SlideShare a Scribd company logo
Mihály Ormos and Dávid Zibriczky
Department of Finance
Budapest University of Technology and Economics
5th international ECEE series conference "Economic Challenges in Enlarged Europe„
Tallinn, Estonia, June 16-18, 2013
 Risk vs. expected return (risk premium)
 Standard deviation (Markowitz, 1952)
 Assumes normal distribution
 CAPM beta (Sharpe, 1964)
 Requires market portfolio
 Assumes linearity
 Measures systematic risk only
 Problems:
 Assumptions
 Low explanatory power
 Alternative single risk measure?
2
mathematically-defined quantity that is generally used for characterizing the probability
of outcomes in a system through a process
Main applications:
 Thermodynamics: Clausius (1867) – distribution of inner energy
 Statistical Mechanics: Boltzmann (1872) – molecular disorder
 Information Theory: Shannon (1948) – message compression
3
 Generalized discrete formula: 𝐻 𝛼 𝑋 =
1
1−𝛼
log 𝑖 𝑝𝑖
𝛼
 𝑝𝑖: probability of discrete outcome Xi
 Special cases
 𝛼 = 1: Shannon entropy (L'Hôpital's rule)
 𝛼 = 2: Rényi entropy
 Problem: Value of return is continuous, discrete formula cannot be applied
 Continuous formula: 𝐻 𝛼 𝑋 =
1
1−𝛼
ln 𝑓 𝑥 𝛼
𝑑𝑥 (differential entropy)
 𝑓 𝑥 : probability function
 𝑓 𝑥 = ?
4
 Density estimation: 𝑓𝑛 𝑥 ~ 𝑓 𝑥
 Methods
 Histogram (fixed bin width)
 Kernel density estimation (sum of core weights)
 Sample spacing (fixed number of elements in one bin)
Histogram Kernel Sample spacing
5
 Original formula: 𝐻 𝛼 𝑋 =
1
1−𝛼
ln 𝑓 𝑥 𝛼
𝑑𝑥
 Entropy estimation in single formula using histogram*:
 Shannon entropy: 𝐻1,𝑛 𝑋 = −
1
𝑛 𝑗 𝑣𝑗 ln
𝑣 𝑗
𝑛ℎ
 Rényi entropy: 𝐻2,𝑛 𝑋 = −ln 𝑗 ℎ
𝑣 𝑗
𝑛ℎ
2
 Entropy risk measure**: 𝜿 𝑯 𝜶
𝑺𝒊 = 𝒆
𝑯 𝜶,𝒏 𝑹 𝒊−𝑹 𝑭
* 𝑣𝑗: number of elements falling into the jth bin, h: bin size
** 𝜿: risk measure, 𝑺𝒊: Security i, 𝑹𝒊 − 𝑹 𝑭: Risk premium
6
 Standard deviation (Markovitz):
𝜿 𝝈 𝑺𝒊 = 𝝈 𝑹𝒊 − 𝑹 𝑭
 Beta (CAPM):
𝜿 𝜷 𝑺𝒊 =
𝐜𝐨𝐯 𝑹 𝒊−𝑹 𝑭,𝑹 𝑴−𝑹 𝑭
𝝈 𝟐 𝑹 𝑴−𝑹 𝑭
 Shannon entropy:
𝜿 𝑯 𝟏
𝑺𝒊 = 𝒆
𝑯 𝟏 𝑹𝒊−𝑹 𝑭
 Rényi entropy:
𝜿 𝑯 𝟐
𝑺𝒊 = 𝒆
𝑯 𝟐 𝑹𝒊−𝑹 𝑭
7
 Source: The Center for Research in Security Prices (CRSP)
 Series: Daily return
 Market return (value weighted)
 Risk free rate (1-month T-bill)
 150, randomly selected securities from the components of S&P500 index
 Period: 1985-2011 (27 years or 6810 days)
8
0
1
2
3
4
5
6
7
8
1 10 100
Averagerisk
Number of securities in portfolio
Shannon Rényi StDev
0%
10%
20%
30%
40%
50%
1 10 100
Averageriskreduction
Number of securities in portfolio
Shannon Rényi StDev
9
-0,02
0
0,02
0,04
0,06
0,08
0,1
0 2,5 5 7,5 10 12,5 15
E(rp-rF)
Risk (H1)
Shannon entropy
n=1
n=2
n=5
n=10
10
-0,02
0
0,02
0,04
0,06
0,08
0,1
