4_Accuracy_Backtesting.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Yu\\Anaconda3\\lib\\site-packages\\numpy\\lib\\arraysetops.py:522: FutureWarning: elementwise comparison failed; returning scalar instead, but in the future will perform elementwise comparison\n",
      "  mask |= (ar1 == a)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Overall accuracy for K = 10 is 0.5134\n",
      "Overall accuracy for K = 50 is 0.5086\n",
      "Overall accuracy for K = 100 is 0.5072\n",
      "Overall accuracy for K = 150 is 0.5059\n",
      "Overall accuracy for K = 200 is 0.5043\n",
      "Accuracy for sector Industrials is 0.5053 when K = 10\n",
      "Accuracy for sector Health Care is 0.5082 when K = 10\n",
      "Accuracy for sector Information Technology is 0.5087 when K = 10\n",
      "Accuracy for sector Financials is 0.5046 when K = 10\n",
      "Accuracy for sector Consumer Discretionary is 0.5076 when K = 10\n",
      "Accuracy for sector Consumer Staples is 0.5008 when K = 10\n",
      "Accuracy for sector Communication Services is 0.5054 when K = 10\n",
      "Accuracy for sector Materials is 0.5043 when K = 10\n",
      "Accuracy for sector Energy is 0.5081 when K = 10\n",
      "Accuracy for sector Real Estate is 0.4944 when K = 10\n"
     ]
    }
   ],
   "source": [
    "\"\"\"\n",
    "The following code shows how to analyze the accuracy of LSTM based on the size of portfolio K (Top K to long, bottom K to short).\n",
    "We select K = [10,50,100,150,200] and we realize that the accuracy is the highest when K =10. Hence, we will continue our backtestubg using K = 10.\n",
    "Furthermore, we have listed the accuracy of 10 sectors when k = 10 (ignore Real Estate due to small sample size)\n",
    "We have realized that our LSTM model does not have a sector preference.\n",
    "\n",
    "\"\"\"\n",
    "\n",
    "import pandas as pd\n",
    "import statistics\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.ticker as ticker\n",
    "\n",
    "def count_accu(K,df):\n",
    "    count = 0\n",
    "    for i in range(K):\n",
    "        if df.iloc[i,-1] == df.iloc[i,-2]:\n",
    "            count+=1\n",
    "    for i in range(K):\n",
    "        if df.iloc[-(i+1),-1]== df.iloc[-(i+1),-2]:\n",
    "            count+=1\n",
    "    return count\n",
    "\n",
    "def classification(y):\n",
    "    if y >=0.5:\n",
    "        return 1\n",
    "    else:\n",
    "        return 0\n",
    "\n",
    "def rebalance_freq(ds_1,ds_2):\n",
    "    ds = pd.concat([ds_1,ds_2],axis = 0,ignore_index = True).reset_index(drop = True)\n",
    "    unique_length = len(ds.value_counts().keys().tolist())\n",
    "    non_unique = ds.size- unique_length\n",
    "    rebalance_percentage = (ds_1.size-non_unique)/(ds_1.size)\n",
    "    return rebalance_percentage\n",
    "\n",
    "\n",
    "     \n",
    "data = pd.read_csv('C:/Users/Yu/Desktop/pred_result.csv',index_col = 0)\n",
    "data['y_pred'] = data['y_prob'].apply(classification)\n",
    "data = data.sort_values(['Date','y_prob'],ascending = True)\n",
    "date_list = list(data.Date.unique())\n",
    "data = data[['Date','Ticker','Sector','y_prob','y_pred','y_true']]\n",
    "data = data.reset_index(drop = True)\n",
    "# Overall accuracy of LSTM model\n",
    "K = [10,50,100,150,200]\n",
    "for k in K:\n",
    "    accuracy_list = []\n",
