU/mdT03yNnJwc9ejRo+Nrgg0kAlpRmZmCjfOrNSQqNkCwAYQtlT8ImAIagg0kGlO1xqRJkzR+/Pg47CawaFwV7SvYeOeddyI6bjycPHky6GCDVlRAaPwNCKYVlW/+KsGYsxEeWlFZ2V2xYZqxwfBwpFrFRjAnfAPN15AINpAYCDbMampqLGvnz9eQCDZAsAGELZWDDVpRIRkkUxsqiYqNQF588UWdPXvWa23o0KG6/vrrLfelFRUQGn9VpqaKDVpRtSPYsB/Dw721tLQYq38iCTb69+9vOVnb2tpq/L5meLhzpFqwEW7FxvnzNSQpMzPTMti8paVFzc3NIe4SCJ+vE/FOv7g00OBwiVZUINgAwkYrKiB+ysvLLVfod+vWLaGDjWi0e0m1YKMzU7WGRMUGEKpQZ2xQsdHOX7BBK6rwMGPDm6laY8SIEcaqlVAE244qlS/Ugrd+/fpZAq9Tp07ZHmQnUrARTMWGRNUG4uvQoUNqa2sz3ub0io1gWlENGzZMXbp4n9o+evSompqaoro3JA6CDSBMqfxBgFZUSHS+qjUyMzPjsJvg2H3y8N133/X5hu3AgQM+204kIpfLpT/+8Y+WdV/BRrdu3SwnfdxuNx9CAR/8XZU9cOBAy9Wqx48fT+nWP8Hyd0KBio3wEGx4MwUbkczX8Agm2Kivr7ecTOvXr5/X/AGkllhUbZje20YabPTr10/dunXzWjt58mTAE5emYKNzxYZEsIH48hcuOjnYOHv2bFCtqCTaUTkdwQYQhpaWFssHrPT0dMvww2RFKyoksl27dqmystKynsjVGpL9wYZpcPr5kqlqw1StceGFF+raa6/1+RjaUQHBC9RHnwHiVm1tbX5PbNXW1lra5yGwYIaHZ2ZmKj093WutublZp0+fjure4sHu+RoewczZoA2V88Qi2IhGxYYUXtWGqRUVFRtINAQbZrW1tXK73V5rQ4cONX6/mtpREWw4B8EGEIZUrtaQaEWFxGaq1hgzZoymTp0ah90Ez+7h4b7aUHkk0wDxdevWWdZ8VWt4DBgwwLJGOyrALFD7TLt/PqWCYIJSqjZCF0zFhuScqo1oBRumio09e/Z4fZ3qn2dg5bRgg1ZUSAb+TsC7XC6dOXMmhrtJHMG0ofIwVWwwZ8M5CDaAMKT6BwFaUSGRlZeXW9YSvVpDav/Q1KtXL6+15ubmsD84BQo2kqVio66uzliBEyjYYM4GELxAV2ZTsWEVTLDBnI3QtLW1GX9Om05YmmZMpGJ7tHi2ogpUyYXUE+1go7Gx0dIeKiMjwxgchCrUYOOjjz6yVNVlZ2cbQ1OCDcRToAtJnHoeJtg2VBIVG05HsAGEIZUHh0u0okLievXVV/Xhhx9a1pMh2JDsuyr64MGDxn+H8yVLsGGq1hg1apQmTZrk93EEG0DwqNgIHRUb9vPVhiotLc2y7oSKjaamJmOrnGhVbNCKCtEONqJVrSGZP5/6a9MT7HwNiWAD8RXo/ZZT21GZ3mP5CjaYseFsBBtAGFK9YoNWVEhUpjZUN910k3JycuKwm9DZNWfDVK3RtWtXr69bWlr03nvvhXzsWDPN15g1a1bAxzFjAwgeMzZCF8yJBIKN0Jj+TU3vOSVnBBumao2LL77YUt0ZDl8zNs7vV57qn2dgFe1gIxqDwz1C/Xwa7HwNiWAD8RXoBDzBxjm0ooIJwQYQhlT/IJCZmWn5UHXq1Ck1NDTEaUdA+wcMU7Dxla98JQ67CY9dJw9NwYbpDV2iz9n45JNPtGHDBst6oDZUEjM2gGDV1dVZhi9mZWWpe/fuHV/bFbqmElpR2S/Y+RqSM4KNnTt3WtbsqNaQ2t/Ld/5s4na7veZspHoFOqxSrWIjUCuqzqjYQCKiYsOMVlQIFsEGEAYnfBBgzgYSjSnUGDx4sPLz82O/mTDZdfJw69atlrUrrrjCshbNdlSnTp3SjTfeqH/5l3/R5z//eZWVlamlpSWkY5iqNcaMGaNPf/rTAR9LKyogOMG0mzGFrrSiohWV3UIJNpwwYyNa8zU8ArWjSvULtWA1ZMgQ9ezZ02utsbHRtvdPiRRsULGBZMGMDauDBw9avgczMjKMv9ckWlE5HcEGEAYnfBBgzgYSjSnYSJbZGh7RbEU1Y8YMy1o0g41vfetb2rhxo95//3397W9/08KFC3XRRRdp0aJFAQebe5iCjWCqNSSCDSBYwVyMQcWGlekKyfT0dK+v6+vrHXslZTh8zdgwcULFhinYsKtiQyLYgFk0qzZMwYbp90s47KjYINhAomlubg74+cWJ7zNCaUMlmSs2Dh8+rFOnTtm6LyQmgg0gDE74IMCcDSSSt99+21ilkArBRqhXRW/dulWtra1ea0OGDNHNN99sue+OHTuiciJo3759WrNmjWW9ublZpaWluv766wNWcezevVsbN260rEcSbDBjA7AKNF9DYsaGienniem9HlUbwaMVlbd4BxtOqECHVayDDbsqNkwhaDRbUdGCGbEQTFWBE4ONUNpQSVKXLl2M72WZs+EMBBtAGJzwQYBWVEgkpmqNSZMmafz48XHYTfjsuCraVA1x9dVXa8CAAcYrWaJRtVFVVRXwPoGqOEzVGpdffnnQrykzNoDgBPOepV+/fpbZWk1NTY6+YtUUbOTk5FjWmLMRvFCGh6d6K6rjx497hQwedrai8jVA3MMJF2rBKlmDjVAuutu/f7+am5u91vr16+fz5w0VG4iXYC5wc2KwYbpoxF+wIdGOyskINoAwOOGDAK2okEhSoQ2VZM9V0b6CDUnG2RTRGCD++uuvB31fX1UckbShkmhFBQQrmBkbkj0VZanEdCLB1MaEio3gUbFxjqla45JLLlGPHj1sew5TxYZnePixY8csLTr69OljCTiRekyBV6oFG6HM15AINhA/wbzPcuLFpaaKDX+tqCQGiDsZwQYQolOnTllOnqWlpTki2HDiL1XEX3l5ueV7rlu3bkkZbESrYmPixImSpKuuuspyWzQqNkzBRudhlCaeKo4RI0bob3/7m+X2SIMNWlEBVsG0opJoR9WZ6eeJqU0QwUbwCDbOiXYbKsl/KyonVJ/DLJoVG6bfGXYFGwMGDLDMOTp+/LilPasU2nwNiWAD8UPFhhkVGwgFwQYQIqd8EKAVFRKFr2qNzMzMOOwmMn379rXsu6WlJegT8p988omxdYW/ig27g42///3vln6l6enp+uijj/TTn/5Un/3sZwMewzR346qrrgqpBQcVG0Bwgn3fwgBxb6afy+PGjbOs0YoqeKEMD0/1VlTxCjb279+v1tZWR1Sfwyxawcbp06eNJ2DtCjak4C+8M1Vs+JqvIUm9e/dWt27dvNaampp8zokD7MKMDavm5mbj591wgg1mbDgDwQYQIqd8EKAVFRLBrl27VFlZaVlPxmoNj0iuijZVa3z605/uqJa46qqr1LVrV6/bDxw4YHxzGC5TtcbUqVOVmZmpO+64Q+vXr9df/vIXzZ8/P6TwKZRqDam9aqfziS+3280VdkAnwQYbpp9NTm5FZQo2TFVxVGwEj4qNc3bu3GlZs3O+hiR17drVOBfm448/DrqSC6knWsGGrzZUXbrYd8op2GAj1IoNiaoNxEciVGxs375dc+fO1ahRo3TLLbfEvQLe1IZq5MiRysjI8Ps4WlE5F8EGEKJge1UnO1pRIRGYqjXGjBmjqVOnxmE39ojkqmh/8zUkqUuXLlGv2jANDr/22mu9vv70pz+tn/zkJ/r444+DruIINdiQaEcFBCOSGRtOrdg4duyYzp4967XWt29fDR482PJv19bWRtVGkEIZHp7qwUYsKjYk33M2nPJ5BlamYOPw4cOWYduhiuZ8DQ9ThVewwYa/ig2JYAPxYQo2Os9aam1tjWrF4kMPPaSXXnpJ9fX1evXVV3XbbbdF7bmCEU4bKolWVE5GsAGEyClXONGKComgvLzcspbM1RpSZAN6AwUbUvQHiPuq2DDp3r17UFUc+fn5uuSSS0Ley4ABAyxrtKMCvAX7voXh4eeYAlLPzxvTh2uqNgI7cuSIJSzq16+fpf2LRyoHGy6XS/v27bOsxyrY8FWxQbDhDGlpaVEZIB6LYCOSVlRUbCARmU68my7citZ5mKamJr388steaxs3btQvfvGLqDxfMMINNkwVG7SicgaCDSBETvkgQCsqxNurr76qDz/80LKeisFGMFdFt7W1BRVsRHOA+BtvvKGmpiavtUGDBhnDlM5MVRxjx47Vddddp6eeeiqs/TBnA/CvsbHR0iM8IyPDOL+A4eHnmCoLPFcKm0JYKjYCC6UNlZTaMza2b99uWRszZozPkCcSoQQbqXihFsyi0Y4qmoPDPYIJNurr69XQ0OC11rNnT+NJz/OZgo3OxwHsZrqAxHRuKVrtqEyBpCQ9+OCDtrSoC4epFdXo0aMDPs5UsXHgwAG1tbXZsi8kLoINIEROGR7es2dPy4fKtrY22rwgZkxtqG666SZjv+hkEu7Jwy1btljemA0fPlwjR470WotmKypTG6pQ24KdX8Xx1ltvad26dWHvh2AD8C+U9yy0ojqHig37hdKGSkrtio0dO3ZY1vLy8qLyXL6CDVpROVs0go1EqdgIZ76GRMUGYq+urk6nT5/2WuvXr5/xfVq0gg1TyC1JLS0tevDBB6PynIGEW7HRvXt3478d7ahSH8EGECKnVGxIzNlA/LhcLmOwkezVGlL4Jw+3bt1qWetcrSFJn/rUpyxv6lpaWvTee+8Fv0kfTG2oOs/XiCVmbAD+hXLykuHh5/gLNkwVGwQbgYVasdG9e3fLoNAzZ87o5MmTtu8t1kwVG5dddllUnsvUcmjPnj2O+jwDq1QKNjqf8DW1oQo0X0Mi2EDsmd5jDRs2zDhLJtYVG5L03HPPqaKiIirP60+4wYbEAHGnItgAQuSkDwLM2UC8mEKNwYMH60tf+lIcdmOvcPvYB9OGyiMaczZcLpf++te/WtbjOcidGRuAf6G0m+ndu7elUvPUqVOO/L3vrxWV6cM1ragCCzXYkFK3HVWiVmykYgU6zEzBxp49eyI6pukEqek9bySo2EAqMZ1wHzZsWEwvLvVVseHx4IMPxrSV0+7du9Xc3Oy11r9//6B/lpjaUTFnI/URbAAhctIHAeZsIF5StVpDCr9iI5RgIxpzNkzVGrm5uUFdBRcttKIC/Av1PQvtqNoFakXVpYv3R6hDhw7xsycA00kZ01Wp50vVdlSxrNi48MIL1bVrV6+1+vp640noVL1QC1apVLERTLBBxQYSUSJUbAQKNnbt2hXTllSRVGtIVGw4FcEGECInVWzQigrx0NbWZmy75ORg4+OPP9a+ffu81rp06RJSxUakwYZpvkY821BJtKICAgm1jz4DxNv5CzYkqjbCEeqMDSk1g42PPvrIUnXSr18/y7wsO5mqNlpaWry+7tWrl7FCBqkplYMNUysqKjaQiEzBxvDhwxOmFZXHY489pr/97W9Ref7OIg02TBUbBBupj2ADCEFbW5vxxH6qBhu0okI8mE5a9OvXT+PHj4/DbuzXp0+fkNu9+KrW6HwVpocp2NixY0dEJ4RMFRvxbEMlUbEBBGJHxYYT52yYTiAECjaYs+EfrajaxbJaw8M0Z6OzVK0+h5mvYMPtdod9zEQJNmhFhWQRSiuqeFVseDzwwANRef7OampqLGujR48O+vG0onImgg0gBKYTBAMHDlR6enocdhN9tKJCPJhOWqTaVYShVm2Ygo2JEyf6vP+AAQOMbwLDrdrYvXu3PvjgA8t6vIMNZmwA/oUyY0OiYsPDVLFx/hWUDBAPXTjBRipWbMRyvoaHqWKjs1S9SAtmGRkZxtc83DkbR44c0alTp7zW+vTpo8zMzLCO54vpSnaXy6XTp09Lav8M0fn3XlpaGq2okJBCaUUVrYtLTYHknXfeaVnbuHGjVqxYEZU9nI9WVAgHwQYQAie1oZJoRYX4MJ20MJ3cSGahnjw0teby1YbKw84B4qY2VFdffbWxYiKWqNgA/GPGRnhoRWU/Zmy0i0fFBsEGTOxsRxWLweFSe0jh72r2cOdrSAQbiL1EaEVlOr919913G+c1PvDAAxG3rAvEFGxEWrFBsJH6CDaAEDgt2KAVFeKBig1vp06dCmlwuIedA8QTsQ2VxIwNIBA7Zmw4sRVVoGCDio3Q0YqqXaJWbNCKynmiHWzY3YbKw9+Fd+HO15AINhB7iTA83Nf3rmlgeEtLS1QHiR87dszyebhLly5UbCAggg0gBKFe+ZjsaEWFeCDY8GYKNUaMGGH8QHo+OweImyo2EjXYoGIDOCfUVlRUbLQznUA4/0QDFRuhY3h4+0khU1vHRJixkcoXasEsFYMNKjaQLFpaWiy/F9PS0nzO2Dh27Jja2tps3YPL5VJra6vXWmZmpnr37q1p06Zp/vz5lsc899xzqqiosHUfHr7aUKWlpQV9jIyMDEur4ra2NsKNFEewAYTAaRUbtKJCPDihFVUoA3rDqdaQ2is2Og8XP3DggD7++OMgd9nu3Xfftfzs69Gjh6699tqQjhMN3bp1s4RebrebD6KApNbWVjU0NHitdenSxe/JZCo22n8Hefq1e/Tq1Us9e/bs+HrQoEGWf8dTp04ZrxaG1NDQYOm/n5mZqYyMDL+PS7Vgw1StMXLkSPXr1y+qz0srKpjYGWyYAvBkq9jo06eP5X3zyZMnLT+7ADv4qtaQ2gMO0xxBu6s2AgWSDz74oPEz64MPPmh7yCJF3obKg3ZUzkOwAYTAacEGragQD1RseAs32OjSpYstVRu+2lCFcvVMNNGOCjALp8qUio3Ag8M9TO2oqNowC6cNlZR6wUY85mtI7SeqevXq5fc+qVyBDjOnVGwEG2xIVG0gdnzN1/CIxQWmgc5t9enTx9h6ateuXVFpSVVTU2NZC6UNlQftqJyHYAMIgdNaUXXr1s3yBs/tdhNuIKpMJy1SLdgIZXi4KdiYOHFiUM9jxwDxRG1D5WG6ool2VEDo8zWk9hL+zr/329rajCetUlWg+Roepg/bzNkwC6cNlZR6MzbiMV/DI1DVRipfqAWzZA02TEGzHa2oJIINxI6/ig3J/N95rCs2JOmrX/2qvvjFL1ru99hjj+lvf/ubrfvx1YoqVFRsOA/BBhACp1VsSMzZQOyZTlo4oRWVKdjYtWuXZb1bt25BVWxIkQ8Qb2trM1ZsJEIbKg/mbABmoc7X8AgleE1FwQYbDBAPHhUb7eJVsSEFDjZS+UItmEU72DC917WDryvZT506Zdw/FRtIRKYT7bEONoI9t+WrOsPuqo1otqL65JNPwtoTkgPBBhCCcK5+THbM2UCsObUVlenKHV9tqIJtAxVpK6qqqipLb+GhQ4fqiiuuCPoY0UawAZiFW2UaygygVBRocLgHA8SDR7DRLp4VG4EGiKf65xlY9e/f3/I91tLSElaFXiK0ojLN18jJyVGPHj2CPjbBBmIlUCuqeAUbpu/bSy+9VA888IBl/c0339SKFSts2cvZs2dpRYWwEWwAIQj36sdkxpwNxJoTWlFlZmZahoWeOXPGciIy3PkaHp/61Kcs38MtLS167733gnq8r/kaiYQZG4BZuMEGFRu0orKb6X2j6aRNZ6nUiurAgQOWzxHdu3fXmDFjYvL8/io2evToYTyhi9RnV9XGhx9+aFmLdbAR6XwNiWADsROoFVUsLi41BZK+Qu7vfve7xk4ADzzwQNiVXuerra2V2+32Whs6dGhYv5toReU8BBtACGhF1Y5WVIgmJ7SikoJrRxVpsCFFNmcj0dtQSczYAHwJt8rU6QPEI2lFdeDAATU0NERlX8mMio34VmtI/oONVL9IC77ZEWzcf//9amtr81rr2rVrUN/j4TAd98iRI8aKjVDma0gEG4idRG1F5S+QNLWeamlpsaUllV1tqCRzxQatqFIbwQYQpLq6OkuKnJWVpe7du8dpR7FBKyrEmhMqNqTA7V6ampr07rvvWu4TarAR7pyN+vp6bd261bKeDBUbBBuAvTM2nN6KyhRspKen044qSOEOD0+lYCOe8zUk/8FGql+kBd8iDTYqKyv105/+1LJ+ww03RLQvf6jYQCpI1OHh/n4fTJs2TfPnz7esP/fcc6qoqIhoL3a1oZKo2HAigg0gSE6s1pBoRYXYc8KMDSnwVdGmao1Ro0YZTzr6E+6cDVO1xtixYwP26Y41WlEBZnbO2HB6xYavtkmmqg2CDSs7KzZOnDihs2fP2rKvWIp3xYa/391O+DwDs0iCjdOnT+u+++4z3uZr3Q5UbCAVBJqx4eu/czuFc37rwQcfNL5PfPDBBy2VW6EwVWyEG2z06dPH8v7h1KlTxr8vUgPBBhAkpwYbtKJCrDmlFVWgPvamaomJEyeG/DymYGPHjh0Br3qtqqqyrCVaGyqJig3AF1pRhSfYVlQSczaCFe6MDSl1qjbiXbHRv39/n73KaUXlXJEEG/fdd5/x593//M//6Iorroh4b76kp6cb3/uZQmUqNpCIjhw5otbWVq+1Pn36eF3IZ/odGYsZG4Fm4/Tp08fYemrXrl0RtaSysxWVxABxpyHYAIIU7pWPyY5WVIg1J7eiClSxEWobKqn9hJzpjWGgqo1kGBwuMWMD8MXO4eG0ojIHG6aKDYINq3ArNqTUDjZiWbEh+W5H5YQLtWAWbrDx4osv6oknnrCsX3/99brnnnts2Zs/pp8fe/bssawRbCARBZqvIUW/FdXhw4ctbdb79+8fVJv1r371q/riF79oWX/sscf0t7/9Laz92NmKSjK3o2LORuoi2ACC5NSKDVpRIdac3Irq/JOHdgUbUugDxP/5z3/qww8/tKwnYrBBxQZgFu6MDV+hazK2/wlHKK2omLERnHBnbEjm3/+m9wmJbMeOHTpz5ozX2pAhQwJeGWs3gg10Fk6w0djYaGw11bVrVz366KO27c0f08+Pzr+jsrOzQ674JthALASaryFFP9gIp1rjfL6qM7773e+qubk5pL0cPHjQ8n2WkZHhdzZUIFRsOAvBBhAkpwYbtKJCLJ09e1YnT570WktLS0vJVlT+KjZqamos32cZGRnGQeDBCHWAuKkN1aRJkxIyYGLGBmBVV1dnuRIvKysrqCvxunfvbvxA7ZR2VLSistfJkyfV1NTktdazZ8+gf5+kQsVGvOdrePias+GECnSYDRs2zPJ7weVyqaGhwedj7rvvPu3bt8+y/sgjj0TUOiYUwQSjoc7XkAg2EBuB5mtIUu/evdWzZ0+vtZaWFp04ccKWPUR6buvSSy/VAw88YFnftm2bccC4P6ZqjUh/ljBA3FkINoAg0YrqnKNHj0Y0HArwxSltqCT/wYad1RpS6APEk6UNlUQrKsAk3PkaHoFmAKWqpqYmy5WGPXr0UO/evY33Hzx4sCVcbWlp0e7du6O1xaQTyXwNKTWCjXjP1/CgYgMmoVRtrF69Wr/97W8t6zfffLPuuusu2/fmSzDBRqhtqCSCDcRGMK2opOhWbZiCjVCrCL/73e8aL5578cUXtWTJkqCPY+fgcA+CDWch2ACC5NSKjfT0dOOJQ9pRIRqc0oZK8n/i0O5g46qrrlLXrl291g4cOKCPP/7YeH9TxUaiBhtdu3a1/Dfidrv5IApHi/RijECt8lJVKNUaHrSj8i+SNlRSagQbiVKxQbABk2CDjUOHDhlbUPXp00ePPPJIVPbmCxUbSGbBtKKSojvr1NSKKpzfBcuXL7dUlkjSz372M61atSqoY0Qj2DC1omLGRuoi2ACC5NRgQzKfDKEdFaLBdLIiFdtQSe2tpTp/gGpra9OhQ4e0detWy/0nTpwY9nN16dIl6KqNLVu2WE7u9e7dW1OmTAn7+aONdlSAt3Dna3g4tWIjnGDDNECcYOOcSAaHS6kxYyPRKzacUIEO34INNr7//e8bg8pHHnnEeIxoilbFRt++fZWenu61duLECZ0+fTrkYwG+BNOKSkr8ig1JmjBhgp544gnjbffee69eeeWVgMegFRUiRbABBMmpraik6F4tAJzPSRUbkvnk4Ycffqjq6mrLeiQVG1LwA8RN1RrXXnttRM8dbbSjArxRsREe0wmDQG2TmLPhX6TBRrJXbLhcLu3Zs8eynigVG127dg2pNRhSTzDBxpNPPqkXXnjBcr85c+bo61//etT25ku0gg2Jqg1EXyK0orKrYkOSZs2apYcffth42/z58/X+++/7fXysKjYINlIXwQYQJCdXbBBsIFacFmyYTh6++eablrXc3NyIf94EO0A8meZreJgqNgg24GTM2AiPXa2oCDbOcfqMDVO1xpgxYywDm2OhV69e+vKXv+y1dtttt8V8H0gsgYKN3bt3G1tQDR48OOYtqDyC+RkSTisqiWAD0RdsK6pkqNjwuPvuu41Dw10ul+bPn2+ZX+bR3NxsbI0cabCRlZWlzMxMr7Wmpibb/v2QWAg2gCAcOXJEZ86c8Vrr27evMjIy4rSj2KIVFWLFSa2oJHOw8e6771rWIq3WkIIbIN7a2kqwAaSASFtRmX42EWyY0YrKP6e3okqU+Roe3/3udzVnzhylp6frtttu03/8x3/EbS9IDCNGjLCsnR9sfP/731dTU5PlPo8++mhI38t2CvS8/fr1C3tvBBuIplOnThl/L5paJyXDjI3z/eQnP9GMGTMs69u2bTOGHpK5DdXIkSNtOc9GOyrnINgAguDkNlQSFRuIHSo2ZCzXtSPY+NSnPmX5udXS0qL33nuv4+uqqiqdPXvW6z45OTlx6QUeCmZsAN4ifd9iunKQVlRmpqsK9+3bpxMnTti2r2Tm9OHhiTJf4/zn/uUvf6ljx45pxYoVGjt2bNz2gsRgqtjwtE9bsWKFXnrpJcvt3/jGNzR79uyo782XQD9Dwm1DJRFsILp8VWukpaVZ1mNdsWFHN5JVq1bp0ksvtay/+OKLWrJkiWU9Gm2oPGhH5RwEG0AQnNyGSiLYQOyYTlY4LdjYt2+fZc2OYEMKPGcjGas1JGZsAJ1FY8YGFRtm3bp1M7Y8oR1VO6fP2Ei0ig2gM1OwcejQIW3bts3Yguqiiy6KWwsqj0A/Q8JtQyURbCC6gp2vIUUv2Dh16pTl/U5aWlpErag8BgwYoCeeeEI9e/a03Pazn/1Mq1at8lqLZrBBxYZzEGwAQYi0V3WyoxUVYsVUsZHKrahMb2RbW1u9vu7du7fGjx9vy/MFmrNhGhyeDMEGragAb9GYsVFXV6fTp09HtK9EF06wIZk/hNOOqp3TW1ElWsUG0Fl6errxyuZ7771Xbrfbsv7oo49aetfHWvfu3dWvXz+ft1OxgUQV7HwNyfy70o5gI9oX7U6YMEFPPPGE8bZ7771Xr7zySsfXplZUo0ePtmUfpmDjk08+seXYSCwEG0AQIu1Vneyo2ECs0IrKyq5qDcn/nI2DBw9q27ZtltuvvfZa254/WmhFBXiL9H1Lly5djFfupXrVhumEQTDBhmnOBhUb7Zw8PPyjjz6yvK/p16+fRo4cGZ8NAT6Yqjbeeusty9qCBQt00003xWJLAfkLSAk2kKhMwYbpBLxk/l1pxzkY03wNO6o1zjdr1iw9/PDDxtvmz5/f0XqZVlSwA8EGEARaURFsIDZoRWUV7WBjx44damxsNLahuvzyy32+2U4kVGwA5zQ2NqqlpcVrLSMjI+SfpaafT6k+Z8MUiAZzEp6KDd+cPGODag0kC9MA8c7GjBkT9xZU5/P3s5lWVEhUoVRsRKsVVazObd19993GoeEul0vz589Xc3OzMdiIZsUGwUZqItgAguD04eG0okKsOK0VVayDjQEDBhjfLL799ttJ24ZKYsYGcD673rOYPminesWGna2oqNhob63YOYTo2rWrMYz2JZmDDeZrIFmYKjY6e/TRR9WlS+KcPqJiA8ko0hkbR48e1dmzZyPagynYsLtiw+MnP/mJZsyYYVnftm2bbrvtNjU3N3ut9+/f37a9mCo2aEWVmhLnNxOQwKjYGKS0tDSvtYaGBp06dSpOO0KqclrFRs+ePQOeNLMz2JB8DxA3VWwkQxsqiYoN4Hx2zQVz4gDxcIMNUysqKjYib0MlJfeMDSo2kCwCBRuLFi3SddddF6PdBMdXsNGzZ0/jCc1gEWwgmkKp2EhPTzd+xom0asPUiiqa57ZWrVqlSy+91LL+6quvWtbsakMlUbHhJAQbQBCcHmxItKNCbDitYkPyX7UxduzYoE6qhcI0QHzDhg36+OOPLevJUrHBjA3gHLvmgpk+aEfaiurvf/+75syZo5ycHH3pS1/Svn37IjqenU6dOmX5HZSenm48ydXZsGHDLINsT548qb1799q5xaQTaRsqqb2NWvfu3b3WWltbLe3WEpEp2KBiA4nIX7Bx5ZVX6qGHHorhboLj62dJJNUaEsEGoiuUGRtSdNpRxbJiQ2q/QOSJJ55Qz549A97XrjZUUvvPiB49enitHT9+XA0NDbY9BxIDwQYQBKe3opLMf1+CDdjNaRUbkv9gY+LEibY/n6li491337WsXXvttcrIyLD9+aOBVlTAOXa9Z4lGxcaDDz6ol19+WY2NjXrttdf0ta99LaLj2SncweEeDBC3Mr1PDDXYkJKzaqOlpUU1NTWWdSo2kIj8BRuPPvpoDHcSPF8/SyKZryERbCC6QmlFJUXn4tJYV2xI0oQJE/TEE08EvJ+dFRsSA8SdgmADCAIVG+ZfqszZgN1MJypSPdjw92bW7jZUUnuw0bVrV681U6CULG2opPae7Z3/O3G73XwQhSPZ1YrK7oqNjz76SJWVlV5r77zzjhYtWhT2Me0U7uBwDwaIW9kVbCTjnA3TfI2RI0daKnuAROAr2PjBD36gz372szHeTXCo2ECyOXr0qKXaMDMz0+/vhVSo2PCYNWuWHn74Yb/3sTvYMFXDMGcj9RBsAAG4XC61trZ6rWVmZqp3795x2lF80IoKsUArKm/RCDbS0tKMVRudJUsbKg/aUQHtErVi47XXXjOul5aW6te//nXYx7VLuPM1PKjYsLJjxoaUnMEG8zWQTDIzM3XTTTd5rU2ePFn33XdfnHYUGBUbSDahzNfwiEawEY+KDY+7775b8+fP93m7na2oJCo2nCJmwUZZWZnS0tIi+lNWVhar7SIB3HPPPVq4cGG8t0Ebqv+HYAPRdvbsWZ04ccKynuoVG76CjaysrKj14g4UbGRlZekzn/lMVJ47WmhHBbSza8ZGrIINqf09X3V1ddjHtkOkwQYVG1ZUbHhjvgYS2QMPPKCZM2dqyJAhuvHGG/X000/He0t+Ratio1+/fkpLS/NaO378uM6cORPRcYFQ52tIsavYiGU3kp/85CeaMWOGZb1Lly4xqdgg2Eg9XQPfBYiPBx98MN5bkBT/H/yJwnRShFZUsJOvNlSdP1ykGl/BRjSqNTxMA8TPl0xtqDxMFRsEG3AiOys2unTporNnz3asedooBDMA8nxut9tvsHH27Fndfffd2rBhQ8j7tIvpREGkraicXrFhx/BwKTlnbFCxgWQzduxYPfPMM/HeRtCiFWxI7Rf4dH4P6XK5wvr5BXiEOl9Dsv/i0hMnTlguJOzRo4fxc1Q0rVq1SjfccIM++OCDjrVbbrnF9s/9BBvOQCsqIACCjXZUbCDanDhfQ4pPsBGoYiPZ2lBJBBuAh10zNiTzz6dw5my89tprlr7SncV73kY0WlHt2bNHTU1NEe0rmTm5YsMUbFCxAdjHFDynpaVF3IpKMr+npB0VIpUIragS5dzWgAED9Mwzz2jGjBkaPHiwvvSlL+n++++3/XlMraiYsZF6YlaxMWPGDG3bts2yXltbq1tvvVV5eXkqKirSxIkTO2577bXXVFxcLElat26dJk2aFKvtAh1oRdWOYAPRZjpJkerzNSTfb2jP/31ot0996lPKzs72WXWVKsEGMzbgRHa1opLag43OV7YdPHgw5CtiX331Vcta//79LeFjaWmpJkyYoG984xuhbzZCkQYbPXv21MiRI7V7926v9X/+858aP358pNtLSk4NNg4cOGD5/dq9e3eNGTMmTjsCUk9GRoZuvfVWPf/88x1rt99+uy3HZs4GoiERWlGZ5mvEYnC4SW5urtf3bzRQseEMMQs2cnJylJOTY1m/++67lZeXpzfffNPyC2T8+PG68847NXr0aM2aNUs1NTXGXzJANCVKqh1vtKJCtFGx4S2aFRtSe9XGyy+/bFkfOXKk7YPbYoEZG4DU2tqqhoYGr7UuXbqE3T7DFLyGM2fD1IbqP/7jP7R06VLL+j333KMrr7wy5mGA6URBKMGG1F610TnYqK2tJdg4TzjDw5OtFRXzNYDYKCwsVNeuXbVu3Trl5+fre9/7ni3HJdhANITTiipVKzZihWDDGeLaimrz5s2qqqpSUVGRz8AiKytLixcvVl1dnbZs2RLbDQJy3g9/X6jYQLSZrr50QrDRvXt3zZkzx2tt7ty56tevX1Sf19ecjWSs1pBoRQVI9leZ2tGKqra21quHskdBQYGWLVtmWffM24g1U8VGqCfhGSDuzakVG8zXAGLj6quv1qpVq3To0CE9/vjjGjFihC3HJdhANITTisruczCmc1vxqtiIhaFDh6prV+/r+Y8dO2aZM4LkFtdgY+fOnZJkubLMl71790ZzO0hANTU1uu222zRq1CjNnDkzonQ6XLSiakewgWgzXX3phFZUUvuVy3PnzlXv3r11yy236H/+53+i/py+5mykUrBBKyo4jZ3zNSR7KjZM1RqTJk3SoEGDVFBQYGw7FY95G5G2opIYIH6+trY2Y7jshGCDig0guRFsIBoStRVVql+0S9VG6kuI4eHFxcU+3/TX1tZ2zNmA8/zwhz/U73//e9XX1+v111/Xv//7v8d8D1RstBswYIAl7T5+/HjAYaBAsJzaikpqr5544okntH//fj355JM+21PZyQnBBhUbcBo752tI5ooNO4KN66+/vuP/P/bYY8ZWTaWlpfr1r38d0nNFwq5WVJ05tWLDVxuqtLS0kI+VbMEGFRtAciPYgN3OnDljfI8WqGKjT58+6tGjh9dac3OzTp48GdY+nFaxIZkHiBNspJa4BhvXXHONpPary0aPHq3CwkKVlZV1/CksLNTo0aM7rj7z3B/O0NbWphdffNFr7a9//at+9atfxXQfdl/9mMxMV9kxZwN2cerw8HgZMGCAvvSlL3mt3XLLLUlbkcaMDcD+KlPTB+5QWlGdOnUqYLDRpUsX/fSnPzU+/p577lF1dXXQzxeJaLWicmrFhl1tqKTkm7FhCjao2ACSB8EG7GY6kT5kyBClp6cHfKydVRtUbLQj2EgtcQ02cnNztWDBgo6vV65cqYKCgo4/K1eu7LitpKTE+GEBqcvXlVj/+7//G9MqAbuvfkxmtKNCNDm5YiNe/uu//ku33HKLevToofz8fD3yyCPx3lLYqNgAYjNjI5SKjddee01nzpzxWhs2bJilYuyqq66K67yNtrY240mrUCs2LrjgAvXu3dtr7fjx4/rkk08i2V5SMp10CTfYSKaKjR07dlj+mx8yZEjKXxELpBKCDdgtnPkaHnYGG07sRmIKNpz4viyVxb0VVXFxsWVoamclJSUx77OL+PP1geXQoUP63//935js4fjx42pqavJa69mzZ9SH+iYq08kRgg3YxanDw+Ppkksu0ZNPPqm6ujo99dRTMWmBFS3M2ABiM2MjlIoNU7XG5z//eeN94zlvw9d8jXDaJjFAvJ2dFRvJFmx0RrUGkFwINmC3cOZreNh5campYiPVg3daUaW+uAcbWVlZWr16tbZt26aSkhLNmTOn409paalqamoINRzK3weW5cuXdwyfjyYGh3ujFRWiycnDwxE5WlEB9leZDho0SN26dfNaa2xsDLq3c6A2VJ3Fa96GHYPDPUxzNpzYjsqpraiYrwEkP4IN2I2KjfihFVXq6xr4LrExfvx44wcZOFegK7GKi4v11FNPRXUPTvzB7w+tqBBNtKJCJLp27ao+ffp4/XfkdrvlcrmMH1CBVBSNCzKGDh2qvXv3eq0dOHDAeAL/fDt27NCHH35oWfcXbHjmbUybNs1y2z333KMrr7wyKp8X7Aw2qNho52t4eDiSqWKD+RpA8iPYgN1MJ9JjHWwcPXpUp06d8lrr06ePMjMzQz5WMjFVbNCKKrXEvWID8CXQB5aKigq99NJLUd0DwYY3WlEhmmhFhUjRjgpOZ3crKsn8wTuYORumao1rrrkmYDvPeMzbMJ0gCPckPAPE2zl5xkZnVGwAyYVgA3ZLhIoNp57bomIj9SVExYbL5dKePXt83n7y5En98Y9/1JVXXqnZs2fHcGeIp2A+sBQXF+vGG2+M2h5oReWNVlSIJlpRIVIDBgywvJ+gHRWcJFoVG50FM2cj1DZU5ysoKNB7771naT/lmbdhCj4iEe1WVFRstAs32EhPT1dmZqZXCzS3263jx48n1AUQvj7TUrEBJBeCDdgtEWZsOHG+hmT+d66vr1dLS4t69uwZhx3BbnEPNiorK3XbbbcFdXK0tLQ0BjtComhoaAh4n/fee08rVqxQYWFhVPbg1FTbF1pRIZqo2ECkTBUbBBtwErtnbEjhVWw0NTXpz3/+s2Xd1+Bwk8cee0zvvvuuqqurvdZLS0s1YcIE46DxcEW7FdVHH32k1tZW9ejRI6xjJiM7gw2p/UKHzrNdGhsbE+p9gqkN1ZgxY9S9e/c47AZAuEzBRmNjo9ra2pSenh77DSHpJUIrKqee20pLS9MFF1xgaT+1f/9+jRo1Kk67gp3i2orK5XIFHWrAeUwnOUeMGGFZKy4ujtqJK6f+8PeFVlSIJmZsIFIEG3Cyuro6ud1ur7WsrKyIT6qaKjYCBRuvvvqqZW3EiBEhzcfwzNswueeeeyyBRyRMJwjCDTZ69eqlCy+80LLutHZUds7YkJKjHZWpDRXVGkByomoDdkrUVlROqNiQzFUbzNlIHXENNnbu3Bl0qLFkyRLNmDEjyjtCIjF9WDFd6edyufTII49EZQ+0ovJGKypEE62oEClmbMDJojFfQzJ/8A7UiiqSNlTn8zdv43//939DPp4vpp8TkZyEZ4C4/RUbpgsdTO8b4slUscF8DSA5EWzALi6XS01NTV5rGRkZxs8tJnYFG6ZWVE65aNc0QJw5G6kj7sGGJM2ZM0fHjh3Tpk2bJEkLFiyQ2+2W2+3W888/L0l68803lZOTE7e9IvZMwca4ceN05513WtZXrFihd9991/Y9ULHhze5WVAcOHNCtt96qgQMH6utf/7rjrmbEOZ5e2Z1RsYFQmK6wpmIDThGtizHCqdiwK9iQ2udtmNpO/elPf1JbW1tYx+zMzlZUEgPEJXuHh0vJUbFhCjao2ACSE8EG7BLJfA3JvnMwVGx4I9hIHXENNjxuuOEGZWVladKkSZKkNWvWdNw2e/ZsLVmyRFVVVbaWnCPxmT6s9O3bV4sXL1ZGRobltuLiYtv3QLDhrV+/fpaWFk1NTZaex8G65557tH79ep0+fVrr1q3Tww8/bMc2kYRMoUbv3r3VpUtC/JpCkqAVFZwsGvM1pNBnbFRXVxsHKIcbbEjt8za6detmWf/oo4/CPub57GxFJTFA/MiRIzp79qzXWr9+/YyvYbCSIdgwtaKiYgNITgQbsEsk8zUkKjbsQLCR2hLijNGuXbs6/v/UqVNVV1enzZs3d6x5vpG3bt0a870hfnwFG0OHDtXixYstt7388stau3atrXugFZWVXXM2ampq9PLLL3utrV27Nqih8Ug9tKGCHWhFBSeLZcWGv1ZUpmqNf/3Xf1VmZmbYe+jSpYsmTpxoWbcr2IhFKyonVWzY3YZKMr8nSKRWVB999JFlP/369dPIkSPjsyEAETG9pyTYQDgima8hSV27djUGbaGeg3HyRbumVlTM2EgdcQ02PB9Qli5dqvvvv1+1tbWaNWuWJGnx4sWqrq5WZWVlVK7ER+LzFWxI0sKFCzV27FjL7Xb+t9LU1GT5gNKtW7eIruBLBXbN2fjjH/9oXH/77bdDPhaSn+n7nTZUCBUVG3CyaM3Y6N+/v3r27Om1dvLkSZ8XItjZhup8F198sWVt9+7dER9Xsr8VFRUb9rahkszvCRKpYoP5GkBqoWIDdom0FZVkT9WGqWKDVlRIBXENNsaPH6+pU6dKag833njjDX32s5+VJFVVVemKK67QjBkzOj6omU5kI3X5CzYkGas23n//feOQyXA4OdH2x64ej76CjXfeeSfkYyH5MV8DdmDGBpwsWq2opODbUTU0NOj111+3rEcr2IhmxUYkwcaIESPUq1cvrzWXyxVwNkmqiFXFRiIFG6Y2VMzXAJIXwQbsEmkrKinyYMPtdjv6/BbBRmqLeyuqiooKzZkzp+PrSZMmdYQd55s6dWrHDA44g+lKwPM/1OTn5+vGG2+03Ke4uNhvi4Rg0YbKzI5WVB999JH+9re/GW+jYsOZAgWZQDCo2ICTRfN9S7ADxE3VGqNGjbLlynVTSx87go1jx45Z5kH07ds3onkQkrlqwyntqJwYbFCxAaQWgg3YJdJWVFLkF5eaqjUGDhwY8XudZGFqRXXw4EGdPn06DruB3eIebGRlZWn16tXatm1bR2uqiooKLViwoOM+CxYsUEVFRZx2iHgJ5kSnqWqjpaVFjzzySMTP7+RE2x87WlH5qtaQCDaciooN2IEZG3CyWAcbpg/qpmDj85//vC17iFYrKrurNTxMczac0o7KdLIlkpklkvk9QSLN2DAFG1RsAMmLYAN2SYRWVE4/t9W1a1dj2y2qNlJD3IMNj/Hjx2v8+PGS2n+JrFixQm63W263WytWrDD+YkHqOnv2rPHDSr9+/by+njBhggoLCy33+9WvfuU1gD4c0epVnezsaEXlL9g4ePCgbT2zkTwINmAHWlHByaL5viXYVlTRmq8hRa8VlenEQKQn4SVnDxCPxoyNRK7YaGlpUU1NjWWdig0geRFswC52VGxEGmw4eb6Gh6lqg2AjNSRMsAGcz1e1RlpammV98eLFxqt0Ix0kHs1e1cks0lZU+/bt08aNG/3ehzkbzkMrKtiha9eulkDM7XbzQRSOEM33LcFUbLz99tuWD4jp6em2BRvZ2dmW7++mpqaI249Gq2LDyQPEndaKyjRfY+TIkZYLsgAkD4IN2KGtrc14IUisgw2nV2xIzNlIZQkTbOzbt09lZWWaO3eu5s6dq8rKSlVXVzvmyiZ4C+UkZ1ZWlrEl1YYNG/Tss8+GvQd++JtF2orKX7WGB+2onIeKDdiFdlRwosbGRrW0tHitZWRk2PZzNJiKDV/VGt27d7dlD1J0qjZi2YrKKZ9rohFsJHIrKuZrAKmHYAN2MF18kZ2dHfJsi0i7ZpjObTmtYsMUbHzyySdx2AnslhDBxrJly3ThhReqoKBA5eXlKi8v1969e7V161aNHj1ay5Yti/cWEWOhXr29YMECXXnllZb1SKo2GB5uFukvVYINmJi+5wk2EA7aUcGJot0+M5jh4dFsQ+URjTkb0WpFZarYsKN1VjKIxoyNRK7YYL4GkHoINmAHO+ZrSNFpReW0i3ZpRZW64h5sLFu2TEVFRX7vU1RUlBThxpo1azRu3DilpaUpLS1NhYWFjrkyy27htKUxVW3s3r077HCDig2zSFpRHTp0SBs2bAh4P4IN5zFddUkrKoTDVLFBsIFUF+2LMUwVG+d/WD9y5Ihxtlksgo1Erdjo06ePpRXR2bNnQ6pyTVa0oqJiA0h2BBuwgx3zNaTotKKiYoNgI1XENdjYt29fR6iRl5en0tJSzZkzp+P2Cy+8sONDWVFRkfbt2xeXfQbj/vvv16233ipJKi0t1ZIlS7Ry5UpNmTKFcCMM4QQbX/jCF/SlL33Jsl5cXBzWh16CDbNIWlGZqjVGjx5t+XdtbW3Vu+++G94GkZSo2IBdCDbgRNGeCxaoYuPVV1+13H7ppZdq9OjRtu1Bap9b0FmiBhuS+X2jE4INpw0Pp2IDSD2mYKOhoUFnz56N/WaQtEwnzuMRbFCxYa7YoBVVaohrsLFz505J7W/83nzzTc2bN0833HBDx+3Tp09XTU1NxxvDt956Ky77DKS6ulpLly7VnDlz9I9//EPz5s3Tww8/rG3btqmurk73339/vLeYdMIdJGyq2pCkRx55JOQ90IrKrHfv3srIyPBaa21tDerDpSnY+OIXv6hPf/rTlnUGiDsLMzZgF2ZswImi/Z6lb9++yszM9FpraWnp+N6KRRsqKXatqKIZbJhCqFTS0NCgU6dOea1lZmZa3juGKjMzU+np6V5rzc3NOn36dETHjdSBAwcs33/du3fXmDFj4rQjAHZIS0uzVN1JVG0gNHZVbERjxobTgg0qNlJXXIONvXv3SmqvxjAl4lJ7Uu6p6mhoaIjV1kJSXl4uSVq4cKHX+vjx47VgwQKVl5dTtREi02sdTLAxZswYLVq0yLL+7LPP6rHHHgv6+VtbWy1vWrp06UKw8f+E84v16NGjqqystKx/8Ytf1FVXXWVZpx2Vs9CKCnZhxgacKNozNiT/A8TjGWxEo2LDjhkbkjODjWi0ofJIxKoNUxsqqjWA1EA7KkTKrhkbffv2Vffu3b3Wmpqa1NzcHNTjTRUbtKJqDzbcbnccdgM7xX3GhiTt2rXL7+2vvPJKjHYSnoqKCknSpEmTLLd9/vOfl9Re1YHgmT6kmK6YMFm8eLHxw/d//dd/6Tvf+U5QHyhJtP0LZ86GqVrj4osv1r/8y78YKzYINpyFVlSwC62o4ESxqDI1taM6cOCA3nrrLcvz9+zZMyrBxsiRI9Wli/fHl7q6uohObse6FVWqBxvRaEPlYXpfYLowIpZMbaiYrwGkBoINRMquVlSS+aKLYKo2TBftpqenO+6i3Z49e1rej7jdbqo2UkBcg42JEydKkpYuXaqysjLLN5vL5VJZWVlHRYTn/olm+/btmjp1qvG23NxcSYlbbZKowm1FJbX/wPLVkurXv/61JkyYoBUrVvg9Bm2o/AtnzoavNlSSjMHGzp07+b5xEFpRwS60ooITRXvGhuS7YsNXtUbnAMIudldt0IrKXk6r2GC+BpC6CDYQKbtaUUnhz9lgvsY5pqoN5mwkv7gGG+PHj+8IBAoKCtS/f38VFBQYv87Ly9P48ePjttdAApVxJXrVSaKJJNiQpDvuuMNYQSNJJ06c0Pe//319/vOfV1VVlfE+VGz4F2orquPHj+tPf/qTZd0TbPTv31+XXnqp5XaqNpyDVlSwCxUbcKJ4VWz4Czaixe45G7SispfTgg1TKyoqNoDUQLCBSNnVikoKP9jg3NY5pgHiVGwkv7i3onriiScCfvDKzs7W2rVrY7SjxOJ2uy1/nCDSYEOSnnzySU2ePNnn7X/729908803G9tT8cPfv1BbUZmqNXJycjRlypSOrxkg7my0ooJdmLEBJ4rXjI0PP/xQW7ZssazHOtgIt2KjsbHRMny6V69e6tmzZ1jH64xgo51dQVEiBhtUbACpi2ADkWhsbNSJEye81nr27Bl2VWi4A8SZr3EOA8RTU9d4byA3N1fvvPOOVq5cqaVLl1puX7BggZYsWaKcnJw47C7+HnnkEZ+3FRcXx3AnsWU6ob1+/fqA81g6mzp1qnr37q3XX39dLS0txvv8+te/1rPPPqtrr71WV199tSRp06ZNlvvt2rUrpf/NQ2GaGfOnP/1Jp06dMt7fFEwOHTrU69/T9Ev52WeftZxwQOppbW21rHXv3l0/+tGP4rAbJDvTz5IPP/yQn99IaXv37rWsPfPMM+rVq5dtz2G6Mv3VV1+1rA0ePFi/+93vbHvezkzvBdeuXRv0AM3zmU5Qde3a1bafF6YrNf/+97+n9M8jUwWPXX9n03/nTz/9tLZt2xbxscNRV1enM2fOeK317t1bZWVlcdkPAHt98MEHlrX/n727D4+6vvP9/woJCSFAEkhQJCqiwUoQENHKTQtuC9azVlFosLU3lggLe7pWsl5nZWGP67ngsD2XhraeXSk01NN22wMrFt1aFXZ/hfUGV6hIJKJEKJUgSLiZcJMQyM3vD86MmXw/k7n73s3M83FdXi2fzHznA8lkZr6v7/v93rRpk/FEMdCT6TNJfn5+wq+HpurU9evXRz1H9u6771rWjh49mtbvRSL56KOPLGu/+c1vaIHuY5HGDHTnabARCARUVFSksrIyLV++XMuXLw87Yern1lM9RXtxu+OOOxI6blZWlmUtE6o2TCc68/LyEjrWzTffrNGjR+s//uM/IlYAXLhwQf/2b/+mvXv36otf/KJaWlostykoKEjo8dOR6d/C9G8mSe3t7dq3b59lPTh/Jsh0JajphADSjykQy83N9WAnSAf5+fmWtUROeAKpoqOjw3LxRlZWlq2hhnTphG1Ppg+CpooKO5muoE20Ksv0u8H0OyRRpvdL586ds+34fmR6P2jXz6LpvUGki2rcwEw+IL2ZqvciXSwJ9NSzWkMyv5eKlem1NNI5mGj7yNRzW6aOEKaW2EgtWV0eniXfvn275s+fr1deeSWlKzLGjBmj+vp6Y+Cwbt06VVVV6bnnntPs2bNtebxgshpLcpWqJk2aZCntfuONN3TjjTcmddzXX39dy5cvN1ZkdHfNNddY2hqsWbNG999/f1KPny42b96sOXPmhK392Z/9mTZt2mS57caNG/Xd7343bO2yyy5TQ0ND2FpXV5dKSkosFRp1dXUaMWKELfuGP33wwQe69dZbw9ZGjRqlnTt3erQjpLL29nZLiXdWVhZX4iBtNTY2Wnr6m15nk7V//37ddNNNYWt9+vRRZ2dn2NqLL76o6dOn2/rY3dXX11vmqI0YMcJYTRrNv/3bv+m+++4LW7v99tv1wgsvJLXHoLa2NsuJ7j59+qR1K5PZs2dry5YtYWsbNmzQV77ylaSP/fjjj2vVqlWWtb/+679O+tiJeOKJJ/TUU0+FrX3ve9/T//yf/9OT/QCw17p16/TII4+ErT344IP68Y9/7M2GkFJ+9atfaeHChWFrc+bMSbiqb82aNXr00UfD1ubNm6cf/vCHvd6vurpaP/3pT8PWfvCDH2jRokUJ7SOVbd26VXfffXfY2qRJk5iJnOI8nbGxd+9e1dfXa/PmzV5uI2mzZs2SZG7PE+w7nErVJ35gOgFlxyDhqVOn6pVXXtGTTz5pvOIvyNSrmSuwPmPq72i6ak0yz9cIDg3vLisrizkbGcp0lQTzNZConJwcy89PV1dXWp9IjGT//v164IEHNGjQIH3nO98xll8j9bkxX0MyV1b2DDUGDBjgaKghRR4e3tHREfexTEM37ZoHIV2qNi4sLAxb6+zsjPieKR04OTzcb1daMl8DSG/M2EAyTLMbTO+lYpXojA3TbC9mbHyGGRupz9NgI/hGP9WvoqysrJQkPfzww2EvdHV1dVq9erUqKystbXfQO9MgwJ4fDJOxYMEC7dq1S1VVVTHfh+Hhn4n1RbWjoyPmYEMyDxD/wx/+kMAOkUpMz3c7gkxkruLiYsvayZMnPdiJt5544gn967/+q6RL/WN7m9uF1GX6wOrExRj9+/eP+l7MyaHh3fdx+eWXW9YTGSBu+r2Q6FDPSEzvHzMt2EjX4eGmuTM9q6cApC6CDSTD1FY7mWDD9FpqukCjJ1Pb/Ew9tzV8+HDL2uHDhz3YCezkabDx+c9/XqWlpVq5cmVC5eN+MXbsWC1dulTbtm3T1KlTtW7dOi1btkzjxo1TaWmpli9f7vUWU0pXV5crJzqHDBmiVatW6Xe/+50mT54c9faZ+svfxHTCxPRB9qWXXrL0rx4yZIhmzJhhPO6ECRMsawQb6Y+KDdjNdGIy0R78qerEiROW9oAbNmxg3kgacrPPvylQ6M6NYEMyV22kUrBhCqPShekki10VG34KNlpbW/Xxxx9b1qnYANIHwQaSYQo2TBUDsUo02DC958jUc1sFBQWWi3Ta29tjqnyBf3k6PPzkyZNasGCBVqxYEQoBbr/99oi3f+ihhyKeEPXa8uXLNXLkSNXU1ISqABYuXKjq6mqqNeJk+oAycOBA9enjTA4XbE+1Zs0aLV++POKbFVpRfSY/P18DBgwIG0R18eJFBQKBsDeApmqN//Jf/kvE41KxkZkiPeeBRJkqNjIt2IjUxq++vl4TJ050eTdwkpvBxrBhw/Thhx9G/Lqbwcb27dvD1g4ePBj3cUwnBAg2Enfu3DnLINN+/frZ9ppuCja8akVlugJ2+PDhxgHnAFITwQaS4eeKjUxtRSVdukinZ9ego0eP2nYRBtznabCxc+dOrVixIvTnpqYmbdiwIeLt77jjDje2lbB58+Zp3rx5Xm8j5XnVlmbBggWaPXu2li9frtra2rCvzZw507FgJVWVlJSEBRvSpedwtGAjUhsqSRo5cqSGDh0a9oG/ra1Nu3btsgwsRfownZSgFRWSQSuqyMHG+++/T7CRZtyasSH1XrExfvx4XXnllY48bk9OVmzYOWNDyqxgw8k2VJL5ogevKjaOHj1qWcvkE0VAOiLYQDL8MGPjzJkzlgsO8vPzbW3znmouv/xyy0U6R48e1ZgxYzzaEZLFmVr4jpf99ru3p/rSl76kvLw8ffGLX9TTTz/tyuOnkmgvrC+//LLlezlo0KBeKzYkc9UGA8STs3v3bn3729/W5ZdfrtmzZ/uutzetqGA3KjZ6r9hAenFrxobU+wdyt6o1JGnEiBGWNVpRec/JNlSSv1pRma6AjdaqDUBqiRRsdHV1ub8ZpJSuri7bW1H17dvXEkh0dXX1evEW8zWsTBchmC5WQOrwtGJj4sSJlivjo90e6c8Pg4SnTp2qqVOnuvqYqSbanI1421AFTZgwQS+//HLY2h/+8Ie4Br0j3FNPPRXqtb9lyxZ973vf0/r1673dVDe0ooLdmLHRe8UG0otfZmy4GWyYKjZoReU905Wj6RpsULEBpL8+ffpo0KBBlt8zgUDAeBENEGQKNUpKSpJuVzhkyBBLG6Xjx49HfO/CfA0r03tZgg1/2bdvn37wgx/o5Zdf1qxZs/RP//RPvd7e02Bj7NixGjt2rJdbgA/5IdhAdKYPqt1PrsTbhiqIORv2OnfunGWA8Msvv6x9+/Zp1KhR3myqB1pRwW6ZXrHxxz/+0XiFlkTFRjryQyuq4uJiVy8IoRWVPzkdbJguevDTjA2CDSD9FBUVEWwgbna3oQoaMmSIDhw4ELbW25wNXqusCDb870c/+pH+5V/+RZL0y1/+MmqwQSsq+E7PBFriJKcf9daK6t/+7d8sL7D5+fkJBxt79+41/lwgukhDXl988UWXdxIZFRuwW6bP2Oitfd/x48eNH7aQutweHm7iZrWGdOnv1/N1oqWlxXiFZG9oRWUvp2ds+L1ig1ZUQPphzgYSYffg8KB4B4hTsWFFsOFvHR0d+sUvfhHXfXwTbNTV1WnZsmWaO3du6L9Vq1aprq7O663BZVRspIbeWlFFakOVkxO9SKy4uFjXX3+9ZZ2qjcSkQrBBxQbslukVG9HmElG1kV7cnLER6cSt28GGlHzVRktLi1pbW8PW8vLyNGDAgKT31l2mBxt2Vmzk5uYqPz8/bK29vV3nzp2z7TFixVWwQGYg2EAi7J6vERTvAHHT+41Mf60i2PC33n6eI/E82AgEApo7d67GjRunFStWaMOGDaH/qqurNW7cOC1atIgXjwxiCjZ6DkmC93prRZVoG6ogBojbZ9++fcb1d9991zcnNxkeDrtl+oyNXbt29fp15mykj6amJssQ06KioqR7OEfil4oNKfk5G25Ua0iZFWw4PTxc8k87KoaHA5mBYAOJ8EvFBsPDrQg2/M1UiR6N58HGkiVLtGHDhl5vs3r1ai1ZssSlHcFrVGykhkhXC2zbts3ywpCTkxNXsDFhwgTLGhUbiYlUsSH5p2qDVlSwGxUbVGxkCjfna0iXKhpmz54dtjZ79mxPTugmW7HhVrCRl5dnuUCns7MzoQ9ufud0xYbkn3ZUDA8HMgPBBhLh5IyNnuJtRZXpr1UEG/6WSMWGp8PDGxoatHr1akmXSuYXLFigkSNHhr5+4MABrVmzRk1NTVq9erWqq6tVXl7u1XbhEoKN1BCpFVWkNlQ9Wwf0hgHi9okWbPghNKYVFeyWyTM26urq1NLS0uttqNhIH27O1whatGiRsrOztXnzZn3hC19QTU2No48XyYgRIyxr8QQbphMBdg8ODxo6dKhlVlhTU5Pj3yu3ZUqw0dHRYTxZRMUGkH4INpAIKjb8a8CAARo4cGDYOYgLFy7o5MmTjlzggvgkcuGPp8HGa6+9JunSB7A33njDGFosXLhQEyZMUFNTk1577TWCjQxAsJEaIrWiSrYNlXSpYqNv3766ePFiaO3o0aM6ePCg8UQGzNrb29XQ0BDx6/X19Xr33Xc1fvx49zZlQCsq2C2TW1HF0raPio304eZ8jaBbb71Vt956q6OPEYtUaUUlXTqJ0PP1+NixYxo9erQjj+cVp4eHS/5oRWU6UVRSUhLTLDkAqYVgA4nw84yNTA82pEsXIvR873D06FGCDR9IyVZU0qV2VJECi7KyMl9cUQz3EGykhkgvqocOHbKsxxtsZGVlUbVhg0jzNbrzQzsqWlHBbjk5OZafoa6uroz4IGoKNnoOQ+7o6CDcSBNeVGz4Raq0opIyZ85GplRsmFpWUK0BpCeCDSTCL62oTEF8preikmhH5WcpNzz8hhtuiOl2wSfqxIkTndwOfIJgIzXk5eVZvi+dnZ2W2915550Jff8YIJ683tpQBb3wwgsu7CSys2fPWgbf9u/fn6sekbRMbUdl+j157bXXWtYINtKD2zM2/GTEiBHq0yf8o0xTU1PMJ7ndbkXVU7oFG62trZarH3Nycoy/i5Phh2CDE0VA5jD9DiPYQG/OnDljeT3Mzc21JeiPJ9g4ceKE2tvbw9YGDRoUV4vwdEWw4V8pV7ExadIkVVRUaOXKlRHbpdTV1WnFihVauHChxo4d6/IO4YWePYglgg2/iuWq0HirNYIYIJ68WIKNhoYG7dixw4XdmBFkwimZ2I6qtbVVdXV1lnXThSHM2UgPXrSi8pNkqjao2LCXW8O0/dqKiooNID1RsYF4OTVfQ4ov2CCEj4xgw79Sbnh4Y2OjnnjiCc2ZM0ejRo3SwoULdcstt4S+vmPHjtBw8VGjRmndunWWY0ycOJHAI81wojN1lJSUaP/+/b3eJtFgg1ZUyTO1orr++ustgccLL7wQ9rvXTczXgFNMV9ile7Bhqta45pprdMstt6i2tjZsnYqN9JDJraikSz/fPd+HHDx4UOPGjYt6X4INezl5Iqc7P1RsuBXiAPAewQbi5dR8DenSz2NOTk5YJcbZs2d1/vx59evXL+y2zNeIjGDDv1JuePjmzZtVVVUV+vPq1atDQUZP1dXVxvXa2lqCjTRj+nBSWFjowU4QTbRyyi9/+csJt3UYOXKkhg4dGvaC3NbWpl27dummm25K6JiZxlSxMWPGDMv6iy++qOXLl7u1rTAEG3BKJraiMgUbEyZMMA4opmIjPRBsJF6xYbrCkWAjcZkUbFCxAWQOgg3Ey6n5GkElJSWWk/AnTpzQ8OHDw9ao2IiMYMO/Uq4VFWBCxUbqiHbyJNFqjSDmbCTHVLHxne98x7J28OBBvfHGG25syYLnO5xCxcYlEyZMUEVFhWX90KFDaf/vkQkyecaGZH8rKmZsJC6Tgg0qNoDMQbCBeDn9emh6r2Jq30PFRmSm12yCDX9IuVZUV155pSorK5M+RiZavXq1Nm/erJdfftlScpbKTp8+bRkkPGDAAGVnZ3u0I/QmWsVGssHGhAkT9PLLL4et/eEPfwir9ILZ/v371dbWFrZWUlKi66+/Xl/5ylf0yiuvhH3txRdf1JQpU9zcoiQqNuCcTJyxESnY6Nu3r2644Qbt3bs37Gv19fWaOnWqW9uDAzJ9xsaIESMsawcPHozpvrSispfphIATVQx+nbFBsAGkJ4INxMvJVlRS7HM2TO8zeK26hIoNf2ptbbW8p+vbt2/U+3kabMyYMUMzZszwcgsp69SpU/rDH/6g7373u/r1r3/t9XZsw9XbqaW3YGPatGlJf6BlzkbiTNUao0aNkiTdfffdxmDjBz/4gSt7647nPJySaRUbn376qfFK9QkTJkiSKioqCDbSzOnTp3X+/Pmwtfz8/IwKh2lF5R+ZXrFBKyogPWVnZ2vgwIFhJ9u6uroUCASMoQfgdCuqWIMNUwhPxcYlBBv+ZKrWiHYxtUQrqpT30ksvGfvopypOcqaW3q4KTbZaQzIHG3v37lVzc3PSx053pt8L3YONng4fPqytW7c6vS0LKjbglEybsWGq1hg7dqzy8/MliTkbaSjT21BJ5mDj4MGD6ujo6PV+Fy5c0NmzZ8PWsrOzHTtJlZeXZ5kX19nZmVAfYb/KpGCDig0gs1C1gXg4/XpoOtEbaysqXqsuGTRokAoKCsLWzp8/n9YXwaWCRGcH+jLYqKurU11dnRobG73eSkr47W9/6/UWbEOwkVp6S0/tCDaKi4t1/fXXW9ap2ojOVLER/LccNGiQvvrVr1q+/uKLLzq+r55Mz3mCDdgh0yo2IrWhCjLN2aivr3d0T3BWpg8Ol6T+/fsbr7qLVrXhZrVGkCl0SqdgI1NaUR0/flwXL14MWzOdIAGQPgg2EA+/tKKiYqN3VG34j+l9sW8rNhoaGrRu3TrV1dWFrW/cuFFDhw7VuHHjNG7cOF155ZWaO3euGhoavNhmyiDYgFci/ZKZPHmybfNvGCCemN4qNiRz1YYXwYbpZATPedgh02ZsRAs2qNhIP5k+XyMoUtVGb9wcHB6U7u2oMqViw3SiiDZUQHoj2EA8vBgeHuuMDYKNzxBs+I+p8siXFRurVq3SqFGjVFVVpZ07d4bW161bpzlz5lgSmg0bNmjKlCmEG734wx/+YLw6OxWZWgxxktO/Iv2SsaNaI6j7ibkgKjaiMwUb3atf7r77buXkhI9ZOnbsmLZs2eL43rqjFRWckmkVG7t27bKsdQ+Gr776aksrnLNnz2r//v2O7w3OoGLjkkTmbLg5ODwonYON5uZmnTt3LmwtPz/f+Hs4WV4HG6aTHrT2ANIbwQZideTIEXV2doatDR48WP369bPtMWIJNjo6OozvE3m9+gzBhv+kRCuqdevWqbq62rLe2NioqqqqiPdramrSsmXLnNxayvvXf/1Xr7dgC9MHk54nYuAfkSo27Aw20mmA+Mcff6z7779fhYWF+ta3vuXYCcXDhw9bnksDBgzQVVddFfpzfn6+L6o2aEUFp2TSjI2PPvrIcoVL//79deONN4atmdpRUbWRupixcQnBhvfcqtaQzMHG2bNnLSeSnMJ8DSDzEGwgVm68HsYyY8P0/qKkpETZ2dm27iWVZUqwsXv3bn3729/WoEGDNH/+fF9fFO/74eGBQECPPfZY6M+lpaW64YYbJEn/+Z//GXbbffv2qaurS88991xobcOGDczc6EW6tKOiFVVqycnJ0de+9rWwtbvvvlsjR4607TEmTJigvn37hq0dPXo0apsJP3rsscf0u9/9Tl1dXXrhhRe0fPlyRx7H9GLVvQ1VkB+CDVpRwSmZ1IoqWhuqIOZspBcqNi4ZMWKEZS2RGRu0okqcm8GG5G3VhluzRAD4B8EGYuX0fA0ptooNQvjoMiXYeOqpp7Rp0yZJ0vr16/XjH//Y2w31wvcVGx9//HFok6Wlpdq3b58mTZokSXr66adDt1u4cKHKy8slSbNnz9a0adNCX9u7d69b20056dKOimAj9fzVX/2V5s6dqyFDhujOO+/U//7f/9vW42dlZaVF1UZjY6MlgNy4caMjH8SjtaEKuueee5Sfnx+2durUKf3ud7+zfU+R0IoKTsnJybH8LHV1daXlB9FYgw3mbKQXZmxcYteMDSo2Euf2yX4vgw1OFgGZh2ADsXIj6I8l2GC+RnSZEGw0NzeHQo2gX/3qV95sJga+Hx7efZ7GkiVLQi8OgUBA27ZtC33tvvvuC7vft7/97dD/P3TokLObTCGm1Dcd2lERbKSe8ePHa+3atfrjH/+o9evXG9/4JSsdBohv3brVuG7qi5+saIPDg7Kzsz2v2qAVFZyUKe2oqNjITFRsXEIrKu+5XbFhep9gulDCCVRsAJmHYAOx+uSTTyxrXgQbhPDRZUKw8ac//cmy1t7ebvz58IOUGR4uhc9M6FmFEWxPhd6ZTlKmQzsqgg2YpMMA8e4BbndOBDSm6i1TxYZ0qWqjpxdffNG1XtW0ooKTMqUdVTIVGw0NDWptbXVkX5nigw8+0Le//W0VFhbqoYce0nvvvefK4zJj45LS0lLLie6WlhbjyfYgUysqgo3EZVIrKk4WAZmHYAOxcuP1MC8vz/K+p6OjI+wzDhUb0WVCsBGpgtkUwPmB71tRddfc3Bz6/y+99FLo/1dUVKisrCzstgcOHHBtX6nEdJIyHdpREWzAJB1aUfmxYkOS7rrrLstz7OzZs65VgNGKCk4yVWykW7Cxa9cuXbhwIWztsssuM84dGDRokHGdqo3k/OAHP9CmTZvU1dWlDRs26B//8R9deVxaUX0m3nZUpooNZmwkznQiJ5NaUVGxAaQ3gg3Eyo0ZG5K5PU/3izZM7y8I4cNlQrBhqtiQ/Bts+H54ePdKjNraWkmXXgzWrFkTWp81a1bYfXp+feLEic5uMoUMHjzY+O+R6u2oCDZgMnLkSMsJgba2NkdCASfs2rUrYrmf3X+HEydOWN7I5OTkRAw2pMhVG26gFRWclAmtqGKt1ggytaNizkbizpw5o40bN4at/frXv3b8cdva2sIuFJKkPn36xPTmPx3F246KVlT2Mp0ISNeKDdPflZNFQHoj2ECs3GhFJZkvxuh+Uth07oGKjXCFhYXq379/2Fpra6vl/XUqS6WKjebmZrW1tYWtFRQUqKCgIOp9PQk26uvrNX36dE2dOjWs1OQ73/lO6P9v37497OulpaUaO3asW9tNCV/96lcta6nejsr0S4RgA1Jqz9mIVK0hXUrR7bwyIJ5qjSDTnI0XXnjB8sJit3PnzllaXvXv3199+/Z19HGROTKhYiPeYMPUjoqKjcQ1NDRY1rq6uhzvXcsH1nDxBhtetKLKy8sLa8crSZ2dncay+1STKTM2Tp8+rXPnzoWt9evXz/haAyB9EGwgVm69Hkabs0HFRmzSvWojUsXG4cOHXd5JdIlWa0guBhtFRUWqqakJ/Xnbtm1hH6QrKytVXl4uSVq3bp0mT54c9vV/+Id/cGurKeOuu+6yrKV6OyrT1VY9PwQiM6XynI3egg3J3qqNeOZrBN1xxx2WF40LFy44XrVBtQaclgkzNqjY8Fak91xOX4lPn/9wphZr8VZsON2KSjKHTwQb8fOqYoPnHZCZCDYQi9OnT1su1M3JyXHkwpNowQYXwMQm3YONVKrYSHS+huTyjI3vfve7WrhwoWW9srJSP/nJTyLeb+nSpZo3b56TW0tJ5eXladeOilZUiCRV52y0tbVFHBweZGewkUjFhmSu2nD6dwnzNeC0dK/YOHPmjLHaIt5gg4qNxJkqNiTzB0o7MRQyXDwzNjo6OownpJyu2JDSsx3VsWPH1N7eHrZWVFRkae9gJ6+CDdPJDuZrAOmPYAOxML3vuPrqqx15LNOV7N2veOd9YmzSPdhIpRkbpmDDdxUb0qUXhGeeeUa7d+9WbW2tamtr9eabb2r9+vXGF4uFCxfqzTff1PLly93cZkpJt3ZUBBuIxBRs7N271/c9ELdu3Wppt9STnS21TMFGtIoNKXI7qrNnz9qyLxOe73Baus/YMP3uuO6663q98vz6669XXl5e2Nrx48d9+QY3FUQKNqjYcFc8ragizdfIysqyfV89pWOw4Xa1huRdKyqed0Bm6tu3rwYMGBC21tnZ6fvPoXCX6X2H6f2JHXqr2DDNicjJycnYOWy9Sedg4/DhwxFbi/vxc5+pFZUvKzaCxo4dq3nz5mnevHmaNGmS5eszZ85UV1eXnnnmGePX8Zl0akd15swZywnggoIC5eTkeLQj+ElxcbHxBL3fqzaiVWtI3reikqQ/+7M/s5yI6OrqcrQdFRUbcFq6V2zE24YqiDkb9on0fsuLio1MPsE6YsQI9ekT/rGmqanJGKB7MTg8KB2DDS+qGPzUioqKDSAzULWBaPwSbFCtETvTe+d0CTYiVS5L/gw2UqYVVazKysq83kLKSKd2VFy9jWhScYC4ab5Gz+HYTU1NOnDgQNKPde7cOX388ceW9VhaUUnmqg2CDaSydJ+xkWiwwZwN+/ipYiPTP7TG2o6KYMNeXlRs+KkVVSYHikAmIdhANKZgwzQDzA69BRtUF8YunSs2IrWhkqSWlhbfdTFIieHhcE66tKMi2EA0qTZAvLGxUXv27LGsjxs3zrJmR9WGqQ3Vtddea2k7E4kp2Pjd737n2IlgnvNwWiZWbJgC4J6o2LDH/v37dfHiRePXnK7Y4EOrVaztqLoP1wxyY3C4RLBhF9N7Ba9aUVGxAWQGgg1E42bFRm8zNqjYiF06Bxu9VWxI/qvaSLuKDcQnXdpRcZIT0aTaAHFTtcZNN92kz3/+85Z1p4KNWKs1JOkLX/iCrrrqKsu6U1UbVGzAaek8Y+OTTz7RoUOHLOtUbLint/dZDA93X6zBBhUb9vKiFZXpvQIVGwCcRLCBaEwnkr1oRcXFL7FL52Cjt4oNKTWCDSo2Mki6tKMyDd8i2EB3EyZMUG5ubtja0aNHo6bRXjHN15g+fbpuuukmy7odLbUSna/RnZvtqEwnIQg2YKd0bkUVqQ1Vz1Z3JqZgg4qN+EVqQyUxPNwLBBve8EvFhlczNjL9eQdkCoIN9Kajo8N4TsKLVlRc/BK7dA42Uq1iI+WGh8N+6dCOiooNRJOVlZVS7ahMFRvTpk0z/h38ULEhmYONLVu2OHLixVSxwXMedsrJybGEZV1dXWnxQTTR+RrSpQ83Pd/Id3R0EG7Eqbdgg4oN95lOHpg+1NGKyl6ZFGx4UZ0CwB8INtAb04UUw4YNU0FBgSOPV1xcrOzs7LC1M2fOqK2tzfi+ghDerLi4WP369QtbO3funCvvKZyWasEGraiQFu2oTL88CgsLPdgJ/CxVBojv2rXLcmIrLy9P06dP13XXXWe5OvTcuXPGeRzxsKNi47bbbtO1115rWXeiaoNWVHBDurajSibYkJizYYfe3mMFAgG1tbU58rjNzc1qbW0NW8vPz8/490xUbHjDFGx40YrK6RkbbW1tloq/Pn36cLIIyBAEG+iNm9UaQZGqNkwX12T6xS+9SceqjfPnzxvfn3V3+PBhl3YTG4aHIy3aUVGxgVikSsVGpGqNPn0u/dq1u2qjvb3deJIt3ooNSbrnnnssa04EG7SighvStR1VssEGczaS11vFhuRc1QbtcMxMwcbBgwfV0dERtuZlsJGXl2cJoDo7O41XqaWC9vZ2YzDjdMVGfn6+pTVpW1ubzp8/79hjMl8DyGwEG+iNm4PDg0wnfU+cOEHFRpxMwYbTlddOi6VVu58qNpqamtTV1RW2VlxcHFOLZYlgI62kejsqgg3EIlUGiJuCjenTp4f+v2nORjLBhinUGD58eEJX8JraUW3dulWNjY0J7S0SWlHBDaaKjVQPNj788EPLh+mBAwcaqzAioWIjOcePHzdeWdSdUx+KaENl1r9/f+OH054nG0ytqNwKNiTz9ypVgw3T1YBDhw5VTk6O44/tdjsq0/OZNlRA5iDYQG+8CDZMFRvHjx+nYiNO6VixEW1wuOSvYCOZag2JYCOtpHo7KoINxGLkyJGWF+a2tjZbZlTYpa2tLeLg8CC7gw075msETZgwQTfccINl3e4KMFpRwQ3p2Ioq2WoNiYqNZMXy3sqpFkNUbEQWqWqjO9Pz360ZG1J6taMyffB3ulojyO12VFRsAJmNYAO9MQUbXrWi4gKY+KRjsJGKFRs9xTpfQyLYSCup3o6KYAOx8nvVxtatW9XZ2Rm2VlZWpjFjxoT+bAo23nnnHbW3tyf0mKZgI975Gt2ZqjbsbkdFKyq4IR0rNuwINkwVG4cOHUr5fxu3RGtDJblbscEJ1ktimbPhZSsqKb2CDS8Ghwf5oWKD5x2QOQg20BvTiWQvKjYOHz5smcPWv39/zqv1Ih2DDVPFRs9h883NzTp79qxbW+qVKdigYiODpXI7KoINxMp0As9PA8SjVWtIl9pElZWVha11dXUlXLVh13yNIFOw8cYbbxivRkkUrajghnScsWFHsJGbm2uszKIdVWz8VrHBlXiXEGy4K5OCDdNJDlpRAZmDYAO98aJiw3Ti9+OPP7as8R6xd+kYbJiCNtPFfn6p2jC1oqJiI4Olcjuq5uZmyxonOWGSChUbPU2bNs2yFqlqIxF2V2zceOONGjt2rGXdzqoNKjbghnSr2Ojo6LAl2JCYs5EMLys2uHI8MtNJhO4nG06dOmWpqBw0aFDMwwntkE7Bhpcn+91uRcXzDshsBBuI5MiRI2ppaQlbGzRokOOBgqliw3Simteq3mVKsGG68OTw4cMu7CY6WlEhTKR2VKlQtUHFBmJlCjb27t3rwU6sGhsbtWfPHst6z4oNyXwi0s6KjWSCDclctWFnazsqNuCGdJux8c4776ijoyNs7YorrtCVV14Z97GYs5E4v7Wi4mq8S6LN2PB6cLiUXsFGJlVsMDwcyGwEG4jEdBLZ6WoNyRxsmE4Q8x6xd+kYbJhaUV199dWWNT9XbNCKKsOZ2lGlwpwN04eRwsJCD3YCvysuLjb+bPjhZKWpWuOmm24yXilh1wDx/fv3q62tLWxtyJAhSb+JMQUbb7/9tuVq10S0tLRYTs7m5+e7etUsMkO6VWzYVa0hUbGRqAsXLmj//v1Rb8fwcPdFa0Xl9eBwiWDDLn5oRcXzDsgcubm5KigoCFvr6Ohw9PcOUoOpDZXT8zWkyMPDe+K1qnemf59UDjaamposF5AWFBRo5MiRltv6JdigYgMWqdqOiooNxMOvffNjma8RZAo29u7dG/cbZCeqNSTpc5/7nPEFxXQiI160oYJb/Pq7IlF2BhtUbCTGVK1hutqLig33lZaWWl5LWlpaQq9bXs/XkNIr2Mj0VlRUbACZhaoNmHgVbJiuaDe1d+c9Yu+GDBmi3NzcsLUzZ874ZrB2vCJVa1xxxRWWdT8HG1RsZLhUbUdFsIF4+LW9TKzzNaRLfwfTgO94qzZM8zWSGRzenenKSzuCDdPJB4INOCETKjZM7flicfXVV1uq386ePRtTNUImMwUb5eXlljUqNrzRWzsqgg17ZUrFRldXFxUbAAg2YOTF4HDJXLFx7tw5yxqvVdGlUzuqSK3R/BxsMDwcRqnWjurs2bOWtjT9+/enLQ0i8uPJyl27dllOOOXl5UWs2JDsGSDuVMWGZD5BYccLIEEm3OLXEDQRgUDAGGQmWrEhUbWRCNPv3NGjRys/Pz9srbW11XjlXDI+/fRTdXV1ha0VFxcrLy/P1sdJZb21ozK1aKAVVWJaWlosJ/RycnJcuzLUzWDDdHLDdIUngPRGsAET04lkr1pRnT9/3rJGxUZ06R5sRKrYYHg4fC3V2lFxkhPxMl1h6fXJykjVGn36RP5Va8cAcSo2gMjSqRWVKfT83Oc+l9Q8KuZsxC9SxYbpiji721HRhiq63oINP1Rs5OXlWZ6znZ2dxg91fuZltYbkbrBBlRQAiWADZl61ourXr58GDBgQttbz4heJ16tYpFOwYWpF5eeKjQsXLlguBOvTp09cFx4RbKSpVGtHRbCBePmxYsMUbPRWrSHZM0Dc7YoNZmwgleTk5Fh+trq6ulLyg6id8zWCqNiIX6Rgw40r8TnBGp2p/YOfgg3JHEYRbMTHzRkbzNcAIBFswOr06dOW1+/s7GxXWlFJsVWdcgFMdOkUbESq2Bg+fLhl/fjx42pra3NhV5ElW60hEWyktVRqR0WwgXj5rWKjra0trsHhQaZg409/+lPML6SHDx+2JNwFBQW66qqrYrp/NG5WbPCch1PSpR2VE8EGFRvxM4XJo0aNomLDJ3qbsWFqReWXYCPV2lGZ3gu4ebLf61ZUBIpA5iHYQE+R5hm4JZYBy7xPjC6dgo1IFRtZWVm+rNpIdnC4RLCR1lKpHZWpBzUnOdEbv1VsbN26VZ2dnWFrZWVlGjNmTK/3y8vL0/jx4y3rsVZtOFmtIdGKCukhXdpRuVWx0dDQoNbW1qSOm64aGxt19uzZsLWBAwdq+PDhxhOdVGy4L95WVG7P2JDSI9gwfeDPpFZUVGwAmYdgAz151YYqKNp7mKKiIvXr18+l3aSudAk2Ojs7I1ZsSDJWbXgdbCQ7OFwi2EhrqdSOiooNxMtvFRuJVGsEJTNA3Mn5GhKtqJAe/BaEJuLjjz+2vPHMzs5OOtgYNGiQ8coyqjbMTG2ogr9zTSer3ajYINgIN2LECMtsq6amJp0+fdrXrahSLdjIpFZUVGwAkAg2YOX3YINqjdikS7BhCjUuu+wyFRQUSJIvB4hTsYGoUqUdlekkZzLDUJH+/HaiMtLg8FgkM2fDFGykasUGYSackg6tqEy/EyZMmGA5gZsI5mzEzlQlV15eLsl8otPuYMN0PD60WkVqR0UrKvt4HWxQsQHAbQQb6MkUbLjZiipasEEIH5t0CTYitaEK8mMrKio2EFWqtKOiYgPx8lPFRmNjo/bs2WNZj7Viw3TFtV9aUZWWlio3Nzds7fTp05ZWLPGiFRXc5LcgNBFOtKEKYs5G7CINDpfcOVlNK6rYRGpHRSsq+5g+8DNjA0A6I9hAT15XbES7sp2LX2KTLsFGb22oJH8GGwwPR1Sp0o6KYAPx8tOJSlO1xk033RTzh96xY8dael82NTXpwIEDUe/rdCsqyZmqDZ7zcFM6zNhwMtigYiN2vbWiYni4f5hOKnzwwQe6ePFi2Fr//v096T2dDsGG1xUb2dnZodYKQV1dXY60oyJQBCARbMDKdCLZT62oeK2KTUlJifr27Ru2dvr0abW0tHi0o8SkYsUGragQk1RoR8VJTsTLT61lkpmvEZRI1caJEycsJ0JycnJSItigYgNu8lMQmigqNvyht1ZUVGz4h+mkgulCAC+qNSSCDbu4VbXhdXUKAH8g2EB3HR0dxmDDT62ouPgldulQtZGKFRu0okJMUqEdFcEG4jV48GBlZWWFrZ0+fVrt7e2u7yWZ+RpBiczZcKNaQzKfqEj2BZBgA27yUxCaiPr6estzpqioyLa2c5/73OeUl5cXtnb8+HHP3+j6zZkzZ4wD9tyasdHW1mY5gdKnTx8+tBqYTiqYPux5MV9DSv1g49SpU2ptbQ1bKygocH0+nhvBxsmTJ3XhwoWwtYEDB2rAgAG2Pg4A/yPYQHemNlTDhg2zVBM6iYoN+6RDsJFIxUY6DA/PsWsz8K9gO6qdO3eGrf/2t79VdXW1R7sKR7CBRAwePNgyCPTUqVNxJ7zJ2LVrl+XEVV5eXtwVG6Zgw3SFdndOz9cIohUVUp2dFRvt7e0KBAI6deqUAoFAxP9MXz9//rwmTJigb3zjG/rGN74RcwscJ6s1gkaPHm0JU+vr641vgDOVqQ3VddddFypdz8vLU3FxcdjPVldXlz799FNbPlgyODx2poqNQ4cOWdYINhLjh2oNyXxBhN2tqJivASCIYAPdeV2tITFjw07pEGxE+5k0fa47cuSIOjs71aePN3UPdlRsEGxkiK9+9auWYONf//VffRNsNDc3W9Y4yYloiouLLcHGyZMnXQ02IlVrxPvCkEgrKi8rNmhFhVRix4yN9957T6tWrdJzzz2X1F7efvttvf3221qyZEko4Ljlllt6vY8bwUZFRYXld87777+vGTNm2Po4qay3NlRBQ4cOtfxsHTt2zLFggxOsZqZgw/Tv51Urqry8PBUWFoa9/+3s7FRTU5Or72ES5ZfWTG5UbJh+bmhDBWSmvLw89e/fP6zvfnt7u86cOcPnmAzk9eBwiYoNO6V6sNHc3Gw5N5aTk6Mrr7wy9Oe8vDyVlpZaqiQ++eQTlZWVubLPnhgejphFakfV8wffK6YPIW6XsyP1+GEgsCnYiLdaQ7p01W/Pv8+5c+e0Z8+eiPdJt4oNPhDAKclWbLz33nuaM2dO0qFGd62traqtrdWXvvQlffnLX9a6det0/vx5421NwcbNN99s214k5mzEwlSx0TPYcLIdFYPDY9e/f/+YTj57VbEhmb93pg93fuSXig03gg0qNgB0R9UGgvwQbAwePLjXCyp5nxi7VA82orWhCvLTnI2zZ89aWpv269cv7vNCBBsZory8XBUVFZb1t956y4PdWNGWBonwum9+W1ubLYPDg+Kt2jBVbKRKsGGq2OA5D6ck87siGGok+zPfm7fffluPPPKIrr76ai1evFg7duwIfe3ChQvG3wNOVGz09P7779v6GKnOFGz0rJJzssUQFRvxieXkgt+CjVRpR5VJwQbPOwDdEWwgyBRsuN2KSuq9aoNgI3apHmxEGxwe5Kdgw45qDYlgI6PcdtttlrXt27d7sBMrgg0kws6++YnYunWrOjs7w9bKyso0ZsyYhI4XzwDxc+fO6eOPP7asp0IrqpaWFsuQ9379+ik3NzfhYwK9SbS6y41QoztTFYfpAoSrrrrK9pOIVGxEF0srKrcrNjjBGlkswYZXraik9As2vGjP5NWMDVpRAZmLYANBphPJbldsSJHfywwdOtSzuQmpKNWDjVgrNoYPH25Z82qAuB2DwyVmbGSU2267TbW1tWFrVGwglZlOVrpZsWFntYYU3wBxU7XGyJEjlZeXl/DjR2J3sMF8DbgtJydHAwcODPvZ6+rqUiAQMH5AlXoPNfr06aMRI0aoqKgo9F9xcXHYn3v+V1xcrI8++ki/+tWv9Ktf/Upnz57tdc/BWRzBwdTd2V2tIV06QX7ZZZeFnYTv6OhQfX29sZojEyXaisrJig2uxIuMig3nmD7oU7EBIBN4GWwcPXpUVVVV2rVrl77yla/or/7qr4yfH+EOv1RsRDoRzHvE+Jhe21Mp2Ih1mL2fKjbsGBwuEWxkFFPFxttvv63z58+rX79+HuzoknPnzlmu3s7Pz+fqbUTlh4qNnqZNm5bw8SK1ompvb1dOTviva7faUElSQUGBioqKwt60t7e369ixYwm9YSLYgBeKi4stP3snT540fkDtLdS45ppr9Nxzz1lOaMdiwoQJmjBhglauXBkKOKJVTl68eNF4HCdUVFRYTuIRbFyyf/9+y/eitLTU8mHS9DvRrooNTrDGJ5aTCwQbicn0VlRUbACZy8tgo6qqSq+99pok6bnnnlNOTo7WrFnjymMj3JEjR8KGyEuXXpO8CBMiVWzwHjE+6VixkYqtqBKp2KAuKYNcffXVxh9sr6s2qNZAorys2GhsbDQO9k6mYuOKK65QWVlZ2FpXV5exasOtweFBdlZt8JyHF2JtR+VUqNFd37599Z3vfEevvvqqtm7dqgULFmjAgAEx39/JYKMn5mxcEksbKonh4X5CxYZz/BJseNWKipNFQObyKth48803Q6FG0IYNGxx/XJjFenW8G3prRYXYDR06VNnZ2WFrgUDAMtzarzK5YoNgI8OYqjYINpCqvKzYMFVr3HTTTUl/2I11gLipYsOJ+RpBphMWib4AUrEBL8Ty+8KNUKOnCRMm6Mknn9Sf/vQnPf3005o0aVJM93ECczYii6UNlcTwcD9hxoZz/DJ3gooNAG7zKtj49a9/bVnr7OxUY2Oj448NK1MbKi/ma0hUbNgplas20qVig2ADUZlOmHgdbDQ3N1vWCDYQCy8rNkzBRjLVGkGxDhCnYgOIjynY6P77wotQo7tYqzjmzZsXV3VHPKjYiMz0O9cUJlOx4R+lpaVRQ3MqNuJ39OhRdXR0hK0NHjxY+fn5ru/F6WDj7NmzlosxcnNzPf25AeAtL4KNlpYW/d//+3+NXzt06JCjjw0zPwUbzNiwT6oGG4cPH1ZbW1vYWnFxsfHzrynYSPXh4QQbGYaKDaQTLys2TIPDk5mvERRLsNHe3h7zSTa72BlsULEBL/T2+8LrUKOnYBXHwYMHQ1Uc/fr107Rp0/TDH/7Qscc1VWwcOnTI1dlFfhVPxUafPuFvrwOBgOXDRryam5stpfD5+fkqLCxM6rjprreTDHl5eY6FhLFI5WCjJ68qGJxuReWnvysAf/Ai2Pj1r38d8X0EwYY3/DI4XKJiw06pGmzE0xptwIABlt9j7e3ttl2IFQ9aUSEho0ePtiRgLS0t2rFjh0c7MgcbfFBHLKJdge2UXbt2WX7x5+XlOVaxsXfv3rDniSnUGD58uKPPG4INpLpIMzb8Fmp0l5ubG6riOHbsmP71X//V8cf73Oc+Z1mnHVXswYbkzABx2lAlprdgw+ur7lM12PDLfA3J+YoNnncAevIq2IiEYMMbfqrYYMaGfUzBhhcn/OMVaxuqIL+0o6JiAwnzW9UGFRtIVKzDgO1makM1bdo0y1W6iSguLjZWXnSv2nB7voZEKyqkPlMQ2tDQ4NtQwyumdlSZHmw0NTVZrijKzc3Vtddea7y9E+2oaEOVGD8HG3l5eZYLEjo7O40f8vwk04MNKjaAzOZ2sLFz5069/fbbEb9OsOEN0xXyfgs2COLjlwkVG5J/gg0qNpAwgg2ki0GDBqlv375ha+fOnUu65Uc0Ts3XCDJVbbzzzjuh/+/2fA2Jig2kPlOw8dvf/pZQowdTO6pMn7MRT7WG5MyV+Fw5npjePtR5OTg8yPSz4vdgw0/tmZwONkx/V553QGZzO9jorVpDEsPDPXD69GnLa3V2drZnraiYsWGfVA020qlig2ADMSHYQDpxux1VW1ubcb6GncHGhAkTLGvRKjYINoDemX5X9JxZIGV2qCFRsWESb7DhVsUGJ1ij83PFhpSa7aj8VLFRUFCg7OzssLXW1lZdvHjRluNTsQGgJ1OwYbpa2g7t7e0Rh4YHUbHhvnivjnea6UKN3NxcX7zPSTWpGmzYUbHh9gDxkydPqqOjI2xt0KBBysvLi/tYBBsZ6NZbb1V+fn7YWlNTk2dXZBJsIBlut6PaunWrOjs7w9bKyso0ZswY2x4j2gBxtweHS5dOWmRlZYWtnThxIqHqGJ7z8EIsb+4zPdSQzMFGpldsxPs71xQ4OFGxwZV40RFs2M9PwYbkbNUGFRsAejIFGxcvXtQrr7xi+2P9+te/Nl4Q1h3Bhvv8NF9DkvLz8/W1r30tbO2b3/ymR7tJbekUbPRWsTF8+HDLmtsVG3ZVa0gEGxnLT1UbnOREMtyu2HC6WkMyBxt/+tOfQi+qXlRsZGVl2Va1QcUGvGD6XdEdocYlV199taXv/9mzZ7V//36PduQ9O1pRMTzcGyNGjIg4/8qvrahSMdjwsorB9P4h2onAWPG8A9BTv379NHHiRMv6//7f/9v2xzK1ocrJyQn789mzZ3XixAnbHxuR+S3YkKRFixbp/vvvV25urr71rW/pe9/7nqf7SVWpGGycP3/e+N7M7zM27BocLhFsZCw/BRvNzc2WNYINxMp0stLpio2epk2bZutj5OXlafz48Zb1Xbt2af/+/ZYqiSFDhrhy5a4p2EjkBdAUZhJswGm9BRuEGuGYsxHOr62oqNiITaSTDVRsJMb0AT+TKjZoRQXgr/7qryxr//Ef/2Fr1UZdXZ1ef/11y7rphCRzNtxlCja8bEUlSRMnTtSaNWt0/Phx/eM//qOuu+46T/eTqi677DLLBTEnT550fIZrMkzVGldddZUlBO3OD8GGXYPDJYKNjOWnYIOKDSTDdGLCqYqNixcvas+ePZZ1uys2pMgDxL0YHB7kZMUGz3k4LdJJTEINK+ZsfObChQvGapXeWlExPNxfCDbsc+HCBeMVdukabPC8A2By7733Ol61YarWmDBhgvH9B+2o3OXHig3YJ9WqNuIdHC75I9igYgNJMwUbBw8eND4pnGb68NGzDQYQiZsVG5FSZSc+5EYaIG5qQ+X0fI0gWlEhleXk5OgrX/lK2BqhhhkVG58xVWuUlZVpwIABEe9DxYa/EGzYJ1IFQ6R2X25wKti4cOGCsb0LFRsAJOerNkzBxte//nWVlZVZ1gk23GW6Qp5gI32kWrCRyDD74uJiFRQUhK21tLQ42tK9Jyo2kLT8/HzdeuutlnUvqjao2EAy3KzYsHPAUTSRBoinQ8UGrajglaeeekpf/vKXVVpaqlmzZhFqREDFxmfibUMlUbHhN5FONjBjI35+GxwuOTdjw/Scu/zyy5WVlZX0sQGkPierNtavX2/5PJudna2vf/3ruvLKKy23J9hwT0dHR0InkpE6TO+vUy3YiFaxIXlftcHwcNjCL+2oCDaQDDcrNtwMNsaOHav8/HzL49fV1Vlumw4VGzzn4YYrr7xSzz//vPbv36+f//znhBoRmIKNhoYGtbS0eLAbb5nC5Gi/cwsLCy2/v1tbW40zxWLx6aefqqurK2ytuLhYeXl5CR0v00Q62UDFRvz8GGw4VbFhOolBmAigO6eqNkzVGvfff78GDRpEsOExUxuqYcOGWa5+R+pKtYoNU9edWII2PwYbtKJC3Ag2kA7StWJDMldtfPTRR5a1VKrYaG1t1cWLF8PW8vLyOEEH+MigQYN07bXXWtZ3797twW68lUjFhmRvOyraUCWHVlT28eMwbaeCDaqkAETjRNXGBx98oP/v//v/LOtf//rXJYlgw2NUa6S/VAs2UrVig1ZUsIUp2Kivrzf2k3VKS0uL5SRnv379OMmJmHldsZFoqhwLU7DR84rpgoICXXXVVY7toTs7gg3mawCpYezYsZY1U8VYujMFG7FUydl5wpoTrMkxBRvZ2dkqKipyfzM95OXlWebKdXZ2Gt9v+IEfKzacakXlxxAHgP/YXbVhqtYYM2aMvvjFL0oyBxuNjY0JPRbix+Dw9JdqwQYVGwQbGa2kpMTYbsLNqg2qNZAsNys27EyVY2EKNnpyq1pDsifY4DkPpAaCjUtMraio2Egt/fv319y5c8PWHnjgAY92Y+VE1cb+/ftDbUu+/e1vGz/0JsKPwQYVGwC8ZHfVRqSh4UGmYOPYsWNqbW1N6PEQH4KN9JdKwUZTU5PlYo6CgoKY3q+Ygo3Dhw/btrdoqNiAbUxVG9u3b3ft8TnJiWR5XbHhZLAxYcKEqLdxa76GJBUVFVn6h7a2tsb1703FBpAaxo0bZ1nLtFZUjY2NOnv2bNjawIEDNXz48Kj3NX2goGLDO48++qgqKyuVm5urb37zm1q4cKHXWwoxBRvJVmz83d/9nX73u99JkjZt2qT/8T/+R1LHC/JjFYObwYbXf1cA/mRX1cZvfvMb4+/Z7sFGVlYW7ag8ZAo2aEWVXlIp2Ei0DZXkbcVGe3u78WJkgg0kxOs5G6ZBmgQbiIcp2EiXGRvXXXdd1B7gblZsSMlXbRBsAKkhUsVGe3u7B7vxRqJtqCTzyWo7KzYINuJz/fXX66c//amOHz+uf/qnf9KYMWO83lKI3RUbR44c0W9/+9uwtX/5l3+xhHSJHrundK3YYHg4gFjZVbVhqtb42te+ZmnPQrDhHdOJZCo20ksqBRuJtqGSZLxQy61gw1StUVJSoqysrISOR7CR4UzBxttvv63z58+78vhUbCBZBQUF6tevX9haW1ubzp07Z/tjuR1sSNGrNvwQbMTzAshzHkgNQ4cONX5wzqSqjUTbUEn2tqIy3Y9WVOnD7mAjUuX1H/7wh4SPGeTHYMOpGRtUbACIR7JVGwcOHDDetnu1RhBzNrxDxUb6M73WnzhxQhcuXPBgN71L1YoNu8+rEWxkuKuvvtr4g+9W1YbpJGfPIYpANKaqBifaUbk9PFyKPmfDzVZUEhUbQCbJ9DkbpoqNWIMNhocjVqkSbJw9e9byvj03N9fx90HRULEBwA+SrdowVWuMGjVKX/7yly3rVGx448iRI2ppaQlbGzRoEBebpJmsrKyUqdpIpmJj6NChys3NDVtrbm62pcI3GrvPqxFsQJMmTbKseRlscPU24uVWOyq3h4dLvQcb2dnZBBsAHEOwkXgrKoaHI1ZuBRvvvPNOwseU/FmtITFjA4B/JFO1EW1oeHcEG94wXR1PtUZ6Mr3eJ/o+3knJVGxI3lVt2H1ejWADns7ZINiAHdwYIH7mzBm1traGreXn5zt+Ur63VlRut6GSkg82eM4DqSPTB4gn04qKig3Eys6flZMnT0YMH5Ot2DC91vvhRL8Trag+/fRTdXZ2hq0VFxcrLy8vqeMCSG+JVm289NJL+vjjjy3rkYKNsrIyyxrBhvNMbaiYr5GeMqFiQzIHG4cPH05mSzGhYgO2I9hAqjO1orK7YsOL+RrSpRcb05tXKTWDDSo2gNSRyRUbZ86cMb6xd3vGxvnz5xUIBMLW+vTpQ8VGGrEz2IhUrSFd+qCazIkv0wd6v1ZsnDx5Mqn3gVRrAEhUIlUbpmqNe+65x3jSUaJiwysEG5kjFYKNzs7OlK3YYMYGbDd69GhLOtbS0qIdO3Y4/tgEG7CDGxUbXszXCIpUteF2GyrJmYoNgg3An8rKyiwnXS9cuKA9e/Z4tCP3mNpQXXfdderbt29M98/Ly7O8NnV1dcUdbtCGKv3ZGWxEuzApmaoNv7aiys3N1ec+9znL+ptvvpnwMZmvASBR8VZtNDY26sUXX7SsR6rWkAg2vMLg8MyRCsGGKdS47LLLVFBQEPMxaEWFtOJV1QbBBuzgRsWGF/M1giLN2UiXig2e84B/ZWo7qmTaUAXZccKaNlTpz62KDSk9gw1JmjJlimXtjTfeSPh4VGwASEY8VRumao0RI0bov/yX/xLx+AUFBZYL7Do7O9XY2JjAbhErKjYyRyoEG8m2oZL8VbFBKyokzatgo7m52bLGSU7Ey6uKDYKNS44ePWrpRR0JraiA1JKp7ahMFRvxBht2tKOiYiP95eXlqbCwMGyts7PT+L6jNy0tLXr77bd7vU0yA8RNH+j9crLf7mCDig0AyYinaiOeoeHdMWfDfaYr5Ak20lMqBBvJtqGSvJuxQcUGHOGnio2eH+6AaNJ5xoYUOdjwohVVbm6u8e8da9UGraiA1EKw8Zl4f+eaToRSsQETO6o2olVrSJlVsfHuu+8m/F6Qig0AyYqlamPLli366KOPLLeLJdigHZW7Tp8+bTkfkJ2dTSuqNJUKwQYVG58h2IAk6dZbb1V+fn7YWlNTk95//31HH5dWVLBDus/YKC4u1pw5c8LW7r//fuXl5bny+D2ZTmTE+gJIKyogtdCK6jN2tKKyo2KDYCP9mH5W4q3YMAUbPd8ntLS0JBxM+jnYGDZsmK1zNqjYAJCsWKo2TNUad955Z0wnJwk23GW6Op5QI32lQrDhVMUGw8OR0ryo2iDYgB3SvWJDkh5//HHNnj1bubm5mjt3rv7bf/tvrj12T8nM2aAVFZBaRo4cqaKiorC1s2fPGk/8pxOnWlHZUbFBK6r041TFxnXXXWdZS7Rqw8+tqCR721FRKQXADr1VbRw7dkzPPfec5euxVGtI5mCDGRvOYb5GZjG95jc1Nam9vd2D3ZjZUbExfPhwy9rx48fV1taW6Laiamlp0blz58LW+vbta/m8GQ+CDYQQbCBVuVGx4eXwcOlS+v6zn/1Mx48f19q1a40nK9ySTLDBcx5IPZlWtfHRRx/p4sWLYWulpaVxV+nZUbHBCdbMkGyw0dHRYQw2TCf7Ewk2Tpw4ofPnz4etDRw40Fev33YGG34PcQCkht6qNkzVGldccYVmzZoV07Gp2HAXwUZmyc7ONr4381PVhh1VRFlZWcZww8mqDScuGCbYQAjBBlKVKdhIt4oNP6FiA8gsmTZnw45qDYnh4YhdssHGW2+9ZbmK8IorrtCMGTMst01kgHgqnOi3c84GgSIAu0Sq2li7dq1lPdZqDYlgw22mYINWVOnNz+2ompubdeLEibC1nJwc4++FaNxuR+XEBcMEGwgxBRsHDx40ljjZobW1VRcuXAhby8vLU79+/Rx5PKQvUysqN2ZsEGx8JpZg4/z585bnfG5uLs95wOcINhILNuxoL8QJ1syQ7M+KqVrjtttu080332xZ37Nnj86ePRvX/vw8XyPIrjkbp06dslSnFBQUcOEVgIREqtr4+OOPLWsEG/5FxUbm8XOwYUcbqiBTsHH48OGEjhULJ2bXEmwgJD8/X7feeqtl3amqDao1YJe8vDwNGDAgbK29vd34M5aIrq4uT4eH+02iwQbVGkBqMrWiSudgwzQ/ZNSoUXEfh4oNxMqJYGPSpEkqKSmxZc5GKgQbkj3tqEzPUb9VpwBILaaqjZ6+/OUvx/Veo6SkRP379w9bO3v2rOUqbtjD1PaHYCO9+TnYsGNweBAVG0g7brajam5utqwRbCBRTrajMv3yLS4uVt++fW05fqpJNNggzARS0+c+9znLh+cTJ04Y31SnAzsrNvr0CX+rHQgELFeDR9Lc3KzW1tawtfz8fBUWFsa9F/ibU8GGJGPVhh3Bhh9P9tsRbJhOWlAlBSAZkao2urv//vvjPi4DxN3R0dFhyzwDpBY/BxtOV2y4PWODig3Yys1gg5OcsJOT7ahoQxWOig0g82RSOyq7gg0puRPWtKHKHMn8nOzcudPSWmrw4MGh56wp2Ih3zobpg3yqVGzEO2eD5x0AJ/RWtVFSUqLKysq4j0k7KneY2lANGzZMBQUFHuwGbjEFG/FWXjvFzqDND8EGFRuwlSnYqK+vd6Sk0RRscBUiEuVkxQbBRrjS0lLl5uaGrZ0+fTpqz26CDSB1mdpR7d6924OdOKupqclSpZebm6trr702oeMl046KNlSZI5lgo7dqDUmaMGGC5evp2ooq0pyNeKo2UmFQOoDU01vVRjyzNborKyuzrBFs2I9qjcyUahUbtKIC/p+SkhJVVFRY1p2o2qBiA3YyBRtOVmxk6nyNoESqNnjOA6krUyo27KzWkKjYQGzy8vIsF/d0dnYa33/0FC3YuPnmm5WdnR329cOHD8d18itVgg3JXLURzwBxnncAnBKpaiPRYIOKDXcwODwz+TnYcLpig+HhSHmmqg3Th6ZkcZITdjK1onJyxkYmV2xI5hMa0ZJ9KjaA1EWwkRi7KzY4wZq+Eg3BTBcfdQ82srOzk67aSKUqhmTnbDA8HIBT7r33Xt1yyy1ha3feeafGjBmT0PEINtxBsJGZ/BxsOF2xceTIEXV2diZ0vGio2IAr3JqzQbABO7ldsUGwYU/FBsEGkBrGjRunnJycsLUjR444WqrshX379lnWRo0alfDxTEFEMhUbtKJKX6bvbbSKjT179lg+IPbv399y4iyZAeJdXV0pX7ERz5wNhocDcNIvf/lL/dmf/Zny8vL01a9+Vf/0T/+U8LEYHu4OU7BBK6r0F2nGhlMn/GN1+PBhtbW1ha0VFxcbz4fFIi8vz3huy6nPeFRswBWmYOPtt9/W+fPnbX0cgg3YycmKDYINq0SCDVPFBs95IHVkQtWGG62oYq3YoCVOZkmkYiNaG6qgZAaIHz16VF1dXWFrQ4YMUV5eXkz3d1uyczao2ADgpGHDhmnTpk1qamrSP//zP2vIkCEJH4uKDXeY2v5QsZH+cnJyjOd8vK7acGLmi5tzNhgeDldcffXVxieG3VUbBBuwExUb7rIr2KBiA0gdmTBA3BRs2F2xwfBwmDgZbCTTiiqVqjWCkpmzQcUGgFRhCjaOHTum1tZWD3aTvmhFlbn82I7KzjZUQcOHD7esORFsBAIBXbx4MWytoKBA/fv3T+q4BBswcqMdVXNzs2WNYAOJcrtig+HhDA8HMk26V2xcuHBB+/fvt6wzPBxucDLYKC8vt1wV3NLSEtPzNxWCjeLiYmVlZSkQCEhKfM7GuXPnLO9V+vbtm/Hv+QD4U1ZWFlUbDjty5IhaWlrC1gYNGpTxFzlmCj8GG25VbDgxQNypC4YJNmDkRrDBSU7YycmKDYaHW1GxAWSedA82TNUaZWVlGjBgQMLHpGIDsYo32GhoaLB86MzOzjYGG1LiczZSaXB4UKJzNggTAaQagg1nOXESGanDj8GGExUbbrWicuq8GsEGjLwKNgoLC219DGQOZmy4i2ADyDymYONPf/qT8U1qKrJ7voZExQZiF+/PSqRqjZycHOPtEw02UqFio6dE52wwXwNAqmGAuLNoQ5XZ/BhspPKMDac6oRBswGj06NGWH7CWlhbt2LHDtsegYgN2cqpio7W11XJCvm/fvsbHyyS0ogIyT15ensaMGWNZT5eqjX379lnWkpmvIV26YCM/Pz9srbW11diOs7tPP/3UMrC5uLjYtwObkbx4gw3TBUeRqjUk85yNWAaIp2KwISU2Z4P5GgBSDRUbziLYyGym9wB+DDao2AAicLpqg5OcsJMpaLCjYoP5GmYFBQUqKioKW2tvb+/1JAwVG0DqS+d2VE5UbEiJtaOiDVXmsaNiw/TePchUsbFnzx6dPXu2132Zgo1UqGJIZM4GFRsAUg3BhrNMwQatqDKH3yo2zp8/b3xfRsUGEAHBBlJJTk6OsZVZsuEG8zUiM12x2dsLIMEGkPrGjRtnWdu9e7cHO7GfKdhItmJDSqwdFW2oMk88PyeHDh0yDrrvrWKjpKRE1113nWU9Wjsq0wf4VK3YiDZng4oNAKmmrKzMskawYR8qNjKb34INU7XGVVddFbENaawYHo60RbCBVGOas5FsOyrma0QWbzsqnvNA6kvnig1TKyo/VWxwgjW95eXlWS7Q6OzsNL4PMVVrTJw4Meqg+0TmbKRqK6pE5mwQKAJINVRsOMt0IplgI3P4LdhwYnC4JA0YMMDYjSPa55V40YoKrrv11lstfaGbmpr0/vvvJ33s8+fPq62tLWwtNzfX8nhAPJxoR0WwEVm8wQYVG0DqMwUbDQ0NxuAyEUeOHNHdd9+t0tJSfetb3zJWUTihsbHR0pJn4MCBGj58eNLHNp0YTaRig1ZU6S/Wqo1Ig8OjMQUbvc3ZaGtr04kTJ8LWsrKyUiLYkOKfs2E6WUErKgB+xvBw55w+fdpyLiA7O5tWVBkkUrDRcw6eW5wYHB7kRjsqWlHBE05VbXDlNpzgxABxgo3I4gk22traLGFm3759CTOBFDNw4EBjeya7qjYWLFigrVu3qq2tTS+88IJWrFhhy3GjcaoNlWQ+WU3FBkxMPyuxVmzEEmyYBoj3VrGRqtUaQfHO2aBiA0CqKSgosJwY7OjoINywgZMnkZEacnNzNWTIEMu6V1UbTgwOD3Ij2KBiA55wKthobm62rBFsIFmmVlROVGwwPPySeIINqjWA9OFUO6odO3Zo27ZtYWvPP/+88feH3ZxqQyUl1oqKio3MFEvFRqTq6VgrNvr0Cf/4d/jw4YhtS9Ix2OhtzgYVGwBSEXM2nMF8DUj+akdlakVFxQbBBqKgYgOpxImKDYaHRxZPsMFzHkgfpgHidgQb69evN6731irHLqaKDbuCDYaHI1ax/KyYqjVuvPFG4xWFPWVnZ8c1ZyPVT/THM2ejvb3d+J6P5x0Av2POhjMINiD5K9hwsmLD1H7XzgHiXV1dVGzAG6Zg4+DBg8akMB6mk5w9ByYC8XKrYoNg4xIqNoDMZKrY2L17d9LH9VuwYVcrKruGh1Oxkf4SDTZM79cjiSfYSPWKDSn2ORumkxRDhw5Vdna2I/sCALswZ8MZpmCDVlSZxxRs2D1UO1apXLFhOq9WXFysnJycpI9NsIFe5efn69Zbb7WsJ1u1wdXbcAIzNtyVbMUGwQaQmkzBRn19vWWOTjzWr19vbFMpuRNsONmKiooNxCrRYCOWNlRB8QwQT9dgw1SxwXMOQKqiYsMZVGxA8k/FRlNTk+Vi0YKCAtveq3gRbNh1Xo1gA1E50Y6KYANOoGLDXcOGDVNWVlbY2okTJ4wnN00VGzzngdQ0ZMgQ49VByVRt/N//+38jfs3pYKOpqclYau3VjI3z588rEAiErfXp04eKjQwQLdg4ffq08fkQT7ARzwDxVG9FJcU+ZyMd/q4AMhPBhjNMbX8INjKPX4INJ9tQSc4HG062eCfYQFQEG0gVblVsMDz8kqysrJirNmhFBaQXOweI79+/X//+7/8e8euHDh2yfXhddzt27LCs3Xjjjerbt68tx8/Ly7O8PnV1dUUMN2hDlbmiBRumao3y8nJjX+RIysvLLfM4WlpajM/fdKjYiHXOBhUbAFIVwYb9Ojo6jCeSaUWVefwSbDjZhkrypmLDrvNqBBuIyhRs1NfXq7W1NeFjEmzACXZXbJw6dUrt7e1hawMHDlR+fn7Cx0w3phMcphdAnvNAerEz2Ig0W6M7J6s2TMHGLbfcYutjxNOOihOsmSuRYCOeao2gWOdspEOwIcU2Z4OKDQCpimDDfqY2VMOGDVNBQYEHu4GX/BJsOF2xUVxcbPn5bmlpSboDShCtqOCpkpISYxJo6kcdK05ywgl2V2zQhio6KjaAzDRu3DjLWqKtqDIh2IinHRUVG5nLb8FGupzsj2XOBoEigFRVUlKi/v37h62dPXtWJ06c8GhHqY9qDQT5JdhwumJDcrZqw9SKiooNuGrUqFGWtWSCDdOAUIINJMsUbCSTMBNsREewAWSmSBUbXV1dcR1n8+bNxqviekrHYIOKDfSUl5enwsLCsLXOzk41NTXpwoULxlawiQQbpjkbPZ9jp0+ftrx29+vXz9LGKhXEMmfD9LxLxRAHQGYyVW00NjZ6sJP0wOBwBPkl2HC6YkNyNtigYgOeszvYMFVs9PwgB8Rr8ODBlmHWp0+ftrSTihXzNaKLNdigSgtIL8OGDbO8+e3o6Ii7HZWpWmP69OmWNaeCjXfeecfSWrOkpMT4vicZpivx46nYINjIHJGqNrZv324JDq+88kqNHDky7scwVWzs2bNHZ8+eDf05Xao1pNjmbJj+vjzvAKQK2lHZi2ADQaZZeZ2dnRHfxzslHSs2CDbgKtMH/A8//DDh43GSE04xzdlItB2Vk7980wUVG0DmMlVtxNOO6uTJk/qXf/kXy/pDDz1k+f0QCAT0wQcfxL/JKHbu3GlZs7taQ0q+YoNWVJnD9L1uamqyrQ2VdCm8u+666yzr3dtRpct8jaBoczao2ACQysrKyixrBBuJMwUbtKLKXF5XbXR2dqZlxQatqOCq66+/3rLGjA34kZ3tqGhFFR3BBpC5kh0gbqrWGDZsmO6+++6YWuXYwY02VFJ8FRu0ospsvVVs9JRosCFFn7ORCcEGFRsA0gUVG/aiYgPdeR1smEKNyy67zPZh9qZg4/Dhw7Ycm1ZU8JwbragINmAHOys2CDaioxUVkLlMA8STDTYqKyslxTYDwA5uBRsMD0esCDac0ducjWPHjqmjoyPsa0VFRcrPz3drewCQFGZs2Mt0IplgI3N5HWy40YZKkoYPH25ZY3g40saQIUMsH7Ta29vV0NCQ0PE4yQmnOF2xwYyNcFRsAJkrmVZUO3bsMAYVc+fOlWQ+6Wp3sHHkyBEdOHDAsu5WxQbDw2Fi+ll59913df78ecvtRo8enfDjRAsP02nGhtT7nA2ecwBSHRUb9jly5IhaWlrC1gYNGsQFjhnM62DDjTZUknOtqM6fP285B5ydna0hQ4YkfWyJYANxsLNqg2ADTjEFG8zYcE5RUZGlBLK1tdXyb06wAaSfq6++2vKGtLW1VXv37o16X1O1xpQpUzRmzBhJ5pOuu3btslxVnQxTtcZNN92k/v372/YYQclWbHCSNXOYgg3T++1kqjWkS+Fhnz7hHwUPHz4cOhGWbhUbUuQ5G8zXAJDqCDbsYzqJzHyNzOZ1sOFWxYZTwYaT1RqSlNXV1dVl29HgipUrV+r999+P6cP91Vdfrerq6phOxjY1Nammpsb4pDGpr6/Xgw8+qMWLF0e97S9+8Qtt27bNknybTJw4UdXV1THt4f3339cvfvGLmPZcWlqqb33rW5o4cWLU27a0tGj16tV6//33bd/zyy+/rI0bN8Z0XCe/f9/61rd05513xnTbVPr+PfbYY1q7dq3GjRunIUOGqG/fvgnveeLEiZaTCT/+8Y/1+uuv8/1TfN+/trY2vffee2FvABoaGown63j+hfPD94/fn+H4/n2mo6NDL7/8ctjamjVrdP/991tuG+/376WXXrJ8uPz3f/93S0VFot+/v/u7v9OPfvSjsK8tWLBATz75ZOjPdn7/ioqK1NnZKelSJWxFRUVMF3WcP39ef/u3f8vz7/9J9+ffE088oQMHDiT1/sUk3u/frl27woZrS9JLL72kL3zhC2Frfvj+bdy4URcuXNCpU6dUVFQU8XbPPfec5s2bF9fzj9e/cOn+/OP79xm+f59Jhe9fV1eXCgsLQ3+uqKhQWVmZ7a8lmfD9u/baay3Hvueee/SLX/wibI3n32f89P1z8vxn99krf/7nf65f//rXxtva/f178MEH9fzzz6fs+5cNGzaora0t6m3j+f51R8VGior1isU//elPeuWVV2K67c6dO2N+UkvSyJEjY6rYaGlp0csvvxzTEyS4j507d8Z021deeSXmPTc1NcX8b3Hw4EHt3Lkzrj3HettYfylLzn7/Yj1uqn3/iouLNWjQIF1++eUxvZGTIn//TK2oXnvtNb5/3fYQ657z8vIsfUkjVWzw/PuMX75//P4Mx/fvM9nZ2ZbftZHmbMT7/QtWb3RnakeV6PfP9G/d/c2/3d+/7lfiX3PNNTFXqvbr14/nXzfp/vw7dOhQ0u9fTOL9/gVDuO5MVyz65fsnXarq6k2wYiOe5x+vf+HS/fnH9+8Svn/hUuH7l5WVFara6Nu3r6655pq4Xkv4/l3ypz/9ydhW1TRfg+ffZ/z0/XPy/Gd3kSqvnfj+BfeZqu9fYgk1pPi+f90RbKS50tJSY59qk4kTJ8acjF28eFGHDh3Shx9+GPW2/fv31xe/+MWYjitJo0ePjrln8Fe+8pWYW0bEs48RI0bElGwGjztx4sSY9zF79uyYbic59/2L598i1b5/gwcP1unTp2MuDYz0/bt48aKxhdWcOXNiOq7E96+74O+MoJycnIj35fmX2G0l759/Et+/RG8rpeb374orrtDFixfD1iMFG/F+/0wtME3BRqLfv2iDw+3+/nWvUPvjH/9o+XeLpKuri+ff/5MJz7+KioqYj+vk96/7VYlBplZUfvn+STLO0OguOGcjnucfr3+fyYTnH9+/+G8r8f3rzsvvXzDY6Pm5Kxq+f58pLS3V2bNnjfvrieffZ/z0/XPi+devXz/LcyrS+SYnvn/BsI33L2a0okpBK1eulCQtWbLE1cc9dOiQ5cPWoEGD1NjYGNdx3nnnHU2fPj1sbfz48fqP//iPZLcIhNoMdHfvvffq//yf/xPXcY4cOaLrr78+bK20tFT79+9Peo/pxtTOZdmyZfpv/+2/SZJOnDhhucpl8ODBxv6lAFJLQ0OD5Q1oYWFhxA/U+/fv10033WRZN7W4+f3vf6977rknbO366683BhLx2rFjh770pS+FrV1++eUJzw6LxezZs7Vly5awtV//+tf68z//89CfN27cqO9+97tht/nqV7+qf/7nf3ZsX/CX8+fPG+dsdHfHHXfoX/7lX5J+LNPzt3///tqzZ4/lysRE3vO7pbi4WIFAIGorKklavHixamtrw9auvfZay/u7n/70p6qsrLR7qwDgmPnz51tmmD399NP6zne+49GOUtOXvvQly3vNF154QbfffrtHO4LXWltbLS20c3JydPLkSccf++c//7m+973vufbY8+bN03PPPRe29swzz+iBBx5I+Jg/+tGP9Hd/93dha3/5l3+pf/iHf0j4mN1RsYGYXXnllRowYEDY2unTp3X48OG4jsPgcDhp8ODBlrVEfumb2lAxONzMdAVn96Gjpuc8g8OB9FBeXm55Pjc3N0cMgU1Dw0ePHm0JNSTzAPEPP/xQgUAgsc12E61awwmxDBBncDj69esX1ifdJNnB4UHl5eUaMmRI2FpLS4vxYqNUHxweZBogbrrqkuHhAFINA8TtYbr4ztSKCpkjPz/fcuFEe3u78X273Z566inL2rRp0xx7PCcGiDs9PJxgA3ExtYWI9+pGgg04qbi42LJmaikVDcFG7KIFG2fOnLF8nWADSB/jxo2zrEVqR2UKNubOnWu8bWFhobG1jKkdVbz8Emz0/EBk6tcb7ep9pJ9o33O7gg1JxpL/7du3W9bSOdg4d+6cZY1AEUCqIdhI3qeffmo5D5CdnW1sRYXMYrrgIdKcDbv8+Mc/NrYGXbx4sWOP6USw4fS5NYINXSr7HzNmjLKyspSVlaVFixapoaEhqWMuWrQo4gf1VEawAb+jYsN9iVRs8JwH0sfYsWMta6bBi5s3bza+Oe/t/ZKpaiNVgw3TyeqeH4hMH5A4wZp5egs28vLyHA82TMFkulQwBOdsRJMuf18AmaOsrMyyRrARH9N7zDFjxniwE/iN6X1BrHNdE9HS0qKamhrL+r33tqaXrQAASbZJREFU3hv3DLJ4mIKNeLv09GSq2CDYsNGyZctCw4Bra2u1dOlSrV69WlOmTEk43NiyZYtWr15t5zZ9g2ADfudkxYad5XLphIoNILOZgg3TiVFTtcZXv/pV4xvoICeCjUOHDunjjz+2rMc6NC9RibaiomIj8/T2Pbcz1JDMzzFTK7l0qdiQzFUb3fXv3z9qOzAA8BtTxYZfZyP5lek9pul1EpnH7WDjqaeeMl6gW11d7dhjSu5VbNCKyiZ1dXVasWKFKisrtWfPHs2bN0/Lly/X7t271dTUpGXLlsV9zI0bN2rmzJkO7NYfCDbgd4MGDVLfvn3D1s6dO6e2tra4juN0qpxOTCc7jh49qs7OTkkEG0C6i6UV1cmTJ43DjqNVtzoRbOzcudOyNnHiROXl5SV13GhMJ6tjaUVFxUbm6S3YuO2222x9LFPFhilgy6Rgg+ccgFREK6rkEWwgEjeDjU8++cRYrfHggw8aP3fZafjw4cb9JINWVA7asGGDJOmRRx4JWx87dqwWLlyoDRs2xFy10djYqEWLFmnOnDmODnLxminY+PDDD+M6RnNzs2WNYAN2MlVtxNuOilZUscvNzTX+2wSrNggzgfRWUVFhCQWOHTsW9mHaVK0xbNgw3X333b0ee8KECcrOzg5b++STT4wVF7Hyog2VRMUGYudmxUZJSYmuu+66qLcj2AAAfysoKLBcBd3R0UHVRhwINhCJm8FGTU2NOjo6LOtOV2tI5oqN48ePx32hcM/790TFhk02bdokyfwB4ctf/rKkyMMve/rrv/5rrV69WkuXLg0dNx2NGjVKffqE/9gcPXo0rlY/nOSE00xzNuJtR0WwER/TCY9gsk/FBpD+orWjMgUblZWVUY/bp08f26s2/BRsULEBEzeDDclctdFTOs2ciDZnI53+rgAyC3M2EtfQ0KATJ06ErfXv31833nijRzuCn7gVbOzdu1dr1qyxrH//+993ZYh9VlaWrVUbp0+f1vnz58PW8vPzbT0flNHBRn19fcTqivLycknm6oJIt3/zzTe1fPlyFRUV2bVF3+nTp0/S7ahMwQZ9bGEnKjbc19ucDYINIP2ZyqKDA8R37NhhDCKitaEKsjPYaG9v9yzYGDRokPLz88PWWltbQ+81P/30U3V1dYV9vbi42PEWWfCfSMHGbbfdpn79+tn+eLFcjZpOFRtS71UbhIkAUhXtqBJHtQZ641awYWpBNWDAAFeqNYLsnLPhdLWGlOHBhhT9jeurr74a03GWL1/uyBVUfuREsEHFBuxkxwBxhofHp7dgg+c8kP56q9gItv7sbsqUKRozZkxMx7Yz2NixY0do/k9QWVmZrr766oSOF6/e2lHRhgpBkb7vTn3WiKViI5OCDSo2AKQqBognbteuXZY1gg0Emd7D7969Wy0tLbY9xltvvWWscq+urjae43KKKdg4fPhwQsdy44LhjA82/K6rq8vyn9cINuB3plZU8VZsMDw8PlRsAJmtt2DD9AY91moNyXzSNZlgo6eJEycmdKxE9DZAnDZUCPIi2OjZara70tJS9e3b15HH9goVGwDSERUbiaNiA70xXfRw4cIFLViwwLbHeOqppyxrw4cPd7VaQ3K+YsPu82o5th4NtvuHf/gHr7dgQbABv0u2YuPMmTNqbW0NW7O7D2C6IdgAMtu4ceOUlZUVdgFGY2Oj1qxZo0AgEHbbvLy8uIKNUaNGqaioKOw4Z8+eVX19vSoqKuLa586dOy1rbrShCoq3YoMTrJkp0gc+p4KN7Oxs3XzzzcbgT0q/ag3pszkbH3zwgeVrVGwASFUEG4kzBRuxVDQiMxQUFOgrX/mKXnnllbD1F198UUuXLtWKFSuSOv4rr7xi7Bj013/9171efOIEO4MNNzqhpHXFRl1dnbKysiz/dR9mabo6rrs77rjD6W32yrR/rxFswO+Srdhgvkb8aEUFZLbs7Gxj1cY///M/W9bmzp1rmTURjV1VG17N1wjqbYC46T0pragyU79+/TRnzpywtfvvv9/RmXS9nbxJ1xP9kao2CBQBpCqCjcS88847unDhQtjaZZdd5lqrUqSGv/3bvzXOOnv66aeNA7/jYZqtUVFRoYceeiip4ybC6WCDVlQ2qqio0LZt24xfC17R5/VQ68cee8zyn9dMwcb+/fvV1tYW0/05yQmnJVuxwXyN+FGxAcA0QNzUrzieao0gO+ZsHDx40NIfNjs729VgwxRUULEBk7/4i78IhRn33Xef/uf//J+OPl5v7TbSsWJDihxspGuQAyD9EWwkhmoNxGL8+PFau3at8WuPPvpozDOae1q/fr3eeusty7rbLaiC7JyxwfDwJI0dO9Y4oyJ4ReGsWbMkKayCIyh4RZ/p6sNMN2DAAF111VWW9ViqNi5cuGBp8ZOTk6OCggLb9gckW7HBfI34EWwAiOU90+jRo/WFL3wh7mPbEWxEqtbIzs6Oez+J6q0VFRUb6O7zn/+81qxZo0OHDunZZ591/AKL3k7gZFKwkZ2dzXs+ACmrpKRE/fv3D1s7e/asTpw44dGOUgPzNRCre+65R8uXLzd+bf78+cYWl9GYqjUmT56sr33ta3Efyw5UbKSQyspKSdLDDz8c1re5rq5Oq1evVmVlpcrLyz3anb8l2o6qubnZska1BuzmRMUGH3J7V1paqtzc3LC106dP6+zZs1RpARkilmAjkWoNyfzhcteuXZa2Ab3xug2VxPBw+Fd5ebmGDBli/Fq6VjAMGzYsdKFb0Ne//nVvNgMANjFVbTQ2Nnqwk9RBsIF4PPzww5o/f75lPRAIaP78+WppaYn5WGvWrNHevXst615Va0jmYOPIkSPq7OyM+1huXDSc0cHG2LFjtXTpUm3btk1Tp07VunXrtGzZMo0bN06lpaXGFM4vcy68lmiwwQlOuIEZG94wXdH5ySefULEBZAhTK6qeEg02hg0bZqwWjadqww/BRrzDw6nYgJsiVW2ka8WGJK1cuVL33HOPBg4cqAceeECLFi3yeksAkBTaUcXnzJkzev/99y3rBBvozVNPPaWZM2da1nfv3q0FCxbEdIyOjg5jtcadd95pPLZb8vLyjOe/EqnaYHi4C5YvX67a2lpJUlVVlVasWKGFCxfqjTfeoFqjF3YGG17PMUH6oWLDG6YTH4cOHbK0n8vOzqb9HJCG8vPzdcMNN0T8+le/+lXjFUCxSqYdVVtbm/7whz9Y1idOnJjwfhJBxQb8LNJJnHQONoYPH65f/OIXOnz4sJ555hndeOONXm8JAJJSVlZmWSPYiMz0XrK8vNx4sSTQ3dq1a3X99ddb1l988UUtXbo06v1ramqMYYGX1RpBdrWjohWVS+bNm6c9e/aEZnA888wzEUON4G2i6erq0vr16+3eqm9QsQE/MwUbyVZsMDw8OtOJjz/+8Y+WNZ7zQPrqrR1VotUaQckEG6Zqjauvvtr44d9JkSo2zp8/H9YWVZL69OlDxQZclYkVGwCQbqjYiA9tqJCo4uJirV27Vv369bN87emnn9aaNWsi3vfkyZPGao37779fn//8523dZyKGDx9uWUsk2GB4OHyLYAN+VlBQYHlxaWtr07lz52K6P8PDE2M68fHxxx9b1mhDBaSvSO2ohg0bprvvvjupY9sdbLjdhkq6VNrdM3zv6uoytkAg1IDbTMFGdnY2lUMAkEKYsREfgg0kY/z48Vq7dq3xa48++qheffVV49dqamqM56f8UK0hmSs2Dh8+HNcxjh8/bpnLUVhYqLy8vKT21hPBBhIydOhQS8rW1tam/fv393o/gg24xVQ6Gms7KlpRJcYUbJhe/Ag2gPQVqWKjsrIy6WObPmR+9NFHxjC6J78EG5I5sDBdHMLJZLitpKRE9913X9jaN77xDY92AwBIBBUb8SHYQLLuuece44xmSZo/f74++OCDsLU//vGP+vGPf2y57V/8xV/oc5/7nCN7jJcdrajcqNaQCDaQhESqNgg24JZk2lERbCTGFGwcOXLEssZzHkhfkYKNZNtQSdKAAQNUUVFhWY+lasNPwYYpsDhw4IBljYoNeOGJJ57Qfffdp/79++uBBx7QX/7lX3q9JQBAHAg2Ynf48GHjvw3BBuL18MMPa/78+Zb1QCCg+fPnq6WlJbT21FNPWW7Xt29f/fVf/7Wje4yHHcGGW+fVCDaQMIIN+FmiFRtdXV3M2EiQKdgIDsXtjooNIH0VFRVp1qxZYWt33nmnxowZY8vxE2lH9dFHH1kGc+fm5ro+ODzIFGyYPlRTsQEvXH311Xr22Wd19OhRPfPMM8YwEQDgX6Zg49ixY2ptbfVgN/4WqVqjb9++HuwGqe6pp57SzJkzLeu7d+/WggULJEm7du3Sz3/+c8ttqqurdfnllzu+x1g5VbFBsAFfSSTYaG5utqwRbMAJiVZsmH75FhcX8+YmBqZg48SJE5Y1gg0gvf3N3/yNZs+ercLCQn3lK1/pdXBevBIJNnbu3GlZ86paQzJXYpg+KBBsAACAeGVlZVG1ESPaUMFua9eu1fXXX29Zf/HFF7V06VLjwPCSkhLfzNYIMgUbf/jDH8IqT6Jx64Jhgg0kzK6KjcLCQtv2BASZgo1YKjZoQ5U4U7Bh+jcnzATSW0VFhX72s5/p0KFD2rBhg62v86bhxtGCDT+1oZLMgYXptYdWVAAAIBEEG7HZtWuXZY1gA8koLi7W2rVr1a9fP8vXnn76ab3wwguW9erqauXn57uxvZiZgo0LFy6EKk9iQSsq+B6tqOBnplZUsVRsEGwkrqCgQEVFRWFrnZ2dlttRsQEgUePHj1dubm7Y2rFjx/THP/4x4n38FmyYAgtTCEzFBgAASIQp2GhsbPRgJ/5GxQacMH78eK1duzam21577bX63ve+5/CO4jdgwADdcccdlvVg5UksGB4O3xsxYoQlVTx16pSOHj0a8T4EG3CLnRUbzNeInalqoyeCDQDJiKdq49y5c3r33Xct636r2Dhz5oxljYoNAACQCCo2ovvggw8UCATC1gYOHKjRo0d7syGklXvuuUfLly+Peju/taDqbunSpRErT2JpNUzFBlKCqWrjww8/jHh7gg24JdGKDbcGHKWrWIINnvMAkhHPnA1Ttca1117r6XA+U2BhGuhJxQYAAEgEwUZ0pveOpotngEQ9/PDDmj9/fsSvT5gwQd/61rdc3FF8eqs8efTRR/Xqq6/2en+GhyMlmIbi9NaOimADbmHGhjeo2ADgtGSDDS+rNSRzYNHe3h7T7QAAAKIpKyuzrBFshKMNFdzw1FNPaebMmcav+blaI6i3ypP58+frgw8+iHhfhocjJcQ7Z4NgA25hxoY3qNgA4LR4go2dO3da1iZOnGj7nuIxdOhQ9ekT/ha8q6sr7M/5+fn8rgQAAAlhxkZ0BBtwy9q1ay0Xhd9+++26++67PdpRfCJVngQCAc2fP18tLS3G+9GKCikh3mCjubnZssYHdziBig1vULEBwGnXXnut5Wqf1tZW1dXVWW7rx4oNKfr8DKo1AABAomhF1buOjg6CDbimuLhYv/rVr/TFL35RAwYM0J133qna2lqvtxWXSJUnu3fv1oIFCyzrFy9etMywkajYgA/FE2xcuHDB0kM6OztbAwYMcGRvyGymYCPRig2Gh8eOYAOAG2Kp2vjwww8tvV379++vm266ydG9xSJacMHgcAAAkKiCggLLZ9iOjg6qNv6fd955R52dnWFrw4cPN7bwAuxQXl6u3/72t/rkk0+0fv36lDzHZKo8kaQXX3xRS5cuDVuLdMFwVlaW7fsi2EBSTMHG4cOHjS2naEMFN5laUcVSscHw8OTQigqAG2IJNvxarSFRsQEAAJzFnI3IqNYA4ldcXKy1a9eqX79+lq89/fTTWrNmTejPpvNqToU5BBtISk5OjsrLyy3rpqoNU7BRWFjoyL6AvLw8SzVQe3u78eewO1pRJYeKDQBuSPVgI1pwQbABAACSQTuqyAg2gMSMHz9ea9euNX7t0Ucf1auvvirJ3fNqBBtImqlq48MPP7SsUbEBt8Xbjqq1tVVnzpwJW+vbt6/xODAbNmxY1PJCgg0AyTJ9+KyrqwsbXpfKwQatqAAAQDIYIB4ZwQaQuHvuuUfLly83fm3+/Pn64IMPCDaQWkw91mKt2CDYgJPibUfFfI3kZWVl9Vq10adPH+bqAEja0KFDdc0111jWgx9UT58+rT179li+PnHiRMf3FgtaUQEAACdRsWF26tQp44W4BBtA7B5++GHNnz/fsh4IBDR//nx98sknlq/Rigq+FesAcYINuC3eig3ma9ijt2CD5zwAu/TWjspUrTFq1Cjf/E6nYgMAADiJYMPMVK3xuc99jjbpQJyeeuopzZw507K+e/du/fKXv7SsU7EB3yLYgF+Zgo14Kzb8chIslfQWbNCGCoBd4g02/NKGSqJiAwAAOItgw4w2VIB91q5da+zi89FHH1nWqNiAb5mCjYaGBrW3t4etEWzAbaZWVL1VbBBs2INgA4Abbr75ZstaqgQbVGwAAAAnMWPDbNeuXZY1gg0gMcXFxVq7dq369esX9bZUbMC3Bg0apOHDh1vWe1ZtNDc3G+8LOIWKDW/QigqAG0wfQg8ePKhPP/005YMNKjYAAEAySkpK1L9//7C1M2fO6MSJEx7tyB+o2ADsNX78eK1duzbq7Qg24GuxtKOiYgNus6Nig+Hh8aNiA4Ab+vXrp3HjxlnWN23apEAgELY2aNAg3XjjjS7tLLpBgwYpPz/f+LXi4mLl5eW5vCMAAJBuqNoI9/HHH1uGGmdnZxurgAHE7p577tHy5ct7vQ2tqOBriQYbDGiCk+Kt2GB4uD2o2ADgFtMVdv/+7/9uWZs4caIb24lLpKoM2lABAAA7MGcjnKla4+abb1ZWVpYHuwHSy8MPP6z58+dH/DoVG/A107CYDz/8MOzPVGzAbczY8AYVGwDcYgo23nvvPcuan9pQBUUKMGhDBQAA7ECwEY42VICznnrqKc2cOdOynpub69iF7QQbsAWtqOBHzNjwBsEGALeYPowePXrUsubHYCNSgEGwAQAA7FBWVmZZI9gIR7AB2Gvt2rWWi9/vuecexx6PYAO2INiAH1Gx4Y2ioiIVFBQYv8ZzHoCdbrzxRstgzI6ODsvtUinYoBUVAACwAzM2whFsAM4rLi7Wr3/9a9155526/PLLNWvWrKjzN5KR49iRkVEuv/xyFRcXh10N39raqoMHD2rEiBGSCDbgPjsqNhgenphhw4bpo48+sqxTsQHAbhMmTNDrr78e8eujR482vh54jVZUAADASbSi+kx9fb3Onj0btlZcXGy8SBdAcq677jqtX7/elceiYgO2iVa1QbABt5lOZEWq2Dh16pTa29vD1gYOHKj8/HxH9pbuIrWjItgAYLdoV9r5sVpDomIDAAA4i2DjM1RrAOmJYAO2iRZsNDc3W75OsAEn5eTkGAcUmcIN2lDZK1KwwXMegN2ifSidOHGiSzuJDxUbAADASaZg49ixY2ptbfVgN94i2ADSE8EGbNNbsHHx4kW1tLSEfa1Pnz5cvQ3HmeZsmNpREWzYi4oNAG6hYgMAAMAqKyuLqo3/h2ADSE8EG7CNKdj48MMPJdGGCt6JtR0V8zXsRbABwC0jRoyIGBIUFxdr9OjRLu8oNlRsAAAApxFsSG1tbdq1a5dlnWADSH0EG7DN9ddfb1kLVmyYgg1TiyDAbrFWbBw/ftyyRsVG4mhFBcBNkT6Y+rVaQzIHGH369KFiAwAA2MYUbDQ2NnqwE++YqjWuuuqqiJ9ZAaQOgg3YZuTIkcrLywtbO3HihI4dO0bFBjyTTMUGwUbiqNgA4KZUDDby8vJ0//33h61985vf9Gg3AAAgHVGxQRsqIJ0RbMBWkeZsEGzAK6ZggxkbziPYAOCmm2++2bju52BDkr73ve/p61//uvr27atvfvOb+t73vuf1lgAAQBoh2CDYANJZjtcbQHoZNWqU3nvvvbC1ffv2GdstEGzADaZWVMzYcJ4p2OjTpw/BBgBHRPpwOnHiRJd3Ep+xY8fqJz/5iX7yk594vRUAAJCGysrKLGvBWaiZgvkaQPqiYgO2omIDfkPFhjdyc3P1jW98I2ztgQce8Gg3ANLd4MGDdd9994WtzZ07l/caQBo4ePCgbrrpJm3dujXu+7777ru6/fbbE7ovAKQDU8VGXV1dxoQbx48f10cffWRZj1TtCyC1EGzAVpGCjebmZss6JxvghlgrNhgebr/vf//7+vrXv668vDx961vf0vz5873eEoA09t//+3/X7NmzlZubq6997Wt64oknvN4SABscPHhQ7777ru6999647hcIBLR48WJt3bpVBw8edGZzAOBzpmCjvb1dc+bMyYhww9SGqqKiQgUFBR7sBoDdCDZgK1Ow8eGHH1KxAc9QseGdG264QT/5yU/U1NSkf/zHf9T48eO93hKANDZy5Ej97Gc/0/Hjx1VbW6srrrjC6y0BsMH06dM1fvx4BQIBffe73435fs8++6y2bt2qESNGaNasWc5tEAB8rKCgQPfcc49l/U9/+lNGhBumYINqDSB9EGzAVqZg49ChQzpx4oRlvbCw0I0tIcPFUrFx8eJFY9hBsAEAAOC93/zmN5IuhRWxVF8EAoFQ1dbjjz+uoqIiB3cHAP722GOPafjw4Zb1TAg3GBwOpDeCDdgqLy9P1157rWX90KFDljUqNuCGWCo2qNYAAADwrxEjRujBBx+UpJiqNp544gkFAgHNmjUrdD8AyFQVFRV67rnnMjLcINgA0hvBBmxnqto4evSoZY1gA24wBRs9KzaYrwEAAOBvq1atUlFRkbZu3apNmzZFvN27776rZ599VkVFRfr+97/v3gYBwMcyMdz44x//qGPHjoWt5eXl0SIZSCMEG7CdKdgwtaIi2IAbBg8erKysrLC106dPq729PfRnKjYAAAD8raioSKtWrZLUe9XG4sWLFQgE9OCDD2r69Oku7Q4A/C/Twg2qNYD0R7AB25mCjUAgYFkj2IBbTHM2urejItgAAADwvwcffFAjRoxQIBDQ3//931u+vmnTJm3dulVFRUV6/PHH3d8gAPhcJoUbBBtA+iPYgO1MwUZLS4tljWADbonWjsoUbJSUlDi6JwAAAMTvZz/7mSTpRz/6UdggcQaGA0BsMiXcINgA0h/BBmxnCjba2tosawQbcEu0ig1mbAAAAKSG6dOna9asWQoEAlq8eHFo/Uc/+pHeffddTZ8+nYHhABBFJoQbBBtA+iPYgO2Ki4s1bNiwqLcj2IBbEqnYINgAAADwp2DVxqZNm0Itb5999llJVGsAQKzSOdzYvXu3Wltbw9ZKSkp07bXXerQjAE4g2IAjTFUb3WVlZRFswDWmYIMZGwAAAKnJNEMjEAho1qxZDAwHgDika7hBtQaQGQg24IhowUZhYaFLOwHMraio2AAAAEhdjzzyiEaMGBH6MwPDASAx6RhuEGwAmYFgA46IFmxQrQE3JVKxwfBwAAAA/yoqKtKqVatCf/7+97+v8ePHe7chAEhh6RZuEGwAmYFgA44g2ICfRKvYYHg4AABA6gm2nhoxYoQeeeQRr7cDACktlnBj3759HuwsPi0tLXrvvfcs6zfffLMHuwHgJIINOIJgA37SW8XGmTNnLEPF8vPzNXDgQFf2BgAAgMT97Gc/Y2A4ANgkWrjx1FNPebCr+JiqNa655houXgTSEMEGHDF8+PBewwuCDbipt4oN5msAAACkrhEjRujBBx/0ehsAkDZ6Czc2bNjgwY7iYwo2qNYA0hPBBhxz/fXXR/wawQbc1FvFBvM1AAAAAAD4TDDc6NMn/LRhR0eHjhw54tGuYsN8DSBzEGzAMb21oyLYgJt6q9hgvgYAAAAAAOEqKio0evRoy/rhw4c92E3sCDaAzEGwAccQbMAv4q3YINgAAAAAAGQ6UzuqTz75xIOdxOb48eM6ePCgZZ1gA0hPBBtwTG/BRmFhoYs7QaYbNGiQ+vbtG7Z27tw5tbW1EWwAAAAAAGBgCjb8XLFRV1dnWRszZoz69evnwW4AOI1gA46hYgN+YqraOHnyJMEGAAAAAAAG6RBsjB071oOdAHADwQYcU15erpycHOPXCDbgNtOcjVOnTjE8HAAAAAAAgyuuuMKy5udg47333rOsEWwA6YtgA46KVLVBsAG3RarYYHg4AAAAAABWqTZjw1SxceONN3qwEwBuINiAo66//nrjOsEG3BZpgDitqAAAAAAAsEqlVlTnzp3Thx9+aFmnYgNIXwQbcBQVG/ALUysqZmwAAAAAAGCWSsGGqVrj2muvVWFhoQe7AeAGgg04imADfhFPxQYzNgAAAAAAma5///6Wz9IdHR06cuSIRzuKjPkaQOYh2ICjCDbgF6aKjcbGRstacXGx+vbt68aWAAAAAADwtVSp2mC+BpB5CDbgqEjBBqWAcJupYuPo0aOWNdpQAQAAAABwSaoMEDcFG1RsAOmNYAOOys/P19e//vWwtQceeMCj3SCTmSo2mK8BAAAAAEBkqVCx0dnZSbABZKAcrzeA9LdkyRJ1dHTopZde0r333qt58+Z5vSVkIFPFxsmTJy1rBBsAAAAAAFySCsFGXV2dOjs7w9auuOIKXX755R7tCIAbCDbguBEjRuinP/2p19tAhjNVbDQ3N1vWGBwOAAAAAMAlV1xxhWXNb8EGg8OBzEQrKgAZwVSxcfbsWcsaFRsAAAAAAFySCjM2GBwOZCaCDQAZwRRstLa2WtYINgAAAAAAuCRVWlH1RMUGkP4INgBkhIKCAvXr1y9sraOjw3I7gg0AAAAAAC4h2ADgVwQbADKGac5GT8zYAAAAAADgkv79+1s6IHR0dOjIkSMe7Sjcvn37dO7cubC14uJiXXPNNR7tCIBbCDYAZAxTO6qeqNgAAAAAAOAzfp6zwXwNIHMRbADIGLFUbBBsAACAdDF37lxlZWUZ/5s+fbrWrVunQCDg6B6ysrI0d+7cXm9TV1enrKwsrVu3ztG9AAAS4+d2VLShAjJXjtcbAAC3RKvY6Nu3b0xVHQAAAKmktrbWsvbzn/9cVVVVevXVV7V+/XrX9tLY2KjVq1ersrKSE08AkCIINgD4EcEGgIwRLbRgvgYAAEhH8+bNM67NnTtXGzZs0COPPKJJkyY58thdXV1hf968ebNWrFihysrK0NrYsWMttwMA+McVV1xhWfNLsPHee+9Z1gg2gMxAKyo4rri4WFlZWY6XuQPRRGtFRRsqAACQSR555BFJ0t69e73dCADA1/xasXHo0CE1NTWFreXl5Wn06NEe7QiAmwg2AGSMaBUbBBsAACCTFBQUGNfXrVunMWPGhOZxLFq0SA0NDVFvN3fuXG3fvj3sNt1nbMydO1dVVVWSpHHjxikrK0tS5Bkb27dv1/Tp08PmgvQ8fvC+dXV1WrdunYYOHaqsrCyNGTNGGzduDLttIBDQsmXLQrcZOnRoxL+byZYtW8LmlkS6f/DvsmrVqtBegrdpaGjQokWLQscw7RMA/Mavw8NpQwVkNoINABmDig0AAIDP7Ny5U5JUWFgYWps+fbqqqqpUUVGh2tpa1dTUaOPGjRo1alTYCaR169ZZbvf73/9ekydPNp5okqSHHnoo1IJq6dKlxtkfQatWrdLkyZN1/Phx1dbWqra2VsePH9fkyZONQ8Yffvhh1dTUaMmSJaqpqdGxY8c0Z84cbdmyJXSbv/iLv9CKFSu0YMEC1dbWasGCBVq9erWmTJkStbp83bp1mjlzpj799NPQfm6//XatXr1a9957r+X2jz32WOh2s2bNUnl5uRoaGjRlyhStXr069PevqKjQnDlztGrVql4fHwC85NeKDYINILMxYwNAxqBiAwAA4FLlwo4dO/TYY4+ptLRUX/rSlyRdqkjYtm2bli5dquXLl4duf9ddd2nKlCn6xje+oT179kiSampqNG3atLDB43fddZdGjRqlTz/91Pi4M2bM0KFDh7Rhw4Zeh4c3Njaqurpa06ZN06ZNm1RUVCRJuu+++zRr1ixVVVVp5syZKisrC93n+PHjev3110O3De7l+eef14wZM9TY2KgNGzaopqZGixcvDt3vpptu0qJFi9TU1BS6r8mrr76qiooKbd26NbQ2b948lZeXa8WKFaqrqwv7+zQ1NWnfvn1hx1y2bFlovby8PHSMwYMHq7q6WnfddVdoHQD8hGADgB9RsQEgY0Sr2GB4OAAASEfBtkfB/4qLizVz5kxJ0htvvBE6+f7Tn/5UkvToo4+G3b+8vFxLlixRfX196CRSRUWF3n//fW3ZskWNjY2h23V1dWnGjBlJ7Xfz5s2SpJUrV4YFA0VFRVq5cmXYbYKqq6vDbhsMCE6ePClJoRDkhRde0Pbt20MVGrNnz9axY8eiBgrr168PhTrdjRw50nj7ysrKsP0Eg5WFCxdaHmvp0qWSpN/+9re97gEAvNK/f3/LhYIdHR06cuSIRzu6hMHhQGajYgNAxqBiAwAAZCJTy6cbbrhBkyZNsqxPmzbNWLkwZsyYsD8vX75cU6ZMCQUkFRUVqqqqsrXq4IYbbrCsXXnllQkfb/PmzZo5c6YmT54s6dLf9a/+6q/0+c9/Pqz6ozcNDQ06fvy49u7dqx07dsQ8HyMYsJw8edLYSkuS3nrrrZiOBQBeGD58uE6dOhW29sknn2jYsGGe7Of48eP6+OOPLes33nijB7sB4AWCDQAZgxkbAAAgE82bN8/2Y5aXl2vfvn3asWOHnn/+eW3cuFHV1dWqrq5WbW2tI4+ZrGArrP/8z//Uhg0btGHDBm3btk2StHv37l6v8t2yZYseeOABNTU1Sbr0vvH222/X7bffrg0bNsS8h+DjAkCqGT58uKVy7fDhw7r55ps92Y+pDdWYMWOUm5vrwW4AeIFWVAAyBhUbAAAAvdu2bZtxkLapDVNRUZFmzJihZ555RseOHdO+fftClRt22Lt3r2Xt0KFDSR2zrKxMs2fP1vr169XV1aU333xTkrRixYqI92lsbNTMmTM1evRo7d69W11dXTp27JjWr1+vO+64I6bHzc/Pl3Speqarq8v4X/d5JQDgN36bs8F8DQAEGwAyRl5engYMGBDx6wQbAAAgkz300EOSpCeffDJsvaGhQStXrlRFRYXGjh2rQCCg6dOna9GiRWG3Ky8vV0lJSdLvqYLtrZYsWRIWsgQCAS1ZsiTsNrGqq6vTmDFjLK2jgu2ueqvsDbaRuueee8JOmgUCAdXU1MT0+OXl5aqoqNBjjz1mCY62bNmirKysiC2qAMAPrrjiCsual8EG8zUA0IoKQEYpLi7W2bNnjV9jeDgAAMhkM2bM0LRp07RixQo1NDTojjvuUHNzs1auXKmmpia98MILki5Vatxwww1avXq19u7dq29/+9uSpJ///Ofatm2bcaZHUDBIeOaZZ3TLLbcYW1aVlZWppqZG1dXVmjp1qqqrqyVJNTU1qq+vV01NTcwzMYKuuuoqSdKcOXO0cOFC3XLLLWpubg7ttWdI093gwYNVWlqq6upqnThxQiNHjtSBAwe0Zs2auPbwm9/8RlOmTNGoUaO0YMGC0HFWrFih0tLSuMMaAHBTKlRsMF8DyCxUbADIKJGuxhs4cGCoRQAAAECm2rp1q2pra1VfX6+qqipVV1dr9uzZ2rdvX9iw8ZUrV6q2tlbHjx9XVVVVqP3U5s2be52vccMNN6iyslKrV6/utWXV4sWL9eabb6qkpCR0/JKSEr355ptavHhx3H+voqIivfLKK1q6dKk2btwY+rt94QtfiDpfo6ysTG+88UYo9KmqqtLrr7+uf/iHf9Abb7whSfr3f//3qHsoLy/XG2+8odmzZ2vNmjWqqqrSmjVrtHTpUr3zzjtxhzUA4CZTsPHJJ594sBPp3Llz+vDDDy3rVGwAmSWrq6ury+tNID4rV66UpFAZtt8VFxcrEAjo1KlTKioq8no7yHB33323tm7dalkfOXKk3n33Xdf3AwAAAACA3zU0NFgGhV911VXGGUxO2759u2XG0bXXXqtdu3a5vhcA3qFiA0BGiTRAnPkaAAAAAACY+akVFfM1AEgEGwAyTKRWVMzXAAAAAADArH///pYLBTs6OnTkyBHX92Kar0GwAWQegg0AGYWKDQAAAAAA4ueXORsMDgcgEWwAyDCRKjYINgAAAAAAiMwP7ag6Ozup2AAgiWADQIahYgMAAAAAgPj5Idioq6tTZ2dn2NoVV1yhyy+/3NV9APAewQaAjELFBgAAAAAA8bviiissa24HGwwOBxBEsAEgo0Sq2GB4OAAAAAAAkflhxgbzNQAEEWwAyChUbAAAAHxm3bp1ysrKMp4o8pOsrCzNnTvX620AQEbzSyuqnqjYADITwQaAjMKMDQAAAAAA4kewAcBPCDYAZBRTsJGVlUWwAQAAAABAL7wONvbt26dz586FrRUXF+uaa65xbQ8A/INgA0BGycnJ0Te+8Y2wtW9961se7QYAAAAAgNTQv39/y8WCHR0dOnLkiCuPz3wNAN0RbADIOAsXLtQDDzygvn376pvf/Kb+63/9r15vCQAAwBHbt2/X9OnTlZWVpaysLE2fPl3bt2+37X7r1q3TmDFjQrebO3eu8XYNDQ1atGhR6HZjxozRxo0bLbcLBAJatmyZhg4dGnpcv8//AIBM4uUAcdpQAeiOYANAxhk/fryeeeYZnThxQv/0T/+kG264westAQAA2G7dunWaPHmyjh8/rpqaGtXW1ur48eOaPHlyr+HGqlWrQverra0Nu9+6devCjl9VVaWKigrV1taqpqZGv//97zV58uSwk08NDQ2aMmWKVq9eraVLl6q2tlYVFRWaM2eOVq1aFfbYs2bN0ooVK3T77bertrZWl112mcaNG2f/Pw4AICFetqMi2ADQXY7XGwAAAAAA2CsQCKiqqkrTpk3Tpk2bVFRUJEm67777NHXqVL300kuaNGmS5X6NjY2qrq423m/WrFmqqqrSzJkzVVZWppqaGk2bNk3r168P3f+uu+7SqFGj9Omnn4bWli1bpqamJu3bt0/l5eWSpHnz5mnw4MGqrq7WXXfdpfLycm3ZskXbtm1TbW2t5s2bF7rdHXfcoaqqKof+pQAA8fAy2HjvvfcsawQbQOaiYgMAAAAA0syOHTskSUuXLg2FE5JUVFSkPXv2aPny5cb7bd68WZK0cuVKy/1WrlwZdpuKigq9//772rJlixobGyVJ5eXl6urq0owZMyRdCko2bNighQsXhkKNoKVLl0qSfvvb30qSnn/+eUmXQpTuev4ZAOCdK664wrLmRrBx6NAhNTU1ha3169dPo0ePdvyxAfgTFRsAAAAAkGYOHTokSbrssssSur+pVeeVV14Z9ufly5drypQpmjlzpqRLQUdVVVWoAkOSTp48Gfrf7m2sunvrrbdCt6moqAgLVKRLocq0adMS+nsAAOzl1YwNBocD6IlgAwAAAAAQt/Lycu3bt087duzQ888/r40bN6q6ulrV1dVh7aQkacOGDdqwYUPCj5VoQAMAsJdXraiYrwGgJ1pRAQAAAECaCVZXdJ91EbRo0SKNGTOm1/vv3bvXshasAumuqKhIM2bM0DPPPKNjx45p3759ocoNScrPz5ck1dbWqqury/hfcEbH4MGDVV9fr0AgYHmcZEIRAIB9CDYA+AXBBgAAAACkmVtuuUWStGLFirCgoKGhQRs3btQXvvAF4/2CbaWWLFkSdr9AIKAlS5aEbhMIBDR9+nQtWrQo7P7l5eUqKSlRaWlp6M8VFRV67LHHLIHFli1blJWVFWpRFZyl8eSTT4bdbuPGjXH8zQEATvJqxgaDwwH0RCsqAAAAAEgzRUVF2rx5s2bOnKmpU6equrpazc3NoQHg1dXVxvuVlZWppqZG1dXVoftJUk1Njerr61VTU6OysjJJl+ZwrF69Wnv37tW3v/1tSdLPf/5zbdu2TbW1taFj/uY3v9GUKVM0atQoLViwQCNHjtSBAwe0YsUKlZaWhsKUGTNmqLKyUitWrFBDQ4PuuOMOvfrqq9qwYUMoKAEAeKugoEDFxcU6depUaK2jo0NHjhzRsGHDHHnM48eP6+OPP7asM2MDyGxUbAAAAABAGpoxY4befPNNlZSUqKqqStXV1Zo9e7beeOON0HBvk8WLF4fdr6qqSiUlJXrzzTe1ePHi0O1Wrlyp2tpaHT9+PHQ7Sdq8eXPYfI3y8nK98