optimal_holdingsCVX_tester.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {},
   "outputs": [],
   "source": [
    "import datetime as dt\n",
    "import pandas as pd\n",
    "import alphalens as al\n",
    "import numpy as np\n",
    "import alphalens as al\n",
    "import cvxpy as cvx\n",
    "import os\n",
    "import sys\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.style as style\n",
    "from scipy.stats.stats import pearsonr\n",
    "from tqdm import tqdm\n",
    "from scipy.stats import gaussian_kde\n",
    "style.use('ggplot')\n",
    "plt.rcParams['figure.figsize'] = (14,8)\n",
    "\n",
    "mod_path = '/Users/richard.wuebker/Projects/price_data_utils'\n",
    "if mod_path not in sys.path:\n",
    "    sys.path.append(mod_path)\n",
    "\n",
    "libs_path = '/Users/richard.wuebker/Projects/backtesting/libs'\n",
    "if libs_path not in sys.path:\n",
    "    sys.path.append(libs_path)\n",
    "\n",
    "    \n",
    "from risk_model import RiskModelPCA\n",
    "from optimal_holdings import OptimalHoldings\n",
    "    \n",
    "from data_loader import DataLoader\n",
    "from mp_historical_prices import HistoricalPrices"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Prediction Date: 2019-08-09\n"
     ]
    }
   ],
   "source": [
    "dl = DataLoader(prices_dir='../price_data_utils/prices', info_dir='../price_data_utils/info', \n",
    "                analysis_date_str='2019-08-07', cached_dir='../price_data_utils/cached')\n",
    "dl.load_data()\n",
    "data = dl.data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>index</th>\n",
       "      <th>Date</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Open</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "      <th>Adj Close</th>\n",
       "      <th>ticker</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [index, Date, High, Low, Open, Close, Volume, Adj Close, ticker]\n",
       "Index: []"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hp = HistoricalPrices(prices_dir='../price_data_utils/prices', \n",
    "                      cached_dir='../price_data_utils/cached',\n",
    "                      date_str='2019-08-09')\n",
    "hp.get_prices()\n",
    "prices = hp.prices\n",
    "\n",
    "# look for duplicates\n",
    "p = prices.reset_index()\n",
    "p.loc[p.duplicated(subset=['Date', 'ticker'], keep='first')]"
   ]
  },
  {
   "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>Date</th>\n",
       "      <th>High</th>\n",
       "      <th>Low</th>\n",
       "      <th>Open</th>\n",
       "      <th>Close</th>\n",
       "      <th>Volume</th>\n",
       "      <th>Adj Close</th>\n",
       "      <th>ticker</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1990-01-02</td>\n",
       "      <td>17.947027</td>\n",
       "      <td>17.527565</td>\n",
       "      <td>17.557526</td>\n",
       "      <td>17.947027</td>\n",
       "      <td>626200.0</td>\n",
       "      <td>7.267582</td>\n",
       "      <td>KMB</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1990-01-03</td>\n",
       "      <td>17.976990</td>\n",
       "      <td>17.827181</td>\n",
       "      <td>17.976990</td>\n",
       "      <td>17.887104</td>\n",
       "      <td>749200.0</td>\n",
       "      <td>7.243319</td>\n",
       "      <td>KMB</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1990-01-04</td>\n",
       "      <td>17.976990</td>\n",
       "      <td>17.797220</td>\n",
       "      <td>17.857143</td>\n",
       "      <td>17.857143</td>\n",
       "      <td>439700.0</td>\n",
       "      <td>7.231189</td>\n",
       "      <td>KMB</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1990-01-05</td>\n",
       "      <td>17.857143</td>\n",
       "      <td>17.617449</td>\n",
       "      <td>17.797220</td>\n",
       "      <td>17.737297</td>\n",
       "      <td>712500.0</td>\n",
       "      <td>7.182656</td>\n",
       "      <td>KMB</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1990-01-08</td>\n",
       "      <td>17.917067</td>\n",
       "      <td>17.707335</td>\n",
       "      <td>17.737297</td>\n",
       "      <td>17.917067</td>\n",
       "      <td>495600.0</td>\n",
       "      <td>7.255455</td>\n",
       "      <td>KMB</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         Date       High        Low       Open      Close    Volume  \\\n",
       "0  1990-01-02  17.947027  17.527565  17.557526  17.947027  626200.0   \n",
       "1  1990-01-03  17.976990  17.827181  17.976990  17.887104  749200.0   \n",
       "2  1990-01-04  17.976990  17.797220  17.857143  17.857143  439700.0   \n",
       "3  1990-01-05  17.857143  17.617449  17.797220  17.737297  712500.0   \n",
       "4  1990-01-08  17.917067  17.707335  17.737297  17.917067  495600.0   \n",
       "\n",
       "   Adj Close ticker  \n",
       "0   7.267582    KMB  \n",
       "1   7.243319    KMB  \n",
       "2   7.231189    KMB  \n",
       "3   7.182656    KMB  \n",
       "4   7.255455    KMB  "
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "prices.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def transform_prices(df, end_date_str):\n",
    "    prices_temp = df.pivot(index='Date', columns='ticker', values='Adj Close')\n",
    "    end_date = dt.datetime.strptime(end_date_str, '%Y-%m-%d')\n",
    "    start_date = end_date - pd.DateOffset(years=5)\n",
    "    prices_temp.index = pd.to_datetime(prices_temp.index)\n",
    "    prices = prices_temp.loc[start_date:end_date]\n",
    "    return prices.dropna(axis=1)\n",
    "new_p = transform_prices(prices, '2016-01-05')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# to make the index of data we have with sector the same as in returns and prices\n",
    "\n",
    "sector_data = data['sector']\n",
    "index_ = sector_data.index.intersection(new_p.columns)\n",
    "sector_data = sector_data.loc[index_]\n",
    "p = new_p.loc[:, index_].copy() # to use only the stocks that are in sector_data\n",
    "sectors = sector_data.unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1257, 741)\n",
      "(741,)\n",
      "(1258, 741)\n"
     ]
    }
   ],
   "source": [
    "returns = p.pct_change().drop(p.index[0])\n",
    "print(returns.shape)\n",
    "print(sector_data.shape)\n",
    "print(p.shape)"
   ]
  },
  {
   "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>ticker</th>\n",
       "      <th>CSCO</th>\n",
       "      <th>UAL</th>\n",
       "      <th>TROW</th>\n",
       "      <th>ISRG</th>\n",
       "      <th>NVR</th>\n",
       "      <th>PRGO</th>\n",
       "      <th>LECO</th>\n",
       "      <th>TPR</th>\n",
       "      <th>DVN</th>\n",
       "      <th>CR</th>\n",
       "      <th>...</th>\n",
       "      <th>PGR</th>\n",
       "      <th>WAT</th>\n",
       "      <th>NNN</th>\n",
       "      <th>IEX</th>\n",
       "      <th>LRCX</th>\n",
       "      <th>NUS</th>\n",
       "      <th>NWL</th>\n",
       "      <th>UAA</th>\n",
       "      <th>BLK</th>\n",
       "      <th>PPL</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</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",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2011-01-05</th>\n",
       "      <td>16.319235</td>\n",
       "      <td>25.350000</td>\n",
       "      <td>50.676487</td>\n",
       "      <td>91.209999</td>\n",
       "      <td>692.690002</td>\n",
       "      <td>59.998005</td>\n",
       "      <td>28.536177</td>\n",
