Modern portfolio
theory
I N T R O D U C T I O N TO P O R T F O L I O A N A LY S I S I N P Y T H O N
Charlotte Werger
Data Scientist
Creating optimal portfolios
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
What is Portfolio Optimization?
Meet Harry Markowitz
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
The optimization problem: nding optimal
weights
In words:
Minimize the portfolio variance, subject to:
The expected mean return is at least some
target return
The weights sum up to 100%
At least some weights are positive
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Varying target returns leads to the Ef cient
Frontier
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
PyPortfolioOpt for portfolio optimization
from pypfopt.efficient_frontier import EfficientFrontier
from pypfopt import risk_models
from pypfopt import expected_returns
df=pd.read_csv('portfolio.csv')
df.head(2)
XOM RRC BBY MA PFE
date
2010-01-04 54.068794 51.300568 32.524055 22.062426 13.940202
2010-01-05 54.279907 51.993038 33.349487 21.997149 13.741367
# Calculate expected annualized returns and sample covariance
mu = expected_returns.mean_historical_return(df)
Sigma = risk_models.sample_cov(df)
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Get the Ef cient Frontier and portfolio weights
# Calculate expected annualized returns and risk
mu = expected_returns.mean_historical_return(df)
Sigma = risk_models.sample_cov(df)
# Obtain the EfficientFrontier
ef = EfficientFrontier(mu, Sigma)
# Select a chosen optimal portfolio
ef.max_sharpe()
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Different optimizations
# Select the maximum Sharpe portfolio
ef.max_sharpe()
# Select an optimal return for a target risk
ef.efficient_risk(2.3)
# Select a minimal risk for a target return
ef.efficient_return(1.5)
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Calculate portfolio risk and performance
# Obtain the performance numbers
ef.portfolio_performance(verbose=True, risk_free_rate = 0.01)
Expected annual return: 21.3%
Annual volatility: 19.5%
Sharpe Ratio: 0.98
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Let's optimize a
portfolio!
I N T R O D U C T I O N TO P O R T F O L I O A N A LY S I S I N P Y T H O N
Maximum Sharpe vs.
minimum volatility
I N T R O D U C T I O N TO P O R T F O L I O A N A LY S I S I N P Y T H O N
Charlotte Werger
Data Scientist
Remember the Ef cient Frontier?
Ef cient frontier: all portfolios with an
optimal risk and return trade-off
Maximum Sharpe portfolio: the highest
Sharpe ratio on the EF
Minimum volatility portfolio: the lowest
level of risk on the EF
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Adjusting PyPortfolioOpt optimization
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Maximum Sharpe portfolio
Maximum Sharpe portfolio: the highest Sharpe ratio on the EF
from pypfopt.efficient_frontier import EfficientFrontier
# Calculate the Efficient Frontier with mu and S
ef = EfficientFrontier(mu, Sigma)
raw_weights = ef.max_sharpe()
# Get interpretable weights
cleaned_weights = ef.clean_weights()
{'GOOG': 0.01269,'AAPL': 0.09202,'FB': 0.19856,
'BABA': 0.09642,'AMZN': 0.07158,'GE': 0.02456,...}
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Maximum Sharpe portfolio
# Get performance numbers
ef.portfolio_performance(verbose=True)
Expected annual return: 33.0%
Annual volatility: 21.7%
Sharpe Ratio: 1.43
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Minimum Volatility Portfolio
Minimum volatility portfolio: the lowest level of risk on the EF
# Calculate the Efficient Frontier with mu and S
ef = EfficientFrontier(mu, Sigma)
raw_weights = ef.min_volatility()
# Get interpretable weights and performance numbers
cleaned_weights = ef.clean_weights()
{'GOOG': 0.05664, 'AAPL': 0.087, 'FB': 0.1591,
'BABA': 0.09784, 'AMZN': 0.06986, 'GE': 0.0123,...}
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Minimum Volatility Portfolio
ef.portfolio_performance(verbose=True)
Expected annual return: 17.4%
Annual volatility: 13.2%
Sharpe Ratio: 1.28
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Let's have another look at the Ef cient Frontier
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Maximum Sharpe versus Minimum Volatility
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Let's practice!