cYbmj17ttasWaOqqiqtWbNGS5cu1TvvvBMKSiRp/fr1qq2tVX19vaqqqlRfX68333xTt99+uwP/QgCARLg9QNzUhmrMmDHKzc117DEB+F9WV1dXl9ebQHyCV1kFS8H9rri4WIFAQKdOnVJRUZHX2wEAAAAAAECCvva1r+nVV18NW/vlL3+pu+++25HH++EPf6j//t//e9jaN77xDa1evdqRxwOQGqjYAAAAAAAAABATtweIM18DgAnBBgAAAAAAAICYuD1A3NSKivkaAAg2AAAAAAAAAMTEzRkb586d04cffmhZp2IDAMEGAAAAAAAAgJi42YrKVK1x7bXXqrCw0JHHA5A6CDYAAAAAAAAAxMTNYIP5GgAiIdgAAAAAAAAAEBM3Z2yYKjYINgBIBBsAAAAAAAAAYlRQUKDi4uKwtY6ODh05csT2x2JwOIBICDYAAAAAAAAAxMyNAeKdnZ1UbACIiGADAAAAAAAAQMzcmLNRV1enzs7OsLUrrrhCl19+ua2PAyA1EWwAAAAAAAAAiJkbwQaDwwH0hmADAAAAAAAAQMzcGCDOfA0AvSHYAAAAAAAAABAzN2ZsMF8DQG8INgAAAAAAAADEzK0ZGz0RbAAIItgAAAAAAAAAEDOng419+/bp3LlzYWvFxcW65pprbHsMAKmNYAMAAAAAAABAzJyescF8DQDREGwAAAAAAAAAiFlBQYGKi4vD1jo6OnTkyBFbjk8bKgDREGwAAAAAAAAAiIuTA8QJNgBEQ7ABAAAAAAAAIC5Oztkg2AAQDcEGAAAAAAAAgLg4FWwcOnRIx48fD1vr16+fRo8enfSxAaQPgg0AAAAAAAAAcXFqgDiDwwHEgmADAAAAAAAAQFycmrFBGyoAsSDYAAAAAAAAABAXp1pREWwAiAXBBgAAAAAAAIC4OBVsvPfee5Y1gg0APRFsAAAAAAAAAIiLEzM2Dhw4oI8//tiyzowNAD0RbAAAAAAAAACIS0FBgYqLi8PWOjo6dOTIkYSP+fvf/96y9vnPf165ubkJHxNAeiLYAAAAAAAAABA3uweIb9261bI2ffr0hI8HIH0RbAAAAAAAAACIm91zNkwVG7fffnvCxwOQvgg2AAAAAAAAAMTNzmDj9ddf1+nTp8PWioqKNHny5ISOByC9EWwAAAAAAAAAiJudA8RNbaio1gAQCcEGAAAAAAAAgLjZOWPD1IaK+RoAIiHYkLRx40aNGTNGWVlZysrK0qJFi9TQ0BDXMRobG7Vs2bKw40yfPl1btmxxaNcAAAAAAACAd+xqRdXU1KQdO3ZY1qnYABBJxgcby5Yt05w5cyRJtbW1Wrp0qVavXq0pU6bEHG40NDRowoQJWrFihb7whS+otrZWNTU1ev/99zVz5kytWrXKyb8CAAAAAAAA4Dq7gg1Ttcbo0aM1YsSIRLYFIAPkeL0BL9XV1WnFihWqrKzU+vXrQ+uVlZUaN26cli1bFrYeSU1NjZqamrR582bNmDEjtP7d735XU6dOVXV1te666y6Vl5c78vcAAAAAAAAA3GbXjA3TfA3aUAHoTUZXbGzYsEGS9Mgjj4Stjx07VgsXLtSGDRtiqtp47bXXJCks1JCkoqIiPfHEE2G3AQAAAAAAANJBQUGBiouLw9Y6Ojp05MiRuI5jqtigDRWA3mR0sLFp0yZJ0qRJkyxf+/KXvyzpUlVHNHv27FFXV5fxa4MGDUp8gwAAAAAAAICPJTtAfPfu3ZYqj+zsbIINAL3K6GCjvr5e06ZNM34t2Daqubk5qcf46U9/Kkm64YYbkjqOn7377ru65pprQkFRPLZu3ari4uKE7gsAAAAAAABvJTtnI1Ibqtzc3GS2BSDNZXSwIUmXXXZZr19/9dVXEz72li1btGHDBlVUVBirQtJFIBDQwYMH9d3vflcHDx6M636LFy9WIBBwbG8AAAAAAABwTrJzNmhDBSARGR9sOGXLli2aOXOmSktL9Zvf/Cbh43R2dlr+85vp06dr1qxZCgQCoZkisfjhD3+od999V+PHj2cgFAAAAAAAQApKpmLj/PnzBBsAEpLj9QbSUfdQ44033gi1tUrED37wAxt35pxVq1Zp06ZNevbZZ/X9739f48eP7/X2Bw8e1I9+9KPQfYuKipzfJAAAAAAAAGyVzIyNrVu3WubWXnnllbrxxhtt2RuA9JXWFRt1dXXKysqy/Nd9IPinn37a6zHuuOOOuB5z1apVtoUaktSnTx/Lf340YsQIPfLII5Kk7373u1Fv/8QTTygQCGjWrFlUawAAAAAAAKSoZCo2TNUanCcCEIuMrtioqKjQtm3bjF/buXOnJKmwsDDm4y1btkwrVqzQtGnT9Mtf/lJlZWVJ7/Fv/uZvLGsrV65M+rhOePzxx7Vp0ya9++672rRpk2bNmmW83datW/Xss8+qqKhIjz/+uLubBAAAAAAAgG3sDjZoQwUgFv68/N8mY8eOVVdXl+W/sWPHSlLoxHv3Co6gHTt2hI4Ri+6hxqZNm2wJNVJN96Cit6qN4ByOWFpWAQAAAAAAwL8SHR5+4MABffDBB5Z1gg0AsUjrYCOayspKSdLDDz+sQCAQWq+rq9Pq1atVWVkZUyupdevWacWKFaqsrNTWrVszel7Egw8+qPHjxysQCOjv//7vLV//4Q9/qK1bt4a1rgIAAAAAAEBqKigoUHFxcdhaR0eHjh492uv9tm7dalm79dZbNWTIEDu3ByBNZXQrqrFjx2rp0qVasWKFpk6dqurqah04cEArVqxQaWmpli9fbrlPVlaWJIUGGwUCAVVVVUm6NK9j7ty5xsd66KGHNGPGDIf+Jv7ym9/8Rtdcc42eeOIJPfjgg6H1QCAQqtZ4/PHHMzoAAgAAAAAASBfDhw/XqVOnwtYOHz6syy+/POJ9aEMFIBkZHWxI0vLlyzVy5EjV1NSEAoqFCxequro6pmqNjz/+OPT/I83rkOIfQp7KRowYoQcffFDPPvusFi9eHFr/0Y9+FBoY3j3wAAAAAAAAQOoaPny49uzZE7Z2+PBh3XzzzRHvQ7ABIBkZH2xI0rx58zRv3ryYbhus1AgKzvFAuFWrVunZZ5/Vpk2bQmvBgeHf//73vdsYAAAAAAAAbBXvnI033nhDp0+fDlsrKirS5MmTbd8bgPSU0TM24JyioiKtWrUqbC0QCOjBBx/U9OnTvdkUAAAAAAAAbDd8+HDLWm/Bhqlag/NFAOJBsAHHPPLIIxoxYkToz0VFRXr88ce92xAAAAAAAABsZwo2Pvnkk4i3pw0VgGQRbMBRP/vZz0L/n4HhAAAAAAAA6Seeio2mpibt2LHDsk6wASAezNiAo6ZPn65Zs2bp4MGDDAwHAAAAAABIQ/EEG1u3brWs3XDDDWFdPwAgGoINOO5nP/uZ3n33Xao1AAAAAAAA0lA8w8NpQwXADrSiguOKiooYAAUAAAAAAJCmCgoKVFxcHLbW0dGho0ePWm5rqtgg2AAQL4INAAAAAAAAAEmJpR1VXV2dGhsbw9b69OnDBbEA4kawAQAAAAAAACApsQQbkdpQ5eXlObYvAOmJYAMAAAAAAABAUmKZs8F8DQB2IdgAAAAAAAAAkJRoFRttbW3M1wBgG4INAAAAAAAAAEkxBRuffPJJ6P///ve/V2dnZ9jXy8rKdOONNzq+NwDph2ADAAAAAAAAQFKiVWzQhgqAnQg2AAAAAAAAACQlWrBBGyoAdiLYyFDr1q1TVlaW8b8xY8Zo2bJlamxsdHQPc+fOVVZWVtTbZWVlae7cuY7uBQAAAAAAAInrbXj4H//4R+3du9fy9enTpzu9LQBpKsfrDcBbS5cu1ciRI8PWduzYoRUrVmjNmjXat2+fioqKXNlLIBDQ888/r8LCQs2ePduVxwQAAAAAAEDyCgoKVFxcrFOnToXWOjo6dPToUWMbqltvvVUlJSVubhFAGiHYyHCVlZUaO3Zs2Nq8efN0yy23qKqqSj/72c+0ePFiRx57/fr1Wr9+fejPH3/8saqqqlRbWxt2u66uLkceHwAAAAAAAPYZPnx4WLAhXaraYL4GALvRigpG9913nyTprbfe8ngnAAAAAAAASAWR5myY5mvQhgpAMgg2YBSp/dT27ds1ffr00DyO6dOna/v27THdbt26dWG36T5jY926dRo3bpwkqaqqSllZWaqrq5NknrHR0NCgRYsWhc0F6Xn8urq60HHWrVunoUOHhm67ceNGy57XrVunMWPGhI45d+5c49/NpK6uTosWLQo9RqT7Z2Vlad26dVq1alVoLw0NDZIuteJatmxZ6BhDhw7VqlWrFAgEYtoDAAAAAACAl0xzNl5//XU1NzeHrRUWFmrKlClubQtAGiLYgFEwVBg8eHBobdWqVZo8ebKOHz+u2tpa1dbW6vjx45o8eXJYqFBXV2e8XVVVlSV8CJo4caKWLl0q6VJ7rNra2rDH7m779u0aNWqUNm7cqJqaGtXW1qqiokJVVVXGIeMPP/ywampqtGTJEtXU1OjYsWOaM2eOtmzZErrNunXrVFVVpYqKCtXW1qqmpka///3vNXny5NC/RW//VuPGjdNrr72mJUuWqLa2VkuXLtWGDRs0efJkyxD2xx57LPTvMmvWLJWXlysQCGjWrFlasWKFbr/9dtXW1mrBggWqrq7WrFmzen18AAAAAAAAPzBVbLzzzjuWNdpQAUgWMzZgsX37di1ZskSStGjRIklSY2OjqqurNW3aNG3atClU0XHfffdp1qxZqqqq0syZM1VWVqYNGzZIupTId7/d1KlTdeDAAeNjBud8rFixQnfccYfmzZsXcX/z589XaWmp3njjDZWXl0u6NBekvLxcK1as0EMPPaQZM2aEbn/8+PGwvdx1110aNWqUnn/++dDtampqNG3atLCZH8Hbffrpp73+e+3cuVOS9Morr6isrCy0Pm3aNM2cOVObN28O+/s0NTVZhrI///zz2rZtm5577rmwwek33XST5syZo40bNzJQHQAAAAAA+Jop2Ni/f79ljWADQLKo2Mhw48aNC7VOCv43efJkbdu2TZs3bw4FDps3b5YkrVy5MuyEfFFRkVauXBl2m5EjR0q6dLI+2GapqKhIe/bs0fLly5Pab11dnerr67VkyZJQqBH06KOPSpJ++tOfhq1XV1eH7Tl4v5MnT4bWKioq9P7772vLli2hCovy8nJ1dXWFhSQm8+bNU1dXV1ioIUmXXXaZ8faVlZWWVl81NTUqLS21hBezZ89WRUWFnn766V73AAAAAAAA4DVTsHHixAnLGvM1ACSLio0Mt3Tp0lAQEXTllVfqlltuMc7ZuOGGGyxrV155Zdif77vvPv385z9XVVWVJKm0tFQLFizQn//5n2vSpEm27Pu2226zrBUVFamioiKh4y1fvlxTpkzRzJkzJSnU2uquu+6yBCiRNDY26uTJk9q5c6cOHDigTZs2xfz49fX1qqioiNiqa9u2bTEfCwAAAAAAwAumGRs93XDDDbrmmmtc2A2AdEawkeEqKytDVRl2KSoq0tatW7V9+3a99NJL2rRpk1asWKEVK1aosrIyrN2TX5SXl2vfvn3asWOHnn/+eW3cuFHV1dWqrq5WbW1tr62xGhoadO+996q+vj60VllZqS984Qtha9HU19eHwiAAAAAAAIBUY6rY6Ik2VADsQCsqxGXv3r2WtUOHDhlvO2nSJC1fvlx79uzRoUOHtHDhQm3YsCHqMO5YvPXWW5a1QCAQV5DQU1FRkWbMmKFnnnlGx44d0759+0KVG7259957dezYMb355ps6deqUurq6tH79+tB8kliUlpaqsrJSXV1dEf8DAAAAAADws4KCAhUXF/d6G9pQAbADwQZiEmzRtGTJEgUCgdB6IBAIDRoP3mbZsmUaM2ZM2O3Kyso0atQoSVJ+fn7C+xg7dqwqKiq0cuXK0PyOoCeffFKS9NBDD8V1zEAgoOnTp1uCiPLycpWUlKi0tLTX+9fX12v06NGaNGlSWPuuZ555JuY9LFiwwBj6NDQ0aOjQoZo7d27MxwIAAAAAAPBKb1Ubffr0oWIDgC1oRYWYlJWVqaamRtXV1Zo6daqqq6slXRp6XV9fr5qamtDw7JtuukkrVqzQ1KlTVVVVpcLCQu3YsUOrV69WZWVlxJkVV111lSTp5z//uaRLQUnPgdyStHbtWk2ePFlTpkzRkiVLVFhYqFdffVUbNmzQtGnTog777qmoqEg33HCDVq9erb179+rb3/52aB/btm1TbW1tr/efNm2atm3bprlz5+qOO+5Qc3OzamtrdezYsZj38Oijj+r111/XuHHjtHDhQt1yyy1qbm7WypUr1dTUpEceeSSuvxMAAAAAAIAXhg8frj179hi/dvvttysvL8/lHQFIR1RsIGaLFy/Wm2++qZKSElVVVamqqkolJSV68803tXjx4tDtZs+erc2bN6ukpETV1dWqqqrSa6+9ptraWv3kJz+JePyioiLV1NRo27Ztqqqq0smTJ423mzRpkvbt26fZs2eHjl9fX6/a2lpt3bo1ob/bypUrVVtbq+PHj4f+bpK0efPmXudrSNKmTZtCbbaqqqpUW1urWbNmad++fSotLQ0FNb0pKirSpk2btHTpUr322muqqqpSdXW1Zs+erd27d9s2dB0AAAAAAMBJvQ0Qp1oDgF2yumjen3JWrlwpSaEWUAAAAAAAAIAf/K//9b+0fPly49def/11jR071uUdAUhHVGwAAAAAAAAAsEWkGRtlZWWEGgBsQ7ABAAAAAAAAwBaRgg3aUAGwE8EGAAAAAAAAAFtEmrExffp0dzcCIK0RbAAAAAAAAACwBRUbANxAsAEAAAAAAADAFgUFBbrnnnvC1r72ta+ppKTEox0BSEc5Xm8AAAAAAAAAQPr4r//1vyonJ0dbtmzRtGnT9OSTT3q9JQBphmADAAAAAAAAgG1uu+023XbbbV5vA0AaoxUV0l5dXZ2ysrK0bt06r7cCAAAAAAAAAEgSwQYAAAAAAAAAAEgZBBsAAAAAAAAAACBlEGwAAAAAAAAAAICUQbCRwRoaGrRo0SJlZWUpKytLY8aMscyhCM6nqKur06pVqzR06FBlZWVp+vTpqqursxwzEAho2bJlodsNHTpUq1atUiAQCLvd3LlzNXfuXDU0NGju3LmhPSxatEiNjY2W427fvl3Tp08P3S7S4/f8Oy1atEiffvppcv9QAAAAAAAAAADfyPF6A/DG9u3bNXnyZJWWlqqmpkaFhYV69dVXVVVVpVdffVXr168Pu/3DDz+s999/X0uWLJEkrVy5UuPGjdObb76pSZMmSboUasyaNUvbtm1TZWWl7rjjDh04cEDV1dV64YUXtHXr1rBj1tfXa8qUKZo9e7Zqa2v16quvavXq1Xrttde0Z8+e0O22bNmimTNnhu21pqbG8viNjY2aMmWKmpqatHTpUo0cOVI1NTVavXq1g/+SAAAAAAAAAAA3EWxkqPnz56u0tFRvvPGGysvLJUnz5s1TeXm5VqxYoYceekgzZswI3f7999/XO++8o7KyMknSXXfdpSlTpmj+/PmhEOL555/Xtm3b9Nxzz2n27Nmh+950002aM2eONm7cGLZeX18fdtt58+Zp8ODBWr16tRoaGlReXq5AIKAHHnhAFRUVev3111VUVCRJuu+++zR16lTdc889OnbsmCRp9erVampq0u7duzV27NjQMefOnasNGzY49C8JAAAAAAAAAHATragyUF1dnerr67VkyZJQqBH06KOPSpJ++tOfhq0vWbIkFGpIUnl5uZYsWaL6+no1NDRIkmpqalRaWhoWXkjS7NmzVVFRoaefftqyl563veWWWyRJra2tkqQdO3aoqalJTzzxRCjUkKSioiKtWrVKTU1N2r59uyRpzZo1mjZtWijUCHrkkUd6/fcAAAAAAAAAAKQOKjYy2G233WZZKyoqUkVFhWX9S1/6kmWtsLBQ0mchRH19vSoqKixzOoK2bdsW9x4PHTokSdq1a5eam5vDvnbgwAFJ0t69ezVp0iQ1NTVp6tSplmPccMMNcT8uAAAAAAAAAMCfCDZgq/r6elVVVdl+3BUrViR83+6VHgAAAAAAAACA1EYrqgz21ltvWdYCgYDq6+st68F2U90FKyYGDx4sSSotLVVlZaW6uroi/hevYFXI7t27Ix5z3rx5ocd//fXXLceoq6uL+3EBAAAAAAAAAP5EsJGBxo4dq4qKCq1cudISWDz55JOSpIceeihs/fHHH1cgEAj9uaGhQWvWrFFlZWVo9saCBQu0YcMGS5DQ0NCgoUOHau7cuXHv9fOf/7wkc8XGsmXLlJWVFXq8BQsWaNu2bdqyZUvY7Z555pm4HxcAAAAAAAAA4E+0ospQa9eu1eTJkzVlyhQtWbJEhYWFevXVV7VhwwZNmzZNM2bMCLt9fX29pk6dqurqajU3N2vlypWSpOXLl4du8+ijj+r111/XuHHjtHDhQt1yyy2h2zY1NSU0xLusrEybN2/WzJkzQ22ueu41OCz80Ucf1Zo1azRz5kwtXbpUI0eOVE1NjY4dO5b4PxQAAAAAAAAAwFcINjLUpEmTtG/fPtXU1Ki6ulqSVFFRodra2lBrp+5qamokKTQ/Y+HChaqurlZ5eXnoNkVFRdq0aZOefPJJbdq0SatXrw7ddtGiRaEAIl4zZszQm2++qR/+8IeWvd53331hj79v3z49+eSTWrNmjZqamlRZWanf/OY3GjVqVEKPDQAAAAAAAADwl6yuRAYfwFPBaoklS5Y4/lh1dXUaN25cxMADAAAAAAAAAAA3MWMDAAAAAAAAAACkDIINAAAAAAAAAACQMgg2AAAAAAAAAABAymB4OHo1duxYMYYFAAAAAAAAAOAXVGwAAAAAAAAAAICUQbABAAAAAAAAAABSBsEGAAAAAAAAgP+/PTsgAQAAABD0/3U7Ar0hwIbYAAAAAAAANsQGAAAAAACwITYAAAAAAIANsQEAAAAAAGyIDQAAAAAAYENsAAAAAAAAGwEXBlmXbP9+8QAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<PIL.PngImagePlugin.PngImageFile image mode=RGBA size=1590x784 at 0x24C33F33088>"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from PIL import Image\n",
    "str_dir = 'C:/Users/Alexandre/Desktop/Quant_Fin/'\n",
    "#str_dir = 'C:/Users/rafac/Documents/Mestrado/6T_Financas_Quant2/Trabalho/'\n",
    "myImage = Image.open(str_dir + 'Figura_Pairs_Trading.png')\n",
    "myImage"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Bibliotecas"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Abaixo estão todas as bibliotecas utilizadas no código."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import scipy as sp\n",
    "from statsmodels.tsa.stattools import adfuller\n",
    "from statsmodels.graphics.tsaplots import plot_acf, plot_pacf\n",
    "from statsmodels.tsa.statespace.sarimax import SARIMAX # Modelos SARIMA \n",
    "from pykalman import KalmanFilter\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Configurações"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As configurações necessárias às bibliotecas empregadas no modelo estão consolidadas abaixo."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Comando para que os gráficos sejam exibidos no corpo do texto\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "plt.rcParams['figure.dpi'] = 200 # Resolução dos gráficos impressos no pdf (dpi)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Estilo dos gráficos\n",
    "sns.set(color_codes = True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Parâmetros"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Esses são os parâmetros referentes aos modelos de pairs trading que são implementados no código, a serem detalhados mais a frente."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "alfa = 0.01 # Nível de significância utilizado nos testes de hipótese \n",
    "janela_dados = 252 # tamanho da janela de dados utilizada para estimação dos modelos\n",
    "janela_coint = 5 # janela de dias que será analisada para determinação da cointegração das ações de um par\n",
    "cost = 0.0005 # custo de execução \n",
    "th_KF = 0.25 # Número de desvios-padrão de distância da média para abertura de posição usando KF\n",
    "delta_beta_lim = 0 #variável para reduzir a frequência de ajustes de posição\n",
    "amostra = 2*252 # amostra usada para estimação do threshold ótimo de cada par"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Importação e Tratamento dos Dados"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Leitura da planilha com as séries de preços diários das ações de cada par\n",
    "\n",
    "str_plan_bd_stocks = 'BD_Stocks_v6'\n",
    "\n",
    "df_prices = pd.read_csv(filepath_or_buffer = str_dir + str_plan_bd_stocks + '.csv', sep=';')\n",
    "df_prices.index = df_prices['Dates']\n",
    "#df_prices.index = pd.to_datetime(df_prices['Dates'])\n",
    "df_prices = df_prices.drop(columns=['Dates'])\n",
    "df_prices = df_prices.dropna()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": false
   },
   "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>PETR3 BZ Equity</th>\n",
       "      <th>PETR4 BZ Equity</th>\n",
       "      <th>ITUB4 BZ Equity</th>\n",
       "      <th>BBDC4 BZ Equity</th>\n",
       "      <th>ITUB4 BZ Equity.1</th>\n",
       "      <th>BBAS3 BZ Equity</th>\n",
       "      <th>BBDC4 BZ Equity.1</th>\n",
       "      <th>BBAS3 BZ Equity.1</th>\n",
       "      <th>GOAU4 BZ Equity</th>\n",
       "      <th>GGBR4 BZ Equity</th>\n",
       "      <th>CSNA3 BZ Equity</th>\n",
       "      <th>USIM5 BZ Equity</th>\n",
       "      <th>ELET3 BZ Equity</th>\n",
       "      <th>ELET6 BZ Equity</th>\n",
       "      <th>BRML3 BZ Equity</th>\n",
       "      <th>MULT3 BZ Equity</th>\n",
       "      <th>MULT3 BZ Equity.1</th>\n",
       "      <th>IGTA3 BZ Equity</th>\n",
       "      <th>TAEE11 BZ Equity</th>\n",
       "      <th>TRPL4 BZ Equity</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Dates</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>07-06-20</th>\n",
       "      <td>3.141995</td>\n",
       "      <td>3.113960</td>\n",
       "      <td>3.322515</td>\n",
       "      <td>3.120160</td>\n",
       "      <td>3.322515</td>\n",
       "      <td>3.545875</td>\n",
       "      <td>3.120160</td>\n",
       "      <td>3.545875</td>\n",
       "      <td>1.996060</td>\n",
       "      <td>2.781920</td>\n",
       "      <td>2.415914</td>\n",
       "      <td>2.029463</td>\n",
       "      <td>3.490733</td>\n",
       "      <td>3.519277</td>\n",
       "      <td>2.391511</td>\n",
       "      <td>3.126761</td>\n",
       "      <td>3.126761</td>\n",
       "      <td>3.586293</td>\n",
       "      <td>3.349202</td>\n",
       "      <td>3.079154</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07-07-20</th>\n",
       "      <td>3.125005</td>\n",
       "      <td>3.101443</td>\n",
       "      <td>3.272227</td>\n",
       "      <td>3.077773</td>\n",
       "      <td>3.272227</td>\n",
       "      <td>3.504957</td>\n",
       "      <td>3.077773</td>\n",
       "      <td>3.504957</td>\n",
       "      <td>1.982380</td>\n",
       "      <td>2.758109</td>\n",
       "      <td>2.392426</td>\n",
       "      <td>2.008214</td>\n",
       "      <td>3.511545</td>\n",
       "      <td>3.531933</td>\n",
       "      <td>2.373975</td>\n",
       "      <td>3.109061</td>\n",
       "      <td>3.109061</td>\n",
       "      <td>3.567277</td>\n",
       "      <td>3.348500</td>\n",
       "      <td>3.044522</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07-08-20</th>\n",
       "      <td>3.149311</td>\n",
       "      <td>3.120160</td>\n",
       "      <td>3.293612</td>\n",
       "      <td>3.111291</td>\n",
       "      <td>3.293612</td>\n",
       "      <td>3.529004</td>\n",
       "      <td>3.111291</td>\n",
       "      <td>3.529004</td>\n",
       "      <td>1.997418</td>\n",
       "      <td>2.770086</td>\n",
       "      <td>2.430098</td>\n",
       "      <td>2.041220</td>\n",
       "      <td>3.521939</td>\n",
       "      <td>3.543854</td>\n",
       "      <td>2.386007</td>\n",
       "      <td>3.128951</td>\n",
       "      <td>3.128951</td>\n",
       "      <td>3.569533</td>\n",
       "      <td>3.362457</td>\n",
       "      <td>3.053529</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07-09-20</th>\n",
       "      <td>3.120160</td>\n",
       "      <td>3.097386</td>\n",
       "      <td>3.293983</td>\n",
       "      <td>3.088767</td>\n",
       "      <td>3.293983</td>\n",
       "      <td>3.525183</td>\n",
       "      <td>3.088767</td>\n",
       "      <td>3.525183</td>\n",
       "      <td>1.991976</td>\n",
       "      <td>2.775709</td>\n",
       "      <td>2.405142</td>\n",
       "      <td>2.010895</td>\n",
       "      <td>3.604410</td>\n",
       "      <td>3.633631</td>\n",
       "      <td>2.400619</td>\n",
       "      <td>3.151453</td>\n",
       "      <td>3.151453</td>\n",
       "      <td>3.585461</td>\n",
       "      <td>3.351657</td>\n",
       "      <td>3.039749</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07-10-20</th>\n",
       "      <td>3.135494</td>\n",
       "      <td>3.113960</td>\n",
       "      <td>3.303217</td>\n",
       "      <td>3.103240</td>\n",
       "      <td>3.303217</td>\n",
       "      <td>3.530470</td>\n",
       "      <td>3.103240</td>\n",
       "      <td>3.530470</td>\n",
       "      <td>1.990610</td>\n",
       "      <td>2.778819</td>\n",
       "      <td>2.418589</td>\n",
       "      <td>2.005526</td>\n",
       "      <td>3.590991</td>\n",
       "      <td>3.613347</td>\n",
       "      <td>2.422144</td>\n",
       "      <td>3.152736</td>\n",
       "      <td>3.152736</td>\n",
       "      <td>3.598408</td>\n",
       "      <td>3.349553</td>\n",
       "      <td>3.042139</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07-13-2020</th>\n",
       "      <td>3.128951</td>\n",
       "      <td>3.098289</td>\n",
       "      <td>3.288402</td>\n",
       "      <td>3.089223</td>\n",
       "      <td>3.288402</td>\n",
       "      <td>3.520461</td>\n",
       "      <td>3.089223</td>\n",
       "      <td>3.520461</td>\n",
       "      <td>2.002830</td>\n",
       "      <td>2.788708</td>\n",
       "      <td>2.457021</td>\n",
       "      <td>2.000128</td>\n",
       "      <td>3.558486</td>\n",
       "      <td>3.589888</td>\n",
       "      <td>2.380472</td>\n",
       "      <td>3.132010</td>\n",
       "      <td>3.132010</td>\n",
       "      <td>3.566147</td>\n",