    "    for date in date_list:\n",
    "        data_date = data[data.Date == date]\n",
    "        data_date = data_date.reset_index(drop = True)\n",
    "        accu_item = count_accu(k,data_date)\n",
    "        accuracy_date = accu_item/(2*k)\n",
    "        accuracy_list.append(accuracy_date)\n",
    "    accuracy = round(sum(accuracy_list)/len(accuracy_list),4)\n",
    "    if k == 10:\n",
    "        overall_accuracy = accuracy\n",
    "    print('Overall accuracy for K = '+str(k)+' is '+str(accuracy))\n",
    "    \n",
    "# Accuracy analysis for each sectors when K = 10\n",
    "sector = list(data.Sector.unique())\n",
    "del(sector[-1])\n",
    "\n",
    "sec_dict = {}\n",
    "for sec in sector:\n",
    "    data_sector = data[data.Sector == sec]\n",
    "    data_sector = data_sector.reset_index(drop =True)\n",
    "    sec_accuracy_list = []\n",
    "    for date in date_list:\n",
    "        data_sector_date = data_sector[data_sector.Date == date]\n",
    "        data_sector_date = data_sector_date.reset_index(drop =True)\n",
    "        if data_sector_date.shape[0] >=20:\n",
    "            accu_item = count_accu(10,data_sector_date)\n",
    "            accuracy_date = accu_item/20\n",
    "            sec_accuracy_list.append(accuracy_date)\n",
    "    sec_accuracy  = round(sum(sec_accuracy_list)/len(sec_accuracy_list),4)\n",
    "    sec_dict.update({sec:sec_accuracy})\n",
    "    print('Accuracy for sector '+sec  +' is '+str(sec_accuracy) +' when K = 10')\n",
    "\n",
    "sec_dict.update({'Overall':overall_accuracy})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot bar chart of overall prediction accuracy and accuracy for each sector when K = 10.\n",
    "accuracy_ar = sec_dict.values()\n",
    "sector_ar = sec_dict.keys()\n",
    "plt.bar(sector_ar,accuracy_ar,align='center', alpha=0.5)\n",
    "plt.ylim([0.48,0.52])\n",
    "plt.xticks(rotation=90)\n",
    "for a,b in zip(sector_ar,accuracy_ar):\n",
    "    plt.text(a, b+0.002,\"%.4f\"%b, ha='center', va= 'bottom',fontsize=9)\n",
    "plt.ylabel('Accuracy')\n",
    "plt.tight_layout()\n",
    "#plt.savefig('C:/Users/Yu/Desktop/Accuracy For Each Sector And All Stocks When K = 10.png')\n",
    "plt.show()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\"\"\"\n",
    "The following code shows that how we backtest different trading strategies (130-30, market neutral (50-50), long only) with each stock equally weighted.\n",
    "We calculate the profits without transaction cost at first.\n",
    "Then, we assume the transaction cost equals to 2bps for market neutral and 130-30 strategies and 1bps for long only portfolio.\n",
    "\n",
    "\"\"\"\n",
    "return_info = pd.read_csv('C:/Users/Yu/Desktop/return.csv',index_col = 0)\n",
    "return_info = return_info[['Date','New_Ticker','Price_Return']]\n",
    "return_info.columns = ['Date','Ticker','Price_Return']\n",
    "\n",
    "# equal-weighted long only portfolio, prior to transaction cost and after transaction cost\n",
    "long_only_profit_nocost = {}\n",
    "long_only_accuprofit_nocost = {}\n",
    "long_only_nocost_accu_profit = 1\n",
    "\n",
    "long_only_profit_cost = {}\n",
    "long_only_accuprofit_cost = {}\n",
    "long_only_cost_accu_profit = 1\n",
    "for date in date_list[:-1]:\n",
    "    data_date = data[data.Date == date]\n",
    "    long_part = data_date.iloc[-10:]\n",