0 2,5 5 7,5 10 12,5 15
E(rp-rF)
Risk (H1)
Shannon entropy
n=1
n=2
n=5
n=10
11
 Evaluation method
 Long term (1985-2011), for 150 random securities
 Explanatory variable (X): Risk
 Target variable (Y): Expected risk premium
 Linear regression (X,Y)
 Explanatory power: Goodness of fit of regression line (R2)
 Result
 Higher explanatory power
Risk measure R2 long
Standard deviation 7.83%
Beta 6.17%
Shannon entropy 12.98%
Rényi entropy 15.71%
12
-0,02
0
0,02
0,04
0,06
0,08
0,1
0,12
0,14
0 1 2 3 4 5
E(ri-rF)
Risk (std)
Standard deviation
E(ri-rF) = 0.0170 + 0.0085*std
R² = 7.83%
-0,02
0
0,02
0,04
0,06
0,08
0,1
0,12
0,14
0 0,5 1 1,5 2
E(ri-rF)
Risk (beta)
Beta
E(ri-rF) = 0.0209 + 0.0151*beta
R² = 6.17%
-0,02
0
0,02
0,04
0,06
0,08
0,1
0,12
0,14
0 2,5 5 7,5 10 12,5 15
E(ri-rF)
Risk (H1)
Shannon entropy
E(ri-rF) = 0.0091 + 0.0034*H1
R² = 12.98%
-0,02
0
0,02
0,04
0,06
0,08
0,1
0,12
0,14
0 2,5 5 7,5 10 12,5
E(ri-rF)
Risk (H2)
Rényi entropy
E(ri-rF) = 0.0059 + 0.0049*H2
R² = 15.71%
13
0
0,05
0,1
0,15
0 10 20 30 40 50 60 70 80 90 100
Explanatorypower(R-squared)
Number of securities in portfolio
StDev Beta Shannon Rényi
14
-0,3
-0,2
-0,1
0
0,1
0,2
0 5 10 15 20
E(ri-rF)
Risk (H1)
Shannon entropy
bear market
E(ri-rF) = 0.0818 - 0.0150*H1
R² = 0.3961
-0,05
0
0,05
0,1
0,15
0,2
0 5 10 15
E(ri-rF)
Risk (H1)
Shannon entropy
bull market
E(ri-rF) = -0.0116 + 0.0103*H1
R² = 0.4345
Risk measure R2 long R2 bull R2 bear
Standard deviation 7.83% 33.9% 36.7%
Beta 6.17% 36.7% 43.7%
Shannon entropy 12.98% 43.5% 39.6%
Rényi entropy 15.71% 42.4% 38.6%
15
 Evaluation method
 Estimating risk based on a 5-year period (short term)
 Predicting average risk premium for the next 5-year period
 Applying this on several periods
 Predicting power: Average goodness of fit (R2) based on tested periods
 Reliability: Standard deviation of R2 values (lower is better)
 Result
 Better explanatory power for the same first 5 years (R2 short)
 Better predicting power for the next 5 years (R2 pred)
 Higher reliability (σ)
Risk measure R2 long R2 short R2 pred σ short σ pred
Standard deviation 7.83% 7.94% 9.70% 0.73 0.63
Beta 6.17% 13.31% 6.45% 0.95 0.99
Shannon entropy 12.98% 13.38% 10.15% 0.67 0.62
Rényi entropy 15.71% 12.82% 9.34% 0.62 0.60
16
 No assumptions concerning the returns
 Entropy estimation doesn’t require an undefined market portfolio
 Characterizes specific risk and captures diversification effect
 More efficient and reliable risk estimate compared to the standard methods
 If the trend is identified entropy based equilibrium model behaves similarly to the
standard models
Thank you for paying attention!
17