       "      <td>3.348148</td>\n",
       "      <td>3.031099</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07-14-2020</th>\n",
       "      <td>3.162940</td>\n",
       "      <td>3.131137</td>\n",
       "      <td>3.305054</td>\n",
       "      <td>3.105931</td>\n",
       "      <td>3.305054</td>\n",
       "      <td>3.536602</td>\n",
       "      <td>3.105931</td>\n",
       "      <td>3.536602</td>\n",
       "      <td>2.032088</td>\n",
       "      <td>2.809403</td>\n",
       "      <td>2.500616</td>\n",
       "      <td>2.017566</td>\n",
       "      <td>3.549617</td>\n",
       "      <td>3.587124</td>\n",
       "      <td>2.384165</td>\n",
       "      <td>3.129826</td>\n",
       "      <td>3.129826</td>\n",
       "      <td>3.563883</td>\n",
       "      <td>3.342508</td>\n",
       "      <td>3.037354</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07-15-2020</th>\n",
       "      <td>3.186353</td>\n",
       "      <td>3.150169</td>\n",
       "      <td>3.312002</td>\n",
       "      <td>3.122365</td>\n",
       "      <td>3.312002</td>\n",
       "      <td>3.546451</td>\n",
       "      <td>3.122365</td>\n",
       "      <td>3.546451</td>\n",
       "      <td>2.036012</td>\n",
       "      <td>2.806386</td>\n",
       "      <td>2.479056</td>\n",
       "      <td>2.016235</td>\n",
       "      <td>3.582129</td>\n",
       "      <td>3.627004</td>\n",
       "      <td>2.402430</td>\n",
       "      <td>3.124565</td>\n",
       "      <td>3.124565</td>\n",
       "      <td>3.575151</td>\n",
       "      <td>3.343568</td>\n",
       "      <td>3.047850</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07-16-2020</th>\n",
       "      <td>3.162517</td>\n",
       "      <td>3.123246</td>\n",
       "      <td>3.295096</td>\n",
       "      <td>3.104587</td>\n",
       "      <td>3.295096</td>\n",
       "      <td>3.538057</td>\n",
       "      <td>3.104587</td>\n",
       "      <td>3.538057</td>\n",
       "      <td>2.021548</td>\n",
       "      <td>2.787477</td>\n",
       "      <td>2.458734</td>\n",
       "      <td>2.000128</td>\n",
       "      <td>3.564732</td>\n",
       "      <td>3.597038</td>\n",
       "      <td>2.398804</td>\n",
       "      <td>3.114848</td>\n",
       "      <td>3.114848</td>\n",
       "      <td>3.565581</td>\n",
       "      <td>3.336481</td>\n",
       "      <td>3.047850</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07-17-2020</th>\n",
       "      <td>3.163363</td>\n",
       "      <td>3.124125</td>\n",
       "      <td>3.303952</td>\n",
       "      <td>3.109507</td>\n",
       "      <td>3.303952</td>\n",
       "      <td>3.568405</td>\n",
       "      <td>3.109507</td>\n",
       "      <td>3.568405</td>\n",
       "      <td>2.028148</td>\n",
       "      <td>2.791778</td>\n",
       "      <td>2.476538</td>\n",
       "      <td>2.013569</td>\n",
       "      <td>3.698830</td>\n",
       "      <td>3.708927</td>\n",
       "      <td>2.427454</td>\n",
       "      <td>3.125444</td>\n",
       "      <td>3.125444</td>\n",
       "      <td>3.573188</td>\n",
       "      <td>3.338967</td>\n",
       "      <td>3.040228</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07-20-2020</th>\n",
       "      <td>3.157000</td>\n",
       "      <td>3.124125</td>\n",
       "      <td>3.306887</td>\n",
       "      <td>3.118834</td>\n",
       "      <td>3.306887</td>\n",
       "      <td>3.564449</td>\n",
       "      <td>3.118834</td>\n",
       "      <td>3.564449</td>\n",
       "      <td>2.038620</td>\n",
       "      <td>2.791778</td>\n",
       "      <td>2.474014</td>\n",
       "      <td>2.009555</td>\n",
       "      <td>3.688879</td>\n",
       "      <td>3.698830</td>\n",
       "      <td>2.442347</td>\n",
       "      <td>3.143721</td>\n",
       "      <td>3.143721</td>\n",
       "      <td>3.589059</td>\n",
       "      <td>3.343568</td>\n",
       "      <td>3.058707</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07-21-2020</th>\n",
       "      <td>3.177637</td>\n",
       "      <td>3.151453</td>\n",
       "      <td>3.321432</td>\n",
       "      <td>3.145445</td>\n",
       "      <td>3.321432</td>\n",
       "      <td>3.569251</td>\n",
       "      <td>3.145445</td>\n",
       "      <td>3.569251</td>\n",
       "      <td>2.024193</td>\n",
       "      <td>2.781920</td>\n",
       "      <td>2.456164</td>\n",
       "      <td>2.000128</td>\n",
       "      <td>3.657389</td>\n",
       "      <td>3.671988</td>\n",
       "      <td>2.441477</td>\n",
       "      <td>3.136798</td>\n",
       "      <td>3.136798</td>\n",
       "      <td>3.604410</td>\n",
       "      <td>3.340739</td>\n",
       "      <td>3.063391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07-22-2020</th>\n",
       "      <td>3.167161</td>\n",
       "      <td>3.137666</td>\n",
       "      <td>3.298426</td>\n",
       "      <td>3.128075</td>\n",
       "      <td>3.298426</td>\n",
       "      <td>3.561046</td>\n",
       "      <td>3.128075</td>\n",
       "      <td>3.561046</td>\n",
       "      <td>2.075684</td>\n",
       "      <td>2.817801</td>\n",
       "      <td>2.468947</td>\n",
       "      <td>2.016235</td>\n",
       "      <td>3.666378</td>\n",
       "      <td>3.681099</td>\n",
       "      <td>2.391511</td>\n",
       "      <td>3.098740</td>\n",
       "      <td>3.098740</td>\n",
       "      <td>3.562466</td>\n",
       "      <td>3.343568</td>\n",
       "      <td>3.068053</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07-23-2020</th>\n",
       "      <td>3.145445</td>\n",
       "      <td>3.116622</td>\n",
       "      <td>3.277145</td>\n",
       "      <td>3.113515</td>\n",
       "      <td>3.277145</td>\n",
       "      <td>3.536602</td>\n",
       "      <td>3.113515</td>\n",
       "      <td>3.536602</td>\n",
       "      <td>2.037317</td>\n",
       "      <td>2.796671</td>\n",
       "      <td>2.455306</td>\n",
       "      <td>1.989243</td>\n",
       "      <td>3.649619</td>\n",
       "      <td>3.669697</td>\n",
       "      <td>2.343727</td>\n",
       "      <td>3.082369</td>\n",
       "      <td>3.082369</td>\n",
       "      <td>3.533395</td>\n",
       "      <td>3.338258</td>\n",
       "      <td>3.055886</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07-24-2020</th>\n",
       "      <td>3.149740</td>\n",
       "      <td>3.123686</td>\n",
       "      <td>3.276012</td>\n",
       "      <td>3.119718</td>\n",
       "      <td>3.276012</td>\n",
       "      <td>3.524889</td>\n",
       "      <td>3.119718</td>\n",
       "      <td>3.524889</td>\n",
       "      <td>2.065596</td>\n",
       "      <td>2.841415</td>\n",
       "      <td>2.451867</td>\n",
       "      <td>2.021548</td>\n",
       "      <td>3.640214</td>\n",
       "      <td>3.664843</td>\n",
       "      <td>2.332144</td>\n",
       "      <td>3.061988</td>\n",
       "      <td>3.061988</td>\n",
       "      <td>3.516904</td>\n",
       "      <td>3.344980</td>\n",
       "      <td>3.061052</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07-27-2020</th>\n",
       "      <td>3.168424</td>\n",
       "      <td>3.144152</td>\n",
       "      <td>3.325036</td>\n",
       "      <td>3.164631</td>\n",
       "      <td>3.325036</td>\n",
       "      <td>3.552773</td>\n",
       "      <td>3.164631</td>\n",
       "      <td>3.552773</td>\n",
       "      <td>2.110213</td>\n",
       "      <td>2.881443</td>\n",
       "      <td>2.504709</td>\n",
       "      <td>2.090629</td>\n",
       "      <td>3.671225</td>\n",
       "      <td>3.683364</td>\n",
       "      <td>2.315501</td>\n",
       "      <td>3.056827</td>\n",
       "      <td>3.056827</td>\n",
       "      <td>3.519573</td>\n",
       "      <td>3.354804</td>\n",
       "      <td>3.080073</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07-28-2020</th>\n",
       "      <td>3.152309</td>\n",
       "      <td>3.126761</td>\n",
       "      <td>3.319264</td>\n",
       "      <td>3.157851</td>\n",
       "      <td>3.319264</td>\n",
       "      <td>3.556205</td>\n",
       "      <td>3.157851</td>\n",
       "      <td>3.556205</td>\n",
       "      <td>2.100469</td>\n",
       "      <td>2.875258</td>\n",
       "      <td>2.509599</td>\n",
       "      <td>2.078191</td>\n",
       "      <td>3.660994</td>\n",
       "      <td>3.680343</td>\n",
       "      <td>2.333114</td>\n",
       "      <td>3.072230</td>\n",
       "      <td>3.072230</td>\n",
       "      <td>3.543275</td>\n",
       "      <td>3.367296</td>\n",
       "      <td>3.068518</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07-29-2020</th>\n",
       "      <td>3.170945</td>\n",
       "      <td>3.142858</td>\n",
       "      <td>3.342508</td>\n",
       "      <td>3.188829</td>\n",
       "      <td>3.342508</td>\n",
       "      <td>3.582129</td>\n",
       "      <td>3.188829</td>\n",
       "      <td>3.582129</td>\n",
       "      <td>2.106570</td>\n",
       "      <td>2.889816</td>\n",
       "      <td>2.564949</td>\n",
       "      <td>2.098018</td>\n",
       "      <td>3.682610</td>\n",
       "      <td>3.703275</td>\n",
       "      <td>2.336020</td>\n",
       "      <td>3.086487</td>\n",
       "      <td>3.086487</td>\n",
       "      <td>3.543275</td>\n",
       "      <td>3.365225</td>\n",
       "      <td>3.066191</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07-30-2020</th>\n",
       "      <td>3.149311</td>\n",
       "      <td>3.127637</td>\n",
       "      <td>3.325396</td>\n",
       "      <td>3.153163</td>\n",
       "      <td>3.325396</td>\n",
       "      <td>3.549617</td>\n",
       "      <td>3.153163</td>\n",
       "      <td>3.549617</td>\n",
       "      <td>2.094330</td>\n",
       "      <td>2.875822</td>\n",
       "      <td>2.547881</td>\n",
       "      <td>2.105353</td>\n",
       "      <td>3.641526</td>\n",
       "      <td>3.668422</td>\n",
       "      <td>2.332144</td>\n",
       "      <td>3.072693</td>\n",
       "      <td>3.072693</td>\n",
       "      <td>3.535145</td>\n",
       "      <td>3.374169</td>\n",
       "      <td>3.080073</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>07-31-2020</th>\n",
       "      <td>3.122365</td>\n",
       "      <td>3.100092</td>\n",
       "      <td>3.291754</td>\n",
       "      <td>3.110845</td>\n",
       "      <td>3.291754</td>\n",
       "      <td>3.513931</td>\n",
       "      <td>3.110845</td>\n",
       "      <td>3.513931</td>\n",
       "      <td>2.084429</td>\n",
       "      <td>2.859340</td>\n",
       "      <td>2.505526</td>\n",
       "      <td>2.095561</td>\n",
       "      <td>3.609295</td>\n",
       "      <td>3.649359</td>\n",
       "      <td>2.292535</td>\n",
       "      <td>3.051167</td>\n",
       "      <td>3.051167</td>\n",
       "      <td>3.513037</td>\n",
       "      <td>3.365916</td>\n",
       "      <td>3.115292</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            PETR3 BZ Equity  PETR4 BZ Equity  ITUB4 BZ Equity  \\\n",
       "Dates                                                           \n",
       "07-06-20           3.141995         3.113960         3.322515   \n",
       "07-07-20           3.125005         3.101443         3.272227   \n",
       "07-08-20           3.149311         3.120160         3.293612   \n",
       "07-09-20           3.120160         3.097386         3.293983   \n",
       "07-10-20           3.135494         3.113960         3.303217   \n",
       "07-13-2020         3.128951         3.098289         3.288402   \n",
       "07-14-2020         3.162940         3.131137         3.305054   \n",
       "07-15-2020         3.186353         3.150169         3.312002   \n",
       "07-16-2020         3.162517         3.123246         3.295096   \n",
       "07-17-2020         3.163363         3.124125         3.303952   \n",
       "07-20-2020         3.157000         3.124125         3.306887   \n",
       "07-21-2020         3.177637         3.151453         3.321432   \n",
       "07-22-2020         3.167161         3.137666         3.298426   \n",
       "07-23-2020         3.145445         3.116622         3.277145   \n",
       "07-24-2020         3.149740         3.123686         3.276012   \n",
       "07-27-2020         3.168424         3.144152         3.325036   \n",
       "07-28-2020         3.152309         3.126761         3.319264   \n",
       "07-29-2020         3.170945         3.142858         3.342508   \n",
       "07-30-2020         3.149311         3.127637         3.325396   \n",
       "07-31-2020         3.122365         3.100092         3.291754   \n",
       "\n",
       "            BBDC4 BZ Equity  ITUB4 BZ Equity.1  BBAS3 BZ Equity  \\\n",
       "Dates                                                             \n",
       "07-06-20           3.120160           3.322515         3.545875   \n",
       "07-07-20           3.077773           3.272227         3.504957   \n",
       "07-08-20           3.111291           3.293612         3.529004   \n",
       "07-09-20           3.088767           3.293983         3.525183   \n",
       "07-10-20           3.103240           3.303217         3.530470   \n",
       "07-13-2020         3.089223           3.288402         3.520461   \n",
       "07-14-2020         3.105931           3.305054         3.536602   \n",
       "07-15-2020         3.122365           3.312002         3.546451   \n",
       "07-16-2020         3.104587           3.295096         3.538057   \n",
       "07-17-2020         3.109507           3.303952         3.568405   \n",
       "07-20-2020         3.118834           3.306887         3.564449   \n",
       "07-21-2020         3.145445           3.321432         3.569251   \n",
       "07-22-2020         3.128075           3.298426         3.561046   \n",
       "07-23-2020         3.113515           3.277145         3.536602   \n",
       "07-24-2020         3.119718           3.276012         3.524889   \n",
       "07-27-2020         3.164631           3.325036         3.552773   \n",
       "07-28-2020         3.157851           3.319264         3.556205   \n",
       "07-29-2020         3.188829           3.342508         3.582129   \n",
       "07-30-2020         3.153163           3.325396         3.549617   \n",
       "07-31-2020         3.110845           3.291754         3.513931   \n",
       "\n",
       "            BBDC4 BZ Equity.1  BBAS3 BZ Equity.1  GOAU4 BZ Equity  \\\n",
       "Dates                                                               \n",
       "07-06-20             3.120160           3.545875         1.996060   \n",
       "07-07-20             3.077773           3.504957         1.982380   \n",
       "07-08-20             3.111291           3.529004         1.997418   \n",
       "07-09-20             3.088767           3.525183         1.991976   \n",
       "07-10-20             3.103240           3.530470         1.990610   \n",
       "07-13-2020           3.089223           3.520461         2.002830   \n",
       "07-14-2020           3.105931           3.536602         2.032088   \n",
       "07-15-2020           3.122365           3.546451         2.036012   \n",
       "07-16-2020           3.104587           3.538057         2.021548   \n",
       "07-17-2020           3.109507           3.568405         2.028148   \n",
       "07-20-2020           3.118834           3.564449         2.038620   \n",
       "07-21-2020           3.145445           3.569251         2.024193   \n",
       "07-22-2020           3.128075           3.561046         2.075684   \n",
       "07-23-2020           3.113515           3.536602         2.037317   \n",
       "07-24-2020           3.119718           3.524889         2.065596   \n",
       "07-27-2020           3.164631           3.552773         2.110213   \n",
       "07-28-2020           3.157851           3.556205         2.100469   \n",
       "07-29-2020           3.188829           3.582129         2.106570   \n",
       "07-30-2020           3.153163           3.549617         2.094330   \n",
       "07-31-2020           3.110845           3.513931         2.084429   \n",
       "\n",
       "            GGBR4 BZ Equity  CSNA3 BZ Equity  USIM5 BZ Equity  \\\n",
       "Dates                                                           \n",
       "07-06-20           2.781920         2.415914         2.029463   \n",
       "07-07-20           2.758109         2.392426         2.008214   \n",
       "07-08-20           2.770086         2.430098         2.041220   \n",
       "07-09-20           2.775709         2.405142         2.010895   \n",
       "07-10-20           2.778819         2.418589         2.005526   \n",
       "07-13-2020         2.788708         2.457021         2.000128   \n",
       "07-14-2020         2.809403         2.500616         2.017566   \n",
       "07-15-2020         2.806386         2.479056         2.016235   \n",
       "07-16-2020         2.787477         2.458734         2.000128   \n",
       "07-17-2020         2.791778         2.476538         2.013569   \n",
       "07-20-2020         2.791778         2.474014         2.009555   \n",
       "07-21-2020         2.781920         2.456164         2.000128   \n",
       "07-22-2020         2.817801         2.468947         2.016235   \n",
       "07-23-2020         2.796671         2.455306         1.989243   \n",
       "07-24-2020         2.841415         2.451867         2.021548   \n",
