AFML 16.1.1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Machine-Learning Asset Allocation\n",
    "\n",
    "Most of the functions below can be found under:\n",
    "\n",
    "* Portfolio_opt/**\n",
    "\n",
    "This exercise is not part of AFML syllabus, but forms part of portfolio optimization series.\n",
    "\n",
    "The focused optimization technique will be Nested Clustered Optimization (NCO), which has to be combined with other portfolio management style to form a complete picture.\n",
    "\n",
    "There are 2 other techniques available; Minimize Variance (IVP) and Maximize Returns.\n",
    "\n",
    "For simplicity, the exercise will minimize variance only instead of return maximization.\n",
    "\n",
    "Like HRP, this technique requires large investment universe or AUM estimated (> 350 millions at least) to implement.\n",
    "\n",
    "There will be 2 innovations to this method.\n",
    "1. Solving convex optimization problems, while controlling for instabilities caused by input noise and complex signal structure\n",
    "2. Evaluate the estimation error associated with a variety of optimization approaches\n",
    "\n",
    "If you are keen on generating synthetic data for your research, copy the code snippets [Generate synthetic raw data](https://gist.github.com/boyboi86/5e00faf48f60abfdbe838fbdee269471) in my gist.\n",
    "\n",
    "Additional resource: \n",
    "* [Critical Line Aglorithm](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2197616)\n",
    "* [Efficient Frontier Robust Estimator](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3469961)\n",
    "* [ML Asset Allocation](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3469964)\n",
    "* [Detection of False Investment Strategies Using Unsupervised Learning Methods](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3167017)\n",
    "* [Estimation of Implied-Correlation Matrix](https://poseidon01.ssrn.com/delivery.php?ID=777119081090115114117002126067074029117046025068004010066117097028007110104100113119029035055009008044005105102126124025012120106071060023014122071102065110095077066011054101118071118101024105021102075086098114101095105016028088066084075003091083020&EXT=pdf)\n",
    "\n",
    "Contact: boyboi86@gmail.com"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Num of CPU core:  4\n",
      "Machine info:  Windows-10-10.0.18362-SP0\n",
      "Python 3.7.4 (default, Aug  9 2019, 18:34:13) [MSC v.1915 64 bit (AMD64)]\n",
      "Numpy 1.18.5\n",
      "Pandas 1.0.4\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import research as rs\n",
    "from scipy.linalg import block_diag\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def corr_mtx_plot(corr: float, path: str = None, labels = None, title = None):\n",
    "    plt.figure(figsize=(8,6))\n",
    "    if labels is None: \n",
    "        labels = []\n",
    "    plt.pcolor(corr)\n",
    "    plt.colorbar()\n",
    "    plt.yticks(np.arange(.5, corr.shape[0] + .5), labels)\n",
    "    plt.xticks(np.arange(.5, corr.shape[0] + .5), labels)\n",
    "    if path is not None:\n",
    "        plt.savefig(path)\n",
    "        plt.clf()\n",
    "        plt.close()\n",
    "    if title is not None:\n",
    "        plt.title(title)\n",
    "    plt.show()\n",
    "\n",
    "def corr2cov(corr,std):\n",
    "    cov=corr*np.outer(std,std)\n",
    "    return cov\n",
    "\n",
    "#------------------------------------------------------------------------------\n",
    "def formBlockMatrix(nBlocks,bSize,bCorr):\n",
    "    block=np.ones((bSize,bSize))*bCorr\n",
    "    block[range(bSize),range(bSize)]=1\n",
    "    corr=block_diag(*([block]*nBlocks))\n",