       "      <td>40.534866</td>\n",
       "      <td>70.415245</td>\n",
       "      <td>34.523262</td>\n",
       "      <td>...</td>\n",
       "      <td>14.976269</td>\n",
       "      <td>77.790001</td>\n",
       "      <td>17.154980</td>\n",
       "      <td>33.694031</td>\n",
       "      <td>43.528183</td>\n",
       "      <td>23.909542</td>\n",
       "      <td>14.824180</td>\n",
       "      <td>6.94250</td>\n",
       "      <td>151.752335</td>\n",
       "      <td>16.587673</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2011-01-06</th>\n",
       "      <td>16.460661</td>\n",
       "      <td>25.750000</td>\n",
       "      <td>50.676487</td>\n",
       "      <td>89.559998</td>\n",
       "      <td>718.400024</td>\n",
       "      <td>62.723892</td>\n",
       "      <td>28.326279</td>\n",
       "      <td>40.797176</td>\n",
       "      <td>69.994743</td>\n",
       "      <td>34.397305</td>\n",
       "      <td>...</td>\n",
       "      <td>14.842355</td>\n",
       "      <td>76.050003</td>\n",
       "      <td>16.839489</td>\n",
       "      <td>33.711502</td>\n",
       "      <td>43.555824</td>\n",
       "      <td>23.958487</td>\n",
       "      <td>14.856651</td>\n",
       "      <td>6.83000</td>\n",
       "      <td>150.116272</td>\n",
       "      <td>16.470205</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2011-01-07</th>\n",
       "      <td>16.476372</td>\n",
       "      <td>25.950001</td>\n",
       "      <td>50.264294</td>\n",
       "      <td>89.133331</td>\n",
       "      <td>735.150024</td>\n",
       "      <td>62.590939</td>\n",
       "      <td>28.334854</td>\n",
       "      <td>40.758602</td>\n",
       "      <td>70.200493</td>\n",
       "      <td>34.976746</td>\n",
       "      <td>...</td>\n",
       "      <td>14.686119</td>\n",
       "      <td>75.209999</td>\n",
       "      <td>16.681742</td>\n",
       "      <td>33.589188</td>\n",
       "      <td>43.472923</td>\n",
       "      <td>24.031902</td>\n",
       "      <td>14.669928</td>\n",
       "      <td>6.69375</td>\n",
       "      <td>148.875351</td>\n",
       "      <td>16.315643</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2011-01-10</th>\n",
       "      <td>16.334949</td>\n",
       "      <td>26.200001</td>\n",
       "      <td>50.342079</td>\n",
       "      <td>90.526665</td>\n",
       "      <td>749.960022</td>\n",
       "      <td>62.932850</td>\n",
       "      <td>28.647541</td>\n",
       "      <td>41.267784</td>\n",
       "      <td>70.057358</td>\n",
       "      <td>35.111115</td>\n",
       "      <td>...</td>\n",
       "      <td>14.358768</td>\n",
       "      <td>74.680000</td>\n",
       "      <td>16.497704</td>\n",
       "      <td>33.807583</td>\n",
       "      <td>44.891319</td>\n",
       "      <td>24.553984</td>\n",
       "      <td>14.678050</td>\n",
       "      <td>6.90250</td>\n",
       "      <td>151.096298</td>\n",
       "      <td>16.290915</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2011-01-11</th>\n",
       "      <td>16.405655</td>\n",
       "      <td>25.690001</td>\n",
       "      <td>50.715374</td>\n",
       "      <td>93.663330</td>\n",
       "      <td>751.900024</td>\n",
       "      <td>63.474232</td>\n",
       "      <td>29.354307</td>\n",
       "      <td>41.653553</td>\n",
       "      <td>72.347855</td>\n",
       "      <td>35.279057</td>\n",
       "      <td>...</td>\n",
       "      <td>14.336452</td>\n",
       "      <td>76.300003</td>\n",
       "      <td>16.333384</td>\n",
       "      <td>33.982292</td>\n",
       "      <td>45.572895</td>\n",
       "      <td>24.456089</td>\n",
       "      <td>14.507563</td>\n",
       "      <td>6.79875</td>\n",
       "      <td>152.874680</td>\n",
       "      <td>16.303284</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 741 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "ticker           CSCO        UAL       TROW       ISRG         NVR       PRGO  \\\n",
       "Date                                                                            \n",
       "2011-01-05  16.319235  25.350000  50.676487  91.209999  692.690002  59.998005   \n",
       "2011-01-06  16.460661  25.750000  50.676487  89.559998  718.400024  62.723892   \n",
       "2011-01-07  16.476372  25.950001  50.264294  89.133331  735.150024  62.590939   \n",
       "2011-01-10  16.334949  26.200001  50.342079  90.526665  749.960022  62.932850   \n",
       "2011-01-11  16.405655  25.690001  50.715374  93.663330  751.900024  63.474232   \n",
       "\n",
       "ticker           LECO        TPR        DVN         CR    ...            PGR  \\\n",
       "Date                                                      ...                  \n",
       "2011-01-05  28.536177  40.534866  70.415245  34.523262    ...      14.976269   \n",
       "2011-01-06  28.326279  40.797176  69.994743  34.397305    ...      14.842355   \n",
       "2011-01-07  28.334854  40.758602  70.200493  34.976746    ...      14.686119   \n",
       "2011-01-10  28.647541  41.267784  70.057358  35.111115    ...      14.358768   \n",
       "2011-01-11  29.354307  41.653553  72.347855  35.279057    ...      14.336452   \n",
       "\n",
       "ticker            WAT        NNN        IEX       LRCX        NUS        NWL  \\\n",
       "Date                                                                           \n",
       "2011-01-05  77.790001  17.154980  33.694031  43.528183  23.909542  14.824180   \n",
       "2011-01-06  76.050003  16.839489  33.711502  43.555824  23.958487  14.856651   \n",
       "2011-01-07  75.209999  16.681742  33.589188  43.472923  24.031902  14.669928   \n",
       "2011-01-10  74.680000  16.497704  33.807583  44.891319  24.553984  14.678050   \n",
       "2011-01-11  76.300003  16.333384  33.982292  45.572895  24.456089  14.507563   \n",
       "\n",
       "ticker          UAA         BLK        PPL  \n",
       "Date                                        \n",
       "2011-01-05  6.94250  151.752335  16.587673  \n",
       "2011-01-06  6.83000  150.116272  16.470205  \n",
       "2011-01-07  6.69375  148.875351  16.315643  \n",
       "2011-01-10  6.90250  151.096298  16.290915  \n",
       "2011-01-11  6.79875  152.874680  16.303284  \n",
       "\n",
       "[5 rows x 741 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "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>ticker</th>\n",
       "      <th>CSCO</th>\n",
       "      <th>UAL</th>\n",
       "      <th>TROW</th>\n",
       "      <th>ISRG</th>\n",
       "      <th>NVR</th>\n",
       "      <th>PRGO</th>\n",
       "      <th>LECO</th>\n",
       "      <th>TPR</th>\n",
       "      <th>DVN</th>\n",
       "      <th>CR</th>\n",
       "      <th>...</th>\n",
       "      <th>PGR</th>\n",
       "      <th>WAT</th>\n",
       "      <th>NNN</th>\n",
       "      <th>IEX</th>\n",
       "      <th>LRCX</th>\n",
       "      <th>NUS</th>\n",
       "      <th>NWL</th>\n",
       "      <th>UAA</th>\n",
       "      <th>BLK</th>\n",
       "      <th>PPL</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</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",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2015-12-29</th>\n",
       "      <td>24.640928</td>\n",
       "      <td>59.299999</td>\n",
       "      <td>66.267441</td>\n",
       "      <td>184.456665</td>\n",
       "      <td>1650.000000</td>\n",
