iPython/MCRisk1-LargePoolCap.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#iPython Cookbook - Monte Carlo Risk Analysis - Large Pool Capital Model\n",
    ">Looking at the Large Pool Capital Model and idiosyncratic risks on top of it"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Set-up"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Model calculator factory\n",
    "We define a *factory* function that then provides us with a calculation function for our specific model. The factory function most importantly takes the number of Monte Carlo samples to be drawn `N`. For convenience it also takes a default payoff function as often the payoff will not be varied over the course of one analysis.\n",
    "\n",
    "Importantly, the vector of Standard Gaussian random variables `z` that will be the basis of the Monte Carlo simulation is drawn immediately, and it is remembered in the closure This allows us to run the calculation function (the one returned by the factory) with different parameters but against the same sampling set, thereby greatly reducing Monte Carlo errors.\n",
    "\n",
    "**TODO** operator module / partial"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from scipy.stats import norm\n",
    "from functools import partial\n",
    "from operator import le\n",
    "\n",
    "def calc_factory (N=1000, d=10, defaul_payoff=None):\n",
    "    \n",
    "    z = np.random.standard_normal((N,d+1))\n",
    "    \n",
    "    \n",
    "    def run (pd=0.05, lgd=1, rho=0):\n",
    "        \n",
    "        m = np.vstack((rho * np.ones(d),sqrt(1-rho*rho) * np.identity(d)))\n",
    "        x = np.dot(z,m)\n",
    "        \n",
    "        ppf_pd = norm.ppf(pd)\n",
    "        \n",
    "        def lossf (x):\n",
    "            if x < ppf_pd:\n",
    "                return 1\n",
    "            else:\n",
    "                return 0\n",
    "        \n",
    "        lossf = partial(le,ppf_pd)\n",
    "        lossv = list(map(lambda xx: list(map(lossf, xx)), x))\n",
    "        nloss = list(map(lambda xx: sum(xx), lossv))\n",
    "        loss  = list(map(lambda xx: xx*lgd, nloss))\n",
    "        return (loss, nloss)\n",
    "    \n",
    "    return run"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "calc = calc_factory(N=1000, d=1000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "loss, lossn = calc (pd=0.01, lgd=0.4*1000, rho=0.1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY0AAAEZCAYAAABrUHmEAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAGqlJREFUeJzt3X2YJFV96PFvswuCsMMyrvIOo0SCKOYBFXgSNjTxjahB\nHqMoGgJKJMYoaDS6G5PsXn30gu96TeIruHoF4SaEQK5XWQktJBBQgbC8w8rKi2FRF9gRE1nYuX/8\nTtM1Pd0zZ2f6pbr7+3mefqaqpqrO6Zqa+tU5p84pkCRJkiRJkiRJkiRJkiRJkiSV1FeBD2Wu++vA\nDcBm4B0LTHc18PUFbF8F7i3M3wT89gL2V/Qm4DuF+a3Aszq0b4BJYKKD+9OI2a7fGdBIm0qfHO8D\nLgPGgM91IN26CeLCvJD/hecBV8yxTm463wBevoC8FNWAU5uWLQE2dGj/GkEGDfVbJXO9/YFbuphm\nbj66kXbdog6nlRuQpWwGDfXSocB1RBXTN4Edm37/KqIK6iHg34BD0vJ/IaqEPpe2fTbwSuB64BHg\nHmBVYT9VplcfQdxd/05hvn5BrZcQHiaqbo5oke+diKq0TcDNwItm2ffhwA9Svh4APt4inc3AkcAp\n6Xt+EvgZUW12CnBl0/5fCawHfgp8lEbgWc30arYJojSzCPgwsJw4ZpPAZ9M6xequXYGvAQ+m7/CB\nwr5PAf4V+Fj63j8CjkWSemQH4MfAGcRF7feBx4APpt8fCmwkLsgV4A+Bu4Ht0+8vB95S2N/RwHPT\n9CHEBfrVab7KzKBxN40L+2oaF9v9mbva6Ezge8BSYB+iDeOeNvu+mmiXAHgqjSDUKp1TgC3An6bl\nOzIzaGwlquWWAvsCt9OoclpF66BRT6P5mNX3Vw8aXwP+Edg55e/2wvqnEH+fU4m/x9uA+9HIs6Sh\nXjkSWAx8BngC+Afg+4XfnwZ8IS2bIi5ov0rb1RWrdr5H3PUDrCNKLkfPI1851VKvI+7cHwbuI75D\nu+0eI0pCy4BfAtfMkc5PgL8hLub/3Wads1La9wKfBk7chry3W2cR8HpgJfAoEdA/AZxUWOfHwFdo\n/D32BJ6RkaaGmEFDvbIXM+9Uf1yY3h94D1E1Vf/sk7arK9bRH0HcST9IXFD/GHhaZ7P8pL2YXnK5\np92KxJ35gcCtwLVE1dJsmktEc61zD9OPyVzatWssI0pxxb/BPcDehfkHCtO/TD932Ya0NYQMGuqV\n/2T6BQkiUNTdQ9zN71b47AKc32Z/5wIXEYFlKfB5Gufzo0TVUN0i4Olt9pPTWPyfwH6F+f3arQjc\nBbwxpXcW8PdEm0i7dHLSb067Hnybv+ce27DvnxFVYxNN+74vIz8aYQYN9cpVwOPA6cQd7muY3qD8\nJaLe/HCiSmVn4i69eGdbrGrZhSiNPJa2eSONi+QdRPvAK1Jafwk8pU2+fkpUDR0wS94vIKpx6m0a\n75xl3T+gEaAeSXnamplOO++l0aZxOo1Aej3RP2RfolF7ZdN2G2dJ7wnie32YOJb7A+8G/vc88qcR\nYtBQr2whAsUpwM+BE4h2jbofAm8lnvbZBNxJNIYX75aL028nGtE3A3/F9BLJI+n3XybunH/B9Cqe\nYv+QXxIXzn8jgtDhLfL+P4hqnLuBbxP1++3u4l9ONJRPAp8C3kC0zRTT2URUr7Xqp9Jq2T8Rx+d6\n4J+Bs9Py76bvfSPRFnRJ07afAV6b0vt0i7y+kyit/IhofP8GcM4s+fARXnXV2cSdzrrCso8Rdb3/\nAVxI3B3VrSQuFLcBL+tRHiVJJbGceIyyGDReSqN0c2b6ABxMPJ+/PVHHeheWgiSpdLp5Yb6SKO4X\nrSXqdSEeRdwnTb8aOI+owthABI1W1QSSpD7q5938W4Bvpem9mP7Uxn3MfNJGktRn/QoaHyCeejl3\nlnVsdJOkklnchzRPIR6FfHFh2f3EY4N1+9BiyIIDDjhgav369V3NnCQNofXAr3ViR70uaRwL/DnR\nhlEcMuFi4tHEHYBnEsMwXNu88fr165mamvIzNcWqVav6noeyfDwWHguPxewf5tc/qKVuljTOI8YC\nWkY8I7+KeKx2B6JBHGJwt7cTQ15fkH4+npZZPSVJJdPNoHFii2Vnt1hW95H0kSSVlH0hBlS1Wu13\nFkrDY9HgsWjwWHRHr95W1ilTqX5OUp+NjY0zOdncFQuWLNmNzZs39SFHaqdSqUCHrvcGDUnzEhei\nVv+PFfw/LZdOBg2rpyRJ2QwakqRsBg1JUjaDhiQpm0FDkpTNoCFJymbQkCRlM2hIYmxsnEqlMuMz\nNjbe76ypZOzcJ2leHfXs3Dc47NwnSeoLg4YkKZtBQ5KUzaAhScpm0JAkZTNoSJKyGTQkSdkMGpKk\nbAYNSVI2g4YkKZtBQ5KUzaAhScpm0JAkZTNoSJKyGTQkSdkMGpKkbN0MGmcDG4F1hWXjwFrgDuBS\nYGnhdyuBO4HbgJd1MV+SpHnqZtA4Bzi2adkKImgcCFyW5gEOBl6ffh4L/G2X8yZJmoduXpivBB5q\nWnYcsCZNrwGOT9OvBs4DtgAbgLuAw7uYN0nSPPT6bn53osqK9HP3NL0XcF9hvfuAvXuYL0lShn5W\nAU3R+q30xd9LkkpkcY/T2wjsATwA7Ak8mJbfD+xbWG+ftGyG1atXPzldrVapVqtdyKYkDa5arUat\nVuvKvitd2WvDBHAJcEia/yjwc+AsohF8afp5MHAu0Y6xN/Bd4NeYWdqYmpqyACJ1WqVSoXXhvkK7\n/7n5bKP+iL9VZ6733SxpnAccDSwD7gX+GjgTuAA4lWjwPiGte0tafgvwOPB2rJ6SpNLpdkmj0yxp\nSF1gSWO4dbKkYV8ISVI2g4YkKZtBQ5KUzaAhScpm0JAkZTNoSJKyGTQkSdkMGpKkbAYNSVI2g4Yk\nKZtBQ5KUzaAhScpm0JAkZTNoSJKyGTQkSdkMGpKkbAYNSVI2g4YkKZtBQ5KUzaAhScpm0JBKamxs\nnEqlMuMzNjbe76xphFX6nYFtNDU1NdXvPEg9UalUgFbne4VO/x/MJ61e5k8LE3+rzlzvLWlIkrIZ\nNCRJ2QwakqRsBg1JUjaDhiQpm0FDkpTNoCFJytavoLESuBlYB5wLPAUYB9YCdwCXAkv7lDdJUhv9\nCBoTwFuBw4BDgEXAG4AVRNA4ELgszUuSSqQfQWMzsAV4KrA4/fwJcBywJq2zBji+D3mTJM2iH0Fj\nE/AJ4B4iWDxMlDB2BzamdTameUlSiSzuQ5oHAO8iqqkeAf4P8AdN60zRelAbVq9e/eR0tVqlWq12\nIYuSymRsbJzJyYdmLF+yZDc2b97UhxyVW61Wo1ardWXf/Riw8PXAS4E/SvMnAUcCvwMcAzwA7Alc\nDhzUtK0DFmpkOGBhf9IaRoM+YOFtRJDYifgSLwFuAS4BTk7rnAxc1Ie8SQPN4dTVbf0aGv19RGDY\nClxHlDqWABcA+wEbgBOI9o4iSxoaGb28+7ekMdw6WdLwfRpSSQ1r0JhP+4RBY2EMGtIIGNagUfZS\nzTAa9DYNSdKAMmhIkrIZNCRJ2bY1aCwCxrqREWmY+SishkVO0DiPCBQ7E6PS3ko8MispUzwtNDXj\n0+opIqnMcoLGwcQgg8cD/48Y/uOkLuZJklRSOUFjMbA9ETQuIUao9Rk3SRpBOUHjC0QP7V2AK2gM\nNChJGjHz6exRIRrEH+9wXnLYuU8DqVcd2gahw52d+3qv15379gC+Anw7zT+HxsCCkqQRkhM0vkq8\ns3uvNH8n8O5uZUiSVF45QWMZcD7wRJrfQn+qpiRJfZYTNH4BPK0wfyQ2hEvSSMp53et7iEdtnwVc\nBTwdeG03MyVJKqfc1vTtgV9P699GVFH1g09PaSD59FR/0lLo9dNT7yD6aNxEDCOyC/D2TiQuSRos\nOUHjrUBxgJyHgNO6kx1JUpnlBI3tmtZbRFRXSZJGTE5D+HeAbxLDiVSAP6bR0U+SNEJyGkYWEdVR\nL07za4Ev0+i30Us2hGsg2RDen7QUOtkQ3pGd9JBBQwPJoNH7tMbGxlu+r2TJkt3YvHlT2/wNo04G\njZzqqaOAVcTotvX1p4h+G5JUSo0XXzUvH7R75XLJCRpfAd4FXEd/qqQkSSWREzQeJt7YJ0kacTlB\n43LgY8CFwK8Ky6/rSo4kSaWVEzSOJCoGX9i0/JjOZ0eSVGY5QaPa7UxIkgbDfN7cdzBw6gLTXQr8\nPXArcAtwBDBO9AG5g3jp09IFpiFJ6rB+vbnvM8C3iFfHPp8YOXcFETQOBC5L85KkEunHm/t2BZYD\nZ6f5x4mXOh0HrEnL1gDHLyANSVIX9OPNfc8EfgqcQzyB9SVgZ2B3YGNaZ2OalySVSD/e3LcYOIx4\nT8f3gU8zsypqitZjBrB69eonp6vVKtVqdQFZkaThU6vVqNVqXdn3XP3pFwGnA/8LOCitfzvw2ALS\n3AO4mihxQAxTspIISscADwB7Ev1DDmra1rGnNJAce6r3aTnIYUMv39z3BPBGot2h/ua+hQQMiKBw\nL9HgDfAS4GaiNHNyWnYycNEC05EkdVhO5PkU8dKl84FH0zZTLKxH+G8Qw6vvAKwH3kyUai4A9gM2\nACcQQ5gUWdLQQCr73fUwpmVJo6HXQ6PXaH3k+9Ej3KChgVT2C+UwpmXQaOjl0OiLgIuBT3YiMUnS\nYMtp0zixFxmRJJVfziO3/wp8js62aUiSBpBtGlIPlL0efxjTsk2jodeve612IiFJ0uDLCRqriHDd\nHLY/2JUcSZJKKydoPEojWOwEvIoYzlySNGLmU8f1FGKo9KM7nJcctmloIJW9Hn8Y07JNo6GXw4i0\nsjOwdycSl/ppbGycSqUy4zM2Nt7vrEmllVM9ta4wvR3wDGzP0BCYnHyIVneik5MduSGThlJO0Pi9\nwvTjxLsutnQnO1K5jY2Np2Az3ZIlu7F586Y+5Ejqrdx3hG8iBhG8j2gMP6KLeZJKq1E6mf5pFUik\nYZQTND5PvL2v7tG0TJI0YnIbwrcWpp8gBjKUJI2YnKBxN/H2vu2J91+cAfyom5mSJJVTTtB4G/Bb\nwP1Em8aRwGndzJQkqZwG7dlCO/epY+zQNtxp2bmvoded+74GLC3M7wac3YnEJUmDJSdoPJ/p7+p+\nCDisO9mRJJVZTtCoAMVxFcbx6SlJGkk5PcI/AVwNXEAEkNcBH+5mpiRJ5ZTbMPKbwAvT9A+Aq7qT\nnTnZEK6OsfF3uNOyIbyh1w3hZwBfBJYBTwe+QPTbkCSNmJzIs47om/Fomt8Z+HfgkG5lahaWNNQx\n3l0Pd1qWNBr68T6NrW2mJUkjJKch/BzgGuBCIlIdj/00JGkk5RZXXgAcRZT1rgSu71qOZmf1lDrG\nKpnhTsvqqYZOVk85jIhGlhfK4U7LoNHQ73eEd8oiosRySZofB9YCdwCXMn3oEklSCfQzaJwB3ELj\nVmAFETQOBC5L85KkEulX0NgHeAXwZRpFpuOANWl6DdHgLkkqkX4FjU8Bf870x3d3Bzam6Y1pXpJU\nIjmP3Hbaq4AHifaMapt1pmjdgsXq1aufnK5Wq1Sr7XYhSaOpVqtRq9W6su9+PD31EeAk4HFgR2CM\n6APyIiKIPADsCVwOHNS0rU9PqWN8Ymi40/LpqYZBf3rqL4B9gWcCbwD+hQgiFwMnp3VOBi7qQ94k\njbixsXEqlcqMz9jY+Nwbj4B+VE81q4f8M4nh108FNgAn9CtDkkbX5ORDtCqhTE4OWre27hi0o2D1\nlDrGKpnhTquX+Su7Qa+ekiQNKIOGJCmbQUOSlM2gIUnKZtCQJGUzaEiSshk0JEnZDBqSpGwGDUlS\nNoOGJCmbQUMDzwHmpN4pw4CF0oI4wJzUO5Y0JEnZDBqSpGwGDUlSNoOGJCmbQUOl4pNQUrn59JRK\nxSehpHKzpCFJymbQkCRlM2hIkrIZNCRJ2QwakqRsBg1JUjaDhiQpm0FDkhZolDql2rlPkhZolDql\nWtKQJGXrR9DYF7gcuBm4CTg9LR8H1gJ3AJcCS/uQN0nSLPoRNLYA7waeCxwJ/CnwHGAFETQOBC5L\n85KkEulH0HgAuCFN/wK4FdgbOA5Yk5avAY7vfdYkSbPpd5vGBHAocA2wO7AxLd+Y5iVJJdLPoLEL\n8A/AGcBk0++maPUogiSpr/r1yO32RMD4OnBRWrYR2IOovtoTeLDVhqtXr35yulqtUq1Wu5hNSRo8\ntVqNWq3WlX334yHiCtFm8XOiQbzuo2nZWUQj+FJmNoZPTU1ZABlmlUqF1oXMCu3+9vPZZljTKnv+\neplW2fPXS5G/zlzv+xE0jgKuAG6kcZRXAtcCFwD7ARuAE4CHm7Y1aAy5Yb04eKHsfVplz18vDXrQ\nWAiDxgAZGxtPPWWnW7JkNzZv3tRym2G9OHih7H1aZc9fL3UyaDiMiLpmlIZWkEZFvx+51QAYpcHY\nJM3OkobmZIlBUp0lDUlSNoOGJCmbQUOSlM2gIUnKZtCQJGUzaEiSshk0JEnZDBqSpGwGDUlSNoPG\niHFIEEkL4TAiI8YhQSQthCUNSVI2g4YkKZtBQ5KUzaAhScpm0JAkZTNoSJKyGTQkSdkMGpKkbAaN\nAWXPbmmwDer/sD3CB5Q9u6XBNqj/w5Y0Omw+dw+DeschafRY0uiw+dw9DOodh6TRY0lDkpTNoNGG\nVUaSBk2761YnlS1oHAvcBtwJvL+fGWlUGU3/xHJJKp92161OKlPQWAR8jggcBwMnAs9pXsm7f0nq\nnzIFjcOBu4ANwBbgm8CrZ67m3b8k9UuZgsbewL2F+fvSsgWzfUKSOqNMj9x2tuKtwEdaJakzyhQ0\n7gf2LczvS5Q2itZD5YBWG8/9hEDr38++3Xy26WVaZc9fL9Mqe/56mVbZ89fLtIY1f7Npue36Beyw\ntBYTX2wC2AG4gRYN4ZIk1f0ucDvRIL6yz3mRJEmSNEzOANYBN6VpgPOB69Pn7vQToirrvwq/+9te\nZrQHWh2Lw4Frie/7feBFhfVXEh0jbwNe1rts9sS2HIsJhve8aHUcfgO4GrgRuBhYUlh/1M6Jdsdi\nguE6J84GNhLfv24cWAvcAVwKLC38rt158IK0jzuBz3Qxv13zPOIL7Eh09lsLNDd8fxz4yzQ9wfSD\nNkzaHYsa8PK0zu8Cl6fpg4l2oO2J43IX5XqkeiG29VhMMJznRbvj8H1geVrnzcAH0/QonhPtjsUE\nw3VOLAcOZfp3+ijwvjT9fuDMNN3qPKi3ll9L3HwBfIvoXD2rsp1ABwHXAP8NPAF8D3hN4fcV4ATg\nvN5nrefaHYufALumdZYST51BdIQ8j+gYuYE4MQ5nOGzrsRhWrY7D7wPPBq5M63w3LYPROydmOxbD\n5kqguVfzccCaNL0GOD5NtzoPjgD2JEpi16b1vlbYpq2yBY2biAg6DjwVeCWwT+H3y4kiWfHxsWcS\nxc0acFRPctkb7Y7FCuCTwD3Ax2g8MLAX0x9R7ljnyBLY1mMBw3leNB+HVxDH4SYaoye8jsaj66N0\nTsx1LGA4z4mi3YnrI+nn7mm63XnQvPx+Ms6PMvXTgKhvO4uoj3uU+ANvLfz+RODcwvxPiJPiIeAw\n4CLgucBkLzLbZe2OxVeAdwL/SPxTnA28tM0+utZhsse29VgM63nRfBxuIO6yTwU+C/wVUY//2Cz7\nGNZzYq5jMaznRDudH6lwQHwEeFuaXgw8QETHdi4nTohh9GHgT4DNhWUV4JE0vSJ96r5NFEGH0VzH\notmwnhfF/4+6A4lqGxitc2KuY9FsGM6JCaa3adwG7JGm90zz0P482AO4tbD8RODz3chotz0j/dyP\n+EJjaf5YGg2ddcuIRjCAZxFFraUMj+ZjsStwHXB0Wv5iouEPGo1dOxDF8PW06246mLblWAzzedHq\n/+Ppadl2RL30KWl+1M6J2Y7FMJ4TE8xsCK+/UmIFMxvCW50H1xABpEJmQ3gZXQHcTHzJYwrLzwFO\na1r3NUQd5vXAD4m67mHS6li8kPhD30A8WnhoYf2/IBq5bqPxVNGw2JZjMcznRavjcDrRKfZ24o67\naNTOiXbHYtjOifOIKrfHiIFe30y073yX1o/ctjsP6o/c3kVU60mSJEmSJEmSJEmSJEmSJEmSlKNG\nPJPebacDtwBfn2O9GnPnZznRD+E64CnzyMtXaQzY9y5gp3nsQ5qmbAMWSt2ykHF4tmWMtj8BXgKc\nNMd6OWMDvYnooHYY8KttyEOrNM4gBvaTFsSgoTKZIIaD+CLRe/c7xPsSYPqd+TLiZVwQw0RcRPSA\nvRt4B/Be4u78amC3wv5PInoEr6PxwqadiYEOr0nbHFfY78XAZcS7Gpr9WdrPOhovAPo8MUTFt4k7\n+6KdgG8SpZALmX7X/zLgKqKn8gUpT39EDML4IaLUsjPR2/eHxAuG6vmcYPpQEu8FVhXmK8SgjnsR\nw/BcRvzffzVtd2OLvErSQJggxvx/fpo/n7jbhukDzDUHjTuJi+oyYtDC+nAzn6RxQa8BX0jTy2lc\naD9SSGMpMfzEU9N+76X1+EQvIC62O6V0byLeGEfK13iLbf4M+HKaPiR9z8NSnr9HI4i8nxihFWLo\nnPr7ZBbReAvdsvSdYWbQeA/w1y22L+brBUSQrdsVKVPZhkaX7iYuyBB31RMZ21xODI/9KPAwcEla\nvo5GAJqi8fKuK4nB7XYl7vJ/j7hDh2g72C+tvzbtr9lRRGnhv9L8hcBvA/8xSx6X03idZv0OH+BI\nYkC5q9L8DoVpaAwstx3wP9N+thIlh2fQ2lyDEq4nSkSfBf4v0wOINCuDhsqmWHf/BI3qqcdpVKfu\nyHTFbbYW5rcy+zler+9/DY0797ojiCDUbrvihblCXptJ88W8Pr8WeOMceXwTUcI4jDgudxPHoXhc\nIEosc+XlYSKYHksMJ34C8R4KaU62aajs6hfWDcSotgCv3cZt69OvT9NHERfOzUS7yemF9Q4trN/O\nlcRrMevVU8fTeMVoO1fQCAzPIy7aU8C/A79FvN+atL9nt/gOY8CDRMA4Btg/Ld9IlDjGiVLSq9qk\nP0njNQNPI4LphURV2KC/V0I9ZElDZdN8l1yf/zjRSHwaUaUyVfj9VIv1m383RbxP+jrivH9LWv4h\n4NNEddF2wI+IRubZnm66nmhIrr9b+Us0qqbabfN3RBvDLURj/w/S8p8R7Sfn0Xis9gM0Sj71/X2D\nqHa7MW1bf3nOFuCDKS/3p/238kWigf5+4N0pL/WbxhVttpEkSZIkSZIkSZIkSZIkSZIkSZIkSVK/\n/X/gkBNcaO/3eQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f82198ad080>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "hgn=np.histogram(lossn, bins=50)\n",
    "xvalsn = np.delete(hgn[1], -1)\n",
    "plt.bar(xvalsn,hgn[0], width=0.9*(xvalsn[1]-xvalsn[0]))\n",
    "plt.title('default distribution')\n",
    "plt.xlabel('number of defaults')\n",
    "plt.ylabel('occurrences')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZMAAAEZCAYAAABSN8jfAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAGYRJREFUeJzt3XuYXGV9wPHvkHATsiRrFBIuBrGooFjxRhVl8Ep9FFOt\nCCqCd0ut4J1Ya1bbikqtWq2XimBEDaJFHmi9EJERFEUpKIEQgUiEBAlKIgl4IYTpH793mLOzs8lk\nz86cOTPfz/PMk3PeOZd3T3bP77zXA5IkSZIkSZIkSZIkSZIkSZIkTbvVwLMKOveJwGWZ9U3Agmk6\n9iLg82l5AXA/sMM0HXs/Iq+VaTqeJJXezcAzCzr3iYwPJp2oArdu5z4LyBdMVlPcNZLamq4nI0mT\nmzHNx6tjKUR9xmCifrYz8HFgbfp8DNgpfTcX+B9gA3AncGlmv3cDa4CNwEomf4p/MHABcBdwBXBA\ny/f3Aw9Py88HrkvHXAO8DXgQ8G1gPlHNtBGYB4wB3wDOTsc+MaWd3XL816af6zbg7Zn0LwL/nFmv\n0iz9nE1Ua12YzvkOJpZ05qef607gRuB1mWONAecCS1J+rwWegCQNmGw11weAy4nAMRf4UUoDOA34\nDPHUPwN4Wkp/JHALsFda349mQGh1TvrsChxMBIlsUMoGk99kzrEH8Pi0fAQTq7nGgHuBo9P6LsBi\nmsFkQTr2V9K5HwPcQbOt6KzMzwkTq9JaqwIbx2sEk0uBTxGB93Hp2Edm8vZH4CiidPNB4MdIOVky\nUT97OXFT/V36vB84Pn13L1EKWABsIQINaXlnIjjsSASWX7U59gzgxcD7iJvrdcTT+mTVR/emY44Q\npY2rU/pk219OlA4A/jTJdu9P576WCCDHZb6bajXWvsBTidLZvcAvgDOAV2W2uQz4DlFd9mUi4Ei5\nGEzUz+YDv86s35LSAE4HbgIuAlYRN09S2inEE/g6YCkRdFo9BJjJ+Cf+W7aSl5cQVV2rgRpw2Dby\nvmYb39Pm3PMn23A7zAfWA/e0HHvvzPq6zPIfiJKT9wLl4i+Q+tltjO+au19KA7ibaC84gKhOehvN\nqp+lwNOBhxFP3x9uc+zfAvelY2aPP5krgYVEEDqfaHcgHb9VvU16u+1az702Ld9DtMc07MV47Y7V\ncBswCuzecuxOgps0ZQYT9bOlwHtptpm8j2a7wwuARxDVQRuJ6q0twIFEUNkZ+DNRxbSlzbG3AOcR\nJZhdgYOAEybJx47AK4i2ki1Ew3fjmOuIhvyRzPbtqqjapb2XZnvNicDXUvrPiVLQHCKQnNKy3zom\ndhZouJWoYjuNuAaHAK8hqrOkrjGYqJ/9C1EiuCZ9rkxpEIFkGXFjvxz4T+AHxA30NKLk8RsiCC2a\n5PhvJp7gbwfOTJ/sU392+ZVEw/ddwBuI4ALRW2wp0S6znqhSm6xk0nrsHxDVct8jqu2+l747m2jr\nWE20bZzTsu9pRCDaQJTIWvN6HFGiu40ImO8Dvj9JPlr3lfrOmcQT1PJM2unA9cQfynnEk17DIqIb\n40rguT3KoySpzz2d6D6ZDSbPoVka+lD6QFQx/JyoTlhAPK1