More Related Content

Similar to Entropy based asset pricing

1CHAPTER 6Risk, Return, and the Capital Asset Pricing Model.docx
1CHAPTER 6Risk, Return, and the Capital Asset Pricing Model.docx1CHAPTER 6Risk, Return, and the Capital Asset Pricing Model.docx
1CHAPTER 6Risk, Return, and the Capital Asset Pricing Model.docx
hyacinthshackley2629
 
financial management chapter 4 Risk and Return
financial management chapter 4 Risk and Returnfinancial management chapter 4 Risk and Return
financial management chapter 4 Risk and Return
sufyanraza1
 
Fm11 ch 04 risk and return the basics
Fm11 ch 04 risk and return the basicsFm11 ch 04 risk and return the basics
Fm11 ch 04 risk and return the basics
Nhu Tuyet Tran
 

Similar to Entropy based asset pricing (20)

1CHAPTER 6Risk, Return, and the Capital Asset Pricing Model.docx
1CHAPTER 6Risk, Return, and the Capital Asset Pricing Model.docx1CHAPTER 6Risk, Return, and the Capital Asset Pricing Model.docx
1CHAPTER 6Risk, Return, and the Capital Asset Pricing Model.docx
 
Spillover dynamics for sistemic risk measurement using spatial financial time...
Spillover dynamics for sistemic risk measurement using spatial financial time...Spillover dynamics for sistemic risk measurement using spatial financial time...
Spillover dynamics for sistemic risk measurement using spatial financial time...
 
HLEG thematic workshop on Measuring Inequalities of Income and Wealth, Alfred...
HLEG thematic workshop on Measuring Inequalities of Income and Wealth, Alfred...HLEG thematic workshop on Measuring Inequalities of Income and Wealth, Alfred...
HLEG thematic workshop on Measuring Inequalities of Income and Wealth, Alfred...
 
Econometric (Indonesia's Economy).pptx
Econometric (Indonesia's Economy).pptxEconometric (Indonesia's Economy).pptx
Econometric (Indonesia's Economy).pptx
 
Spillover dynamics for systemic risk measurement using spatial financial time...
Spillover dynamics for systemic risk measurement using spatial financial time...Spillover dynamics for systemic risk measurement using spatial financial time...
Spillover dynamics for systemic risk measurement using spatial financial time...
 
Bab 2 risk and return part i
Bab 2   risk and return part iBab 2   risk and return part i
Bab 2 risk and return part i
 
Looking for cooperation on working paper - Expenditure model
Looking for cooperation on working paper - Expenditure modelLooking for cooperation on working paper - Expenditure model
Looking for cooperation on working paper - Expenditure model
 
1635 variance portfolio
1635 variance portfolio1635 variance portfolio
1635 variance portfolio
 
financial management chapter 4 Risk and Return
financial management chapter 4 Risk and Returnfinancial management chapter 4 Risk and Return
financial management chapter 4 Risk and Return
 
Ch 02 show. risk n return part 1
Ch 02 show. risk n return part 1Ch 02 show. risk n return part 1
Ch 02 show. risk n return part 1
 
Chapter 08
Chapter 08Chapter 08
Chapter 08
 
Risk and Return: Portfolio Theory and Assets Pricing Models
Risk and Return: Portfolio Theory and Assets Pricing ModelsRisk and Return: Portfolio Theory and Assets Pricing Models
Risk and Return: Portfolio Theory and Assets Pricing Models
 
Fm11 ch 04 show
Fm11 ch 04 showFm11 ch 04 show
Fm11 ch 04 show
 
Fm11 ch 04 risk and return the basics
Fm11 ch 04 risk and return the basicsFm11 ch 04 risk and return the basics
Fm11 ch 04 risk and return the basics
 
Presentation june 2020
Presentation june 2020Presentation june 2020
Presentation june 2020
 
Probability
ProbabilityProbability
Probability
 
Optimum Algorithm for Computing the Standardized Moments Using MATLAB 7.10(R2...
Optimum Algorithm for Computing the Standardized Moments Using MATLAB 7.10(R2...Optimum Algorithm for Computing the Standardized Moments Using MATLAB 7.10(R2...
Optimum Algorithm for Computing the Standardized Moments Using MATLAB 7.10(R2...
 
