06c-rank-and-trade.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\keb7\\repos\\mgmt767-binder\\venv\\lib\\site-packages\\sklearn\\base.py:288: UserWarning: Trying to unpickle estimator PolynomialFeatures from version 1.1.2 when using version 1.2.0. This might lead to breaking code or invalid results. Use at your own risk. For more info please refer to:\n",
      "https://scikit-learn.org/stable/model_persistence.html#security-maintainability-limitations\n",
      "  warnings.warn(\n",
      "c:\\Users\\keb7\\repos\\mgmt767-binder\\venv\\lib\\site-packages\\sklearn\\base.py:288: UserWarning: Trying to unpickle estimator LinearRegression from version 1.1.2 when using version 1.2.0. This might lead to breaking code or invalid results. Use at your own risk. For more info please refer to:\n",
      "https://scikit-learn.org/stable/model_persistence.html#security-maintainability-limitations\n",
      "  warnings.warn(\n",
      "c:\\Users\\keb7\\repos\\mgmt767-binder\\venv\\lib\\site-packages\\sklearn\\base.py:288: UserWarning: Trying to unpickle estimator Pipeline from version 1.1.2 when using version 1.2.0. This might lead to breaking code or invalid results. Use at your own risk. For more info please refer to:\n",
      "https://scikit-learn.org/stable/model_persistence.html#security-maintainability-limitations\n",
      "  warnings.warn(\n"
     ]
    }
   ],
   "source": [
    "\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.preprocessing import QuantileTransformer\n",
    "from joblib import load\n",
    "\n",
    "from sqlalchemy import create_engine\n",
    "import pymssql\n",
    "from datetime import datetime\n",
    "\n",
    "from alpaca.trading.client import TradingClient\n",
    "from alpaca.data import StockHistoricalDataClient\n",
    "from alpaca.data.requests import StockLatestQuoteRequest\n",
    "from alpaca.trading.requests import MarketOrderRequest, GetAssetsRequest\n",
    "from alpaca.trading.enums import OrderSide, TimeInForce, AssetClass\n",
    "\n",
    "KEY = \"PKX210XR6N17LF7WIWC8\"\n",
    "SECRET_KEY = \"Cc6VXW58xRrsMQ98kbkZvw5dAOKKPMZaiRpLGraG\"\n",
    "numstocks = 200\n",
    "\n",
    "qt = QuantileTransformer(output_distribution=\"normal\")\n",
    "pipe = load(\"files/linear_model_2023-01-20.joblib\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Rank stocks"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Get features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "server = 'fs.rice.edu'\n",
    "database = 'stocks'\n",
    "username = 'stocks'\n",
    "password = '6LAZH1'\n",
    "string = \"mssql+pymssql://\" + username + \":\" + password + \"@\" + server + \"/\" + database \n",
    "conn = create_engine(string).connect()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_sql(\n",
    "    \"\"\"\n",
    "    select ticker, date, bm, mom12m, roeq, mve, famaindustry\n",
    "    from today\n",
    "    where price > 5\n",
    "    \"\"\", \n",
    "    conn\n",
    ")\n",
    "conn.close()\n",
    "\n",
    "\n",
    "df = df.dropna()\n",
    "df = df.set_index(\"ticker\")\n",
    "df = df.sort_values(by=\"mve\")\n",
    "df = df.iloc[:-500]\n",
    "\n",
    "features = [\"bm\", \"mom12m\", \"roeq\"]"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Predict and rank"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "trans_features = qt.fit_transform(df[features])\n",
    "trans_features = pd.DataFrame(trans_features, columns=features)\n",
    "df[\"predict\"] = pipe.predict(trans_features)\n",
    "df[\"rnk\"] = df.predict.rank(method=\"first\")\n",
    "\n",
    "for f in features:\n",
    "    df[\"t\"+f] = qt.fit_transform(df[f].to_numpy().reshape(-1,1))"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Get data from Alpaca"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Get account equity"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "trading_client = TradingClient(KEY, SECRET_KEY, paper=True)\n",
    "account = trading_client.get_account()\n",
    "equity = float(account.equity)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Get tradeable and shortable stocks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "assets = trading_client.get_all_assets()\n",
    "\n",
    "assets = [\n",
    "    x for x in assets \n",
    "    if (x.asset_class[:]=='us_equity') \n",
    "    and (x.symbol in df.index) \n",
    "    and (x.status[:]=='active')\n",
    "]\n",
    "symbols = [x.symbol for x in assets]\n",
    "tradable = [x.tradable for x in assets]\n",
    "shortable = [x.shortable for x in assets]\n",
    "\n",
    "df[\"tradable\"]= pd.Series(tradable, index=symbols)\n",
    "df[\"shortable\"] = pd.Series(shortable, index=symbols)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Get quotes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_client = StockHistoricalDataClient(KEY, SECRET_KEY)\n",
    "params = StockLatestQuoteRequest(symbol_or_symbols=df.index.to_list())\n",
    "quotes = data_client.get_stock_latest_quote(params)\n",
    "\n",
    "df[\"ask\"] = [quotes[x].ask_price for x in df.index]\n",