    "    long_df = long_part.merge(return_info,how ='left',on = ['Date','Ticker'])\n",
    "    daily_profit = long_df['Price_Return'].mean()\n",
    "    long_only_nocost_accu_profit += daily_profit\n",
    "    long_only_profit_nocost.update({date:daily_profit})\n",
    "    long_only_accuprofit_nocost.update({date:long_only_nocost_accu_profit})\n",
    "    \n",
    "    date_index = date_list.index(date)\n",
    "    long_part_next = data[data.Date == date_list[date_index+1]].iloc[-10:]\n",
    "    ds_1 = long_part['Ticker']\n",
    "    ds_2 = long_part_next['Ticker'] \n",
    "    daily_profit_cost = daily_profit - 0.0002*rebalance_freq(ds_1,ds_2)\n",
    "    long_only_cost_accu_profit += daily_profit_cost\n",
    "    long_only_profit_cost.update({date:daily_profit_cost})\n",
    "    long_only_accuprofit_cost.update({date:long_only_cost_accu_profit})\n",
    "long_only_nocost_return = (long_only_nocost_accu_profit - 1)/(len(date_list)-1)\n",
    "long_only_nocost_sd = statistics.stdev(long_only_profit_nocost.values())\n",
    "\n",
    "long_only_cost_return = (long_only_cost_accu_profit - 1)/(len(date_list)-1)\n",
    "long_only_cost_sd = statistics.stdev(long_only_profit_cost.values())\n",
    "\n",
    "# equal-weighted market neutral portfolio （50-50）， prior to transaction cost and after transaction cost\n",
    "neutral_profit_nocost = {}\n",
    "neutral_accuprofit_nocost = {}\n",
    "neutral_nocost_accu_profit = 1\n",
    "\n",
    "neutral_profit_cost = {}\n",
    "neutral_accuprofit_cost = {}\n",
    "neutral_cost_accu_profit = 1\n",
    "\n",
    "for date in date_list[:-1]:\n",
    "    data_date = data[data.Date == date]\n",
    "    long_part = data_date.iloc[-10:]\n",
    "    short_part = data_date.iloc[:10]\n",
    "    long_df = long_part.merge(return_info,how ='left',on = ['Date','Ticker'])\n",
    "    short_df = short_part.merge(return_info,how ='left',on = ['Date','Ticker'])\n",
    "    long_daily_profit = long_df['Price_Return'].mean()\n",
    "    short_daily_profit = (-1)*(short_df['Price_Return'].mean())\n",
    "    daily_profit = (long_daily_profit + short_daily_profit)/2\n",
    "    neutral_nocost_accu_profit += daily_profit\n",
    "    neutral_profit_nocost.update({date:daily_profit})\n",
    "    neutral_accuprofit_nocost.update({date:neutral_nocost_accu_profit})\n",
    "    \n",
    "    date_index = date_list.index(date)\n",
    "    long_part_next = data[data.Date == date_list[date_index+1]].iloc[-10:]\n",
    "    short_part_next = data[data.Date == date_list[date_index+1]].iloc[:10]\n",
    "    \n",
    "    ds_1 = long_part['Ticker']\n",
    "    ds_2 = long_part_next['Ticker']\n",
    "    ds_3 = short_part['Ticker']\n",
    "    ds_4 = short_part['Ticker']\n",
    "\n",
    "    daily_profit_cost = daily_profit - 0.0002*((rebalance_freq(ds_1,ds_2)+rebalance_freq(ds_3,ds_4))/2)\n",
    "    neutral_cost_accu_profit += daily_profit_cost\n",
    "    neutral_profit_cost.update({date:daily_profit_cost})\n",
    "    neutral_accuprofit_cost.update({date:neutral_cost_accu_profit})\n",
    "neutral_nocost_return = (neutral_nocost_accu_profit - 1)/(len(date_list)-1)\n",
    "neutral_nocost_sd = statistics.stdev(neutral_profit_nocost.values())\n",
    "\n",
    "neutral_cost_return = (neutral_cost_accu_profit - 1)/(len(date_list)-1)\n",
    "neutral_cost_sd = statistics.stdev(neutral_profit_cost.values())\n",