       "      <td>141.577072</td>\n",
       "      <td>48.883698</td>\n",
       "      <td>29.317432</td>\n",
       "      <td>31.023300</td>\n",
       "      <td>45.610756</td>\n",
       "      <td>...</td>\n",
       "      <td>29.112070</td>\n",
       "      <td>136.169998</td>\n",
       "      <td>34.340061</td>\n",
       "      <td>73.815681</td>\n",
       "      <td>75.976669</td>\n",
       "      <td>35.107853</td>\n",
       "      <td>39.816341</td>\n",
       "      <td>41.125000</td>\n",
       "      <td>314.309814</td>\n",
       "      <td>29.093164</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-12-30</th>\n",
       "      <td>24.427971</td>\n",
       "      <td>57.509998</td>\n",
       "      <td>65.351830</td>\n",
       "      <td>184.169998</td>\n",
       "      <td>1649.540039</td>\n",
       "      <td>140.273254</td>\n",
       "      <td>48.753433</td>\n",
       "      <td>29.024263</td>\n",
       "      <td>29.830099</td>\n",
       "      <td>45.245197</td>\n",
       "      <td>...</td>\n",
       "      <td>28.986509</td>\n",
       "      <td>135.639999</td>\n",
       "      <td>34.271770</td>\n",
       "      <td>73.255249</td>\n",
       "      <td>75.430962</td>\n",
       "      <td>35.071419</td>\n",
       "      <td>39.375340</td>\n",
       "      <td>40.480000</td>\n",
       "      <td>312.891907</td>\n",
       "      <td>29.008694</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015-12-31</th>\n",
       "      <td>24.099659</td>\n",
       "      <td>57.299999</td>\n",
       "      <td>64.807915</td>\n",
       "      <td>182.053329</td>\n",
       "      <td>1643.000000</td>\n",
       "      <td>139.751724</td>\n",
       "      <td>48.278931</td>\n",
       "      <td>29.077566</td>\n",
       "      <td>30.792358</td>\n",
       "      <td>44.842140</td>\n",
       "      <td>...</td>\n",
       "      <td>28.520142</td>\n",
       "      <td>134.580002</td>\n",
       "      <td>34.186417</td>\n",
       "      <td>72.770805</td>\n",
       "      <td>74.725288</td>\n",
       "      <td>34.515739</td>\n",
       "      <td>39.672337</td>\n",
       "      <td>40.305000</td>\n",
       "      <td>311.492279</td>\n",
       "      <td>28.831297</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-04</th>\n",
       "      <td>23.616772</td>\n",
       "      <td>55.610001</td>\n",
       "      <td>63.629417</td>\n",
       "      <td>182.463333</td>\n",
       "      <td>1555.660034</td>\n",
       "      <td>137.626907</td>\n",
       "      <td>48.325451</td>\n",
       "      <td>29.886017</td>\n",
       "      <td>31.100277</td>\n",
       "      <td>44.710915</td>\n",
       "      <td>...</td>\n",
       "      <td>27.910280</td>\n",
       "      <td>130.350006</td>\n",
       "      <td>33.768162</td>\n",
       "      <td>71.944405</td>\n",
       "      <td>73.135201</td>\n",
       "      <td>34.415535</td>\n",
       "      <td>39.096333</td>\n",
       "      <td>39.830002</td>\n",
       "      <td>304.704834</td>\n",
       "      <td>28.451166</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016-01-05</th>\n",
       "      <td>23.509462</td>\n",
       "      <td>55.060001</td>\n",
       "      <td>63.892326</td>\n",
       "      <td>183.986664</td>\n",
       "      <td>1594.089966</td>\n",
       "      <td>137.800766</td>\n",
       "      <td>48.167278</td>\n",
       "      <td>30.179193</td>\n",
       "      <td>31.591042</td>\n",
       "      <td>44.542198</td>\n",
       "      <td>...</td>\n",
       "      <td>28.170368</td>\n",
       "      <td>129.809998</td>\n",
       "      <td>34.382748</td>\n",
       "      <td>71.241478</td>\n",
       "      <td>73.530373</td>\n",
       "      <td>33.559250</td>\n",
       "      <td>39.078335</td>\n",
       "      <td>40.075001</td>\n",
       "      <td>305.491516</td>\n",
       "      <td>28.755270</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 741 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "ticker           CSCO        UAL       TROW        ISRG          NVR  \\\n",
       "Date                                                                   \n",
       "2015-12-29  24.640928  59.299999  66.267441  184.456665  1650.000000   \n",
       "2015-12-30  24.427971  57.509998  65.351830  184.169998  1649.540039   \n",
       "2015-12-31  24.099659  57.299999  64.807915  182.053329  1643.000000   \n",
       "2016-01-04  23.616772  55.610001  63.629417  182.463333  1555.660034   \n",
       "2016-01-05  23.509462  55.060001  63.892326  183.986664  1594.089966   \n",
       "\n",
       "ticker            PRGO       LECO        TPR        DVN         CR    ...      \\\n",
       "Date                                                                  ...       \n",
       "2015-12-29  141.577072  48.883698  29.317432  31.023300  45.610756    ...       \n",
       "2015-12-30  140.273254  48.753433  29.024263  29.830099  45.245197    ...       \n",
       "2015-12-31  139.751724  48.278931  29.077566  30.792358  44.842140    ...       \n",
       "2016-01-04  137.626907  48.325451  29.886017  31.100277  44.710915    ...       \n",
       "2016-01-05  137.800766  48.167278  30.179193  31.591042  44.542198    ...       \n",
       "\n",
       "ticker            PGR         WAT        NNN        IEX       LRCX        NUS  \\\n",
       "Date                                                                            \n",
       "2015-12-29  29.112070  136.169998  34.340061  73.815681  75.976669  35.107853   \n",
       "2015-12-30  28.986509  135.639999  34.271770  73.255249  75.430962  35.071419   \n",
       "2015-12-31  28.520142  134.580002  34.186417  72.770805  74.725288  34.515739   \n",
       "2016-01-04  27.910280  130.350006  33.768162  71.944405  73.135201  34.415535   \n",
       "2016-01-05  28.170368  129.809998  34.382748  71.241478  73.530373  33.559250   \n",
       "\n",
       "ticker            NWL        UAA         BLK        PPL  \n",
       "Date                                                     \n",
       "2015-12-29  39.816341  41.125000  314.309814  29.093164  \n",
       "2015-12-30  39.375340  40.480000  312.891907  29.008694  \n",
       "2015-12-31  39.672337  40.305000  311.492279  28.831297  \n",
       "2016-01-04  39.096333  39.830002  304.704834  28.451166  \n",
       "2016-01-05  39.078335  40.075001  305.491516  28.755270  \n",
       "\n",
       "[5 rows x 741 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(249, 741)\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>ticker</th>\n",
       "      <th>CSCO</th>\n",
       "      <th>UAL</th>\n",
       "      <th>TROW</th>\n",
       "      <th>ISRG</th>\n",
       "      <th>NVR</th>\n",
       "      <th>PRGO</th>\n",
       "      <th>LECO</th>\n",
       "      <th>TPR</th>\n",
       "      <th>DVN</th>\n",
       "      <th>CR</th>\n",
       "      <th>...</th>\n",
       "      <th>PGR</th>\n",
       "      <th>WAT</th>\n",
       "      <th>NNN</th>\n",
       "      <th>IEX</th>\n",
       "      <th>LRCX</th>\n",
       "      <th>NUS</th>\n",
       "      <th>NWL</th>\n",
       "      <th>UAA</th>\n",
       "      <th>BLK</th>\n",
       "      <th>PPL</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</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",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2011-01-06</th>\n",
       "      <td>0.008666</td>\n",