    "    return corr\n",
    "#------------------------------------------------------------------------------\n",
    "def formTrueMatrix(nBlocks,bSize,bCorr,std0=None):\n",
    "    corr0=formBlockMatrix(nBlocks,bSize,bCorr)\n",
    "    corr0=pd.DataFrame(corr0)\n",
    "    corr_mtx_plot(corr = corr0, title = \"Correlation Matrix Heat-Map\")\n",
    "    cols=corr0.columns.tolist()\n",
    "    np.random.shuffle(cols) #this portion is actually not necessary for this example\n",
    "    corr0=corr0[cols].loc[cols].copy(deep=True)\n",
    "    if std0 is None:std0=np.random.uniform(.05,.2,corr0.shape[0])\n",
    "    else:std0=np.array([std0]*corr0.shape[1])\n",
    "    cov0=corr2cov(corr0,std0)\n",
    "    mu0=np.random.normal(std0,std0,cov0.shape[0]).reshape(-1,1)\n",
    "    return mu0,cov0\n",
    "\n",
    "#=======================================================\n",
    "def create_true_mtx(nBlocks: int = 10, bSize: int = 5, bCorr: float = .5, seed: int = 0):\n",
    "    np.random.seed(seed)\n",
    "    mu0,cov0=formTrueMatrix(nBlocks,bSize,bCorr)\n",
    "    return mu0, cov0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ave_return, cov0 = create_true_mtx(nBlocks = 10, bSize = 5, bCorr = .5, seed = 0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Note**\n",
    "\n",
    "The above Heat-Map reflects Correlation Matrix based on the synthetic data generated.\n",
    "\n",
    "We can resample to a lower frequency (From daily to weekly) to improve the heat map distribution. [AFML 16.1](https://github.com/boyboi86/AFML/blob/master/AFML%2016.1.ipynb)\n",
    "\n",
    "Comparing the above matrix to HRP's. It seems that the above has went through quasi-diagonization.\n",
    "\n",
    "**Note**\n",
    "\n",
    "If you feed the algorithm with sub-zero numbers, it will prompt errors. Alot of times.\n",
    "\n",
    "For now, we can still \"cheat\".\n",
    "\n",
    "Convexity optimization byitself is supposed to be a \"long\" technique.\n",
    "\n",
    "Adjustments will be necessary if bi-directional approach were used.\n",
    "\n",
    "**Note**\n",
    "\n",
    "Moreover, we are only using 50 assets which is considered very small.\n",
    "\n",
    "**Note**\n",
    "\n",
    "Threshold is automatically set at 0.00001"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings(action = \"ignore\",\n",
    "                        category=RuntimeWarning) # too annoying\n",
    "\n",
    "w0, w1, w1_d, err, err_d = rs.NCO(mu0 = ave_return,\n",
    "                                 cov0 = cov0).solve( n_runs = 30,\n",
    "                                                     bWidth = .25,\n",
    "                                                     minVarPortf = True,\n",
    "                                                     shrink = False)\n",
    "\n",
    "def report_(allocation):\n",
    "    tw = allocation[\"True Weights\"].mul(allocation[\"Average Return\"]) * 10\n",
    "    cvo = allocation[\"Covexity-Optimized Weights\"].mul(allocation[\"Average Return\"]) * 10\n",
    "    nco  = allocation[\"Nested-Cluster Weights\"].mul(allocation[\"Average Return\"]) * 10\n",
    "    er0, std0 = np.sum(tw), np.std(tw)\n",
    "    er1, std1 = np.sum(cvo), np.std(cvo)\n",
    "    er2, std2 = np.sum(nco), np.std(nco)\n",
    "    print(\"\\nExpected Return\\n{0}\\nTrue Weights: {1:.3f}\".format(\"=\" * 33, er0))\n",
    "    print(\"Covexity-Optimized Weights: {0:.3f}\".format(er1))\n",
    "    print(\"Nested-Cluster Weights: {0:.3f}\".format(er2))\n",
    "    \n",
    "    print(\"\\nStandard Deviation\\n{0}\\nTrue Weights: {1:.5f}\".format(\"=\" * 33, std0))\n",
    "    print(\"Covexity-Optimized Weights: {0:.5f}\".format(std1))\n",