Introducing R package ESG at Rmetrics Paris 2014 conference
Introducing R package ESG at Rmetrics Paris 2014 conferenceIntroducing R package ESG at Rmetrics Paris 2014 conference
Introducing R package ESG at Rmetrics Paris 2014 conference
 
Efficient Numerical PDE Methods to Solve Calibration and Pricing Problems in ...
Efficient Numerical PDE Methods to Solve Calibration and Pricing Problems in ...Efficient Numerical PDE Methods to Solve Calibration and Pricing Problems in ...
Efficient Numerical PDE Methods to Solve Calibration and Pricing Problems in ...
 
Optimal Multisine Probing Signal Design for Power System Electromechanical Mo...
Optimal Multisine Probing Signal Design for Power System Electromechanical Mo...Optimal Multisine Probing Signal Design for Power System Electromechanical Mo...
Optimal Multisine Probing Signal Design for Power System Electromechanical Mo...
 

More from David Zibriczky

More from David Zibriczky (10)

Highlights from the 8th ACM Conference on Recommender Systems (RecSys 2014)
Highlights from the 8th ACM Conference on Recommender Systems (RecSys 2014)Highlights from the 8th ACM Conference on Recommender Systems (RecSys 2014)
Highlights from the 8th ACM Conference on Recommender Systems (RecSys 2014)
 
Predictive Solutions and Analytics for TV & Entertainment Businesses
Predictive Solutions and Analytics for TV & Entertainment BusinessesPredictive Solutions and Analytics for TV & Entertainment Businesses
Predictive Solutions and Analytics for TV & Entertainment Businesses
 
Improving the TV User Experience by Algorithms: Personalized Content Recommen...
Improving the TV User Experience by Algorithms: Personalized Content Recommen...Improving the TV User Experience by Algorithms: Personalized Content Recommen...
Improving the TV User Experience by Algorithms: Personalized Content Recommen...
 
Recommender Systems meet Finance - A literature review
Recommender Systems meet Finance - A literature reviewRecommender Systems meet Finance - A literature review
Recommender Systems meet Finance - A literature review
 
A Combination of Simple Models by Forward Predictor Selection for Job Recomme...
A Combination of Simple Models by Forward Predictor Selection for Job Recomme...A Combination of Simple Models by Forward Predictor Selection for Job Recomme...
A Combination of Simple Models by Forward Predictor Selection for Job Recomme...
 
Fast ALS-Based Matrix Factorization for Recommender Systems
Fast ALS-Based Matrix Factorization for Recommender SystemsFast ALS-Based Matrix Factorization for Recommender Systems
Fast ALS-Based Matrix Factorization for Recommender Systems
 
EPG content recommendation in large scale: a case study on interactive TV pla...
EPG content recommendation in large scale: a case study on interactive TV pla...EPG content recommendation in large scale: a case study on interactive TV pla...
EPG content recommendation in large scale: a case study on interactive TV pla...
 