       "07-27-2020         2.881443         2.504709         2.090629   \n",
       "07-28-2020         2.875258         2.509599         2.078191   \n",
       "07-29-2020         2.889816         2.564949         2.098018   \n",
       "07-30-2020         2.875822         2.547881         2.105353   \n",
       "07-31-2020         2.859340         2.505526         2.095561   \n",
       "\n",
       "            ELET3 BZ Equity  ELET6 BZ Equity  BRML3 BZ Equity  \\\n",
       "Dates                                                           \n",
       "07-06-20           3.490733         3.519277         2.391511   \n",
       "07-07-20           3.511545         3.531933         2.373975   \n",
       "07-08-20           3.521939         3.543854         2.386007   \n",
       "07-09-20           3.604410         3.633631         2.400619   \n",
       "07-10-20           3.590991         3.613347         2.422144   \n",
       "07-13-2020         3.558486         3.589888         2.380472   \n",
       "07-14-2020         3.549617         3.587124         2.384165   \n",
       "07-15-2020         3.582129         3.627004         2.402430   \n",
       "07-16-2020         3.564732         3.597038         2.398804   \n",
       "07-17-2020         3.698830         3.708927         2.427454   \n",
       "07-20-2020         3.688879         3.698830         2.442347   \n",
       "07-21-2020         3.657389         3.671988         2.441477   \n",
       "07-22-2020         3.666378         3.681099         2.391511   \n",
       "07-23-2020         3.649619         3.669697         2.343727   \n",
       "07-24-2020         3.640214         3.664843         2.332144   \n",
       "07-27-2020         3.671225         3.683364         2.315501   \n",
       "07-28-2020         3.660994         3.680343         2.333114   \n",
       "07-29-2020         3.682610         3.703275         2.336020   \n",
       "07-30-2020         3.641526         3.668422         2.332144   \n",
       "07-31-2020         3.609295         3.649359         2.292535   \n",
       "\n",
       "            MULT3 BZ Equity  MULT3 BZ Equity.1  IGTA3 BZ Equity  \\\n",
       "Dates                                                             \n",
       "07-06-20           3.126761           3.126761         3.586293   \n",
       "07-07-20           3.109061           3.109061         3.567277   \n",
       "07-08-20           3.128951           3.128951         3.569533   \n",
       "07-09-20           3.151453           3.151453         3.585461   \n",
       "07-10-20           3.152736           3.152736         3.598408   \n",
       "07-13-2020         3.132010           3.132010         3.566147   \n",
       "07-14-2020         3.129826           3.129826         3.563883   \n",
       "07-15-2020         3.124565           3.124565         3.575151   \n",
       "07-16-2020         3.114848           3.114848         3.565581   \n",
       "07-17-2020         3.125444           3.125444         3.573188   \n",
       "07-20-2020         3.143721           3.143721         3.589059   \n",
       "07-21-2020         3.136798           3.136798         3.604410   \n",
       "07-22-2020         3.098740           3.098740         3.562466   \n",
       "07-23-2020         3.082369           3.082369         3.533395   \n",
       "07-24-2020         3.061988           3.061988         3.516904   \n",
       "07-27-2020         3.056827           3.056827         3.519573   \n",
       "07-28-2020         3.072230           3.072230         3.543275   \n",
       "07-29-2020         3.086487           3.086487         3.543275   \n",
       "07-30-2020         3.072693           3.072693         3.535145   \n",
       "07-31-2020         3.051167           3.051167         3.513037   \n",
       "\n",
       "            TAEE11 BZ Equity  TRPL4 BZ Equity  \n",
       "Dates                                          \n",
       "07-06-20            3.349202         3.079154  \n",
       "07-07-20            3.348500         3.044522  \n",
       "07-08-20            3.362457         3.053529  \n",
       "07-09-20            3.351657         3.039749  \n",
       "07-10-20            3.349553         3.042139  \n",
       "07-13-2020          3.348148         3.031099  \n",
       "07-14-2020          3.342508         3.037354  \n",
       "07-15-2020          3.343568         3.047850  \n",
       "07-16-2020          3.336481         3.047850  \n",
       "07-17-2020          3.338967         3.040228  \n",
       "07-20-2020          3.343568         3.058707  \n",
       "07-21-2020          3.340739         3.063391  \n",
       "07-22-2020          3.343568         3.068053  \n",
       "07-23-2020          3.338258         3.055886  \n",
       "07-24-2020          3.344980         3.061052  \n",
       "07-27-2020          3.354804         3.080073  \n",
       "07-28-2020          3.367296         3.068518  \n",
       "07-29-2020          3.365225         3.066191  \n",
       "07-30-2020          3.374169         3.080073  \n",
       "07-31-2020          3.365916         3.115292  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Log preços - serão usados nos testes de cointegração\n",
    "df_log_prices = np.log(df_prices)\n",
    "df_log_prices.tail(20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Planilha com índices\n",
    "str_plan_indices = 'Indices'\n",
    "df_indices = pd.read_csv(filepath_or_buffer = str_dir + str_plan_indices + '.csv', sep=';')\n",
    "df_indices.index = pd.to_datetime(df_indices['Dates'])\n",
    "df_indices = df_indices.drop(columns=['Dates'])\n",
    "df_indices = df_indices.dropna()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Cálculo dos Spreads"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Regressão Usando Filtro de Kalman"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A função abaixo implementa a regressão do log preço de uma ação contra o log preço da outra ação do par utilizando o filtro de Kalman. Ela retorna os coeficientes estimados ao longo do tempo, o resíduo (spread), as matrizes de covariância dos coeficientes e um objeto contendo o filtro de Kalman ajustado.\n",
    "\n",
    "Essa função é empregada na função spread_KF para computar o spread de cada par de ações da base."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "'''Parâmetros da função:\n",
    "-> y, x : log preços do par de ações.\n",
    "-> delta: parâmetro relacionado à variância. Quanto mais próximo de 1, mais voláteis são os betas, melhor o ajuste\n",
    "porém pior a capacidade preditiva do filtro. Quanto mais próximo de zero, mais estáveis são os betas, porém a capacidade \n",
    "preditiva é ruim. Há então um ponto ótimo no meio desse intervalo. \n",
    "-> obs_cov: matriz de covariância das observações - variáveis de estado.\n",
    "-> init_mean, init_cov: são os valores iniciais da média e da matriz de covariância dos betas para inicialização do filtro.\n",
    "-> intercept: se você deseja rodar regressão com (TRUE) ou sem (FALSE) intercepto.\n",
    "-> em: se você deseja utilizar o algoritmo \"em\" para estimar os valores iniciais da média e da matriz de covariância dos betas.\n",
    "''' \n",
    "\n",
    "def KFReg(y, x, delta, obs_cov, init_mean, init_cov, intercept = False, em = False):\n",
    "    if intercept == True: #caso regressão utilize intercepto\n",
    "        obs_mat = np.vstack([x, np.ones(x.shape)]).T[:, np.newaxis] \n",
    "        if em == False:\n",
    "            trans_cov = delta / (1 - delta) * np.eye(2)\n",
    "            kf = KalmanFilter(n_dim_obs = 1, \n",
    "                              n_dim_state = 2, \n",
    "                              initial_state_mean = init_mean,\n",
    "                              initial_state_covariance = init_cov,\n",
    "                              transition_matrices = np.eye(2), # Essa é a matriz que define o random walk\n",
    "                              observation_matrices = obs_mat,\n",
    "                              observation_covariance = obs_cov,\n",
    "                              transition_covariance = trans_cov)\n",
    "            state_means, state_covs = kf.filter(y.values)\n",
    "        else:\n",
    "            kf = KalmanFilter(n_dim_obs = 1, \n",
    "                              n_dim_state = 2, \n",
    "                              initial_state_mean = init_mean, \n",
    "                              initial_state_covariance = init_cov,\n",
    "                              observation_matrices = obs_mat)\n",
    "            state_means, state_covs = kf.filter(y.values)\n",
    "        beta0 = state_means[:,1]\n",
    "        beta1 = state_means[:,0]\n",
    "        resid = y - beta0 - beta1*x\n",
    "        return beta0, beta1, resid, state_covs, kf\n",
    "    else:\n",
    "        obs_mat = np.vstack([x]).T[:, np.newaxis]\n",
    "        if em == False:\n",
    "            trans_cov = delta / (1 - delta)\n",
    "            kf = KalmanFilter(n_dim_obs = 1, \n",
    "                              n_dim_state = 1, \n",
    "                              initial_state_mean = init_mean,\n",
    "                              initial_state_covariance = init_cov,\n",
    "                              transition_matrices = 1,\n",
    "                              observation_matrices = obs_mat,\n",
    "                              observation_covariance = obs_cov,\n",
    "                              transition_covariance = trans_cov)\n",
    "            state_means, state_covs = kf.filter(y.values)\n",
    "        else:\n",
    "            kf = KalmanFilter(n_dim_obs = 1, n_dim_state = 1, \n",
    "                              initial_state_mean = init_mean, \n",
    "                              initial_state_covariance = init_cov,\n",
    "                              observation_matrices = obs_mat)\n",
    "            state_means, state_covs = kf.em(y.values).filter(y.values)\n",
    "        beta1 = state_means[:,0]\n",
    "        resid = y - beta1*x\n",
    "        return beta1, resid, state_covs, kf"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Cálculo do Spread Usando Filtro de Kalman"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Os parâmetros são os mesmos utilizados na função da regressão do filtro de Kalman, com exceção de:\n",
    "# janela_dados: define o tamanho da janela móvel\n",
    "\n",
    "def spread_KF(log_price1, log_price2, janela_dados, delta, obs_cov, init_mean0, init_cov, intercept = False, em = False):\n",
    "    \n",
    "    y = log_price1\n",
    "    x = log_price2\n",
    "    T = len(y)\n",
    "    \n",
    "    s = pd.DataFrame(np.NaN, index = y.index, columns = ['Spread'])\n",
    "    adf_res = pd.DataFrame(np.NaN, index = np.arange(0,T), columns = ['ADF'])\n",
    "    beta0 = pd.DataFrame(np.NaN, index = np.arange(0,T), columns = ['Beta0'])\n",
    "    beta1 = pd.DataFrame(np.NaN, index = np.arange(0,T), columns = ['Beta1'])\n",
    "    \n",
    "    inicio = 0\n",
    "    fim = janela_dados\n",
    "    i = 0\n",
    "    init_mean = init_mean0\n",
    "    \n",
    "    while fim <= T:\n",
    "        \n",
    "        x_array = x.iloc[np.arange(inicio,fim)]\n",
    "        y_array = y.iloc[np.arange(inicio,fim)]\n",
    "        res = KFReg(y = y_array, x = x_array, \n",
    "                    delta = delta, obs_cov = obs_cov, \n",
    "                    init_mean = init_mean, init_cov = init_cov, \n",
    "                    intercept = intercept, em = em)\n",
    "        if intercept == False:\n",
    "            beta1.iloc[np.arange(inicio,fim),:] = pd.DataFrame(res[0]).values\n",
    "            s.iloc[np.arange(inicio,fim),:] = pd.DataFrame(res[1]).values\n",
    "            adf_res.iloc[fim-1,0] = 1*(adfuller(x = res[1], \n",
    "                                            regression = 'c', \n",
    "                                            autolag = 'BIC')[1] < alfa)\n",
    "        else:\n",
    "            beta0.iloc[np.arange(inicio,fim),:] = pd.DataFrame(res[0]).values\n",
    "            beta1.iloc[np.arange(inicio,fim),:] = pd.DataFrame(res[1]).values\n",
    "            s.iloc[np.arange(inicio,fim),:] = pd.DataFrame(res[2]).values\n",
    "            adf_res.iloc[fim-1,0] = 1*(adfuller(x = res[2], \n",
    "                                            regression = 'c', \n",
    "                                            autolag = 'BIC')[1] < alfa)\n",
    "        \n",
    "        inicio = inicio + 1\n",
    "        if intercept == False:\n",
    "            init_mean = beta1.iloc[inicio,0]\n",
    "        else:\n",
    "            init_mean = np.array([beta1.iloc[inicio,0],\n",
    "                                    beta0.iloc[inicio,0]])\n",
    "        fim = fim + 1\n",
    "        i = i + 1\n",
    "    \n",
    "    s = s.dropna()\n",
    "    \n",
    "    if intercept == False:\n",
    "        return s, adf_res, beta1\n",
    "    else:\n",
    "        return s, adf_res, beta0, beta1"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Backtest"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "O objetivo dessa seção é efetuar um backtest da estratégia de pairs trading para verificação de seu retorno no período analisado. Para realização do backtest da estratégia, utilizaremos de um modelo de Thresholds com método de estimação via Filtro de Kalman (backtest_KF), que permite estimação diária dos coeficientes da regressão. \n",
    "\n",
    "Como preocupação para **não se utilizar de dados futuros** para a decisão de trading e cálculo dos retornos, foram **deslocados um dia para trás**:\n",
    "\n",
    "**1.** A série de spread, que atua como sinal;\n",
    "\n",
    "**2.** Os betas;\n",
    "\n",
    "**3.** Os resultados dos testes de cointegração.\n",
    "\n",
    "Tal regra pode ser observada nos índices dos dateframes que são percorridos nos loops - **ex:** fim_est-1 e fim_est-janela_coint-1."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    " ### Função para Realização do Backtest Usando Filtro de Kalman e Threshold Predeterminados"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Função que informa em que intervalo do spread se encontra\n",
    "\n",
    "A função retorna a posição onde o spread se localiza, podendo retornar ao limite superior de cima, limite superior entre as bandas, limite inferior entre as banda e limite inferior de baixo.\n",
    "\n",
    "Importante ressaltar que os resultados iniciais correspondentes a primeira janela móvel não são relevantes, pois, na função de backtest, onde os resultados serão utilizados, essa primeira janela não é levada em consideração, visto que o trade só começa após a mesma."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "## A função retorna a posição onde o spread se localiza\n",
    "## Limite superior de cima: resultado = 3\n",
    "## Limite superior entre as bandas: resultado = 2\n",
    "## Limite inferior entre as banda: resultado = 1\n",
    "### Limite inferior de baixo: resultado = 0\n",
    "\n",
    "def intervalos(df_spread, janela_dados, th):\n",
    "    \n",
    "    avg_spread = df_spread.rolling(window = janela_dados).mean()\n",
    "    vol_spread = df_spread.rolling(window = janela_dados).std(ddof = 1)\n",
    "                        \n",
    "    lim_sup = avg_spread + th*vol_spread\n",
    "    lim_inf = avg_spread - th*vol_spread\n",
    "    \n",
    "    resultado = (df_spread > lim_sup)*1 + (df_spread > avg_spread)*1 + (df_spread > lim_inf)*1\n",
    "\n",
    "    return resultado"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Função do Backtest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def backtest_KF(log_price1, log_price2, df_spread, df_interv, df_beta1, df_coint_res, \n",
    "                janela_dados, janela_coint, cost, delta_beta_lim):\n",
    "    \n",
    "    y = log_price1\n",
    "    x = log_price2\n",
    "    T = len(y)\n",
    "    \n",
    "    lag = 1\n",
    "    #inicio_est = 2 # começa 2 para frente pois iremos aplicar um shift de 2 dias\n",
    "    #fim_est = janela_dados + 2 # notar que, pelo uso do np.arange, último dia eh fim_est - 1\n",
    "    inicio_est = lag # começa lag para frente pois iremos aplicar um shift de lag dias\n",
    "    fim_est = janela_dados + lag # notar que, pelo uso do np.arange, último dia eh fim_est - 1\n",
    "    \n",
    "    df_log_retornos = pd.DataFrame(np.NaN, index = y.index, columns = ['Retornos'])\n",
    "    \n",
    "    # Variável que informa se há posição aberta ou não\n",
    "    par_aberto = False\n",
    "    \n",
    "    # -1 se o par foi aberto quando o spread estava na região inferior de fora\n",
    "    # 1 se o par foi aberto quando o spread estava na região superior de fora\n",
    "    \n",
    "    pos_aberto = -1\n",
    "    beta1_ant = df_beta1.iloc[fim_est-lag,0]\n",
    "    beta1 = beta1_ant\n",
    "    \n",
    "    while fim_est < T:\n",
    "        \n",
    "        # Regra: se nos últimos (janela_coint) dias houver cointegração, então o par é negociado\n",
    "        # inicio = fim_est - janela_coint\n",
    "        # fim = fim_est - 1\n",
    "        # tam = (fim_est - 1 - 1) - (fim_est - janela_coint - 1) + 1 = janela_coint\n",
    "        if np.min(df_coint_res.iloc[np.arange(fim_est-janela_coint-(lag-1), fim_est-(lag-1)), 0]) == True:\n",
    "            \n",
    "            # Acompanha a variação do beta de um dia para o outro para ajustar a posição de acordo\n",
