| Overall Statistics |
|
Total Trades 2282 Average Win 0.25% Average Loss -0.16% Compounding Annual Return 6.597% Drawdown 13.100% Expectancy 0.331 Net Profit 91.671% Sharpe Ratio 0.896 Probabilistic Sharpe Ratio 35.493% Loss Rate 47% Win Rate 53% Profit-Loss Ratio 1.53 Alpha 0.057 Beta -0.023 Annual Standard Deviation 0.061 Annual Variance 0.004 Information Ratio -0.342 Tracking Error 0.151 Treynor Ratio -2.383 Total Fees $11102587.01 |
from Execution.ImmediateExecutionModel import ImmediateExecutionModel
from datetime import timedelta
from QuantConnect.Data.UniverseSelection import *
from Selection.FundamentalUniverseSelectionModel import FundamentalUniverseSelectionModel
from pypfopt.hierarchical_risk_parity import HRPOpt
import pandas as pd
class smartBeta(QCAlgorithm):
def Initialize(self):
self.SetStartDate(2010, 1, 1) # Set Start Date
self.SetCash(3e8) # Set Strategy Cash
self.SetBrokerageModel(BrokerageName.AlphaStreams)
self.spy = self.AddEquity("SPY", Resolution.Daily)
self.SetBenchmark(self.spy.Symbol)
self.lastMonth_ = -1
self.UniverseSettings.Resolution = Resolution.Daily
self.SetUniverseSelection(SP500SectorsETFUniverse())
self.Schedule.On(self.DateRules.EveryDay("SPY"), self.TimeRules.AfterMarketOpen('SPY'), self.Rebalance)
self.SetExecution(ImmediateExecutionModel())
self.SetRiskManagement(TrailingStopRiskManagementModel(0.02))
def Rebalance(self):
"""
"""
if self.lastMonth_ == self.Time.month:
return
else:
self.lastMonth_ = self.Time.month
# 1) Create a dictionary keyed by the symbols in the insights with an pandas.Series as value to create a data frame
df = self.History(self.Securities.Keys, 252, Resolution.Daily)# Get train/test data
returns = df.unstack(level=1).close.transpose().pct_change().dropna()
port = self.getHRP(returns)
weights = port.values
symbols = port.index.values
selected = zip(symbols, weights)
for symbol, weight in selected:
self.SetHoldings(symbol, weight)
def getHRP(self, x):
# Construct a hierarchical portfolio based on the synthetic returns
hrp = HRPOpt(x)
weights = hrp.hrp_portfolio()
return pd.Series(weights)