       "      <td>0.015779</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>-0.018090</td>\n",
       "      <td>0.037116</td>\n",
       "      <td>0.045433</td>\n",
       "      <td>-0.007356</td>\n",
       "      <td>0.006471</td>\n",
       "      <td>-0.005972</td>\n",
       "      <td>-0.003648</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.008942</td>\n",
       "      <td>-0.022368</td>\n",
       "      <td>-0.018391</td>\n",
       "      <td>0.000519</td>\n",
       "      <td>0.000635</td>\n",
       "      <td>0.002047</td>\n",
       "      <td>0.002190</td>\n",
       "      <td>-0.016205</td>\n",
       "      <td>-0.010781</td>\n",
       "      <td>-0.007082</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2011-01-07</th>\n",
       "      <td>0.000954</td>\n",
       "      <td>0.007767</td>\n",
       "      <td>-0.008134</td>\n",
       "      <td>-0.004764</td>\n",
       "      <td>0.023316</td>\n",
       "      <td>-0.002120</td>\n",
       "      <td>0.000303</td>\n",
       "      <td>-0.000946</td>\n",
       "      <td>0.002939</td>\n",
       "      <td>0.016846</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.010526</td>\n",
       "      <td>-0.011045</td>\n",
       "      <td>-0.009368</td>\n",
       "      <td>-0.003628</td>\n",
       "      <td>-0.001903</td>\n",
       "      <td>0.003064</td>\n",
       "      <td>-0.012568</td>\n",
       "      <td>-0.019949</td>\n",
       "      <td>-0.008266</td>\n",
       "      <td>-0.009384</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2011-01-10</th>\n",
       "      <td>-0.008583</td>\n",
       "      <td>0.009634</td>\n",
       "      <td>0.001548</td>\n",
       "      <td>0.015632</td>\n",
       "      <td>0.020146</td>\n",
       "      <td>0.005463</td>\n",
       "      <td>0.011035</td>\n",
       "      <td>0.012493</td>\n",
       "      <td>-0.002039</td>\n",
       "      <td>0.003842</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.022290</td>\n",
       "      <td>-0.007047</td>\n",
       "      <td>-0.011032</td>\n",
       "      <td>0.006502</td>\n",
       "      <td>0.032627</td>\n",
       "      <td>0.021725</td>\n",
       "      <td>0.000554</td>\n",
       "      <td>0.031186</td>\n",
       "      <td>0.014918</td>\n",
       "      <td>-0.001516</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2011-01-11</th>\n",
       "      <td>0.004328</td>\n",
       "      <td>-0.019466</td>\n",
       "      <td>0.007415</td>\n",
       "      <td>0.034649</td>\n",
       "      <td>0.002587</td>\n",
       "      <td>0.008603</td>\n",
       "      <td>0.024671</td>\n",
       "      <td>0.009348</td>\n",
       "      <td>0.032695</td>\n",
       "      <td>0.004783</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.001554</td>\n",
       "      <td>0.021693</td>\n",
       "      <td>-0.009960</td>\n",
       "      <td>0.005168</td>\n",
       "      <td>0.015183</td>\n",
       "      <td>-0.003987</td>\n",
       "      <td>-0.011615</td>\n",
       "      <td>-0.015031</td>\n",
       "      <td>0.011770</td>\n",
       "      <td>0.000759</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2011-01-12</th>\n",
       "      <td>0.011494</td>\n",
       "      <td>0.008174</td>\n",
       "      <td>0.021469</td>\n",
       "      <td>0.004342</td>\n",
       "      <td>0.018779</td>\n",
       "      <td>-0.010175</td>\n",
       "      <td>0.002043</td>\n",
       "      <td>0.005927</td>\n",
       "      <td>0.013975</td>\n",
       "      <td>0.014044</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.002076</td>\n",
       "      <td>0.005898</td>\n",
       "      <td>-0.003622</td>\n",
       "      <td>0.013419</td>\n",
       "      <td>0.014956</td>\n",
       "      <td>0.006004</td>\n",
       "      <td>0.019026</td>\n",
       "      <td>0.011583</td>\n",
       "      <td>0.002171</td>\n",
       "      <td>0.008722</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 741 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "ticker          CSCO       UAL      TROW      ISRG       NVR      PRGO  \\\n",
       "Date                                                                     \n",
       "2011-01-06  0.008666  0.015779  0.000000 -0.018090  0.037116  0.045433   \n",
       "2011-01-07  0.000954  0.007767 -0.008134 -0.004764  0.023316 -0.002120   \n",
       "2011-01-10 -0.008583  0.009634  0.001548  0.015632  0.020146  0.005463   \n",
       "2011-01-11  0.004328 -0.019466  0.007415  0.034649  0.002587  0.008603   \n",
       "2011-01-12  0.011494  0.008174  0.021469  0.004342  0.018779 -0.010175   \n",
       "\n",
       "ticker          LECO       TPR       DVN        CR    ...          PGR  \\\n",
       "Date                                                  ...                \n",
       "2011-01-06 -0.007356  0.006471 -0.005972 -0.003648    ...    -0.008942   \n",
       "2011-01-07  0.000303 -0.000946  0.002939  0.016846    ...    -0.010526   \n",
       "2011-01-10  0.011035  0.012493 -0.002039  0.003842    ...    -0.022290   \n",
       "2011-01-11  0.024671  0.009348  0.032695  0.004783    ...    -0.001554   \n",
       "2011-01-12  0.002043  0.005927  0.013975  0.014044    ...    -0.002076   \n",
       "\n",
       "ticker           WAT       NNN       IEX      LRCX       NUS       NWL  \\\n",
       "Date                                                                     \n",
       "2011-01-06 -0.022368 -0.018391  0.000519  0.000635  0.002047  0.002190   \n",
       "2011-01-07 -0.011045 -0.009368 -0.003628 -0.001903  0.003064 -0.012568   \n",
       "2011-01-10 -0.007047 -0.011032  0.006502  0.032627  0.021725  0.000554   \n",
       "2011-01-11  0.021693 -0.009960  0.005168  0.015183 -0.003987 -0.011615   \n",
       "2011-01-12  0.005898 -0.003622  0.013419  0.014956  0.006004  0.019026   \n",
       "\n",
       "ticker           UAA       BLK       PPL  \n",
       "Date                                      \n",
       "2011-01-06 -0.016205 -0.010781 -0.007082  \n",
       "2011-01-07 -0.019949 -0.008266 -0.009384  \n",
       "2011-01-10  0.031186  0.014918 -0.001516  \n",
       "2011-01-11 -0.015031  0.011770  0.000759  \n",
       "2011-01-12  0.011583  0.002171  0.008722  \n",
       "\n",
       "[5 rows x 741 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# create a risk model for 2011 data\n",
    "returns_2011 = returns.loc[returns.index.year == 2011]\n",
    "print(returns_2011.shape)\n",
    "returns_2011.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "risk_model = RiskModelPCA(returns_2011, 20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<BarContainer object of 20 artists>"
      ]
     },
     "execution_count": 12,
     "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": [
    "plt.bar(np.arange(risk_model.n_components), risk_model.pca.explained_variance_ratio_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "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": [
    "# looking at factor returns\n",
    "risk_model.factor_returns.cumsum().plot(legend=False)\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def density_plot(data): \n",
    "    density = gaussian_kde(data)\n",
    "    xs = np.linspace(np.min(data),np.max(data),200)\n",
    "    density.covariance_factor = lambda : .25\n",
    "    density._compute_covariance()\n",
    "    plt.plot(xs,density(xs))\n",
    "    plt.xlabel('Daily Returns')\n",
