Examples/Overview.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Portfolio Analyzer overview\n",
    "\n",
    "Portfolio Analyzer is a python library built to make easy portfolio analysis. In this Notebook we are going to see few examples. First things first. You can install Portfolio Analyzer with pip, so"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#%%bash\n",
    "#pip install git+https://github.com/Tokukawa/PortfolioAnalyzer.git"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The first thing we need in order to analyze porfolio is data. We can use the function `yahoo2pandas`. This function take a list of tickers and return a pandas dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from portfolio_analyzer.stocks_data_loader import yahoo2pandas"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "tickers = ['SPXL', 'TMF', 'UGL']\n",
    "data = yahoo2pandas(tickers, frequency='daily')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SPXL</th>\n",
       "      <th>TMF</th>\n",
       "      <th>UGL</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>formatted_date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2009-04-16</th>\n",
       "      <td>2.415000</td>\n",
       "      <td>14.5350</td>\n",
       "      <td>29.920000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2009-04-17</th>\n",
       "      <td>2.470000</td>\n",
       "      <td>13.8850</td>\n",
       "      <td>29.440001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2009-04-20</th>\n",
       "      <td>2.152500</td>\n",
       "      <td>14.7275</td>\n",
       "      <td>30.690001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2009-04-21</th>\n",
       "      <td>2.278333</td>\n",
       "      <td>14.2825</td>\n",
       "      <td>30.559999</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2009-04-22</th>\n",
       "      <td>2.237500</td>\n",
       "      <td>13.7950</td>\n",
       "      <td>30.889999</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    SPXL      TMF        UGL\n",
       "formatted_date                              \n",
       "2009-04-16      2.415000  14.5350  29.920000\n",
       "2009-04-17      2.470000  13.8850  29.440001\n",
       "2009-04-20      2.152500  14.7275  30.690001\n",
       "2009-04-21      2.278333  14.2825  30.559999\n",
       "2009-04-22      2.237500  13.7950  30.889999"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "By default `yahoo2pandas` return all possible data from yahoo. You can specify range of time you want and the frequency. Once we obtained the data, we can optimize our portfolio according to some most common algorithms. Up this moment you can use only `minimal_variance`, `risk_parity` and `approximated_max_kelly`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "from portfolio_analyzer.optimizers import approximated_max_kelly, minimal_variance, risk_parity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SPXL</th>\n",
       "      <th>TMF</th>\n",
       "      <th>UGL</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.301606</td>\n",
       "      <td>0.351565</td>\n",
       "      <td>0.346829</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       SPXL       TMF       UGL\n",
       "0  0.301606  0.351565  0.346829"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "minimal_variance(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Optimization terminated successfully.    (Exit mode 0)\n",
      "            Current function value: 2.2903198400221532e-14\n",
      "            Iterations: 21\n",
      "            Function evaluations: 108\n",
      "            Gradient evaluations: 21\n"
     ]
    },
    {
     "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>SPXL</th>\n",
       "      <th>TMF</th>\n",
       "      <th>UGL</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.312304</td>\n",
       "      <td>0.348332</td>\n",
       "      <td>0.339364</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       SPXL       TMF       UGL\n",
       "0  0.312304  0.348332  0.339364"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "risk_parity(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SPXL</th>\n",
       "      <th>TMF</th>\n",
       "      <th>UGL</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.308406</td>\n",
       "      <td>0.354138</td>\n",
       "      <td>0.337455</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       SPXL       TMF       UGL\n",
       "0  0.308406  0.354138  0.337455"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "approximated_max_kelly(data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can see how our portfolio would perform with `NaiveBackTest`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "from portfolio_analyzer.backtest import NaiveBackTest\n",
    "from portfolio_analyzer.utils import portfolio2dic"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "my_portfolio = portfolio2dic(approximated_max_kelly(data))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "backtest = NaiveBackTest(my_portfolio, data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "max_kelly_backtest = backtest.run()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can perform some measures on the back test and we can plot it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "from portfolio_analyzer.metrics import MainMetrics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x11a546710>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "data.plot(logy=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x12696eb10>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "max_kelly_backtest.plot(logy=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "benchmark_data = yahoo2pandas(['VTI'], frequency='daily')\n",
    "measures = MainMetrics(benchmark=benchmark_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>portfolio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>alpha</th>\n",
       "      <td>0.154421</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>average arithmetic return</th>\n",
       "      <td>0.242679</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>average geometric return</th>\n",
       "      <td>0.244589</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>benchmark correlation</th>\n",
       "      <td>0.425570</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>beta</th>\n",
       "      <td>0.520469</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max draw down</th>\n",
       "      <td>-0.342332</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>relative draw down</th>\n",
       "      <td>0.978082</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>relative return</th>\n",