ZapKkkujmDfsyohietYzoDw8xSGyf\ntPwioqpgM1G0vwl4chfzJkmaRkU+/b8G+FZans/43iZrGN+VUZLUx4oKJv9IDKj66la2sVFQkkpi\nZgHnPJHo9pidZnwtMXK3YR+afe4fcMABB9RXrVrV1cxJ0gBaRfSA7Jpel0yOAt5JtJH8KZN+AXAs\nMZfQ/sBfAD9t3XnVqlXU63U/9TqLFy8uPA/98vFaeC28Flv/MPm4pGnTzZLJUmISvLnEQKrFRPff\nnYiGeIgJ5k4CVhAjilcQo5JPwmouSSqNbgaT49qknbmV7T+YPpKkknEsR0lVq9Wis9A3vBZNXosm\nr0Vvle1tbfVU/yepYCMjo2za1DqUDGbNmsPGjesLyJEmU6lUoMv3e4OJpCmJG1S7v8cK/p32l14E\nE6u5JEm5GUwkSbkZTCRJuRlMJEm5GUwkSbkZTCRJuRlMJEm5GUwkMTIySqVSmfAZGRktOmsqCQct\nSprSAEQHLZaHgxYlSaVgMJEk5WYwkSTlZjCRJOVmMJEk5WYwkSTlZjCRJOVmMJEk5WYwkSTlZjCR\nJOVmMJEk5WYwkSTlZjCRJOVmMJEk5WYwkSTlZjCRJOXWzWByJrAOWJ5JGwWWATcAFwGzM98tAm4E\nVgLP7WK+JEnTrJvB5CzgqJa0U4lgciBwcVoHOAh4Wfr3KODTXc6bJGkadfOGfRmwoSXtaGBJWl4C\nLEzLLwKWApuB1cBNwJO7mDdJ0jTq9dP/nkTVF+nfPdPyfGBNZrs1wN49zJckKYciq5Lq6bO17yVJ\nJTCzx+dbB+wF3A7MA+5I6WuBfTPb7ZPSJhgbG3tguVqtUq1Wu5BNSSqvWq1GrVbr6TkrXT7+AuBC\n4LFp/SPAncCHicb32enfg4CvEu0kewPfAx7BxNJJvV63wCJNt0qlQvvKgAqT/c1NZR8VI/6vunu/\n72bJZClwBDAXuBV4H/Ah4FzgtURD+zFp2xUpfQVwH3ASVnNJUml0u2Qy3SyZSF1gyWSw9aJk4lgO\nSVJuBhNJUm4GE0lSbgYTSVJuBhNJUm4GE0lSbgYTSVJuBhNJUm4GE0lSbgYTSVJuBhNJUm4GE0lS\nbgYTSVJuBhNJUm4GE0lSbgYTSVJuBhNJUm4GE0lSbgYTSVJuBhNJUm4GE6lPjYyMUqlUJnxGRkaL\nzpo0QaXoDGyner1eLzoPUk9UKhWg3e97hen+O5jKuXqZP+UT/1fdvd9bMpEk5WYwkSTlZjCRJOVm\nMJEk5WYwkSTlZjCRJOVmMJEk5VZUMFkEXAcsB74K7AyMAsuAG4CLgNkF5U2StJ2KCCYLgNcDhwKP\nBWYAxwKnEsHkQODitC5JKoEigslGYDPwIGBm+vc24GhgSdpmCbCwgLxJkqagiGCyHvgocAsRRH5P\nlEj2BNalbdaldUlSCcws4JwHAKcQ1V13AV8HXtmyTZ32k/4wNjb2wHK1WqVarXYhi5L6ycjIKJs2\nbZiQPmvWHDZuXF9AjvpbrVajVqv19JxFTPT4MuA5wOvS+vHAYcAzgSOB24F5wCXAo1r2daJHDQ0n\neizmXINoUCd6XEkEj12JH+7ZwArgQuCEtM0JwPkF5E0qNaetV1GKmoL+XUTAuB+4iiilzALOBfYD\nVgPHEO0pWZZMNDR6WVqwZDLYelEy8X0mUp8a1GAylfYPg0k+BpOJDCYaGoMaTPq9FDSIBrXNRJI0\nYAwmkqTcDCaSpNy2N5jMAEa6kRFpkNllV4Ouk2CylAgguxGz/F5PdO2V1KHovVSf8GnXq0kqo06C\nyUHE5IwLgW8T06Ac38U8SZJKppNgMhPYkQgmFxIz/toXT5L0gE6CyeeIEem7A5fSnKBRkiRgaoNY\nKkRD/H3TnJdOOGhRpdSrgXplGEjooMXe65dBi3sBXwC+k9YfTXNCRkmSOgomXyTeyT4/rd8IvLVb\nGZIklU8nwWQu8DVgS1rfTDFVXJKkPtVJMLkbeHBm/TBsgJckZXTy2t63E12CHw5cDjwE+NtuZkqS\nVC6dtu7vCDwybb+SqOoqgr25VEr25irmXAr90pvrzcQYk2uJ6VR2B07qZqYkSeXSSTB5PZCdQGgD\n8IbuZEeSVEadBJMdWrabQVR7SZIEdNYA/13gHGJalQrwRpoDGCVJ6qhBZgZRrfWstL4MOIPmuJNe\nsgFepWQDfDHnUuhFA3xXD94FBhOVksGk9+caGRlt+76YWbPmsHHj+knzN4h6EUw6qeY6HFhMzBbc\n2L5OjDuRpL7UfCFZa3rZnqHLoZNg8gXgFOAqiqnakiT1uU6Cye+JNyxKktRWJ8HkEuB04Dzgz5n0\nq7qSI0lS6XQSTA4jKh6f2JJ+5PRnR5JURp0Ek2q3MyFJKrepvGnxIOC1Oc87G/gGcD2wAngKMEqM\nYbmBeBnX7JznkCT1SFFvWvwE8C3iFcCHEDMRn0oEkwOBi9O6JKkEinjT4h7A04Ez0/p9xMu2jgaW\npLQlwMIc55Ak9VARb1rcH/gtcBbRI+zzwG7AnsC6tM26tC5JKoEi3rQ4EziUeE/Kz4CPM7FKq077\nuRMYGxt7YLlarVKtVnNkRZIGT61Wo1ar9fSc25pXYAbwFuCTwKPS9r8E7s1xzr2AHxMlFIjpWhYR\nwepI4HZgHjG+5VEt+zo3l0rJubl6fy4nh2zqhzctbgFeTrRrNN60mCeQQASLW4mGdoBnA9cRpZ8T\nUtoJwPk5zyNJ6pFOItXHiJdhfQ24J+1TJ98I+McR09jvBKwCXk2Ugs4F9gNWA8cQU7lkWTJRKfX7\n0/ggnsuSSVO/TEFfo/3/SBEj4A0mKqV+v4EO4rkMJk39MAX9DOAC4N+7mQlJUrl10mZyXC8yIkkq\nr066Bv8Q+BTT22YiSRogtplIPdDv7QSDeC7bTJr6oc0EnDVYkrQNnQSTxUR4bw3zH+hKjiRJpdNJ\nMLmHZhDZFXgBMW28JEnA1OrQdiampD9imvPSCdtMVEr93k4wiOeyzaSpH6ZTaWc3YO/pzojUayMj\no1QqlQmfkZHRorMmlU4n1VzLM8s7AA/F9hINgE2bNtDuyXXTpq4+wEkDqZNg8sLM8n3Eu0Y2dyc7\nUn8bGRlNQWi8WbPmsHHj+gJyJPWHTt8Bv56YfHEN0Qj/lC7mSepbzdLM+E+7ACMNk06CyWeJty02\n3JPSJEkCOm+Avz+zvIWYAFKSJKCzYHIz8bbFHYn3j5wM/KqbmZIklUsnweRNwNOAtUSbyWHAG7qZ\nKUlSuZStD6SDFjVtHKg32Ody0GJTvwxa/BIwO7M+BzizO9mRJJVRJ8HkEMa/i30DcGh3siNJKqNO\ngkkFyM4vMYq9uSRJGZ2MgP8o8GPgXCKwvBT4125mSpJULp02yDwVeGJavhK4vDvZ2SYb4DVtbHQe\n7HPZAN/ULw3wJwP/BcwFHgJ8jhh3IkkS0FmkWk6MLbknre8G/AR4bLcytRWWTDRtfBof7HNZMmnq\nl5IJjJ9O5f5Jt5IkDaVOGuDPAq4AziMi20IcZyJJyui02PME4HCizHgZcHXXcrR1VnNp2li1M9jn\nspqrqRfVXE6noqHlDXSwz2UwaeqnNpNumEGUcC5M66PAMuAG4CLGT+EiSepjRQaTk4EVNB8dTiWC\nyYHAxWldklQCRQWTfYDnA2fQLHodDSxJy0uIhn5JUgkUFUw+BryT8d2M9wTWpeV1aV2SVAKddA2e\nbi8A7iDaS6qTbFOnfcsZY2NjDyxXq1Wq1ckOIUnDqVarUavVenrOInpzfRA4HrgP2AUYIcawPIkI\nLrcD84BLgEe17GtvLk0bezAN9rnszdU0qL253gPsC+wPHAt8nwguFwAnpG1OAM4vIG+ShtzIyCiV\nSmXCZ2RkdNs7D7EiqrlaNR4RPkRMc/9aYDVwTFEZkjS8Nm3aQLsSzaZNZRuW11tluzpWc2naWLUz\n2OfqZf763aBWc0mSBozBRJKUm8FEkpSbwUSSlJvBRJKUm8FEkpSbwUSSlJvBRJKUm8FEkpSbwUSS\nlJvBRKXnxHxS8fphokcpFyfmk4pnyUSSlJvBRJKUm8FEkpSbwUSSlJvBRH3FnllSOdmbS33FnllS\nOVkykSTlZjCRJOVmMJEk5WYwkSTlZjCRJOVmMJEk5WYwkSTlZjCRpJwcbOugRUnKzcG2lkwkSdOg\niGCyL3AJcB1wLfCWlD4KLANuAC4CZheQN0nSFBQRTDYDbwUOBg4D/h54NHAqEUwOBC5O65KkEigi\nmNwO/Dwt3w1cD+wNHA0sSelLgIW9z5okaSqKbjNZADweuALYE1iX0teldUlSCRQZTHYH/hs4GdjU\n8l2ddl0jJEl9qaiuwTsSgeRs4PyUtg7Yi6gGmwfc0W7HsbGxB5ar1SrVarWL2ZSk8qnVatRqtZ6e\ns4hO0BWiTeROoiG+4SMp7cNE4/tsJjbC1+t1CyyDrFKp0L5QWmGy//up7DOo5+r3/PXyXP2ev16K\n/HX3fl9EMDkcuBS4hubVXwT8FDgX2A9YDRwD/L5lX4PJgBvUm4Y30N6fq9/z10uDGkzyMJiUyMjI\naBoZPN6sWXPYuHF9230G9abhDbT35+r3/PVSL4KJ06moa5xiQhoeRXcNVgk4iZ2kbbFkom2yhCFp\nWyyZSJJyM5hIknIzmEiScjOYSJJyM5hIknIzmEiScjOYSJJyM5hIknIzmEiScjOYDBmnRpHUDU6n\nMmScGkVSN1gykSTlZjCRJOVmMJEk5WYwkSTlZjCRJOVmMJEk5WYwkSTlZjCRJOVmMCkpR7JL5TZo\nf8OOgC8pR7JL5TZof8OWTKbZVJ42Bu0JRdLwsWQyzabytDFoTyiSho8lE0lSbgaTSVj1JKlsJrtv\n9UK/BZOjgJXAjcC7i8xIs+pp/CfSJan/THbf6oV+CiYzgE8RAeUg4Djg0a0bWVqQpP7TT8HkycBN\nwGpgM3AO8KKJm1lakKR+00/BZG/g1sz6mpSWm+0fktRd/dQ1uGsVe3a9laTu6qdgshbYN7O+L1E6\nyVoFlQPa7bztHgvtv9/6flPZp5fn6vf89fJc/Z6/Xp6r3/PXy3MNav62pu2+q3IcsHRmEj/wAmAn\n4Oe0aYCXJGlb/hr4JdEQv6jgvEiSJEkaNLsAVxBVVyuA01L644AfA9cAFwCzMvssIgYvrgSem0l/\nArA8ffeJTPrOwNdS+k+Ah2W+OwG4IX1eNR0/UA7bey2eA1yZ0q8Ejswca9iuRcN+wN3A2zNpw3gt\nDknfXZu+3ymlD9u12AVYmtJXAKdmjlX2a9EwA7gauDCtjwLLiDxeBMzObDuo984HPCj9O5PI8OHA\nz4Cnp/RXAx9IywcRv0g7Em0oN9FsVfopMSYF4FvEIEeAk4BPp+WXEeNUIC76KuJiz84sF2l7rsVf\nAnul5YMZ3ylh2K5FwzeIX/5sMBm2azET+AXw2LQ+h2ZX/2G7FicSwQRgV+Bm4oEDBuNaALwN+