alpha/strategies/strategy_sv.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "pycharm": {
          "is_executing": false,
          "metadata": false,
          "name": "#%% Trade Strategy 4: Follow the upward trend.\n"
        }
      },
      "outputs": [],
      "source": "# %%file strategy_4.py\n\nimport numpy as np\nimport pandas as pd\nimport matplotlib.pyplot as plt\nimport mpl_finance as mpf\nfrom scipy import stats\nfrom matplotlib.dates import datestr2num\n\nclass TradeStrategySV:\n    \"\"\"\n    Trade Strategy 4: Slope and Velocity.\n    Overview:\n    Use close price to draw a trend, use the derivative to find the slope.\n    Use the derivative again to find the velocity for change.\n    \n    When price raises rapidly, slope \u003e 0, velocity \u003e 0\n    When price raises becomes slowly, slope \u003e 0, velocity close 0\n    When price peaks, slope \u003d 0, \n    When price falls rapidly, slope \u003c 0, velocity \u003e 0\n    When price falls slowly, slope \u003c 0, velocity close to 0 \n    When price bottoms, slope \u003d 0,\n    \n    Therefore, we can use a slope and velocity to find trends and make trades.\n    \n    Long when price is rising (slope \u003e _buy_slope_threshold \u003e 0) \u0026 accelerating (velocity \u003e _buy_velocity_threshold \u003e 0) \u0026 Close when price starting to fall (slope \u003c -_close_slope_threshold \u003c 0 and velocity \u003c -_close_velocity_threshold \u003c 0)\n    Short when price is falling (slope \u003c _sell_slope_threshold \u003c 0) \u0026 accelerating (velocity \u003c _sell_velocity_threshold \u003c 0) \u0026 Close when price starting to fall (slope \u003e _close_slope_threshold \u003e 0 and velocity \u003e _close_velocity_threshold \u003e 0)\n    \n    Assumptions: \n    \n    Requirement: \n    1. This strategy should work with any highly liquid security. Preferably Forex or Crypto.\n    2. This strategy should work whether trend is up or down, volatile or not. \n    3. The more volatile, the better, as we need to quickly close the position after mean reversion, BEFORE moving_average moved too much. \n    Constants: \n    Params:\n    _buy_slope_threshold:       # The slope required to trigger BUY\n    _buy_velocity_threshold:    # The velocity required to trigger BUY\n    _close_slope_threshold:     # The slope required to trigger CLOSE\n    _close_velocity_threshold:  # The slope required to trigger CLOSE\n    _sell_slope_threshold:      # The slope required to trigger SELL\n    _sell_velocity_threshold:   # The velocity required to trigger SELL\n    \"\"\"\n    _buy_slope_threshold \u003d 0.1\n    _buy_velocity_threshold \u003d 0.1\n    _close_slope_threshold \u003d 0.1\n    _close_velocity_threshold \u003d 0.1\n    _sell_slope_threshold \u003d -0.1\n    _sell_velocity_threshold \u003d -0.1\n    \n    def __init__(self, leverage:int, margin:float, pip_cost:float, buy_slope_threshold:float \u003d 0.1, buy_velocity_threshold:float \u003d 0.1, close_slope_threshold:float \u003d 0.1, close_velocity_threshold:float \u003d 0.1, sell_slope_threshold:float \u003d -0.1, sell_velocity_threshold:float \u003d -0.1, should_log:bool \u003d False, should_plot:bool \u003d False, axs \u003d None):\n        \"\"\"\n        Initialize Strategy with given parameters\n        :type leverage: float\n        :type margin: float\n        :type buy_slope_threshold: float\n        :type buy_velocity_threshold: float \n        :type close_slope_threshold: float\n        :type close_velocity_threshold: float \n        :type sell_slope_threshold: float\n        :type sell_velocity_threshold: float  \n        :type should_log: bool\n        :type should_plot: bool\n        :param leverage: The leverage\n        :param margin: The margin interest rate\n        :param buy_slope_threshold: # The slope required to trigger BUY\n        :param buy_velocity_threshold: # The velocity required to trigger BUY \n        :param close_slope_threshold: # The slope required to trigger CLOSE\n        :param close_velocity_threshold: # The slope required to trigger CLOSE \n        :param sell_slope_threshold: # The slope required to trigger SELL\n        :param sell_velocity_threshold: # The velocity required to trigger SELL \n        :param should_log Whether or not log the strategy\n        :param should_plot Whether or not plot the strategy\n        \"\"\"\n        self._leverage \u003d leverage\n        self._margin_interest \u003d margin\n        self._pip_cost \u003d pip_cost\n        self._buy_slope_threshold \u003d buy_slope_threshold\n        self._buy_velocity_threshold \u003d buy_velocity_threshold\n        self._close_slope_threshold \u003d close_slope_threshold\n        self._close_velocity_threshold \u003d close_velocity_threshold\n        self._sell_slope_threshold \u003d sell_slope_threshold\n        self._sell_velocity_threshold \u003d sell_velocity_threshold\n        self._should_log \u003d should_log\n        self._should_plot \u003d should_plot\n        self._axs \u003d axs\n        \n    def trade(self, data_df:pd.DataFrame, symbol:str, lots:float, stop_loss:float\u003d0.0):\n        \"\"\"\n        Perform Analysis, Decide on Trade, and Execute Trade\n        :type data_df: pd.DataFrame\n        :type symbol: str\n        :type lots: float\n        :type stop_loss: float\n        :param data_df: The DataFrame for data\n        :param symbol: The symbol\n        :param lots: The standard lots\n        :param stop_loss: number of times of standard deviation\n        :return: \n        \"\"\"\n        self._symbol \u003d symbol\n        self._lots \u003d lots\n            \n        \"\"\"\n        Firstly, add `slope` and `velocity` columns to DataFrame\n        \"\"\"\n        extra_df \u003d data_df.copy()\n        extra_df[\u0027profit\u0027] \u003d 0\n        extra_df[\u0027slope\u0027] \u003d pd.Series(np.gradient(extra_df.pre_close), extra_df.index, name\u003d\u0027slope\u0027) # Find the slope of yesterday, aka. pre_close\n        extra_df[\u0027velocity\u0027] \u003d pd.Series(np.gradient(extra_df.slope), extra_df.index, name\u003d\u0027velocity\u0027)\n        \n        \"\"\"\n        Secondly, Closing\n        \"\"\"\n        extra_df[\u0027position\u0027] \u003d \"CLOSE\"\n        \n        \"\"\"\n        Thirdly, Buying\n        \"\"\"\n        long_mask \u003d (extra_df.slope \u003e self.buy_slope_threshold) \u0026 (extra_df.velocity \u003e self.buy_velocity_threshold)\n        long_close \u003d (extra_df.slope \u003c -self.close_slope_threshold) \u0026 (extra_df.velocity \u003c -self.close_velocity_threshold)\n        extra_df.loc[long_mask, \u0027position\u0027] \u003d \"LONG\"\n        extra_df.loc[long_mask, \u0027profit\u0027] \u003d (extra_df.loc[long_mask, \u0027pip_change\u0027] - self._pip_cost * 2) * self._lots * 10 # Assuming 2 transaction pip cost for buying and selling.\n        # if stop_loss !\u003d 0:\n            # TODO: Implement Stop Loss.\n        \n        \"\"\"\n        Finally, Selling\n        \"\"\"\n        short_mask \u003d (extra_df.slope \u003c self.sell_slope_threshold) \u0026 (extra_df.velocity \u003c self.sell_velocity_threshold)\n        short_close \u003d (extra_df.slope \u003e self.close_slope_threshold) \u0026 (extra_df.velocity \u003e self.close_velocity_threshold)\n        extra_df.loc[short_mask, \u0027position\u0027] \u003d \"SHORT\"\n        extra_df.loc[short_mask, \u0027profit\u0027] \u003d -(extra_df.loc[short_mask, \u0027pip_change\u0027] + self._pip_cost * 2) * self._lots * 10 # Assuming 2 transaction pip cost for buying and selling.\n        # if stop_loss !\u003d 0:\n            # TODO: Implement Stop Loss.\n        \n        self._data_df \u003d extra_df\n        self._quotes \u003d []\n        for _, (d, o, c, h, l) in enumerate(\n            zip(\n                self._data_df.timestamp, \n                self._data_df.open,\n                self._data_df.close,\n                self._data_df.high,\n                self._data_df.low\n            )):\n            # Draw Candle Sticks\n            d \u003d datestr2num(d)\n            # Date, Open, CLose, High, Low tuple \n            val \u003d (d, o, c, h, l)\n            # Add val to quotes\n            self._quotes.append(val)\n        return extra_df\n\n    def plot_trade(self, axs, color_up\u003d\"green\", color_down\u003d\"red\"):\n        \n        print(\"Plotting {} results for trade days\".format(len(self._data_df.index)))\n        mpf.candlestick_ochl(axs, \n                             self._quotes,\n                             width\u003d0.6,\n                             colorup\u003dcolor_up,\n                             colordown\u003dcolor_down,\n                             alpha\u003d0.75)\n        \n        # Background Color\n        # axs.fill_between(self._data_df.timestamp,\n        #                     min(self._data_df.close),\n        #                     self._data_df.close,\n        #                     color\u003d\u0027yellow\u0027,\n        #                     alpha\u003d.05)\n        \n        # Plot Profits\n        ax2 \u003d axs.twinx()\n        ax2.set_ylabel(\u0027Profit %\u0027, \n                       color\u003d\"blue\")\n        ax2.plot(datestr2num(self._data_df.timestamp),\n                 self._data_df.profit.cumsum()/(self._lots * 100000 / self._leverage),\n                 color\u003d\"blue\",\n                 linewidth\u003d2)\n        ax2.tick_params(axis\u003d\u0027y\u0027,\n                        labelcolor\u003d\"blue\")\n        \n        axs.autoscale_view()\n        axs.xaxis_date()\n        \n        # Plot Trade Shadows\n        current_index \u003d -1\n        current_position \u003d \"\"\n        profit_array \u003d []\n        for i, tradeday_row in self._data_df.iterrows():\n            if current_position \u003d\u003d tradeday_row.position: \n                if self._should_log:\n                    # Today is same as yesterday, do nothing\n                    # print(\"HOLD {}\".format(tradeday_row.timestamp))\n                    pass\n            else:\n                if current_position \u003d\u003d \"\":\n                    # First Day Ever, set initial value\n                    current_index \u003d i\n                    current_position \u003d tradeday_row.position\n                else:\n                    # Plot The Trade\n                    if current_position \u003d\u003d \"CLOSE\":\n                        # If last position was \"CLOSED\", do nothing\n                        if self._should_log:\n                            print(\"CLOSE {}\".format(tradeday_row.timestamp))\n                    elif current_position \u003d\u003d \"LONG\":\n                        profit_array.append(tradeday_row.profit)\n                        if self._should_log:\n                            print(\"LONG {} - {}\".format(self._data_df.timestamp[current_index], tradeday_row.timestamp))\n                            # TODO: Fix Index issue. print(\"LONG {} - {}, Profit: ${} ({} pips)\".format(self._data_df.timestamp[current_index], tradeday_row.timestamp, round(self._data_df.iloc[[current_index,i]].profit.sum(), 2), round(self._data_df.iloc[[current_index,i]].pip_change.sum(), 1)))\n                        axs.fill_between(self._data_df.loc[current_index:i, \u0027timestamp\u0027],\n                                        min(self._data_df.close),\n                                        self._data_df.loc[current_index:i, \u0027close\u0027],\n                                        color\u003dcolor_up,\n                                        alpha\u003d.3)\n                        # axs.annotate(\u0027{} pips\u0027.format(round(tradeday_row.pip_change, 1)),\n                        #              xy\u003d(i, tradeday_row.close+ 0.5),\n                        #              arrowprops\u003ddict(facecolor\u003d\u0027red\u0027),\n                        #              horizontalalignment\u003d\u0027left\u0027,\n                        #              verticalalignment\u003d\u0027top\u0027,)\n                    elif current_position \u003d\u003d \"SHORT\":\n                        profit_array.append(tradeday_row.profit)\n                        if self._should_log:\n                            print(\"SHORT {} - {}\".format(self._data_df.timestamp[current_index], tradeday_row.timestamp))\n                            # TODO: Fix Index issue. print(\"SHORT {} - {}, Profit: ${} ({} pips)\".format(self._data_df.timestamp[current_index], tradeday_row.timestamp, round(self._data_df.iloc[[current_index,i].profit].sum(), 2), -round(self._data_df.iloc[[current_index,i]].pip_change.sum(), 1)))\n                        axs.fill_between(self._data_df.loc[current_index:i, \u0027timestamp\u0027],\n                                        min(self._data_df.close),\n                                        self._data_df.loc[current_index:i, \u0027close\u0027],\n                                        color\u003dcolor_down,\n                                        alpha\u003d.3)\n                        # plt.annotate(\u0027{} pips\u0027.format(round(tradeday_row.pip_change, 1)),\n                        #              