       "      <td>1.857434</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>relative sharpe ratio</th>\n",
       "      <td>1.421196</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>relative variance</th>\n",
       "      <td>1.495398</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sharpe ratio</th>\n",
       "      <td>1.115505</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>theoretical optimal leverage</th>\n",
       "      <td>5.102373</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>variance</th>\n",
       "      <td>0.047328</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                              portfolio\n",
       "alpha                          0.154421\n",
       "average arithmetic return      0.242679\n",
       "average geometric return       0.244589\n",
       "benchmark correlation          0.425570\n",
       "beta                           0.520469\n",
       "max draw down                 -0.342332\n",
       "relative draw down             0.978082\n",
       "relative return                1.857434\n",
       "relative sharpe ratio          1.421196\n",
       "relative variance              1.495398\n",
       "sharpe ratio                   1.115505\n",
       "theoretical optimal leverage   5.102373\n",
       "variance                       0.047328"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "MainMetrics(benchmark_data).estimate(max_kelly_backtest)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "But I want to perform a proper out-of-sample back test before rebalancing my portfolio. We can use the class `OutOfSampleBackTest`. First of all we need to know how many times we can split the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from portfolio_analyzer.utils import factors"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Before starting my out of sample back test I neee to know how many eway I can split my data. I can know it with `factors` function "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[1, 2, 4, 8, 16, 179, 358, 716, 1432, 2864]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "factors(len(data))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I can split my data in all the factors shown above!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "SPLITS = 358"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we can make a proper outofsample backtest."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "from portfolio_analyzer.backtest import OutOfSampleBackTest"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "oos_backtest = OutOfSampleBackTest(data, minimal_variance, SPLITS, past_blocks=3, method='asym-blocks')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "out_of_sample_portfolio = oos_backtest.run()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>portfolio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>alpha</th>\n",
       "      <td>0.136197</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>average arithmetic return</th>\n",
       "      <td>0.208964</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>average geometric return</th>\n",
       "      <td>0.206054</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>benchmark correlation</th>\n",
       "      <td>0.344921</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>beta</th>\n",
       "      <td>0.378661</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max draw down</th>\n",
       "      <td>-0.267612</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>relative draw down</th>\n",
       "      <td>0.808612</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>relative return</th>\n",
       "      <td>1.424514</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>relative sharpe ratio</th>\n",
       "      <td>1.236114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>relative variance</th>\n",
       "      <td>1.223791</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sharpe ratio</th>\n",
       "      <td>1.009862</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>theoretical optimal leverage</th>\n",
       "      <td>4.860687</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>variance</th>\n",
       "      <td>0.042817</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                              portfolio\n",
       "alpha                          0.136197\n",
       "average arithmetic return      0.208964\n",
       "average geometric return       0.206054\n",
       "benchmark correlation          0.344921\n",
       "beta                           0.378661\n",
       "max draw down                 -0.267612\n",
       "relative draw down             0.808612\n",
       "relative return                1.424514\n",
       "relative sharpe ratio          1.236114\n",
       "relative variance              1.223791\n",
       "sharpe ratio                   1.009862\n",
       "theoretical optimal leverage   4.860687\n",
       "variance                       0.042817"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "MainMetrics(benchmark_data).estimate(out_of_sample_portfolio)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x126ccf550>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "out_of_sample_portfolio.plot(logy=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I can even define my own strategy!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "def custom_portfolio(data):\n",
    "    return pd.DataFrame({'SPXL': [0.36], 'TMF':[0.48], 'UGL':[0.16]})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "oos_backtest = OutOfSampleBackTest(data, custom_portfolio, SPLITS, past_blocks=4, method='asym-blocks')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "out_of_sample_portfolio = oos_backtest.run()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can get the actual out of sample performance the same way we did before."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>portfolio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>alpha</th>\n",
       "      <td>0.167857</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>average arithmetic return</th>\n",
       "      <td>0.234202</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>average geometric return</th>\n",
       "      <td>0.237923</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>benchmark correlation</th>\n",
       "      <td>0.310753</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>beta</th>\n",
       "      <td>0.334218</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max draw down</th>\n",
       "      <td>-0.215151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>relative draw down</th>\n",