I N T R O D U C T I O N TO P O R T F O L I O A N A LY S I S I N P Y T H O N
Alternative portfolio
optimization
I N T R O D U C T I O N TO P O R T F O L I O A N A LY S I S I N P Y T H O N
Charlotte Werger
Data Scientist
Expected risk and return based on historic data
Mean historic returns, or the historic
portfolio variance are not perfect estimates of
mu and Sigma
Weights from portfolio optimization
therefore not guaranteed to work well on
future data
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Historic data
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Exponentially weighted returns
Need better measures for risk and return
Exponentially weighted risk and return
assigns more importance to the most recent
data
Exponential moving average in the graph:
most weight on t-1 observation
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Exponentially weighted covariance
The exponential covariance matrix: gives
more weight to recent data
In the graph: exponential weighted
volatility in black, follows real volatility better
than standard volatility in blue
1 Source: http://systematicinvestor.github.io/Exponentially 2 Weighted 3 Volatility 4 RCPP
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Exponentially weighted returns
from pypfopt import expected_returns
# Exponentially weighted moving average
mu_ema = expected_returns.ema_historical_return(df,
span=252, frequency=252)
print(mu_ema)
symbol
XOM 0.103030
BBY 0.394629
PFE 0.186058
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Exponentially weighted covariance
from pypfopt import risk_models
# Exponentially weighted covariance
Sigma_ew = risk_models.exp_cov(df, span=180, frequency=252)
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Using downside risk in the optimization
Remember the Sortino ratio: it uses the variance of negative returns only
PyPortfolioOpt allows you to use semicovariance in the optimization, this is a measure for
downside risk:
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Semicovariance in PyPortfolioOpt
Sigma_semi = risk_models.semicovariance(df,
benchmark=0, frequency=252)
print(Sigma_semi)
XOM BBY MA PFE
XOM 0.018939 0.008505 0.006568 0.004058
BBY 0.008505 0.016797 0.009133 0.004404
MA 0.006568 0.009133 0.018711 0.005373
PFE 0.004058 0.004404 0.005373 0.008349
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Let's practice!
I N T R O D U C T I O N TO P O R T F O L I O A N A LY S I S I N P Y T H O N
Recap
I N T R O D U C T I O N TO P O R T F O L I O A N A LY S I S I N P Y T H O N
Charlotte Werger
Data Scientist
Chapter 1: Calculating risk and return
A portfolio as a collection of weight and assets
Diversi cation
Mean returns versus cumulative returns
Variance, standard deviation, correlations and the covariance matrix
Calculating portfolio variance
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Chapter 2: Diving deep into risk measures
Annualizing returns and risk to compare over different periods
Sharpe ratio as a measured of risk adjusted returns
Skewness and Kurtosis: looking beyond mean and variance of a distribution
Maximum draw-down, downside risk and the Sortino ratio
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Chapter 3: Breaking down performance
Compare to benchmark with active weights and active returns
Investment factors: explain returns and sources of risk
Fama French 3 factor model to breakdown performance into explainable factors and alpha
Pyfolio as a portfolio analysis tool
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Chapter 4: Finding the optimal portfolio
Markowitz' portfolio optimization: ef cient frontier, maximum Sharpe and minimum volatility
portfolios
Exponentially weighted risk and return, semicovariance
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
Continued learning
Datacamp course on Portfolio Risk Management in Python
Quantopian's lecture series: https://www.quantopian.com/lectures
Learning by doing: Pyfolio and PyPortfolioOpt
INTRODUCTION TO PORTFOLIO ANALYSIS IN PYTHON
End of this course
I N T R O D U C T I O N TO P O R T F O L I O A N A LY S I S I N P Y T H O N