xy\u003d(i, tradeday_row.close + 0.5),\n                        #              arrowprops\u003ddict(facecolor\u003d\u0027red\u0027),\n                        #              horizontalalignment\u003d\u0027left\u0027,\n                        \n                        #              verticalalignment\u003d\u0027top\u0027)\n                        \n                    # Reset Current Date \u0026 Position\n                    current_index \u003d i\n                    current_position \u003d tradeday_row.position\n        \n        # Plot Labels\n        axs.text(0.05, \n                 0.15, \n                 \u0027Avg. Daily Profit: ${}\u0027.format(round(np.array(profit_array).mean(), 2)),\n                 verticalalignment\u003d\u0027top\u0027, \n                 horizontalalignment\u003d\u0027left\u0027,\n                 transform\u003daxs.transAxes,\n                 color\u003d\u0027green\u0027, \n                 fontsize\u003d14)\n        axs.text(0.05, \n                 0.1, \n                 \u0027Std. Daily Profit: ${}\u0027.format(round(np.array(profit_array).std(), 4)),\n                 verticalalignment\u003d\u0027top\u0027, \n                 horizontalalignment\u003d\u0027left\u0027,\n                 transform\u003daxs.transAxes,\n                 color\u003d\u0027red\u0027, \n                 fontsize\u003d14)\n        axs.text(0.95, \n                 0.15, \u0027Red represents SHORT position.\u0027,\n                 verticalalignment\u003d\u0027top\u0027, \n                 horizontalalignment\u003d\u0027right\u0027,\n                 transform\u003daxs.transAxes,\n                 color\u003d\u0027red\u0027, \n                 fontsize\u003d14, \n                 bbox\u003d{\n                     \u0027facecolor\u0027: \u0027red\u0027,\n                     \u0027alpha\u0027: 0.3,\n                     \u0027pad\u0027: 10\n                 })\n        axs.text(0.95, \n                 0.1, \u0027Green represents LONG position.\u0027,\n                 verticalalignment\u003d\u0027top\u0027, \n                 horizontalalignment\u003d\u0027right\u0027,\n                 transform\u003daxs.transAxes,\n                 color\u003d\u0027green\u0027, \n                 fontsize\u003d14, \n                 bbox\u003d{\n                     \u0027facecolor\u0027: \u0027green\u0027,\n                     \u0027alpha\u0027: 0.3,\n                     \u0027pad\u0027: 10\n                 })\n        # axs.update_datalim(corners)\n        axs.set_xlabel(\u0027Trade Dates\u0027)\n        axs.set_ylabel(\u0027Exchange Rates\u0027)\n        axs.set_title(\u0027Forex {}. {}% Return, trading {} lots with {} leverage for {} Days\u0027.format(self._symbol, round(self.trade_profit / (self._lots * 100000 / self._leverage), 2), self._lots, self._leverage, len(self._data_df.index)))\n        axs.set_xticks(self._data_df.timestamp.values)\n        axs.set_yticks(np.arange(min(self._data_df.close), max(self._data_df.close), 0.0010))\n        axs.set_xticklabels(self._data_df.timestamp.values, rotation\u003d90)\n        axs.grid(True, alpha\u003d0.1)\n        \n    @property\n    def trade_profit(self):\n        return np.sum(self._data_df.profit)\n    \n    @property\n    def buy_slope_threshold(self):\n        return self._buy_slope_threshold\n    \n    @property\n    def buy_velocity_threshold(self):\n        return self._buy_velocity_threshold\n    \n    @property\n    def close_slope_threshold(self):\n        return self._close_slope_threshold\n    \n    @property\n    def close_velocity_threshold(self):\n        return self._close_velocity_threshold\n    \n    @property\n    def sell_slope_threshold(self):\n        return self._sell_slope_threshold\n    \n    @property\n    def sell_velocity_threshold(self):\n        return self._sell_velocity_threshold\n    \n    @property\n    def data_df(self):\n        return self._data_df"
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "outputs": [],
      "source": "\n",
      "metadata": {
        "pycharm": {
          "metadata": false,
          "name": "#%%\n"
        }
      }
    }
  ],
  "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": "3.7.3"
    },
    "stem_cell": {
      "cell_type": "raw",
      "metadata": {
        "pycharm": {
          "metadata": false
        }
      },
      "source": ""
    }
  },
  "nbformat": 4,
  "nbformat_minor": 1
}