Overall Statistics
Total Trades
52
Average Win
0.00%
Average Loss
0.00%
Compounding Annual Return
0.096%
Drawdown
0.100%
Expectancy
3.353
Net Profit
0.090%
Sharpe Ratio
0.87
Probabilistic Sharpe Ratio
44.551%
Loss Rate
24%
Win Rate
76%
Profit-Loss Ratio
4.71
Alpha
0.001
Beta
0
Annual Standard Deviation
0.001
Annual Variance
0
Information Ratio
-2.148
Tracking Error
0.101
Treynor Ratio
-4.174
Total Fees
$52.00
# QUANTCONNECT.COM - Democratizing Finance, Empowering Individuals.
# Lean Algorithmic Trading Engine v2.0. Copyright 2014 QuantConnect Corporation.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

from clr import AddReference

AddReference("System")
AddReference("QuantConnect.Algorithm")
AddReference("QuantConnect.Common")

from System import *
from QuantConnect import *
from QuantConnect.Algorithm import *

# <summary>
# Regression algorithm to test the behaviour of ARMA versus AR models at the same order of differencing.
# In particular, an ARIMA(1,1,1) and ARIMA(1,1,0) are instantiated while orders are placed if their difference
# is sufficiently large (which would be due to the inclusion of the MA(1) term).
# </summary>
class AutoRegressiveIntegratedMovingAverageRegressionAlgorithm(QCAlgorithm):
    def Initialize(self):
        '''Initialise the data and resolution required, as well as the cash and start-end dates for your algorithm. All algorithms must initialized.'''
        self.SetStartDate(2013, 1, 7)
        self.SetEndDate(2013, 12, 11)
        self.EnableAutomaticIndicatorWarmUp = True
        self.AddEquity("SPY", Resolution.Daily)
        self.arima = self.ARIMA("SPY", 1, 1, 1, 50)
        self.ar = self.ARIMA("SPY", 1, 1, 0, 50)

    def OnData(self, data):
        '''OnData event is the primary entry point for your algorithm. Each new data point will be pumped in here.

        Arguments:
            data: Slice object keyed by symbol containing the stock data
        '''
        if self.arima.IsReady:
            if abs(self.arima.Current.Value - self.ar.Current.Value) > 1:
                if self.arima.Current.Value > self.last:
                    self.MarketOrder("SPY", 1)
                else:
                    self.MarketOrder("SPY", -1)
            self.last = self.arima.Current.Value