    "    print(\"Nested-Cluster Weights: {0:.5f}\".format(std2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "MSE CVO: 0.010\n",
      "MSE NCO: 0.010\n",
      "\n",
      "Total Allocation\n",
      "TW: 1.000\n",
      "CVO: 1.000\n",
      "NCO: 1.000\n",
      "\n",
      "Expected Return\n",
      "=================================\n",
      "True Weights: 0.035\n",
      "Covexity-Optimized Weights: 0.081\n",
      "Nested-Cluster Weights: 0.059\n",
      "\n",
      "Standard Deviation\n",
      "=================================\n",
      "True Weights: 0.00331\n",
      "Covexity-Optimized Weights: 0.00206\n",
      "Nested-Cluster Weights: 0.00248\n"
     ]
    }
   ],
   "source": [
    "allocation = pd.DataFrame({\"True Weights\":w0[0], \n",
    "                           \"Covexity-Optimized Weights\": w1.mean(), \n",
    "                           \"Nested-Cluster Weights\": w1_d.mean(), \n",
    "                           \"Average Return\": ave_return.flatten() * 0.1}) #if not value too large\n",
    "\n",
    "allocation.index = np.arange(1,51)\n",
    "allocation.plot.bar(figsize=(15,8))\n",
    "plt.title(\"Minimum Variance 3 methods weightage comparison\")\n",
    "plt.xlabel(\"Assets\")\n",
    "plt.ylabel(\"Weights/ Returns\")\n",
    "plt.show()\n",
    "\n",
    "print(\"MSE CVO: {0:.3f}\\nMSE NCO: {0:.3f}\".format(err, err_d))\n",
    "print(\"\\nTotal Allocation\\nTW: {0:.3f}\\nCVO: {1:.3f}\\nNCO: {2:.3f}\".format(w0[0].sum(),\n",
    "                                                                           w1.mean().sum(),\n",
    "                                                                           w1_d.mean().sum()))\n",
    "\n",
    "report_(allocation)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### NCO vs CVO vs True Weights\n",
    "\n",
    "True Weights is mostly leader in all techniques.\n",
    "\n",
    "In most if not all cases, where NCO is higher than CVO weightage. True weight will become the \"leader\".\n",
    "\n",
    "**Note**\n",
    "\n",
    "Total allocation for NCO and True weight is 1. This is not correct.\n",
    "\n",
    "By right, allocation must be within range (0,1). Given the fact we are assigning negative weights.\n",
    "\n",
    "Total allocation must be 0. In the event if bi-directional portfolio is applied. There will be a need to adjust the weights to tally."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Total Allocation\n",
      "TW: 0.762\n",
      "CVO: 1.000\n",
      "NCO: 0.944\n",
      "\n",
      "Expected Return\n",
      "=================================\n",
      "True Weights: 0.026\n",
      "Covexity-Optimized Weights: 0.081\n",
      "Nested-Cluster Weights: 0.056\n",
      "\n",
      "Standard Deviation\n",
      "=================================\n",
      "True Weights: 0.00252\n",
      "Covexity-Optimized Weights: 0.00206\n",
      "Nested-Cluster Weights: 0.00234\n"
     ]
    }
   ],
   "source": [
    "\n",
    "sum0 = np.absolute(w0[0]).sum()\n",
    "sum1 = np.absolute(w1_d.mean()).sum()\n",
    "allocation[\"True Weights\"] = allocation[\"True Weights\"].div(sum0)\n",
    "allocation[\"Nested-Cluster Weights\"] = allocation[\"Nested-Cluster Weights\"].div(sum1)\n",
    "\n",
    "allocation.plot.bar(figsize=(15,8))\n",
    "plt.title(\"Minimum Variance with adjusted-weightage comparison (bi-directional)\")\n",
    "plt.xlabel(\"Assets\")\n",
    "plt.ylabel(\"Weights/ Returns\")\n",
    "plt.show()\n",
    "\n",
    "print(\"\\nTotal Allocation\\nTW: {0:.3f}\\nCVO: {1:.3f}\\nNCO: {2:.3f}\".format(allocation[\"True Weights\"].sum(),\n",