       "      <td>0.650097</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>relative return</th>\n",
       "      <td>1.628393</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>relative sharpe ratio</th>\n",
       "      <td>1.402954</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>relative variance</th>\n",
       "      <td>1.173468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>sharpe ratio</th>\n",
       "      <td>1.155374</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>theoretical optimal leverage</th>\n",
       "      <td>5.669702</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>variance</th>\n",
       "      <td>0.041090</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                              portfolio\n",
       "alpha                          0.167857\n",
       "average arithmetic return      0.234202\n",
       "average geometric return       0.237923\n",
       "benchmark correlation          0.310753\n",
       "beta                           0.334218\n",
       "max draw down                 -0.215151\n",
       "relative draw down             0.650097\n",
       "relative return                1.628393\n",
       "relative sharpe ratio          1.402954\n",
       "relative variance              1.173468\n",
       "sharpe ratio                   1.155374\n",
       "theoretical optimal leverage   5.669702\n",
       "variance                       0.041090"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "MainMetrics(benchmark_data).estimate(out_of_sample_portfolio)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x126f850d0>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "out_of_sample_portfolio.plot(logy=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "I can even detect trends. For trend detection we can use the function l1filter."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "from portfolio_analyzer.filtering import l1filter, l2filter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1273e6390>"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "filtered, test = l1filter(out_of_sample_portfolio, delta=1E5)\n",
    "filtered.plot(logy=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x127516610>"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "filtered, test = l2filter(out_of_sample_portfolio, delta=1E5)\n",
    "filtered.plot(logy=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>portfolio_filter</th>\n",
       "      <th>portfolio</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>formatted_date</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2009-06-15</th>\n",
       "      <td>1.034321</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2009-06-16</th>\n",
       "      <td>1.035356</td>\n",
       "      <td>1.015936</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2009-06-17</th>\n",
       "      <td>1.036392</td>\n",
       "      <td>1.005799</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2009-06-18</th>\n",
       "      <td>1.037429</td>\n",
       "      <td>0.991149</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2009-06-19</th>\n",
       "      <td>1.038467</td>\n",
       "      <td>1.013210</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                portfolio_filter  portfolio\n",
       "formatted_date                             \n",
       "2009-06-15              1.034321   1.000000\n",
       "2009-06-16              1.035356   1.015936\n",
       "2009-06-17              1.036392   1.005799\n",
       "2009-06-18              1.037429   0.991149\n",
       "2009-06-19              1.038467   1.013210"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "filtered.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Whenever we perform a trend detection the test of stationarity is performed on the residuals. We can see the statistics as a pandas dataframe."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>portfolio</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>ADF Test Statistic</th>\n",
       "      <td>-2.949423</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>P-Value</th>\n",
       "      <td>0.039895</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lags Used</th>\n",
       "      <td>9.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Observations Used</th>\n",
       "      <td>2461.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Critical Value (1%)</th>\n",
       "      <td>-3.433010</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Critical Value (5%)</th>\n",
       "      <td>-2.862715</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Critical Value (10%)</th>\n",
       "      <td>-2.567396</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        portfolio\n",
       "ADF Test Statistic      -2.949423\n",
       "P-Value                  0.039895\n",
       "Lags Used                9.000000\n",
       "Observations Used     2461.000000\n",
       "Critical Value (1%)     -3.433010\n",
       "Critical Value (5%)     -2.862715\n",
       "Critical Value (10%)    -2.567396"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test['portfolio']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1275e5a90>"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "filtered, test = l1filter(data, delta=[1E4, 1E4, 1E1])\n",
    "filtered.plot(logy=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['SPXL', 'TMF', 'UGL'])"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SPXL</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>ADF Test Statistic</th>\n",
       "      <td>-3.843767</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>P-Value</th>\n",
       "      <td>0.002490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Lags Used</th>\n",
       "      <td>27.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Observations Used</th>\n",
       "      <td>2836.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Critical Value (1%)</th>\n",
       "      <td>-3.432658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Critical Value (5%)</th>\n",
       "      <td>-2.862560</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Critical Value (10%)</th>\n",
       "      <td>-2.567313</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             SPXL\n",
       "ADF Test Statistic      -3.843767\n",
       "P-Value                  0.002490\n",
       "Lags Used               27.000000\n",
       "Observations Used     2836.000000\n",
       "Critical Value (1%)     -3.432658\n",
       "Critical Value (5%)     -2.862560\n",
       "Critical Value (10%)    -2.567313"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test['SPXL']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}