path: root/python/notebooks/Curve Trades.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import analytics.curve_trades as ct\n",
    "import matplotlib.pyplot as plt\n",
    "import pandas as pd\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import graphics as g\n",
    "import globeop_reports as go\n",
    "\n",
    "from ipywidgets import widgets\n",
    "from analytics.scenarios import run_curve_scenarios\n",
    "from db import dbengine"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "w = widgets.Dropdown(\n",
    "    options=['IG', 'EU'],\n",
    "    value='IG',\n",
    "    description='Index:',\n",
    "    disabled=False,\n",
    ")\n",
    "w"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "index = w.value\n",
    "report_date = (pd.datetime.today() - pd.offsets.BDay(2)).normalize()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#On the run spread differences\n",
    "spreads_diff = ct.curve_spread_diff(index, 6)\n",
    "spreads_diff.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Table of Spread Differences, and Z-score of current spread differences\n",
    "ct.spreads_diff_table(spreads_diff)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Theta per unit duration\n",
    "ct.theta_matrix_by_series(index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#on the run theta\n",
    "ct.on_the_run_theta(index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Curve Trade returns\n",
    "ct.curve_returns()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "ct.cross_series_curve(index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Theta with 3-5-10 Strategy\n",
    "df = ct.ratio_within_series(param='duration')\n",
    "s = - df.theta2['3yr'] / df.duration_ratio_to_5yr['3yr'] \\\n",
    "    + 2 * df.theta2['5yr'] \\\n",
    "    - df.theta2['10yr'] / df.duration_ratio_to_5yr['10yr']\n",
    "s.dropna().unstack(-1).plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Theta with 5-10 Strategy: buy sell 5y, buy 10y\n",
    "s = df.theta2['5yr'] - df.theta2['10yr'] / df.duration_ratio_to_5yr['10yr']\n",
    "s.dropna().unstack(-1).plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Relative Spread Difference\n",
    "spread_ratio = ct.ratio_within_series(param = 'closespread')\n",
    "spread_ratio.groupby(level = ['date']).last()['closespread_ratio_to_5yr'].plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = ct.curve_model('5yr', '10yr')\n",
    "model_results = ct.curve_model_results(model[0], model[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = ct.forward_spread(report_date, index)\n",
    "df.plot()\n",
    "plt.ylabel('spread')\n",
    "plt.xlabel('forward spread start date')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = ct.spot_forward(index)\n",
    "df = df.rename(columns={'1yr': 'Spot Spread - 1 Year Forward', 'current': 'Spot Spread - Today'})\n",
    "ax = df.plot(title = 'Credit Curve Roll Down')\n",
    "plt.ylabel('spread (bps)')\n",
    "ax.figure.savefig(\"/home/serenitas/edwin/PythonGraphs/curve_trades_roll_down.png\", bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "portf = ct.curve_pos(report_date, index)\n",
    "shock_min = -.5\n",
    "shock_max = .8\n",
    "spread_shock = np.arange(shock_min, shock_max, 0.05)\n",
    "sql_string = \"SELECT closespread FROM index_quotes where index = %s and series = %s and tenor = %s and date = %s\"\n",
    "spread_df = pd.read_sql_query(sql_string, dbengine('serenitasdb'),\n",
    "                              params=[index, ct.on_the_run(index), '5yr', report_date])\n",
    "spread_range = (1+ spread_shock) * spread_df.iloc[0][0]\n",
    "#need to max it at the closest maturity date\n",
    "date_range = pd.bdate_range(report_date, report_date + 180* pd.offsets.DateOffset(), freq='5B')\n",
    "curve_per = np.arange(.01, .99, .1)\n",
    "\n",
    "df = run_curve_scenarios(portf, spread_range, date_range, curve_per)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_plot = df[df.curve_per == curve_per[5]]\n",
    "g.plot_time_color_map(df_plot, spread_range, attr='pnl')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Scenario Anslysis on current position\n",
    "#curve_positions = ct.curve_pos(report_date, index)\n",
    "#origpv = curve_positions.pv\n",
    "#flat_curve = ct.curve_shape(report_date, index, percentile = .05)\n",
    "#for ind in curve_positions.indices:\n",
    "#    ind.spread = flat_curve((pd.to_datetime(ind.end_date) - report_date).days/365)\n",
    "#PNL in flattening to a 5% case\n",
    "#curve_positions.pv - origpv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "curve_positions = ct.curve_pos(report_date, index)\n",
    "df = ct.pos_pnl_abs(curve_positions, report_date)\n",
    "navs = go.get_net_navs()\n",
    "df_plot = df.pnl/navs.iloc[-1].endbooknav"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fig, ax = plt.subplots()\n",
    "ax.plot(df_plot.index, df_plot.values)\n",
    "ax.set(xlabel='date', ylabel='% of NAV',\n",
    "       title='PNL impact from spread curve scenario')\n",
    "plt.xticks(rotation=90)\n",
    "y_ticks = ax.get_yticks()\n",
    "ax.set_yticklabels(['{:.2f}%'.format(y*100) for y in y_ticks])\n",
    "plt.tight_layout()\n",
    "#ax.figure.savefig(\"/home/serenitas/edwin/PythonGraphs/curve_trades.png\", bbox_inches='tight')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Historical PNL in a 5% case\n",
    "df.pnl.quantile(.05)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "scen_table = ct.curve_scen_table(curve_positions)\n",
    "scen_table.pnl = scen_table.pnl/navs.iloc[-1].endbooknav *100\n",
    "scen_table.pivot(index='tighter', columns='wider')"
   ]
  },
  {
   "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": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}