Black-Litterman Portfolio Rebalancing with Historical Return Signals
Summary
This example shows how to assemble a portfolio algorithm using a historical-returns alpha model and a Black-Litterman portfolio construction model. An unconstrained mean-variance optimizer is supplied to the construction model, which generates portfolio targets for rebalancing. The framework also specifies universe selection, immediate execution, and a null risk-management model.
The sample selects a small, fixed set of U.S. equities and changes the selected set during a short demonstration period by dropping one symbol after a particular calendar day. It uses minute-resolution universe settings and daily-resolution historical return signals, with a short dated run and stated initial capital. These choices illustrate how the components connect in a framework algorithm; they do not establish that the portfolio method is profitable or robust. The example gives no performance analysis, transaction-cost study, risk controls, or explanation of the Black-Litterman equations and view assumptions, so those require separate investigation before adapting it for research or live trading.
Key ideas
- Historical return signals can be paired with Black-Litterman portfolio construction to generate portfolio rebalances.
- The example passes an unconstrained mean-variance optimizer into the portfolio construction model.
- Universe selection, execution, and risk management are configured as separate framework components.
- The sample uses a small U.S. equity universe and a short demonstration run, so it offers no evidence of general performance.
Tags
Full text
# BlackLittermanPortfolioOptimizationFrameworkAlgorithm
# BlackLittermanPortfolioOptimizationFrameworkAlgorithm
Black-Litterman Optimization algorithm.
Black-Litterman framework algorithm Uses the HistoricalReturnsAlphaModel and the BlackLittermanPortfolioConstructionModel to create an algorithm that rebalances the portfolio according to Black-Litterman portfolio optimization
## Source (Apache-2.0)
```python
# 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 AlgorithmImports import *
from Alphas.HistoricalReturnsAlphaModel import HistoricalReturnsAlphaModel
from Portfolio.BlackLittermanOptimizationPortfolioConstructionModel import *
from Portfolio.UnconstrainedMeanVariancePortfolioOptimizer import UnconstrainedMeanVariancePortfolioOptimizer
from Risk.NullRiskManagementModel import NullRiskManagementModel
### <summary>
### Black-Litterman framework algorithm
### Uses the HistoricalReturnsAlphaModel and the BlackLittermanPortfolioConstructionModel
### to create an algorithm that rebalances the portfolio according to Black-Litterman portfolio optimization
### </summary>
### <meta name="tag" content="using data" />
### <meta name="tag" content="using quantconnect" />
### <meta name="tag" content="trading and orders" />
class BlackLittermanPortfolioOptimizationFrameworkAlgorithm(QCAlgorithm):
'''Black-Litterman Optimization algorithm.'''
def initialize(self):
# Set requested data resolution
self.universe_settings.resolution = Resolution.MINUTE
# Order margin value has to have a minimum of 0.5% of Portfolio value, allows filtering out small trades and reduce fees.
# Commented so regression algorithm is more sensitive
#self.settings.minimum_order_margin_portfolio_percentage = 0.005
self.set_start_date(2013,10,7) #Set Start Date
self.set_end_date(2013,10,11) #Set End Date
self.set_cash(100000) #Set Strategy Cash
self._symbols = [ Symbol.create(x, SecurityType.EQUITY, Market.USA) for x in [ 'AIG', 'BAC', 'IBM', 'SPY' ] ]
optimizer = UnconstrainedMeanVariancePortfolioOptimizer()
# set algorithm framework models
self.set_universe_selection(CoarseFundamentalUniverseSelectionModel(self.coarse_selector))
self.set_alpha(HistoricalReturnsAlphaModel(resolution = Resolution.DAILY))
self.set_portfolio_construction(BlackLittermanOptimizationPortfolioConstructionModel(optimizer = optimizer))
self.set_execution(ImmediateExecutionModel())
self.set_risk_management(NullRiskManagementModel())
def coarse_selector(self, coarse):
# Drops SPY after the 8th
last = 3 if self.time.day > 8 else len(self._symbols)
return self._symbols[0:last]
def on_order_event(self, order_event):
if order_event.status == OrderStatus.FILLED:
self.debug(order_event)
```Shown in full with attribution under the source's licence. Licence: Apache-2.0
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.