    "df[\"bid\"] = [quotes[x].bid_price for x in df.index]"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Get positions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "positions = trading_client.get_all_positions()\n",
    "if len(positions) > 0:\n",
    "    positions = {x.symbol: int(x.qty) for x in positions}\n",
    "    positions = pd.Series(positions)\n",
    "    df[\"current\"] = positions\n",
    "    df[\"current\"] = df.current.fillna(0)\n",
    "else:\n",
    "    df[\"current\"] = 0"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Trade"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Calculate target positions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "long_per_stock = 1.3*equity / numstocks\n",
    "short_per_stock = 0.3*equity / numstocks\n",
    "\n",
    "df = df.sort_values(by=\"rnk\")\n",
    "\n",
    "try:\n",
    "    short_cutoff = df[df.shortable & (df.bid>0)].rnk.iloc[numstocks-1]\n",
    "    long_cutoff = df[df.tradable & (df.ask>0)].rnk.iloc[-numstocks]\n",
    "    df[\"target\"] = np.where(\n",
    "        df.shortable & (df.bid>0) & (df.rnk<=short_cutoff),\n",
    "        -short_per_stock/df.bid, \n",
    "        0\n",
    "    )\n",
    "    df[\"target\"] = np.where(\n",
    "        df.tradable & (df.ask>0) & (df.rnk>=long_cutoff), \n",
    "        long_per_stock/df.ask, \n",
    "        df.target\n",
    "    )\n",
    "    df[\"target\"] = df.target.astype(int)\n",
    "except:\n",
    "    df[\"target\"] = 0"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Calculate trades\n",
    "\n",
    "Using a simple but suboptimal protocol: trade to target positions without trying to minimize the number of round trips we might eventually make."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "df[\"trade\"] = df.target - df.current"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Make trades"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "for tick in df.index: \n",
    "    if df.loc[tick, \"trade\"]<0:\n",
    "        try:\n",
    "            market_order_data = MarketOrderRequest(\n",
    "                symbol=tick,\n",
    "                qty=-df.loc[tick, \"trade\"],\n",
    "                side=OrderSide.SELL,\n",
    "                time_in_force=TimeInForce.DAY\n",
    "            )\n",
    "            market_order = trading_client.submit_order(\n",
    "                order_data=market_order_data\n",
    "            )\n",
    "        except:\n",
    "            print(f\"sell order for {tick} failed\")\n",
    "    elif df.loc[tick, \"trade\"]>0:\n",
    "        try:\n",
    "            market_order_data = MarketOrderRequest(\n",
    "                symbol=tick,\n",
    "                qty=df.loc[tick, \"trade\"],\n",
    "                side=OrderSide.BUY,\n",
    "                time_in_force=TimeInForce.DAY\n",
    "            )\n",
    "            market_order = trading_client.submit_order(\n",
    "                order_data=market_order_data\n",
    "            )\n",
    "        except:\n",
    "            print(f\"buy order for {tick} failed\")\n",
    "\n"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Save data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "today = datetime.today().strftime(\"%Y-%m-%d\")\n",
    "\n",
    "df[\"date\"] = today\n",
    "\n",
    "try:\n",
    "    d = pd.read_csv(\"files/trade_data.csv\", index_col=\"ticker\")\n",
    "    d = d[d.date != today]\n",
    "    df = pd.concat((d, df))\n",
    "    df.to_csv(\"files/trade_data.csv\")\n",
    "except:\n",
    "    df.to_csv(\"files/trade_data.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\keb7\\AppData\\Local\\Temp\\ipykernel_107440\\1684087778.py:9: FutureWarning: In a future version, object-dtype columns with all-bool values will not be included in reductions with bool_only=True. Explicitly cast to bool dtype instead.\n",
      "  account = pd.concat((d, account))\n"
     ]
    }
   ],
   "source": [
    "account = trading_client.get_account()\n",
    "account = dict(account)\n",
    "account = pd.DataFrame(pd.Series(account)).T\n",
    "account[\"date\"] = today\n",
    "\n",
    "try:\n",
    "    d = pd.read_csv(\"files/account.csv\")\n",
    "    d = d[d.date != today]\n",
    "    account = pd.concat((d, account))\n",
    "    account.to_csv(\"files/account.csv\")\n",
    "except:\n",
    "    account.to_csv(\"files/account.csv\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "positions = trading_client.get_all_positions()\n",
    "positions = {x.symbol: x.qty for x in positions}\n",
    "positions = pd.DataFrame(pd.Series(positions))\n",
    "positions[\"date\"] = today\n",
    "\n",
    "try:\n",
    "    d = pd.read_csv(\"files/positions.csv\")\n",
    "    d = d[d.date != today]\n",
    "    positions = pd.concat((d, positions))\n",
    "    positions.to_csv(\"files/positions.csv\")\n",
    "except:\n",
    "    positions.to_csv(\"files/positions.csv\")"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "venv",
   "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": "3.10.8"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "24514d992635106b0c080fe619fc409ba929c1b95321a4abd0f8c840fea2ec36"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}