Volatility of log returns.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "from datetime import datetime, date\n",
    "import re\n",
    "from pandas import datetime as dt\n",
    "import numpy as np\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import warnings\n",
    "warnings.filterwarnings(action='ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1    1521332236334\n",
       "2    1521332236375\n",
       "3    1521332236439\n",
       "4    1521332236986\n",
       "5    1521332236986\n",
       "6    1521332237087\n",
       "7    1521332237124\n",
       "8    1521332237233\n",
       "9    1521332237482\n",
       "Name: ExactTimeStamp, dtype: int64"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "trades = pd.read_csv('data/trades/ETHBTC-formatted_dates.csv')\n",
    "trades['ExactTimeStamp'][1:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "First day 2018-03-18 00:17:16\n",
      "Last day 2018-04-26 17:42:24\n"
     ]
    }
   ],
   "source": [
    "firstTime = trades['ExactTimeStamp'][0]\n",
    "firstTimeAsDatetime = datetime.utcfromtimestamp(firstTime/1000.0);\n",
    "N = np.size(trades['ExactTimeStamp'])\n",
    "lastTime = trades['ExactTimeStamp'][N-1]\n",
    "lastTimeAsDatetime = datetime.utcfromtimestamp(lastTime/1000.0);\n",
    "print(\"First day \"+ firstTimeAsDatetime.strftime('%Y-%m-%d %H:%M:%S'))\n",
    "print(\"Last day \" + lastTimeAsDatetime.strftime('%Y-%m-%d %H:%M:%S'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Volatility of log returns:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "$$\n",
    "\\begin{equation}\n",
    "\\sigma^2 = \\frac{1}{N-1}\\sum_{i=1}^N \\frac{(x_i - \\bar x)^2}{\\Delta t_i}\\,\n",
    "\\end{equation}\n",
    "$$\n",
    "\n",
    "where, $x_i = \\ln\\frac{S_i}{S_{i-1}}$ with $S_i$ the original prices, here $\\bar x$ is mean of $(x_i)_i$ and $\\Delta {t_i} = t_i - t_{i-1}$.\n",
    "\n",
    "The \"ExactTimeStamp\" is milliseconds since last UNIX epoch."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Total samples before : 4710672\n",
      "Total samples after : 3144409\n"
     ]
    }
   ],
   "source": [
    "print(\"Total samples before :\", trades.shape[0])\n",
    "\n",
    "#deleting the equal ExactTimeStamps to avoid delta_t = 0\n",
    "trades.drop_duplicates('ExactTimeStamp', keep='first', inplace=True)\n",
    "\n",
    "print(\"Total samples after :\", trades.shape[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Vol is 0.30% per second\n",
      "Vol is 17.71% per hour\n",
      "Vol is 86.75% per day\n",
      "Vol is 1657.98% per year\n"
     ]
    }
   ],
   "source": [
    "def logReturns(S):\n",
    "    return np.log(S/S.shift(1));\n",
    "\n",
    "# assumes times come as unix times i.e. number of seconds from a date in the past\n",
    "def volFromTimeSeriesPerSecond(prices, timeStamps):\n",
    "    if (np.size(prices)<= 2):\n",
    "        raise ValueError('we need more prices than 2')\n",
    "    x = logReturns(prices)[1:]; # the value at the first index is NaN\n",
    "    xBar = np.mean(x);\n",
    "    deltaT = (timeStamps - timeStamps.shift(1))[1:]; # the value at the first index is NaN\n",
    "    deltaT = deltaT/1000.0;     \n",
    "    N = np.size(x);\n",
    "    return np.sqrt(sum((x-xBar)*(x-xBar)/deltaT)/(N-1));\n",
    "\n",
    "def volFromTimeSeriesPerHour(prices,timeStamps):\n",
    "    return volFromTimeSeriesPerSecond(prices, timeStamps)*np.sqrt(3600.0)\n",
    "\n",
    "def volFromTimeSeriesPerDay(prices,timeStamps):\n",
    "    return volFromTimeSeriesPerSecond(prices, timeStamps)*np.sqrt(3600.0*24)\n",
    "\n",
    "def volFromTimeSeriesPerYear(prices,timeStamps):\n",
    "    return volFromTimeSeriesPerSecond(prices, timeStamps)*np.sqrt(3600.0*24*365.25)\n",
    "\n",
    "\n",
    "print(\"Vol is %.2f%% per second\" % (100.0*volFromTimeSeriesPerSecond(trades['Price'], trades['ExactTimeStamp'])))\n",
    "print(\"Vol is %.2f%% per hour\" % (100.0*volFromTimeSeriesPerHour(trades['Price'], trades['ExactTimeStamp'])))\n",
    "print(\"Vol is %.2f%% per day\" % (100.0*volFromTimeSeriesPerDay(trades['Price'], trades['ExactTimeStamp'])))\n",
    "print(\"Vol is %.2f%% per year\" % (100.0*volFromTimeSeriesPerYear(trades['Price'], trades['ExactTimeStamp'])))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Volatility per day:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "trades.drop_duplicates('ExactTimeStamp', keep='first', inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "dates = trades.TimeStamp.unique().tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "grouped = trades.set_index(['TimeStamp'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "grouped['DailyVol'] = 0\n",
    "for date in dates:\n",
    "    grouped.loc[date, 'DailyVol'] = volFromTimeSeriesPerSecond(grouped.loc[date, 'Price'], grouped.loc[date, 'ExactTimeStamp'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "grouped.reset_index(inplace=True)\n",
    "grouped.TimeStamp = pd.to_datetime(grouped.TimeStamp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fb4936ae0f0>"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fb493623208>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "plt.figure(figsize=(10,6))\n",
    "plt.plot(grouped.TimeStamp, grouped.DailyVol, label='Daily Volatility')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "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": "3.6.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}