Running A Backtest

In this example we will demonstrate how to run a backtest for a simple algo example

Loading the data bundle

[95]:

import os

import warnings
warnings.filterwarnings('ignore')

os.environ['ZIPLINE_ROOT'] = os.path.join(os.getcwd(), '.zipline')

os.listdir(os.environ['ZIPLINE_ROOT'])

os.environ['ZIPLINE_TRADER_CONFIG'] = os.path.join(os.getcwd(), "./zipline-trader.yaml")

with open(os.environ['ZIPLINE_TRADER_CONFIG'], 'r') as f:
    data = f.read()
    print(data[:20])

import zipline
from zipline.data import bundles

bundle_name = 'alpaca_api'
bundle_data = bundles.load(bundle_name)
from zipline.pipeline.loaders import USEquityPricingLoader
from zipline.utils.calendars import get_calendar
from zipline.pipeline.data import USEquityPricing
from zipline.data.data_portal import DataPortal

import pandas as pd

# Set the dataloader
pricing_loader = USEquityPricingLoader.without_fx(bundle_data.equity_daily_bar_reader, bundle_data.adjustment_reader)


# Define the function for the get_loader parameter
def choose_loader(column):
    if column not in USEquityPricing.columns:
        raise Exception('Column not in USEquityPricing')
    return pricing_loader

# Set the trading calendar
trading_calendar = get_calendar('NYSE')

start_date = pd.Timestamp('2019-07-05', tz='utc')
end_date = pd.Timestamp('2020-11-13', tz='utc')

# Create a data portal
data_portal = DataPortal(bundle_data.asset_finder,
                         trading_calendar = trading_calendar,
                         first_trading_day = start_date,
                         equity_daily_reader = bundle_data.equity_daily_bar_reader,
                         adjustment_reader = bundle_data.adjustment_reader)

[96]:

import dateutil.parser
from os.path import join, exists
import pandas as pd
import pandas_datareader.data as yahoo_reader
import yaml

import numpy as np


def get_benchmark(symbol=None, start = None, end = None, other_file_path=None):
    bm = yahoo_reader.DataReader(symbol,
                                 'yahoo',
                                 pd.Timestamp(start),
                                 pd.Timestamp(end))['Close']
    bm.index = bm.index.tz_localize('UTC')
    return bm.pct_change(periods=1).fillna(0)

[97]:


from zipline.api import order_target, record, symbol
import matplotlib.pyplot as plt

def initialize(context):
    context.equity = bundle_data.asset_finder.lookup_symbol("AMZN", end_date)

def handle_data(context, data):
    order_target(context.equity, 100)

def before_trading_start(context, data):
    pass

def analyze(context, perf):
    ax1 = plt.subplot(211)
    perf.portfolio_value.plot(ax=ax1)
    ax2 = plt.subplot(212, sharex=ax1)
    perf.sym.plot(ax=ax2, color='r')
    plt.gcf().set_size_inches(18, 8)
    plt.legend(['Algo', 'Benchmark'])
    plt.ylabel("Returns", color='black', size=25)


[98]:

import pandas as pd
from datetime import datetime
import pytz

from zipline import run_algorithm

start = pd.Timestamp(datetime(2020, 1, 1, tzinfo=pytz.UTC))
end = pd.Timestamp(datetime(2020, 11, 1, tzinfo=pytz.UTC))

r = run_algorithm(start=start,
                  end=end,
                  initialize=initialize,
                  capital_base=100000,
                  handle_data=handle_data,
                  benchmark_returns=get_benchmark(symbol="SPY",
                                                  start=start.date().isoformat(),
                                                  end=end.date().isoformat()),
                  bundle='alpaca_api',
                  broker=None,
                  state_filename="./demo.state",
                  trading_calendar=trading_calendar,
                  before_trading_start=before_trading_start,
#                   analyze=analyze,
                  data_frequency='daily'
                  )
fig, axes = plt.subplots(1, 1, figsize=(16, 5), sharex=True)
r.algorithm_period_return.plot(color='blue')
r.benchmark_period_return.plot(color='red')

plt.legend(['Algo', 'Benchmark'])
plt.ylabel("Returns", color='black', size=20)


[98]:

Text(0, 0.5, 'Returns')
../_images/notebooks_backtest_5_1.png

Using Pyfolio to analyze your results

[99]:

import pyfolio as pf
returns, positions, transactions = pf.utils.extract_rets_pos_txn_from_zipline(r)
benchmark_returns = r.benchmark_period_return