    "    plt.ylabel('Density')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "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": [
    "density_plot(risk_model.unwinz_returns.iloc[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "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": [
    "density_plot(risk_model.returns.iloc[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[67476.38326585694,\n",
       " 67476.38326585694,\n",
       " 67476.38326585694,\n",
       " 67476.38326585694,\n",
       " 67476.38326585694]"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "holdings = [50000000 / len(returns_2011.columns)] * len(returns_2011.columns)\n",
    "holdings[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "186099218691818.44"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "risk_model.predict_portfolio_risk(holdings)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "186099218691818.47"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "risk_model.predict_portfolio_risk_opt(holdings)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[0.001349527665317139,\n",
       " 0.001349527665317139,\n",
       " 0.001349527665317139,\n",
       " 0.001349527665317139,\n",
       " 0.001349527665317139]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "holdings = [1.0 / len(returns_2011.columns)] * len(returns_2011.columns)\n",
    "holdings[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.27283637491494306"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.sqrt(risk_model.predict_portfolio_risk(holdings))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.27283637491494295"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.sqrt(risk_model.predict_portfolio_risk_opt(holdings))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "def create_momentum_factor(prices, sector_data):\n",
    "\n",
    "    # demean by sector\n",
    "    df = prices.pct_change(252).dropna()\n",
    "    sectors = sector_data.unique()\n",
    "    \n",
    "    # demean by sector\n",
    "    for sector in sectors:\n",
    "        col_filter = sector_data[sector_data==sector].index\n",
    "        means = df.loc[:, col_filter].mean(axis=1)\n",
    "        df.loc[:, col_filter] = df.loc[:, col_filter].subtract(means, axis=0)\n",
    "        \n",
    "    ranked = df.rank(axis=1, method='first')\n",
    "    means = ranked.mean(axis=1)\n",
    "    stds = ranked.std(axis=1)\n",
    "    demeaned = ranked.subtract(means, axis=0)\n",
    "    zscored = demeaned.div(stds, axis=0)\n",
    "    return zscored"
   ]
  },
  {
   "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>ticker</th>\n",
       "      <th>CSCO</th>\n",
       "      <th>UAL</th>\n",
       "      <th>TROW</th>\n",
       "      <th>ISRG</th>\n",
       "      <th>NVR</th>\n",
       "      <th>PRGO</th>\n",
       "      <th>LECO</th>\n",
       "      <th>TPR</th>\n",
       "      <th>DVN</th>\n",
       "      <th>CR</th>\n",
       "      <th>...</th>\n",
       "      <th>PGR</th>\n",
       "      <th>WAT</th>\n",
       "      <th>NNN</th>\n",
       "      <th>IEX</th>\n",
       "      <th>LRCX</th>\n",
       "      <th>NUS</th>\n",
       "      <th>NWL</th>\n",
       "      <th>UAA</th>\n",
       "      <th>BLK</th>\n",
       "      <th>PPL</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</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",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2012-01-05</th>\n",
       "      <td>-0.471847</td>\n",
       "      <td>-1.326777</td>\n",
       "      <td>-0.228916</td>\n",
       "      <td>1.691173</td>\n",
       "      <td>-0.369068</td>\n",
       "      <td>1.588395</td>\n",
       "      <td>1.130563</td>\n",
       "      <td>0.728793</td>\n",
       "      <td>-1.200640</td>\n",
       "      <td>1.032457</td>\n",
       "      <td>...</td>\n",
       "      <td>0.261618</td>\n",
       "      <td>-0.686747</td>\n",
       "      <td>0.149496</td>\n",
       "      <td>-0.158839</td>\n",
       "      <td>-1.051144</td>\n",
       "      <td>1.621097</td>\n",
       "      <td>-1.130563</td>\n",
       "      <td>1.214655</td>\n",
       "      <td>0.126137</td>\n",
       "      <td>-0.172855</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2012-01-06</th>\n",
       "      <td>-0.476518</td>\n",
       "      <td>-1.382838</td>\n",
       "      <td>-0.196213</td>\n",
       "      <td>1.686502</td>\n",
       "      <td>-0.583969</td>\n",
       "      <td>1.532334</td>\n",
       "      <td>1.130563</td>\n",
       "      <td>0.752152</td>\n",
       "      <td>-1.256701</td>\n",
       "      <td>1.027785</td>\n",
       "      <td>...</td>\n",
       "      <td>0.355053</td>\n",
       "      <td>-0.551266</td>\n",
       "      <td>0.219572</td>\n",
       "      <td>-0.182198</td>\n",
       "      <td>-1.004426</td>\n",
       "      <td>1.588395</td>\n",
       "      <td>-1.065159</td>\n",
       "      <td>1.275388</td>\n",
       "      <td>0.233587</td>\n",
       "      <td>-0.149496</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2012-01-09</th>\n",
       "      <td>-0.513892</td>\n",
       "      <td>-1.448242</td>\n",
       "      <td>-0.294320</td>\n",
       "      <td>1.663143</td>\n",
       "      <td>-0.738136</td>\n",
       "      <td>1.513647</td>\n",
       "      <td>1.172609</td>\n",
       "      <td>0.752152</td>\n",
       "      <td>-1.130563</td>\n",
       "      <td>1.023113</td>\n",
       "      <td>...</td>\n",
       "      <td>0.327022</td>\n",
       "      <td>-0.537251</td>\n",
       "      <td>0.303664</td>\n",
       "      <td>-0.144824</td>\n",
       "      <td>-0.976396</td>\n",
       "      <td>1.588395</td>\n",
       "      <td>-1.027785</td>\n",
       "      <td>1.340792</td>\n",
       "      <td>0.154168</td>\n",
       "      <td>0.070076</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2012-01-10</th>\n",
       "      <td>-0.551266</td>\n",
       "      <td>-1.508975</td>\n",
       "      <td>-0.163511</td>\n",
       "      <td>1.649128</td>\n",
       "      <td>-0.896976</td>\n",
       "      <td>1.555693</td>\n",
       "      <td>1.125892</td>\n",
       "      <td>0.569953</td>\n",
       "      <td>-1.191296</td>\n",
       "      <td>1.027785</td>\n",
       "      <td>...</td>\n",
       "      <td>0.415786</td>\n",
       "      <td>-0.560610</td>\n",
       "      <td>0.336366</td>\n",
       "      <td>-0.228916</td>\n",
       "      <td>-1.097861</td>\n",
       "      <td>1.522990</td>\n",
       "      <td>-1.069831</td>\n",
       "      <td>1.261372</td>\n",
       "      <td>0.116794</td>\n",
       "      <td>-0.004672</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2012-01-11</th>\n",
       "      <td>-0.527908</td>\n",
       "      <td>-1.457586</td>\n",
       "      <td>-0.303664</td>\n",
       "      <td>1.649128</td>\n",
       "      <td>-0.873617</td>\n",
       "      <td>1.504303</td>\n",
       "      <td>1.083846</td>\n",
       "      <td>0.284977</td>\n",
       "      <td>-1.219327</td>\n",
       "      <td>1.018441</td>\n",
       "      <td>...</td>\n",
       "      <td>0.485862</td>\n",
       "      <td>-0.551266</td>\n",
       "      <td>0.355053</td>\n",
       "      <td>-0.228916</td>\n",
       "      <td>-1.079174</td>\n",
       "      <td>1.518319</td>\n",
       "      <td>-0.995083</td>\n",
       "      <td>1.331449</td>\n",