    "\n",
    "# equal_weighted long-short portfolio (130-30), prior to transaction cost and after transaction cost\n",
    "long_short_profit_nocost = {}\n",
    "long_short_accuprofit_nocost = {}\n",
    "long_short_nocost_accu_profit = 1\n",
    "\n",
    "long_short_profit_cost = {}\n",
    "long_short_accuprofit_cost = {}\n",
    "long_short_cost_accu_profit = 1\n",
    "\n",
    "for date in date_list[:-1]:\n",
    "    data_date = data[data.Date == date]\n",
    "    long_part = data_date.iloc[-10:]\n",
    "    short_part = data_date.iloc[:10]\n",
    "    long_df = long_part.merge(return_info,how ='left',on = ['Date','Ticker'])\n",
    "    short_df = short_part.merge(return_info,how ='left',on = ['Date','Ticker'])\n",
    "    long_daily_profit = long_df['Price_Return'].mean()\n",
    "    short_daily_profit = (-1)*(short_df['Price_Return'].mean())\n",
    "    daily_profit = 1.3*long_daily_profit + 0.3*short_daily_profit\n",
    "    long_short_nocost_accu_profit += daily_profit\n",
    "    long_short_profit_nocost.update({date:daily_profit}) \n",
    "    long_short_accuprofit_nocost.update({date:long_short_nocost_accu_profit})\n",
    "    \n",
    "    date_index = date_list.index(date)\n",
    "    long_part_next = data[data.Date == date_list[date_index+1]].iloc[-10:]\n",
    "    short_part_next = data[data.Date == date_list[date_index+1]].iloc[:10]\n",
    "    \n",
    "    ds_1 = long_part['Ticker']\n",
    "    ds_2 = long_part_next['Ticker']\n",
    "    ds_3 = short_part['Ticker']\n",
    "    ds_4 = short_part['Ticker']\n",
    "    \n",
    "    daily_profit_cost = daily_profit - 0.0002*(1.3*rebalance_freq(ds_1,ds_2)+0.3*rebalance_freq(ds_3,ds_4))\n",
    "    long_short_cost_accu_profit += daily_profit_cost\n",
    "    long_short_profit_cost.update({date:daily_profit_cost}) \n",
    "    long_short_accuprofit_cost.update({date:long_short_cost_accu_profit})\n",
    "long_short_nocost_return = (long_short_nocost_accu_profit-1)/(len(date_list)-1)\n",
    "long_short_nocost_sd = statistics.stdev(long_short_profit_nocost.values())\n",
    "\n",
    "long_short_cost_return = (long_short_cost_accu_profit-1)/(len(date_list)-1)\n",
    "long_short_cost_sd = statistics.stdev(long_short_profit_cost.values())\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot bar chart for return, standard deviation for three strategies, prior to transaction cost\n",
    "# Daily Return\n",
    "strategies = ['130-30','Market Neutral','Long Only']\n",
    "return_ar = [long_short_nocost_return,neutral_nocost_return,long_only_nocost_return]\n",
    "plt.bar(strategies,return_ar,align='center', alpha=1.0,width = 0.4)\n",
    "plt.ylim([0.0002,0.0010])\n",
    "for a,b in zip(strategies,return_ar):\n",
    "    plt.text(a, b+0.00004,\"%.5f\"%b, ha='center', va= 'bottom',fontsize=9)\n",
    "plt.ylabel('Average Daily Return')\n",
    "#plt.savefig('C:/Users/Yu/Desktop/Average Daily Return Of Three Strategies Prior To Transaction Cost.png')\n",
    "plt.show()\n",
    "\n",
    "# Standard Deviation\n",
    "sd_ar = [long_short_nocost_sd,neutral_nocost_sd,long_only_nocost_sd]\n",
    "plt.bar(strategies,sd_ar,align='center', alpha=1.0,width = 0.4)\n",
    "plt.ylim([0.005,0.025])\n",
    "for a,b in zip(strategies,sd_ar):\n",
    "    plt.text(a, b+0.0005,\"%.5f\"%b, ha='center', va= 'bottom',fontsize=9)\n",