[100]:

import empyrical
print("returns sharp ratio: {}".format(empyrical.sharpe_ratio(returns)))  # how much is it sensative to divergance. the higher the better
print("bencmark sharp ratio: {}".format(empyrical.sharpe_ratio(benchmark_returns)))
print("beta ratio: {}".format(empyrical.beta(returns, benchmark_returns))) # how much correlation between algo to benchmark. we want it to be clsoe to zero
print("alpha ratio: {}".format(empyrical.alpha(returns, benchmark_returns)))



returns sharp ratio: 1.7147226500268884
bencmark sharp ratio: -6.584559370521319
beta ratio: -0.010817530431739008
alpha ratio: 1.829456229577063

[101]:

pf.create_position_tear_sheet(returns, positions=positions, transactions=transactions)
Top 10 long positions of all time max
sid
AMZN 209.44%
Top 10 short positions of all time max
sid
Top 10 positions of all time max
sid
AMZN 209.44%
../_images/notebooks_backtest_10_3.png 
[102]:

pf.create_returns_tear_sheet(returns,
                             positions=positions,
                             transactions=transactions,
                             benchmark_rets=benchmark_returns)
Start date2020-01-02
End date2020-10-30
Total months10
Backtest
Annual return 150.903%
Cumulative returns 116.026%
Annual volatility 66.542%
Sharpe ratio 1.71
Calmar ratio 3.95
Stability 0.87
Max drawdown -38.166%
Omega ratio 1.35
Sortino ratio 2.67
Skew 0.13
Kurtosis 1.98
Tail ratio 1.17
Daily value at risk -7.931%
Gross leverage 1.57
Daily turnover 0.953%
Alpha 1.83
Beta -0.01
Worst drawdown periods Net drawdown in % Peak date Valley date Recovery date Duration
0 38.17 2020-02-19 2020-03-12 2020-04-14 40
1 21.73 2020-09-02 2020-09-18 NaT NaN
2 11.63 2020-04-30 2020-05-01 2020-05-20 15
3 10.39 2020-07-10 2020-07-17 2020-08-05 19
4 7.56 2020-01-07 2020-01-27 2020-01-31 19
../_images/notebooks_backtest_12_2.png 
[103]:


pf.create_full_tear_sheet(returns,
                          positions=positions,
                          transactions=transactions,
                          live_start_date="2020-06-13",
                          round_trips=True,
                          benchmark_rets=benchmark_returns)
Start date2020-01-02
End date2020-10-30
In-sample months5
Out-of-sample months4
All In-sample Out-of-sample
Annual return 150.903% 222.617% 87.764%
Cumulative returns 116.026% 69.084% 27.763%
Annual volatility 66.542% 75.493% 54.733%
Sharpe ratio 1.71 1.93 1.42
Calmar ratio 3.95 5.83 4.04
Stability 0.87 0.63 0.40
Max drawdown -38.166% -38.166% -21.734%
Omega ratio 1.35 1.42 1.26
Sortino ratio 2.67 3.01 2.22
Skew 0.13 0.05 0.23
Kurtosis 1.98 1.85 0.37
Tail ratio 1.17 1.24 1.10
Daily value at risk -7.931% -8.933% -6.587%
Gross leverage 1.57 1.73 1.40
Daily turnover 0.953% 1.787% 0.0%
Alpha 1.83 3.35 1.42
Beta -0.01 0.00 -0.02
Worst drawdown periods Net drawdown in % Peak date Valley date Recovery date Duration
0 38.17 2020-02-19 2020-03-12 2020-04-14 40
1 21.73 2020-09-02 2020-09-18 NaT NaN
2 11.63 2020-04-30 2020-05-01 2020-05-20 15
3 10.39 2020-07-10 2020-07-17 2020-08-05 19
4 7.56 2020-01-07 2020-01-27 2020-01-31 19
Stress Events mean min max
New Normal 0.45% -15.20% 14.70%
Top 10 long positions of all time max
sid
AMZN 209.44%
Top 10 short positions of all time max
sid
Top 10 positions of all time max
sid
AMZN 209.44%
../_images/notebooks_backtest_14_6.png
../_images/notebooks_backtest_14_7.png
../_images/notebooks_backtest_14_8.png
../_images/notebooks_backtest_14_9.png