       "      <td>0.130809</td>\n",
       "      <td>-0.121465</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 741 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "ticker          CSCO       UAL      TROW      ISRG       NVR      PRGO  \\\n",
       "Date                                                                     \n",
       "2012-01-05 -0.471847 -1.326777 -0.228916  1.691173 -0.369068  1.588395   \n",
       "2012-01-06 -0.476518 -1.382838 -0.196213  1.686502 -0.583969  1.532334   \n",
       "2012-01-09 -0.513892 -1.448242 -0.294320  1.663143 -0.738136  1.513647   \n",
       "2012-01-10 -0.551266 -1.508975 -0.163511  1.649128 -0.896976  1.555693   \n",
       "2012-01-11 -0.527908 -1.457586 -0.303664  1.649128 -0.873617  1.504303   \n",
       "\n",
       "ticker          LECO       TPR       DVN        CR    ...          PGR  \\\n",
       "Date                                                  ...                \n",
       "2012-01-05  1.130563  0.728793 -1.200640  1.032457    ...     0.261618   \n",
       "2012-01-06  1.130563  0.752152 -1.256701  1.027785    ...     0.355053   \n",
       "2012-01-09  1.172609  0.752152 -1.130563  1.023113    ...     0.327022   \n",
       "2012-01-10  1.125892  0.569953 -1.191296  1.027785    ...     0.415786   \n",
       "2012-01-11  1.083846  0.284977 -1.219327  1.018441    ...     0.485862   \n",
       "\n",
       "ticker           WAT       NNN       IEX      LRCX       NUS       NWL  \\\n",
       "Date                                                                     \n",
       "2012-01-05 -0.686747  0.149496 -0.158839 -1.051144  1.621097 -1.130563   \n",
       "2012-01-06 -0.551266  0.219572 -0.182198 -1.004426  1.588395 -1.065159   \n",
       "2012-01-09 -0.537251  0.303664 -0.144824 -0.976396  1.588395 -1.027785   \n",
       "2012-01-10 -0.560610  0.336366 -0.228916 -1.097861  1.522990 -1.069831   \n",
       "2012-01-11 -0.551266  0.355053 -0.228916 -1.079174  1.518319 -0.995083   \n",
       "\n",
       "ticker           UAA       BLK       PPL  \n",
       "Date                                      \n",
       "2012-01-05  1.214655  0.126137 -0.172855  \n",
       "2012-01-06  1.275388  0.233587 -0.149496  \n",
       "2012-01-09  1.340792  0.154168  0.070076  \n",
       "2012-01-10  1.261372  0.116794 -0.004672  \n",
       "2012-01-11  1.331449  0.130809 -0.121465  \n",
       "\n",
       "[5 rows x 741 columns]"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "momentum = create_momentum_factor(p, sector_data)\n",
    "momentum.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "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>ticker</th>\n",
       "      <th>CSCO</th>\n",
       "      <th>UAL</th>\n",
       "      <th>TROW</th>\n",
       "      <th>ISRG</th>\n",
       "      <th>NVR</th>\n",
       "      <th>PRGO</th>\n",
       "      <th>LECO</th>\n",
       "      <th>TPR</th>\n",
       "      <th>DVN</th>\n",
       "      <th>CR</th>\n",
       "      <th>...</th>\n",
       "      <th>PGR</th>\n",
       "      <th>WAT</th>\n",
       "      <th>NNN</th>\n",
       "      <th>IEX</th>\n",
       "      <th>LRCX</th>\n",
       "      <th>NUS</th>\n",
       "      <th>NWL</th>\n",
       "      <th>UAA</th>\n",
       "      <th>BLK</th>\n",
       "      <th>PPL</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Date</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",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2011-12-23</th>\n",
       "      <td>0.018753</td>\n",
       "      <td>-0.020237</td>\n",
       "      <td>0.011614</td>\n",
       "      <td>0.021882</td>\n",
       "      <td>-0.001494</td>\n",
       "      <td>0.001786</td>\n",
       "      <td>0.008702</td>\n",
       "      <td>0.024272</td>\n",
       "      <td>-0.005726</td>\n",
       "      <td>0.008050</td>\n",
       "      <td>...</td>\n",
       "      <td>0.011543</td>\n",
       "      <td>0.000939</td>\n",
       "      <td>0.000381</td>\n",
       "      <td>0.009111</td>\n",
       "      <td>0.004898</td>\n",
       "      <td>0.008929</td>\n",
       "      <td>0.013871</td>\n",
       "      <td>0.036691</td>\n",
       "      <td>0.016639</td>\n",
       "      <td>0.003724</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2011-12-27</th>\n",
       "      <td>0.005956</td>\n",
       "      <td>-0.032746</td>\n",
       "      <td>-0.001218</td>\n",
       "      <td>0.012695</td>\n",
       "      <td>-0.004059</td>\n",
       "      <td>-0.002773</td>\n",
       "      <td>0.003299</td>\n",
       "      <td>0.008989</td>\n",
       "      <td>-0.002399</td>\n",
       "      <td>0.011224</td>\n",
       "      <td>...</td>\n",
       "      <td>0.010373</td>\n",
       "      <td>0.000402</td>\n",
       "      <td>0.013318</td>\n",
       "      <td>-0.008482</td>\n",
       "      <td>0.004874</td>\n",
       "      <td>0.006110</td>\n",
       "      <td>-0.001244</td>\n",
       "      <td>0.007106</td>\n",
       "      <td>-0.003195</td>\n",
       "      <td>0.005059</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2011-12-28</th>\n",
       "      <td>-0.022605</td>\n",
       "      <td>-0.013021</td>\n",
       "      <td>-0.019680</td>\n",
       "      <td>-0.012450</td>\n",
       "      <td>-0.009119</td>\n",
       "      <td>-0.017680</td>\n",
       "      <td>-0.019558</td>\n",
       "      <td>-0.014253</td>\n",
       "      <td>-0.019243</td>\n",
       "      <td>-0.020704</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.006673</td>\n",
       "      <td>-0.018226</td>\n",
       "      <td>-0.015020</td>\n",
       "      <td>-0.009106</td>\n",
       "      <td>-0.013743</td>\n",
       "      <td>0.002304</td>\n",
       "      <td>-0.005604</td>\n",
       "      <td>-0.015604</td>\n",
       "      <td>-0.007872</td>\n",
       "      <td>-0.003691</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2011-12-29</th>\n",
       "      <td>0.004956</td>\n",
       "      <td>-0.005805</td>\n",
       "      <td>0.019009</td>\n",
       "      <td>0.018200</td>\n",
       "      <td>0.028328</td>\n",
       "      <td>0.001415</td>\n",
       "      <td>0.019430</td>\n",
       "      <td>0.011173</td>\n",
       "      <td>0.018476</td>\n",
       "      <td>0.023104</td>\n",
       "      <td>...</td>\n",
       "      <td>0.011886</td>\n",
       "      <td>0.012012</td>\n",
       "      <td>0.012581</td>\n",
       "      <td>0.029518</td>\n",
       "      <td>0.012295</td>\n",
       "      <td>0.012954</td>\n",
       "      <td>0.028178</td>\n",
       "      <td>-0.009786</td>\n",
       "      <td>0.016776</td>\n",
       "      <td>0.002021</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2011-12-30</th>\n",
       "      <td>-0.009315</td>\n",
       "      <td>0.001592</td>\n",
       "      <td>-0.007148</td>\n",
       "      <td>-0.006480</td>\n",
       "      <td>0.001460</td>\n",
       "      <td>-0.017569</td>\n",
       "      <td>-0.005845</td>\n",
       "      <td>-0.008125</td>\n",
       "      <td>-0.004655</td>\n",
       "      <td>-0.004900</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.003575</td>\n",
       "      <td>-0.001214</td>\n",
       "      <td>-0.006777</td>\n",
       "      <td>0.003787</td>\n",
       "      <td>-0.000810</td>\n",
       "      <td>0.001857</td>\n",
       "      <td>-0.016444</td>\n",
       "      <td>-0.000696</td>\n",
