fadytos.py

"""MC4-P6: Manual Strategy - Theoretical Optimal Strategy

Copyright 2018, Georgia Institute of Technology (Georgia Tech)
Atlanta, Georgia 30332
All Rights Reserved

Template code for CS 4646/7646

Georgia Tech asserts copyright ownership of this template and all derivative
works, including solutions to the projects assigned in this course. Students
and other users of this template code are advised not to share it with others
or to make it available on publicly viewable websites including repositories
such as github and gitlab.  This copyright statement should not be removed
or edited.

We do grant permission to share solutions privately with non-students such
as potential employers. However, sharing with other current or future
students of CS 7646 is prohibited and subject to being investigated as a
GT honor code violation.

-----do not edit anything above this line---

Student Name: Fady Ibrahim
GT User ID: fibrahim30
GT ID: 903358183
"""

import datetime as dt
import pandas as pd
import marketsimcode as msc
from util import get_data
import matplotlib.pyplot as plt


def author():
    return 'fibrahim30'


def testPolicy(symbol="AAPL", sd=dt.datetime(2010, 1, 1), ed=dt.datetime(2011, 12, 31), sv=100000):
    # Get start and end dates from sorted dates range
    dates = pd.date_range(sd, ed)
    # From 'CS 7646 Spring 2016: Q&A on Market Simulator Project' - https://www.youtube.com/watch?v=1ysZptg2Ypk
    # Make prices dataframe using util.py get_data
    prices_df = get_data(symbol, dates, addSPY=False).dropna()
    orders_df = prices_df
    orders_df['Future'] = prices_df.shift(-1)
    orders_df['Order'] = 0
    orders_df['Shares'] = 0
    orders_df['Holdings'] = 0

    orders_df.loc[orders_df['JPM'] < orders_df['Future'], 'Order'] = 'BUY'
    orders_df.loc[orders_df['JPM'] > orders_df['Future'], 'Order'] = 'SELL'
    orders_df.loc[orders_df['JPM'] == orders_df['Future'], 'Order'] = 'NOTHING'
    orders_df.loc[orders_df['Future'].isnull(), 'Order'] = 'NOTHING'

    # Initial conditions
    orders_df.loc[orders_df['Order'] == 'BUY', 'Shares'] = 1000
    orders_df.loc[orders_df['Order'] == 'SELL', 'Shares'] = -1000
    if orders_df.ix[0, 'Order'] == 'BUY' and orders_df.ix[0, 'Holdings'] == 0:
        orders_df.ix[0, 'Shares'] = 1000
        orders_df.ix[0, 'Holdings'] = 1000
    elif orders_df.ix[0, 'Order'] == 'SELL' and orders_df.ix[0, 'Holdings'] == 0:
        orders_df.ix[0, 'Shares'] = -1000
        orders_df.ix[0, 'Holdings'] = -1000


    # populate trades dataframe with the resulting trades from orders data
    for row in range(1, orders_df.shape[0]):
        order = orders_df.ix[row, 'Order']
        date = orders_df.index[row]

        if order == 'SELL' and orders_df.ix[row-1, 'Holdings'] == 0:
            orders_df.ix[date, 'Shares'] = -1000
            orders_df.ix[date, 'Holdings'] = -1000
        elif order == 'SELL' and orders_df.ix[row-1, 'Holdings'] == -1000:
            orders_df.ix[date, 'Shares'] = 0
            orders_df.ix[date, 'Holdings'] = -1000
            orders_df.ix[date, 'Order'] = 'NOTHING'
        elif order == 'SELL' and orders_df.ix[row-1, 'Holdings'] == 1000:
            orders_df.ix[date, 'Shares'] = -2000
            orders_df.ix[date, 'Holdings'] = -1000
        elif order == 'BUY' and orders_df.ix[row-1, 'Holdings'] == 0:
            orders_df.ix[date, 'Shares'] = 1000
            orders_df.ix[date, 'Holdings'] = 1000
        elif order == 'BUY' and orders_df.ix[row-1, 'Holdings'] == -1000:
            orders_df.ix[date, 'Shares'] = 2000
            orders_df.ix[date, 'Holdings'] = 1000
        elif order == 'BUY' and orders_df.ix[row-1, 'Holdings'] == 1000:
            orders_df.ix[date, 'Shares'] = 0
            orders_df.ix[date, 'Holdings'] = 1000
            orders_df.ix[date, 'Order'] = 'NOTHING'
        elif order == 'NOTHING' and orders_df.ix[row-1, 'Holdings'] == -1000:
            orders_df.ix[date, 'Shares'] = 0
            orders_df.ix[date, 'Holdings'] = -1000
        elif order == 'NOTHING' and orders_df.ix[row-1, 'Holdings'] == 1000:
            orders_df.ix[date, 'Shares'] = 0
            orders_df.ix[date, 'Holdings'] = 1000