    "            # com a variação do beta\n",
    "            if par_aberto == True:\n",
    "                delta_beta = df_beta1.iloc[fim_est-lag,0] - beta1\n",
    "                if np.abs(delta_beta) > delta_beta_lim:\n",
    "                    beta1_ant = beta1\n",
    "                    beta1 = df_beta1.iloc[fim_est-lag,0]\n",
    "               \n",
    "            # Caso 1: spread na parte superior de fora da banda\n",
    "            if df_interv.iloc[fim_est-lag,0] == 3:\n",
    "                    \n",
    "                # Caso 1.1: sem posição em aberto --> abre posição\n",
    "                if par_aberto == False:\n",
    "                    \n",
    "                    beta1 = df_beta1.iloc[fim_est-lag,0]\n",
    "                    par_aberto = True\n",
    "                    pos_aberto = 1\n",
    "                    \n",
    "                # Caso 1.2: com posição em aberto, pela parte superior --> mantém\n",
    "                if par_aberto == True and pos_aberto == 1:\n",
    "                    \n",
    "                    cost_beta = np.abs(beta1 - beta1_ant)*cost/2\n",
    "                    df_log_retornos.iloc[fim_est,0] = -(y.iloc[fim_est]-y.iloc[fim_est-1]) + beta1*(x.iloc[fim_est]-x.iloc[fim_est-1]) - cost_beta\n",
    "                    \n",
    "                # Caso 1.3: com posição em aberto, pela parte inferior --> fecha\n",
    "                if par_aberto == True and pos_aberto == -1:\n",
    "                    \n",
    "                    df_log_retornos.iloc[fim_est,0] = (y.iloc[fim_est]-y.iloc[fim_est-1]) - beta1*(x.iloc[fim_est]-x.iloc[fim_est-1]) - cost\n",
    "                    par_aberto = False\n",
    "    \n",
    "            # Caso 2: spread na parte superior de dentro da banda\n",
    "            if df_interv.iloc[fim_est-lag,0] == 2:\n",
    "                \n",
    "                # Caso 2.1: sem posição em aberto --> mantém\n",
    "                if par_aberto == False:\n",
    "                        \n",
    "                    par_aberto = False\n",
    "                    \n",
    "                # Caso 2.2: com posição em aberto, pela parte superior --> mantém\n",
    "                if par_aberto == True and pos_aberto == 1:\n",
    "                        \n",
    "                    cost_beta = np.abs(beta1 - beta1_ant)*cost/2\n",
    "                    df_log_retornos.iloc[fim_est,0] = -(y.iloc[fim_est]-y.iloc[fim_est-1]) + beta1*(x.iloc[fim_est]-x.iloc[fim_est-1]) - cost_beta\n",
    "                    \n",
    "                # Caso 2.3: com posição em aberto, pela parte inferior --> fecha\n",
    "                if par_aberto == True and pos_aberto == -1:\n",
    "                        \n",
    "                    df_log_retornos.iloc[fim_est,0] = (y.iloc[fim_est]-y.iloc[fim_est-1]) - beta1*(x.iloc[fim_est]-x.iloc[fim_est-1]) - cost\n",
    "                    par_aberto = False\n",
    "                \n",
    "            # Caso 3: spread na parte inferior de dentro da banda\n",
    "            if df_interv.iloc[fim_est-lag,0] == 1:\n",
    "                    \n",
    "                # Caso 3.1: sem posição em aberto --> mantém\n",
    "                if par_aberto == False:\n",
    "                        \n",
    "                    par_aberto = False\n",
    "                    \n",
    "                # Caso 3.2: com posição em aberto, pela parte superior --> fecha\n",
    "                if par_aberto == True and pos_aberto == 1:\n",
    "                        \n",
    "                    df_log_retornos.iloc[fim_est,0] = -(y.iloc[fim_est]-y.iloc[fim_est-1]) + beta1*(x.iloc[fim_est]-x.iloc[fim_est-1]) - cost\n",
    "                    par_aberto = False\n",
    "                    \n",
    "                # Caso 3.3: com posição em aberto, pela parte inferior --> mantém\n",
    "                if par_aberto == True and pos_aberto == -1:\n",
    "                        \n",
    "                    cost_beta = np.abs(beta1 - beta1_ant)*cost/2\n",
    "                    df_log_retornos.iloc[fim_est,0] = (y.iloc[fim_est]-y.iloc[fim_est-1]) - beta1*(x.iloc[fim_est]-x.iloc[fim_est-1]) - cost_beta\n",
    "                \n",
    "            # Caso 4: spread na parte inferior de fora da banda\n",
    "            if df_interv.iloc[fim_est-lag,0] == 0:\n",
    "        \n",
    "                # Caso 4.1: sem posição em aberto --> abre\n",
    "                if par_aberto == False:\n",
    "                        \n",
    "                    beta1 = df_beta1.iloc[fim_est-2,0]\n",
    "                    par_aberto = True\n",
    "                    pos_aberto = -1\n",
    "                    \n",
    "                # Caso 4.2: com posição em aberto, pela parte superior --> fecha\n",
    "                if par_aberto == True and pos_aberto == 1:\n",
    "                        \n",
    "                    df_log_retornos.iloc[fim_est,0] = -(y.iloc[fim_est]-y.iloc[fim_est-1]) + beta1*(x.iloc[fim_est]-x.iloc[fim_est-1]) - cost\n",
    "                    par_aberto = False\n",
    "                    \n",
    "                # Caso 4.3: com posição em aberto, pela parte inferior --> mantém\n",
    "                if par_aberto == True and pos_aberto == -1:\n",
    "                        \n",
    "                    cost_beta = np.abs(beta1 - beta1_ant)*cost/2 \n",
    "                    df_log_retornos.iloc[fim_est,0] = (y.iloc[fim_est]-y.iloc[fim_est-1]) - beta1*(x.iloc[fim_est]-x.iloc[fim_est-1]) - cost_beta\n",
    "                \n",
    "        # Else para contemplar o caso no qual, no período anterior, havia cointegração e foi terminado\n",
    "        # com posição em aberto, que será fechada imediatamente no período seguinte se não houver cointegração\n",
    "        else:\n",
    "            \n",
    "            if par_aberto == True:\n",
    "                    \n",
    "                if pos_aberto == 1:\n",
    "                        \n",
    "                    df_log_retornos.iloc[fim_est,0] = -(y.iloc[fim_est]-y.iloc[fim_est-1]) + beta1*(x.iloc[fim_est]-x.iloc[fim_est-1]) - cost \n",
    "                    \n",
    "                if pos_aberto == -1:\n",
    "                \n",
    "                    df_log_retornos.iloc[fim_est,0] = (y.iloc[fim_est]-y.iloc[fim_est-1]) - beta1*(x.iloc[fim_est]-x.iloc[fim_est-1]) - cost\n",
    "                \n",
    "                par_aberto = False\n",
    "                \n",
    "        inicio_est = inicio_est + 1\n",
    "        fim_est = fim_est + 1\n",
    "        \n",
    "    return df_log_retornos"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Exemplo de Pair Trading - PETR3 x PETR4"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nessa sessão realizamos um exemplo de implementação da estratégia de pairs trading com cointegração, utilizando os spreads calculados pelo filtro de Kalman e adotando um modelo de thresholds para regrar a negociação do par. Para esta análise foi utilizado o par de ações PETR3 e PETR4, papéis da mesma empresa e que possuem alta correlação entre si. Ao final, também foi construído um gráfico de retornos acumulados da estratégia para ilustrar o desempenho desta.\n",
    "\n",
    "**Nessa etapa, ainda não será realizada a seleção do threshold de tal modo a otimizar o desempenho do par.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Cálculo do Retorno Acumulado para as Séries de Preço"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Cálculo do retorno acumulado para as séries de preço\n",
    "x = df_log_prices['PETR3 BZ Equity']\n",
    "x = x.diff()\n",
    "x[0] = 0 #Para casos de não intercepto na regressão do filtro de Kalman, as séries de preço são normalizadas começando em zero\n",
    "x = x.cumsum()\n",
    "y = df_log_prices['PETR4 BZ Equity']\n",
    "y = y.diff()\n",
    "y[0] = 0 #Para casos de não intercepto na regressão do filtro de Kalman, as séries de preço são normalizadas começando em zero\n",
    "y = y.cumsum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Cálculo do Spread das Séries de Preço usando Filtro de Kalman\n",
    "\n",
    "Foi utilizado Filtro de Kalman como intercepto, em concordância com o modelo que foi apresentado na introdução."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Cálculo do spread usando KF com intercepto\n",
    "# Valores calibrados empiricamente\n",
    "res_KF = spread_KF(y, x, janela_dados, \n",
    "                delta = 1e-2, obs_cov = 0.25, \n",
    "                init_mean0 = np.array([1,0]), init_cov = np.matrix([[0.25,0],[0,0.25]]), \n",
    "                intercept = True, em = False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Variáveis que Guardam Resultados do Cálculo do Spread"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Variáveis que guardam os resultados do cálculo do spread usando KF\n",
    "s_KF = res_KF[0]\n",
    "adf_KF = res_KF[1]\n",
    "beta_KF = res_KF[3]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nessa sessão utilizaremos o Threshold inicial de 0,25:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "interv_KF = intervalos(s_KF, janela_dados, th_KF)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Implementação da Função de Backtest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_log_retornos_KF_th = backtest_KF(y, x, s_KF, interv_KF, beta_KF, adf_KF, \n",
    "                                    janela_dados, janela_coint, cost, delta_beta_lim)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Gráfico dos Retornos Acumulados da Estratégia"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x212b191c608>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.lineplot(data=np.array(np.exp(df_log_retornos_KF_th.cumsum())-1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Pairs Trading com Seleção de Threshold Ótimo"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nessa seção, o threshold será escolhido de modo a maximizar o índice de Sharpe de cada par. Com esse intuito, separaremos uma janela amostral de dois anos, ou 504 dias úteis, para realizar a maximização de forma numérica, variando o threshold em um intervalo predefinido e selecionando aquele que produz o Sharpe mais elevado. Nos períodos seguintes, assumiremos que o threshold ótimo se mantém constante e calcularemos os retornos de modo similar ao adotado no exercício de exemplo anterior."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Definição do Threshold Ótimo"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "sharpe_max = -np.inf\n",
    "th_max = 0\n",
    "rf = np.mean(df_indices['BZDIOVRA Index'].iloc[0:amostra])\n",
    "\n",
    "s_KF_th = s_KF[0:amostra]\n",
    "adf_KF_th = adf_KF[0:amostra]\n",
    "beta_KF_th = beta_KF[0:amostra]\n",
    "\n",
    "x_th = x[0:amostra]\n",
    "y_th = y[0:amostra]\n",
    "\n",
    "for th in np.arange(0.01,1.50,0.01):\n",
    "    #print(th)\n",
    "    interv_KF = intervalos(s_KF_th, janela_dados, th)\n",
    "    df_log_retornos_KF_th = backtest_KF(y_th, x_th, s_KF_th, interv_KF, beta_KF_th, adf_KF_th, \n",
    "                                    janela_dados, janela_coint, cost, delta_beta_lim)\n",
    "    sharpe = (np.mean(df_log_retornos_KF_th)*252 - rf*252) / np.std(252*df_log_retornos_KF_th)\n",
    "    sharpe = sharpe[0]\n",
    "    if sharpe > sharpe_max:\n",
    "        sharpe_max = sharpe\n",
    "        th_max = th\n",
    "        df_log_retornos_KF_th_max = df_log_retornos_KF_th"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.93"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "th_max"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Com isso, temos um resultado, como visto no gráfico abaixo:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x212b641ea08>"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "interv_KF = intervalos(s_KF, janela_dados, th_max)\n",
    "df_log_retornos_KF_th = backtest_KF(y, x, s_KF, interv_KF, beta_KF, adf_KF, \n",
    "                                    janela_dados, janela_coint, cost, delta_beta_lim)\n",
    "sns.lineplot(data=np.array(np.exp(df_log_retornos_KF_th.cumsum())-1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x212b608f348>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1200x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Desempenho com otimização do threshold e eliminando o periodo de amostra\n",
    "interv_KF = intervalos(s_KF[amostra:], janela_dados, th_max)\n",
    "df_log_retornos_KF_th = backtest_KF(y[amostra:], x[amostra:], s_KF[amostra:], interv_KF, beta_KF[amostra:], adf_KF[amostra:], \n",
    "                                    janela_dados, janela_coint, cost, delta_beta_lim)\n",
    "sns.lineplot(data=np.array(np.exp(df_log_retornos_KF_th.cumsum())-1))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Processamento de Todos os Pares da Base\n",
    "\n",
    "Extrapolando agora para os 10 pares, temos que processar o código acima de maneira iterativa, como abaixo. Os pares selecionados e que constam na planilha de preços foram construídos com base em uma argumentação econômica: empresas que atuam em um mesmo setor e possuem modelos de negócio similares apresentam maior probabilidade de cointegração."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Th_max\n",
      "0    0.93\n",
      "1    1.21\n",
      "2    0.53\n",
      "3    0.87\n",
      "4    0.69\n",
      "5    0.91\n",
      "6    1.46\n",
      "7    0.81\n",
      "8    0.46\n",
      "9    0.01\n",
      "dtype: float64\n"
     ]
    }
   ],
   "source": [
    "# O loop abaixo realiza o processamento de todos os pares da base\n",
    "acoes = list(df_prices.columns.values)\n",
    "\n",
    "qty_pairs = int(len(acoes)/2)\n",
    "sharpe_max = -np.inf\n",
    "th_max = pd.Series(np.zeros(qty_pairs))\n",
    "test = pd.Series(np.zeros(3))\n",
    "\n",
    "rf = np.mean(df_indices['BZDIOVRA Index'].iloc[0:amostra])\n",
    "\n",
    "for j in np.arange(0, qty_pairs):\n",
    "    \n",
    "    # Cálculo do retorno acumulado para as séries de preço do par j \n",
    "    x = df_log_prices.iloc[:, 2*j]\n",
    "    x = x.diff()\n",
    "    x[0] = 0 \n",
    "    x = x.cumsum()\n",
    "    \n",
    "    y = df_log_prices.iloc[:, 2*j+1]\n",
    "    y = y.diff()\n",
    "    y[0] = 0 \n",
    "    y = y.cumsum()\n",
    "    \n",
    "    res_KF_aux = spread_KF(y, x, janela_dados, \n",
    "                delta = 1e-2, obs_cov = 0.25, \n",
    "                init_mean0 = np.array([1,0]), init_cov = np.matrix([[0.25,0],[0,0.25]]), \n",
    "                intercept = True, em = False)\n",
    "    \n",
    "    s_KF = res_KF_aux[0]\n",
    "    adf_KF = res_KF_aux[1]\n",
    "    beta_KF = res_KF_aux[3]\n",
    "    \n",
    "    #Separação do periodo de amostra nos vetores e inicialização  \n",
    "    s_KF_th = s_KF[0:amostra]\n",
    "    adf_KF_th = adf_KF[0:amostra]\n",
    "    beta_KF_th = beta_KF[0:amostra]\n",
    "    x_th = x[0:amostra]\n",
    "    y_th = y[0:amostra]    \n",
    "    sharpe_max = -np.inf\n",
    "  \n",
    "    for th in np.arange(0.01,1.50,0.01):\n",
    "        interv_KF = intervalos(s_KF_th, janela_dados, th)\n",
    "        df_log_retornos_KF_th = backtest_KF(y_th, x_th, s_KF_th, interv_KF, beta_KF_th, adf_KF_th, \n",
    "                                        janela_dados, janela_coint, cost, delta_beta_lim)\n",
    "        sharpe = (np.mean(df_log_retornos_KF_th)*252 - rf*252) / np.std(252*df_log_retornos_KF_th)\n",
    "        sharpe = sharpe[0]\n",
    "        if sharpe > sharpe_max:\n",
    "            sharpe_max = sharpe\n",
    "            th_max[j] = th\n",
    "    \n",
    "    interv_KF = intervalos(s_KF, janela_dados, th_max[j])\n",
    "    df_log_retornos_KF_th_max = backtest_KF(y, x, s_KF, interv_KF, beta_KF, adf_KF, \n",
    "                                        janela_dados, janela_coint, cost, delta_beta_lim)\n",
    "    \n",
    "    if j == 0:\n",
    "        df_ret_aux = df_log_retornos_KF_th_max\n",
    "    else: \n",
    "        df_ret_aux = pd.concat([df_ret_aux, df_log_retornos_KF_th_max], axis = 1)\n",
    "\n",
    "print('Th_max')\n",
    "print(th_max)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Armazenagem dos resultados do loop anterior em planilhas no formato .csv\n",
    "df_ret_aux.columns = ['PETR3_PETR4', 'ITUB4_BBDC4', 'ITUB4_BBAS3', 'BBDC4_BBAS3',\n",
    "                  'GOAU4_GGBR4', 'CSNA3_USIM5', 'ELET3_ELET6', 'BRML3_MULT3',\n",
    "                  'MULT3_IGTA3', 'TAEE11_TRPL4']\n",
    "\n",
    "str_Nome_Plan_Retornos = 'Retornos_Pares'\n",
    "df_ret_aux.to_csv(path_or_buf = str_dir + str_Nome_Plan_Retornos + '.csv', sep=';')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Referências Bibliográficas\n",
    "\n",
    "CALDEIRA, J. F.; MOURA, G. V. Selection of a portfolio of pairs based in cointegration: A statistical arbitrage strategy. Revista Brasileira de Finanças, v. 11, n. 1, p. 49-80, mar. 2013.\n",
    "\n",
    "CARHART, M. M., On persistence in mutual fund performance. Journal of Finance, v. 52, n. 1, p. 57-82, mar. 1997.\n",
    "\n",
    "FAMA, E. F., FRENCH, K. F., Common risk factors in the returns on stocks and bonds. Journal of Financial Economics, v.33, n.1, p. 3-56, fev. 1993.\n",
    "\n",
    "GATEV, E., $et.$ $al$. Pairs trading: performance of a relative-value arbitrage rule. Review of Financial Studies, v. 19, n. 3, p. 797-827, mar. 2006."
   ]
  }
 ],
 "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}