       "      <td>-0.006466</td>\n",
       "      <td>-0.011092</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 741 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "ticker          CSCO       UAL      TROW      ISRG       NVR      PRGO  \\\n",
       "Date                                                                     \n",
       "2011-12-23  0.018753 -0.020237  0.011614  0.021882 -0.001494  0.001786   \n",
       "2011-12-27  0.005956 -0.032746 -0.001218  0.012695 -0.004059 -0.002773   \n",
       "2011-12-28 -0.022605 -0.013021 -0.019680 -0.012450 -0.009119 -0.017680   \n",
       "2011-12-29  0.004956 -0.005805  0.019009  0.018200  0.028328  0.001415   \n",
       "2011-12-30 -0.009315  0.001592 -0.007148 -0.006480  0.001460 -0.017569   \n",
       "\n",
       "ticker          LECO       TPR       DVN        CR    ...          PGR  \\\n",
       "Date                                                  ...                \n",
       "2011-12-23  0.008702  0.024272 -0.005726  0.008050    ...     0.011543   \n",
       "2011-12-27  0.003299  0.008989 -0.002399  0.011224    ...     0.010373   \n",
       "2011-12-28 -0.019558 -0.014253 -0.019243 -0.020704    ...    -0.006673   \n",
       "2011-12-29  0.019430  0.011173  0.018476  0.023104    ...     0.011886   \n",
       "2011-12-30 -0.005845 -0.008125 -0.004655 -0.004900    ...    -0.003575   \n",
       "\n",
       "ticker           WAT       NNN       IEX      LRCX       NUS       NWL  \\\n",
       "Date                                                                     \n",
       "2011-12-23  0.000939  0.000381  0.009111  0.004898  0.008929  0.013871   \n",
       "2011-12-27  0.000402  0.013318 -0.008482  0.004874  0.006110 -0.001244   \n",
       "2011-12-28 -0.018226 -0.015020 -0.009106 -0.013743  0.002304 -0.005604   \n",
       "2011-12-29  0.012012  0.012581  0.029518  0.012295  0.012954  0.028178   \n",
       "2011-12-30 -0.001214 -0.006777  0.003787 -0.000810  0.001857 -0.016444   \n",
       "\n",
       "ticker           UAA       BLK       PPL  \n",
       "Date                                      \n",
       "2011-12-23  0.036691  0.016639  0.003724  \n",
       "2011-12-27  0.007106 -0.003195  0.005059  \n",
       "2011-12-28 -0.015604 -0.007872 -0.003691  \n",
       "2011-12-29 -0.009786  0.016776  0.002021  \n",
       "2011-12-30 -0.000696 -0.006466 -0.011092  \n",
       "\n",
       "[5 rows x 741 columns]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "risk_model.returns.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Timestamp('2011-12-30 00:00:00')"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "date_ = pd.to_datetime('2011-12-30')\n",
    "date_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "741"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "alpha_vector = momentum.loc['2012-01-05', :]\n",
    "len(alpha_vector)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "741"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "previous = np.asarray(holdings)\n",
    "len(previous)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "oh = OptimalHoldings(risk_model, alpha_vector, previous, lambda_reg=0.0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "this is av.shape:  (741, 1)\n",
      "this is self.previous.shape:  (741,)\n",
      "(741, 1)\n",
      "(741,)\n",
      "False\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Minimize(Expression(AFFINE, UNKNOWN, (1,)))"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "oh._get_obj()\n",
    "oh.obj_func"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(20, 741)"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "risk_model.Q.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Expression(CONVEX, NONNEGATIVE, (1, 1))"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "oh.get_risk()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True\n"
     ]
    }
   ],
   "source": [
    "result = oh.find()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(741, 1)"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/richard.wuebker/.pyenv/versions/zipline/lib/python3.5/site-packages/matplotlib/figure.py:98: MatplotlibDeprecationWarning: \n",
      "Adding an axes using the same arguments as a previous axes currently reuses the earlier instance.  In a future version, a new instance will always be created and returned.  Meanwhile, this warning can be suppressed, and the future behavior ensured, by passing a unique label to each axes instance.\n",
      "  \"Adding an axes using the same arguments as a previous axes \"\n"
     ]
    },
    {
     "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": [
    "(result).plot.bar(legend=False)\n",
    "x_axis = plt.axes().get_xaxis()\n",
    "x_axis.set_visible(False)\n",
    "plt.tight_layout()\n",
    "plt.style.use('ggplot')\n",
    "plt.rcParams['figure.figsize'] = (14, 8)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "this is av.shape:  (741, 1)\n",
      "this is self.previous.shape:  (741,)\n",
      "(741, 1)\n",
      "(741,)\n",
      "False\n"
     ]
    }
   ],
   "source": [
    "oh = OptimalHoldings(risk_model, alpha_vector, previous, lambda_reg=0.75)\n",
    "oh._get_obj()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True\n"
     ]
    }
   ],
   "source": [
    "result = oh.find()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/richard.wuebker/.pyenv/versions/zipline/lib/python3.5/site-packages/matplotlib/figure.py:98: MatplotlibDeprecationWarning: \n",
      "Adding an axes using the same arguments as a previous axes currently reuses the earlier instance.  In a future version, a new instance will always be created and returned.  Meanwhile, this warning can be suppressed, and the future behavior ensured, by passing a unique label to each axes instance.\n",
      "  \"Adding an axes using the same arguments as a previous axes \"\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "(result).plot.bar(legend=False)\n",
    "x_axis = plt.axes().get_xaxis()\n",
    "x_axis.set_visible(False)\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 126,
   "metadata": {},
   "outputs": [],
   "source": [
    "class OptimalHoldingsCVX:\n",
    "\n",
    "    def __init__(self, risk_model, alpha_vector, previous, risk_cap=0.05, factor_max=10.0, aum=50e6, lambda_reg=0.50,\n",
    "                 factor_min=-10.0, weights_max=0.01, weights_min=-0.01, risk_aversion=1.0e-6):\n",
    "\n",
    "        self.risk_model = risk_model\n",
    "        self.alpha_vector = alpha_vector\n",
    "        self.previous = previous\n",
    "        self.risk_cap = risk_cap\n",
    "        self.factor_min = factor_min\n",
    "        self.factor_max = factor_max\n",
    "        self.weights_min = weights_min\n",
    "        self.weights_max = weights_max\n",
    "        self.risk_aversion = risk_aversion\n",
    "        self.aum = aum\n",
    "        self.notional_risk = risk_cap * aum\n",
    "        self.lambda_reg = lambda_reg\n",
    "\n",
    "    def _get_obj(self):\n",
    "        ra = self.risk_aversion\n",
    "        Q = self.risk_model.Q # Q is k x N\n",
    "        ivv = self.risk_model.i_var_vector\n",
    "        av = self.alpha_vector.values.reshape(len(self.alpha_vector), 1)\n",
    "        h0 = self.previous\n",
    "        lambda_vector = np.ones((len(av), 1)) * 0.1/2000000\n",
    "        lv = lambda_vector\n",
    "        h1 = cvx.Variable(len(self.alpha_vector))\n",