    "plt.ylabel('Standard Deviation')\n",
    "#plt.savefig('C:/Users/Yu/Desktop/Standard Deviation Of Three Strategies Prior To Transaction Cost.png')\n",
    "plt.show()\n",
    "\n",
    "# Plot line chart for accumulative return from 11/2002 - 10/2018 for three strategies, prior to transaction cost\n",
    "date_ar = date_list[:-1]\n",
    "longshort_ar = long_short_accuprofit_nocost.values()\n",
    "marketneutral_ar = neutral_accuprofit_nocost.values()\n",
    "longonly_ar = long_only_accuprofit_nocost.values()\n",
    "plt.plot(date_ar,longshort_ar, color = 'red',label = '130-30')\n",
    "plt.plot(date_ar,marketneutral_ar, color = 'blue',label = 'Market Neutral')\n",
    "plt.plot(date_ar,longonly_ar, color = 'yellow',label = 'Long Only')\n",
    "plt.gca().xaxis.set_major_locator(ticker.MultipleLocator(240))\n",
    "plt.xticks(rotation=90)\n",
    "plt.legend()\n",
    "plt.ylabel('Accumulative Return')\n",
    "plt.tight_layout()\n",
    "#plt.savefig('C:/Users/Yu/Desktop/Accumulative Return Of Three Strategies Prior To Transaction Cost.png')\n",
    "plt.show()\n",
    "\n",
    "# Plot bar chart for return, standard deviation for three strategies, after transaction cost\n",
    "# Daily Return\n",
    "strategies_ar_cost = ['130-30','Market Neutral','Long Only']\n",
    "return_ar_cost = [long_short_cost_return,neutral_cost_return,long_only_cost_return]\n",
    "plt.bar(strategies_ar_cost,return_ar_cost,align='center', alpha=1.0,width = 0.4)\n",
    "plt.ylim([0.0001,0.0007])\n",
    "for a,b in zip(strategies_ar_cost,return_ar_cost):\n",
    "    plt.text(a, b+0.00002,\"%.5f\"%b, ha='center', va= 'bottom',fontsize=9)\n",
    "plt.ylabel('Average Daily Return')\n",
    "#plt.savefig('C:/Users/Yu/Desktop/Average Daily Return Of Three Strategies After Transaction Cost.png')\n",
    "plt.show()\n",
    "\n",
    "# Standard Deviation\n",
    "sd_ar_cost = [long_short_cost_sd,neutral_cost_sd,long_only_cost_sd]\n",
    "plt.bar(strategies_ar_cost,sd_ar_cost,align='center', alpha=1.0,width = 0.4)\n",
    "plt.ylim([0.005,0.025])\n",
    "for a,b in zip(strategies_ar_cost,sd_ar_cost):\n",
    "    plt.text(a, b+0.0005,\"%.5f\"%b, ha='center', va= 'bottom',fontsize=9)\n",
    "plt.ylabel('Standard Deviation')\n",
    "#plt.savefig('C:/Users/Yu/Desktop/Standard Deviation Of Three Strategies After Transaction Cost.png')\n",
    "plt.show()\n",
    "\n",
    "# Plot line chart for accumulative return from 11/2002 - 10/2018 for three strategies, after transaction cost\n",
    "date_ar_cost = date_list[:-1]\n",
    "longshort_ar_cost = long_short_accuprofit_cost.values()\n",
    "marketneutral_ar_cost = neutral_accuprofit_cost.values()\n",
    "longonly_ar_cost = long_only_accuprofit_cost.values()\n",
    "plt.plot(date_ar_cost,longshort_ar_cost, color = 'red',label = '130-30')\n",
    "plt.plot(date_ar_cost,marketneutral_ar_cost, color = 'blue',label = 'Market Neutral')\n",
    "plt.plot(date_ar_cost,longonly_ar_cost, color = 'yellow',label = 'Long Only')\n",
    "plt.gca().xaxis.set_major_locator(ticker.MultipleLocator(240))\n",
    "plt.xticks(rotation=90)\n",
    "plt.legend()\n",
    "plt.ylabel('Accumulative Return')\n",
    "plt.tight_layout()\n",
    "#plt.savefig('C:/Users/Yu/Desktop/Accumulative Return Of Three Strategies After Transaction Cost.png')\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}