Personalized recommendation of linear content on interactive TV platforms
Personalized recommendation of linear content on interactive TV platformsPersonalized recommendation of linear content on interactive TV platforms
Personalized recommendation of linear content on interactive TV platforms
 
An introduction to Recommender Systems
An introduction to Recommender SystemsAn introduction to Recommender Systems
An introduction to Recommender Systems
 
Data Modeling in IPTV and OTT Recommender Systems
Data Modeling in IPTV and OTT Recommender SystemsData Modeling in IPTV and OTT Recommender Systems
Data Modeling in IPTV and OTT Recommender Systems
 

Recently uploaded

US Economic Outlook - Being Decided - M Capital Group August 2021.pdf
US Economic Outlook - Being Decided - M Capital Group August 2021.pdfUS Economic Outlook - Being Decided - M Capital Group August 2021.pdf
US Economic Outlook - Being Decided - M Capital Group August 2021.pdf
pchutichetpong
 
一比一原版Birmingham毕业证伯明翰大学|学院毕业证成绩单如何办理
一比一原版Birmingham毕业证伯明翰大学|学院毕业证成绩单如何办理一比一原版Birmingham毕业证伯明翰大学|学院毕业证成绩单如何办理
一比一原版Birmingham毕业证伯明翰大学|学院毕业证成绩单如何办理
betoozp
 
PD ARRAY THEORY FOR INTERMEDIATE (1).pdf
PD ARRAY THEORY FOR INTERMEDIATE (1).pdfPD ARRAY THEORY FOR INTERMEDIATE (1).pdf
PD ARRAY THEORY FOR INTERMEDIATE (1).pdf
JerrySMaliki
 
what is the future of Pi Network currency.
what is the future of Pi Network currency.what is the future of Pi Network currency.
what is the future of Pi Network currency.
DOT TECH
 
NO1 Uk Black Magic Specialist Expert In Sahiwal, Okara, Hafizabad, Mandi Bah...
NO1 Uk Black Magic Specialist Expert In Sahiwal, Okara, Hafizabad,  Mandi Bah...NO1 Uk Black Magic Specialist Expert In Sahiwal, Okara, Hafizabad,  Mandi Bah...
NO1 Uk Black Magic Specialist Expert In Sahiwal, Okara, Hafizabad, Mandi Bah...
Amil Baba Dawood bangali
 
一比一原版Adelaide毕业证阿德莱德大学毕业证成绩单如何办理
一比一原版Adelaide毕业证阿德莱德大学毕业证成绩单如何办理一比一原版Adelaide毕业证阿德莱德大学毕业证成绩单如何办理
一比一原版Adelaide毕业证阿德莱德大学毕业证成绩单如何办理
zsewypy
 
一比一原版UO毕业证渥太华大学毕业证成绩单如何办理
一比一原版UO毕业证渥太华大学毕业证成绩单如何办理一比一原版UO毕业证渥太华大学毕业证成绩单如何办理
一比一原版UO毕业证渥太华大学毕业证成绩单如何办理
yonemuk
 
USDA Loans in California: A Comprehensive Overview.pptx
USDA Loans in California: A Comprehensive Overview.pptxUSDA Loans in California: A Comprehensive Overview.pptx
USDA Loans in California: A Comprehensive Overview.pptx
marketing367770
 

Recently uploaded (20)

US Economic Outlook - Being Decided - M Capital Group August 2021.pdf
US Economic Outlook - Being Decided - M Capital Group August 2021.pdfUS Economic Outlook - Being Decided - M Capital Group August 2021.pdf
US Economic Outlook - Being Decided - M Capital Group August 2021.pdf
 
9th issue of our inhouse magazine Ingenious May 2024.pdf
9th issue of our inhouse magazine Ingenious May 2024.pdf9th issue of our inhouse magazine Ingenious May 2024.pdf
9th issue of our inhouse magazine Ingenious May 2024.pdf
 
一比一原版Birmingham毕业证伯明翰大学|学院毕业证成绩单如何办理
一比一原版Birmingham毕业证伯明翰大学|学院毕业证成绩单如何办理一比一原版Birmingham毕业证伯明翰大学|学院毕业证成绩单如何办理
一比一原版Birmingham毕业证伯明翰大学|学院毕业证成绩单如何办理
 
The European Unemployment Puzzle: implications from population aging
The European Unemployment Puzzle: implications from population agingThe European Unemployment Puzzle: implications from population aging
The European Unemployment Puzzle: implications from population aging
 
when officially can i withdraw my pi Network coins.
when officially can i withdraw my pi Network coins.when officially can i withdraw my pi Network coins.
when officially can i withdraw my pi Network coins.
 