    orders_df = orders_df[orders_df.Order != 'NOTHING']
    orders_df = orders_df.drop(['JPM'], axis=1)
    orders_df = orders_df.drop(['Future'], axis=1)
    orders_df = orders_df.drop(['Order'], axis=1)
    orders_df = orders_df.drop(['Holdings'], axis=1)

    orders_df.rename(columns={'Shares': 'JPM'}, inplace=True)

    return orders_df

def benchMark():
    dev_start_date = dt.datetime(2008, 1, 1)
    dev_end_date = dt.datetime(2009, 12, 31)
    test_start_date = dt.datetime(2010, 1, 1)
    test_end_date = dt.datetime(2011, 12, 31)
    dev_dates = pd.date_range(dev_start_date, dev_end_date)
    test_dates = pd.date_range(test_start_date, test_end_date)

    symbol = ['JPM']

    benchmark_orders = testPolicy(symbol=symbol, sd=test_start_date, ed=test_end_date)
    benchmark_orders = benchmark_orders * 0
    benchmark_orders.ix[0, 'JPM'] = 1000
    prices_df = get_data(symbol, test_dates, addSPY=False).dropna()
    benchmark_portvals = msc.compute_portvals(benchmark_orders, commission=0., impact=0.)

    print "Benchmark Portfolio Stats: "
    msc.portfolio_stats(benchmark_portvals)

    return benchmark_portvals

# Edited from util.py plot_data in order to save plots.
def save_plot(df, title="Stock prices", xlabel="Date", ylabel="Price"):
    """Plot stock prices with a custom title and meaningful axis labels."""
    ax = df.plot(title=title, fontsize=12, color=['green', 'red'])
    ax.set_xlabel(xlabel)
    ax.set_ylabel(ylabel)
    plt.savefig(title)
    plt.close()


if __name__ == "__main__":
    """"" BENCHMARK """
    benchmark_portvals = benchMark()
    # From Project Wiki - Part 2: Theoretically Optimal Strategy
    # http://quantsoftware.gatech.edu/Summer_2019_Project_6:_Manual_Strategy
    # Benchmark normalized to 1.0 at the start
    normalized_benchmark = benchmark_portvals / benchmark_portvals.ix[0, :]
    normalized_benchmark = normalized_benchmark.to_frame("Bench")

    """" THEORETICALLY OPTIMAL """
    trades_df = testPolicy(symbol=['JPM'])
    optimal_portvals = msc.compute_portvals(trades_df, commission=0., impact=0.)
    # From Project Wiki - Part 2: Theoretically Optimal Strategy
    # http://quantsoftware.gatech.edu/Summer_2019_Project_6:_Manual_Strategy
    # Value of the theoretically optimal portfolio (normalized to 1.0 at the start)
    normalized_optimal = optimal_portvals / optimal_portvals.ix[0, :]
    normalized_optimal = normalized_optimal.to_frame("Optimal")
    print " "
    print "Theoretically Optimal Strategy Portfolio Stats: "
    msc.portfolio_stats(optimal_portvals)

    """ Plot - Benchmark (Green Line), Optimal (Red Line) """
    temp_df = normalized_benchmark.join(normalized_optimal)
    save_plot(temp_df, title='TOS vs Benchmark', xlabel='Date', ylabel='Normalized Price')