AoR\nRAE+ArwrLb8b+FBaHvR75zgPIn4pDgZ+n0nfF7guLS9i/HQq3wEOA+YB12fSjwU+m9nmKWl5JvDb\ntHwc8JnMPp9N+/WDTq5FVgW4k/hFGdZrsZD4Q1pMM5gM47V4PnB2m32H8Vo8j7jJzgDmEu2usxmc\na7EP8D2iVqJRMlkJ7JmW90rrUOC9s5eDFncgIuY64BLiF+E6mqPcX0qza/B8xj+BNwYwtqavpTmw\nMTvo8T7gLuDBWzlWkbbnWmS9BPg/YoaAvRm+a7E78TQ21nKMYbwWBxKDp75D/E68M6UP47X4LrAR\n+A0xg8bpROAZlGvxMeL/9/5M2p7EtSH92wgshd07exlM7ieqbPYBngFUgdcQRawriRvFvT3MT5Gm\nci0OJoqyb+xZLntje67FGPGH9Qcm66BfbttzLWYSVT8vT//+DfBMejWrX/dtz7V4JVG9NQ/YH3hH\n+ncQvAC4g2gvmex3vnezOW5FEdOp3AX8L/BEojj6vLR8Ds2BNa0DGPchouLatNya3tinUU86E9iD\nqBLqZDBkUTq5FhA/53nA8UR9MAzXtbgpbfNkoorrZuBk4D3EzWUNw3MtGr8XtwKXAuuBPxJ14Icy\nnL8XTwW+CWwhqmh+RDQ2D8LvxVOBo4nf+aXEA8PZRGmk0ZY6jwg4MAT3zrk0G252Jf4IngU8JKXt\nAHyJaEiDZiPSTsQTxiqaUfkKon6vwsRGpEb93rGMb0T6VTr/nMxyUbb3WswmGloXtjnWsF2LrMVE\no2TDsF2L2UT11q7EDWAZMU4Lhu9avAU4My3vRlSHPSatl/1aZB1Bs83kIzTbRk5lYgP8IN47gehx\nchXxQ15Ds373ZOJp45fAB1v2eQ/x5LGSeBppaHRvuwn4j0z6zsC5NLu3Lch89+qUfiPR1a1I23st\n3kt0g70685mbvhu2a5HVGkyG8Vq8gugWvJzmzQSG71rsDHyZ+Jmvo32X8bJei6wjaPbmGiUa5dt1\nDR7Ue6ckSZIkSZIkSZIkSZIkSZIkSVN1d9EZ6LI9gL8rOhOSNOg2FZ2BLltADECTJHVRI5hUiFlk\nlxOjqY9J6fOIaTquTt89jZim44uZbU9pc9wXEqOEryKmMHloSh8DlqRjrgZeDPxbOs63iSlPIKYF\nuSqlf4HmC61WE6OaIeaeuiRz3DPT+irgH1L6OcSEl1cDH976pZAkTVUjmLyEmG6iQtz4f01MkPd2\nYtoJ0ne7E9NNXJQ5xh5tjpudtuJ1RMCAuOlfSrxT4xDiRt+YyuI8Yur0XYBbgEek9CXEFCEQE/lN\nFkx+SLzD5sHA79I5HoYlE/WRImYNlnrpcOCrxBTddwA/AJ5EvHXu1cQcX4cQbSyrgIcT8xY9j3hH\nRqt9iYBzDTHV+UEpvU6UQLYQ82XtQLxnA+Kmv4B4B8nNNGe7XUJMr741dWLW3M3ETK53EO+uGMQp\n+FViBhMNujoTb7x14DLiFbBriaqt44kXKj0OqAFvAs5oc7xPEsHmEOLdMrtmvmu8X+N+4uZPZn0m\nE1VovofiPpp/j7u0bJd9t82WSY4lFcpgokF3GfFe6x2IKcyfQZRK9iPefXFG+hxKVCPNIKql/iml\ntRoBbkvLJ2bSOykp/JIooRyQ1o8nSkoQbSZPTMsv6eC4m4BZHZxT6gmfcDSoGk/83wT+ingnTJ2Y\nzvwO4FVpeTNxY34V8UrSs2g+ZJ3a5rhjwNeBDcD3ibaLxvmyb7trffNdHfgzUbX2deJv76c038P9\nfqJBfiNRMqpn9mv3Fr07iZdALSfeTfHuNttIkiRJkiRJkiRJkiRJkiRJkiRJkiRJ5fL/P+Knvbha\nZTcAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f8208b9fda0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "hg=np.histogram(loss, bins=50)\n",
    "xvals = np.delete(hg[1], -1)\n",
    "plt.bar(xvals,hg[0], width=0.9*(xvals[1]-xvals[0]))\n",
    "plt.title('loss distribution')\n",
    "plt.xlabel('loss amount')\n",
    "plt.ylabel('occurrences')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Licence and version\n",
    "*(c) Stefan Loesch / oditorium 2014; all rights reserved \n",
    "([license](https://github.com/oditorium/blog/blob/master/LICENSE))*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3.4.0 (default, Apr 11 2014, 13:05:11) \n",
      "[GCC 4.8.2]\n"
     ]
    }
   ],
   "source": [
    "import sys\n",
    "print(sys.version)"
   ]
  }
 ],
 "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": "3.4.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}