Introduction to Indian Financial System ()
Introduction to Indian Financial System ()Introduction to Indian Financial System ()
Introduction to Indian Financial System ()
 
PD ARRAY THEORY FOR INTERMEDIATE (1).pdf
PD ARRAY THEORY FOR INTERMEDIATE (1).pdfPD ARRAY THEORY FOR INTERMEDIATE (1).pdf
PD ARRAY THEORY FOR INTERMEDIATE (1).pdf
 
what is the future of Pi Network currency.
what is the future of Pi Network currency.what is the future of Pi Network currency.
what is the future of Pi Network currency.
 
what is the best method to sell pi coins in 2024
what is the best method to sell pi coins in 2024what is the best method to sell pi coins in 2024
what is the best method to sell pi coins in 2024
 
NO1 Uk Black Magic Specialist Expert In Sahiwal, Okara, Hafizabad, Mandi Bah...
NO1 Uk Black Magic Specialist Expert In Sahiwal, Okara, Hafizabad,  Mandi Bah...NO1 Uk Black Magic Specialist Expert In Sahiwal, Okara, Hafizabad,  Mandi Bah...
NO1 Uk Black Magic Specialist Expert In Sahiwal, Okara, Hafizabad, Mandi Bah...
 
is it possible to sell pi network coin in 2024.
is it possible to sell pi network coin in 2024.is it possible to sell pi network coin in 2024.
is it possible to sell pi network coin in 2024.
 
一比一原版Adelaide毕业证阿德莱德大学毕业证成绩单如何办理
一比一原版Adelaide毕业证阿德莱德大学毕业证成绩单如何办理一比一原版Adelaide毕业证阿德莱德大学毕业证成绩单如何办理
一比一原版Adelaide毕业证阿德莱德大学毕业证成绩单如何办理
 
一比一原版UO毕业证渥太华大学毕业证成绩单如何办理
一比一原版UO毕业证渥太华大学毕业证成绩单如何办理一比一原版UO毕业证渥太华大学毕业证成绩单如何办理
一比一原版UO毕业证渥太华大学毕业证成绩单如何办理
 
Commercial Bank Economic Capsule - May 2024
Commercial Bank Economic Capsule - May 2024Commercial Bank Economic Capsule - May 2024
Commercial Bank Economic Capsule - May 2024
 
USDA Loans in California: A Comprehensive Overview.pptx
USDA Loans in California: A Comprehensive Overview.pptxUSDA Loans in California: A Comprehensive Overview.pptx
USDA Loans in California: A Comprehensive Overview.pptx
 
Proposer Builder Separation Problem in Ethereum
Proposer Builder Separation Problem in EthereumProposer Builder Separation Problem in Ethereum
Proposer Builder Separation Problem in Ethereum
 
how can I sell my locked pi coins safety.
how can I sell my locked pi coins safety.how can I sell my locked pi coins safety.
how can I sell my locked pi coins safety.
 
how can I transfer pi coins to someone in a different country.
how can I transfer pi coins to someone in a different country.how can I transfer pi coins to someone in a different country.
how can I transfer pi coins to someone in a different country.
 
Empowering the Unbanked: The Vital Role of NBFCs in Promoting Financial Inclu...
Empowering the Unbanked: The Vital Role of NBFCs in Promoting Financial Inclu...Empowering the Unbanked: The Vital Role of NBFCs in Promoting Financial Inclu...
Empowering the Unbanked: The Vital Role of NBFCs in Promoting Financial Inclu...
 