    "        m = ra * 0.50\n",
    "        #func = m * cvx.sum((Q @ h1)**2) + m * ((h1**2) @ ivv) - (h1 @ av) + ((h1-h0)**2) @ lv + self.lambda_reg * cvx.norm(h1, 2)\n",
    "        func = -(h1 @ av) + self.lambda_reg * cvx.norm(h1, 2)\n",
    "        func = cvx.Minimize(func)\n",
    "        prob = cvx.Problem(func, [])\n",
    "        print(prob.is_dcp())\n",
    "        #av = -av\n",
    "        self.h1 = h1\n",
    "        self.ivv = ivv\n",
    "        self.Q = Q\n",
    "        #func = cvx.Minimize(av.T @ h1 + self.lambda_reg * cvx.norm(h1, 2))\n",
    "        self.obj_func = func\n",
    "\n",
    "    def _get_constraints(self):\n",
    "        risk = self.get_risk()\n",
    "        c = []\n",
    "        h1 = self.h1\n",
    "        factor_exposures = self.risk_model.factor_exposures\n",
    "        c.append(risk <= self.risk_cap ** 2)\n",
    "        c.append(h1 <= self.weights_max)\n",
    "        c.append(h1 >= self.weights_min)\n",
    "        c.append(sum(h1) == 0)\n",
    "        c.append(factor_exposures.T @ h1 <= self.factor_max)\n",
    "        c.append(factor_exposures.T @ h1 >= self.factor_min)\n",
    "        c.append(sum(cvx.abs(h1)) <= 1.0)\n",
    "        return c\n",
    "\n",
    "    def get_risk(self):\n",
    "        Q = self.risk_model.Q\n",
    "        h1 = self.h1\n",
    "        ivv = self.risk_model.i_var_vector\n",
    "        result = cvx.sum((Q @ h1)**2) + (h1**2) @ ivv\n",
    "        ivm = self.risk_model.i_var_matrix.values\n",
    "        f = self.risk_model.factor_exposures.T @ h1\n",
    "        X = self.risk_model.factor_cov_matrix\n",
    "        return cvx.quad_form(f, X) + cvx.quad_form(h1, ivm)\n",
    "\n",
    "    def find(self):\n",
    "        h1 = self.h1\n",
    "        prob = cvx.Problem(self.obj_func, self._get_constraints())\n",
    "        print(prob.is_dcp())\n",
    "        prob.solve(max_iters=1000)\n",
    "        optimal_weights = np.asarray(h1.value).flatten()\n",
    "        df = pd.DataFrame(optimal_weights, index=self.alpha_vector.index)\n",
    "        return df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 127,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True\n"
     ]
    }
   ],
   "source": [
    "oh = OptimalHoldingsCVX(risk_model, alpha_vector, previous, lambda_reg=1.0)\n",
    "oh._get_obj()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "True\n"
     ]
    }
   ],
   "source": [
    "result = oh.find()\n",
    "(result).plot.bar(legend=False)\n",
    "x_axis = plt.axes().get_xaxis()\n",
    "x_axis.set_visible(False)\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x14ae30940>"
      ]
     },
     "execution_count": 118,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(oh.alpha_vector, result)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "741\n",
      "741\n",
      "ticker\n",
      "CSCO   -0.471847\n",
      "UAL    -1.326777\n",
      "TROW   -0.228916\n",
      "ISRG    1.691173\n",
      "NVR    -0.369068\n",
      "Name: 2012-01-05 00:00:00, dtype: float64\n",
      "                   0\n",
      "ticker              \n",
      "CSCO    1.333379e-11\n",
      "UAL     3.681273e-11\n",
      "TROW    5.984465e-12\n",
      "ISRG    1.000000e-02\n",
      "NVR     1.068318e-11\n"
     ]
    }
   ],
   "source": [
    "print(len(alpha_vector))\n",
    "print(len(result))\n",
    "print(alpha_vector.head())\n",
    "print(result.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 124,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_transfer_coefficient(a, b):\n",
    "    tc, p = pearsonr(a, b)\n",
    "    return tc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 125,
   "metadata": {},
   "outputs": [
    {
     "ename": "TypeError",
     "evalue": "No loop matching the specified signature and casting was found for ufunc add",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-125-f3c40332b822>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mget_transfer_coefficient\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0malpha_vector\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mresult\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m<ipython-input-124-c9d066f4edbf>\u001b[0m in \u001b[0;36mget_transfer_coefficient\u001b[0;34m(a, b)\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mget_transfer_coefficient\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ma\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m     \u001b[0mtc\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mp\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mpearsonr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ma\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      3\u001b[0m     \u001b[0;32mreturn\u001b[0m \u001b[0mtc\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/zipline/lib/python3.5/site-packages/scipy/stats/stats.py\u001b[0m in \u001b[0;36mpearsonr\u001b[0;34m(x, y)\u001b[0m\n\u001b[1;32m   3408\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mdtype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msign\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m-\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msign\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m-\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m1.0\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   3409\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 3410\u001b[0;31m     \u001b[0mxmean\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmean\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdtype\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdtype\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   3411\u001b[0m     \u001b[0mymean\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmean\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdtype\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdtype\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   3412\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/.pyenv/versions/zipline/lib/python3.5/site-packages/numpy/core/_methods.py\u001b[0m in \u001b[0;36m_mean\u001b[0;34m(a, axis, dtype, out, keepdims)\u001b[0m\n\u001b[1;32m    149\u001b[0m             \u001b[0mis_float16_result\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mTrue\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    150\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 151\u001b[0;31m     \u001b[0mret\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mumr_sum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0marr\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maxis\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdtype\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mout\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkeepdims\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    152\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mret\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmu\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mndarray\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    153\u001b[0m         ret = um.true_divide(\n",
      "\u001b[0;31mTypeError\u001b[0m: No loop matching the specified signature and casting was found for ufunc add"
     ]
    }
   ],
   "source": [
    "get_transfer_coefficient(alpha_vector, result)"
   ]
  },
  {
   "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": "3.5.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}