    "                                                                           w1.mean().sum(),\n",
    "                                                                           allocation[\"Nested-Cluster Weights\"].sum()))\n",
    "\n",
    "\n",
    "report_(allocation)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "Total Allocation\n",
      "TW: 1.000\n",
      "CVO: 1.000\n",
      "NCO: 1.000\n",
      "\n",
      "Expected Return\n",
      "=================================\n",
      "True Weights: 0.072\n",
      "Covexity-Optimized Weights: 0.081\n",
      "Nested-Cluster Weights: 0.067\n",
      "\n",
      "Standard Deviation\n",
      "=================================\n",
      "True Weights: 0.00213\n",
      "Covexity-Optimized Weights: 0.00206\n",
      "Nested-Cluster Weights: 0.00222\n"
     ]
    }
   ],
   "source": [
    "allocation = allocation.apply(lambda x: np.absolute(x))\n",
    "allocation[\"Average Return\"] = ave_return.flatten() * 0.1\n",
    "allocation.plot.bar(figsize=(15,8))\n",
    "plt.title(\"Minimum Variance with adjusted-weightage comparison (Long-only)\")\n",
    "plt.xlabel(\"Assets\")\n",
    "plt.ylabel(\"Weights/ Returns\")\n",
    "plt.show()\n",
    "\n",
    "print(\"\\nTotal Allocation\\nTW: {0:.3f}\\nCVO: {1:.3f}\\nNCO: {2:.3f}\".format(allocation[\"True Weights\"].sum(),\n",
    "                                                                           w1.mean().sum(),\n",
    "                                                                           allocation[\"Nested-Cluster Weights\"].sum()))\n",
    "    \n",
    "report_(allocation)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Conclusion\n",
    "\n",
    "In the event where it is necessary to feed sub-returns, make sure you adjust the weights.\n",
    "\n",
    "The True weight allocation will produce the second highest return and standard deviation.\n",
    "\n",
    "Which Is the sweet spot between hierachical cluster (Correlation-minimization) and convexity optimization (Returns-maximization).\n",
    "\n",
    "In short:\n",
    "\n",
    "* The De-noising algorithm calculates the eigenvalues of the correlation matrix and eliminates the ones that are higher than the theoretically estimated ones, as they are caused by noise.\n",
    "* The Convex Optimization Solution (CVO) can either find optimal allocations for maximum Sharpe ratio or minimum variance.\n",
    "* The Nested Clustered Optimization (NCO) breaks the correlation matrix into clusters and applies CVO individually to each cluster as well as to the correlation matrix where each element represents a cluster.\n",
    "* The Monte Carlo Optimization Selection (MCOS) allows running optimization methods (in our case CVO and NCO) on multiple sets of empirical covariation matrices and empirical vectors of means. It can show which method is more robust.\n",
    "\n",
    "**Note**\n",
    "\n",
    "If you notice the below graphs contains strong resemblance in weight assignment to the HRP graph. (Right to left)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "allocation.drop(columns = [\"Nested-Cluster Weights\",\"Covexity-Optimized Weights\"]).plot.bar(figsize=(15,8))\n",
    "plt.title(\"True Weights and Return Comparison\")\n",
    "plt.xlabel(\"Assets\")\n",
    "plt.ylabel(\"Weights/ Returns\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Note**\n",
    "\n",
    "True weights tend to maximization returns/ minimize risk while striking the balance to diversify portfolio based on \"clusters\".\n",
    "\n",
    "In terms of return, True weights trail CVO but from the previous graphs it mimics NCO weightage very closely."
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}