MERCHANTBANKING-PDF complete picture.pdf
MERCHANTBANKING-PDF complete picture.pdfMERCHANTBANKING-PDF complete picture.pdf
MERCHANTBANKING-PDF complete picture.pdf
 

Entropy based asset pricing

  • 1. Mihály Ormos and Dávid Zibriczky Department of Finance Budapest University of Technology and Economics 5th international ECEE series conference "Economic Challenges in Enlarged Europe„ Tallinn, Estonia, June 16-18, 2013
  • 2.  Risk vs. expected return (risk premium)  Standard deviation (Markowitz, 1952)  Assumes normal distribution  CAPM beta (Sharpe, 1964)  Requires market portfolio  Assumes linearity  Measures systematic risk only  Problems:  Assumptions  Low explanatory power  Alternative single risk measure? 2
  • 3. mathematically-defined quantity that is generally used for characterizing the probability of outcomes in a system through a process Main applications:  Thermodynamics: Clausius (1867) – distribution of inner energy  Statistical Mechanics: Boltzmann (1872) – molecular disorder  Information Theory: Shannon (1948) – message compression 3
  • 4.  Generalized discrete formula: 𝐻 𝛼 𝑋 = 1 1−𝛼 log 𝑖 𝑝𝑖 𝛼  𝑝𝑖: probability of discrete outcome Xi  Special cases  𝛼 = 1: Shannon entropy (L'Hôpital's rule)  𝛼 = 2: Rényi entropy  Problem: Value of return is continuous, discrete formula cannot be applied  Continuous formula: 𝐻 𝛼 𝑋 = 1 1−𝛼 ln 𝑓 𝑥 𝛼 𝑑𝑥 (differential entropy)  𝑓 𝑥 : probability function  𝑓 𝑥 = ? 4
  • 5.  Density estimation: 𝑓𝑛 𝑥 ~ 𝑓 𝑥  Methods  Histogram (fixed bin width)  Kernel density estimation (sum of core weights)  Sample spacing (fixed number of elements in one bin) Histogram Kernel Sample spacing 5
  • 6.  Original formula: 𝐻 𝛼 𝑋 = 1 1−𝛼 ln 𝑓 𝑥 𝛼 𝑑𝑥  Entropy estimation in single formula using histogram*:  Shannon entropy: 𝐻1,𝑛 𝑋 = − 1 𝑛 𝑗 𝑣𝑗 ln 𝑣 𝑗 𝑛ℎ  Rényi entropy: 𝐻2,𝑛 𝑋 = −ln 𝑗 ℎ 𝑣 𝑗 𝑛ℎ 2  Entropy risk measure**: 𝜿 𝑯 𝜶 𝑺𝒊 = 𝒆 𝑯 𝜶,𝒏 𝑹 𝒊−𝑹 𝑭 * 𝑣𝑗: number of elements falling into the jth bin, h: bin size ** 𝜿: risk measure, 𝑺𝒊: Security i, 𝑹𝒊 − 𝑹 𝑭: Risk premium 6
  • 7.  Standard deviation (Markovitz): 𝜿 𝝈 𝑺𝒊 = 𝝈 𝑹𝒊 − 𝑹 𝑭  Beta (CAPM): 𝜿 𝜷 𝑺𝒊 = 𝐜𝐨𝐯 𝑹 𝒊−𝑹 𝑭,𝑹 𝑴−𝑹 𝑭 𝝈 𝟐 𝑹 𝑴−𝑹 𝑭  Shannon entropy: 𝜿 𝑯 𝟏 𝑺𝒊 = 𝒆 𝑯 𝟏 𝑹𝒊−𝑹 𝑭  Rényi entropy: 𝜿 𝑯 𝟐 𝑺𝒊 = 𝒆 𝑯 𝟐 𝑹𝒊−𝑹 𝑭 7
  • 8.  Source: The Center for Research in Security Prices (CRSP)  Series: Daily return  Market return (value weighted)  Risk free rate (1-month T-bill)  150, randomly selected securities from the components of S&P500 index  Period: 1985-2011 (27 years or 6810 days) 8
  • 9. 0 1 2 3 4 5 6 7 8 1 10 100 Averagerisk Number of securities in portfolio Shannon Rényi StDev 0% 10% 20% 30% 40% 50% 1 10 100 Averageriskreduction Number of securities in portfolio Shannon Rényi StDev 9
  • 10. -0,02 0 0,02 0,04 0,06 0,08 0,1 0 2,5 5 7,5 10 12,5 15 E(rp-rF) Risk (H1) Shannon entropy n=1 n=2 n=5 n=10 10
  • 11. -0,02 0 0,02 0,04 0,06 0,08 0,1 0 2,5 5 7,5 10 12,5 15 E(rp-rF) Risk (H1) Shannon entropy n=1 n=2 n=5 n=10 11
  • 12.  Evaluation method  Long term (1985-2011), for 150 random securities  Explanatory variable (X): Risk  Target variable (Y): Expected risk premium  Linear regression (X,Y)  Explanatory power: Goodness of fit of regression line (R2)  Result  Higher explanatory power Risk measure R2 long Standard deviation 7.83% Beta 6.17% Shannon entropy 12.98% Rényi entropy 15.71% 12
  • 13. -0,02 0 0,02 0,04 0,06 0,08 0,1 0,12 0,14 0 1 2 3 4 5 E(ri-rF) Risk (std) Standard deviation E(ri-rF) = 0.0170 + 0.0085*std R² = 7.83% -0,02 0 0,02 0,04 0,06 0,08 0,1 0,12 0,14 0 0,5 1 1,5 2 E(ri-rF) Risk (beta) Beta E(ri-rF) = 0.0209 + 0.0151*beta R² = 6.17% -0,02 0 0,02 0,04 0,06 0,08 0,1 0,12 0,14 0 2,5 5 7,5 10 12,5 15 E(ri-rF) Risk (H1) Shannon entropy E(ri-rF) = 0.0091 + 0.0034*H1 R² = 12.98% -0,02 0 0,02 0,04 0,06 0,08 0,1 0,12 0,14 0 2,5 5 7,5 10 12,5 E(ri-rF) Risk (H2) Rényi entropy E(ri-rF) = 0.0059 + 0.0049*H2 R² = 15.71% 13
  • 14. 0 0,05 0,1 0,15 0 10 20 30 40 50 60 70 80 90 100 Explanatorypower(R-squared) Number of securities in portfolio StDev Beta Shannon Rényi 14
  • 15. -0,3 -0,2 -0,1 0 0,1 0,2 0 5 10 15 20 E(ri-rF) Risk (H1) Shannon entropy bear market E(ri-rF) = 0.0818 - 0.0150*H1 R² = 0.3961 -0,05 0 0,05 0,1 0,15 0,2 0 5 10 15 E(ri-rF) Risk (H1) Shannon entropy bull market E(ri-rF) = -0.0116 + 0.0103*H1 R² = 0.4345 Risk measure R2 long R2 bull R2 bear Standard deviation 7.83% 33.9% 36.7% Beta 6.17% 36.7% 43.7% Shannon entropy 12.98% 43.5% 39.6% Rényi entropy 15.71% 42.4% 38.6% 15
  • 16.  Evaluation method  Estimating risk based on a 5-year period (short term)  Predicting average risk premium for the next 5-year period  Applying this on several periods  Predicting power: Average goodness of fit (R2) based on tested periods  Reliability: Standard deviation of R2 values (lower is better)  Result  Better explanatory power for the same first 5 years (R2 short)  Better predicting power for the next 5 years (R2 pred)  Higher reliability (σ) Risk measure R2 long R2 short R2 pred σ short σ pred Standard deviation 7.83% 7.94% 9.70% 0.73 0.63 Beta 6.17% 13.31% 6.45% 0.95 0.99 Shannon entropy 12.98% 13.38% 10.15% 0.67 0.62 Rényi entropy 15.71% 12.82% 9.34% 0.62 0.60 16
  • 17.  No assumptions concerning the returns  Entropy estimation doesn’t require an undefined market portfolio  Characterizes specific risk and captures diversification effect  More efficient and reliable risk estimate compared to the standard methods  If the trend is identified entropy based equilibrium model behaves similarly to the standard models Thank you for paying attention! 17