Efficient

Diversification

6
 SHOULD WE DIVERSIFY

YOUR VIEWS?

Listen to Raymond Dalio, a billionaire investor,

hedge fund manager, and philanthropist, the
founder of investment firm Bridgewater
Associates, one of the world's largest hedge funds.

“Diversification and sample size, knowing how to

engineer your bet is even more important than
2 knowing what the good bet is.”

https://www.investopedia.com/investing/importan
ce-diversification/
OUTCOMES
CHECK IF YOU CAN DO/UNDERSTAND THE FOLLOWING AFTER
CLASS

• Understand systematic and nonsystematic risk

• Appreciate the benefit of diversification
• Allocate two risky assets
• Efficient diversification with many assets
• Index model

3
6.1 DIVERSIFICATION AND PORTFOLIO
RISK
 Can you eliminate all risks through diversification?

 Market/Systematic/Non diversifiable Risk

 Risk factors common to whole economy
 Example?

 Unique/Firm-Specific/Nonsystematic/ Diversifiable
Risk
 Risk that can be eliminated by diversification
 Example?

4
FIGURE 6.1 RISK AS FUNCTION OF NUMBER
OF STOCKS IN PORTFOLIO

5
FIGURE 6.2 RISK VERSUS
DIVERSIFICATION

How many stocks should be included in a portfolio to eliminate

diversifiable risks?
6
 6.2 ASSET ALLOCATION WITH TWO
RISKY ASSETS
 Covariance and Correlation
 Portfolio risk depends on covariance between returns of
assets
 Expected return on two-security portfolio



E (rp ) = W1r1 + W2 r2

W1 = Proportion of funds in security 1



W2 = Proportion of funds in security 2
r1 = Expected return on security 1
r 2 = Expected return on security 2

7
Can I change r to sigma to get the expected risk of the portfolio?
6.2 ASSET ALLOCATION WITH TWO
RISKY ASSETS
 Covariance Calculations
S
Cov( rS , rB ) =  p(i)[rS (i) − E (rS )][rB (i) − E (rB )]
i =1

 Excel example on the coming slides, spreadsheet 6.4

 Correlation Coefficient
Cov(rS , rB )
ρ SB =
σS  σB

Cov( rS , rB ) = ρ SBσ S σ B
8
 SPREADSHEET 6.1 CAPITAL MARKET
EXPECTATIONS

Spreadsheet exercises Not required in the tests,

but please try to repeat the exercise at home if you have spare
time.

9
SPREADSHEET 6.2 VARIANCE OF
RETURNS

10
SPREADSHEET 6.3 PORTFOLIO
PERFORMANCE

0.4x-37%+0.6x-9 = -20.2%

Calculation steps are mechanical, 11

you will get it after trying it one or two times.
SPREADSHEET 6.4 RETURN COVARIANCE

It shows us how to get the Covariance and the correlation coefficient from
different scenario. Try repeat it yourself.

12
 6.2 ASSET ALLOCATION WITH TWO
RISKY ASSETS
 Portfolio risk and return depend on the means and
variances of the component securities (the two risky
assets in this case)
 How can we get the means and variances?

 Using Historical Data (a common approach)

 Variability/covariability change slowly over time, should be
more reliable when using realized returns to estimate
 However, using realized returns to estimate the averages of
past returns (the mean return) can be very noisy,
 Macroeconomic and security analysis are also used to
estimate the mean
13
6.2 ASSET ALLOCATION WITH TWO
RISKY ASSETS
 Weighted average of returns on components, with
investment proportions as weights

 Weighted average of expected returns on

components, with portfolio proportions as weights

 Variance:

14
6.2 ASSET ALLOCATION WITH TWO
RISKY ASSETS
 Risk-Return Trade-Off
 Should we look for lower risk? But it usually comes with
lower return?
 We can look at the Investment opportunity set
 Available portfolio risk-return combinations
 and use the mean-Variance Criterion

 Mean-Variance Criterion
 If E(rA) ≥ E(rB) and σA ≤ σB
 Portfolio A dominates portfolio B

15
FIGURE 6.3 INVESTMENT
OPPORTUNITY SET

16
SPREADSHEET 6.5 INVESTMENT
OPPORTUNITY SET

17
FIGURE 6.4 OPPORTUNITY SETS: VARIOUS
CORRELATION COEFFICIENTS

See port risk.xls as a

demonstration

18
 Near zero
SPREADSHEET 6.6 OPPORTUNITY SET -VARIOUS
CORRELATION COEFFICIENTS

19
6.3 THE OPTIMAL RISKY PORTFOLIO WITH A
RISK-FREE ASSET
 How to find the optimal risky portfolio?
 i.e. for a given level of risk, you get the highest expected return.
 Refer to Figure 6.3 Investment Opportunity Set

 Slope of CAL is Sharpe Ratio of Risky Portfolio



 Optimal Risky Portfolio

 Best combination of risky and safe assets to form

portfolio 20
6.3 THE OPTIMAL RISKY PORTFOLIO WITH A
RISK-FREE ASSET
 Calculating Optimal Risky Portfolio
 Two risky assets

[ E (rB ) − rf ] S2 − [ E (rs ) − rf ] B S  BS
wB =
[ E (rB ) − rf ] S2 + [ E (rs ) − rf ] B2 − [ E (rB ) − rf + E (rs ) − rf ] B S  BS

wS = 1 − wB

Note: Formula will be given if required

21
FIGURE 6.5 TWO CAPITAL
ALLOCATION LINES

Which line (A or MIN) is better?

22
FIGURE 6.6 BOND, STOCK AND T-BILL
OPTIMAL ALLOCATION

23
FIGURE 6.7 THE COMPLETE PORTFOLIO
Once you have obtained the optimal risk
portfolio O, you may determine the complete
portfolio C based on your risk preference.

24
AN EXPLANATION ON GETTING PORTFOLIO
C ON PREVIOUS SLIDE (REFERENCE)

Your risk aversion

determines your
utility function, u1,
u2, etc.

Like investor A and

B, investor C can
get portfolio C
using the
indifference curve
of C.
25
FIGURE 6.8 PORTFOLIO COMPOSITION: ASSET
ALLOCATION SOLUTION
An example may look like
this.
You can refer to the
text for the calculations

26
 6.4 EFFICIENT DIVERSIFICATION WITH
MANY RISKY ASSETS
 Efficient Frontier of Risky Assets
 Graph representing set of portfolios that maximizes
expected return at each level of portfolio risk
 Maximize risk premium for any level standard deviation
 Minimize standard deviation for any level risk premium

 Maximize Sharpe ratio for any standard deviation or risk

premium

27
FIGURE 6.9 PORTFOLIOS CONSTRUCTED
WITH THREE STOCKS (A, B AND C)

What can be achieved if

we have A, B and C?

28
E from AB and F from BC resulting curve EF, more northwest (i.e.
lower risk and higher return, a better feature)
FIGURE 6.10 EFFICIENT FRONTIER: RISKY
AND INDIVIDUAL ASSETS

29
 6.4 EFFICIENT DIVERSIFICATION WITH
MANY RISKY ASSETS
 Choosing Optimal Risky Portfolio
 Optimal portfolio CAL tangent to efficient frontier
 Similar to the process of choosing an optimal risky portfolio

of two risky assets

30
 6.4 EFFICIENT DIVERSIFICATION WITH
MANY RISKY ASSETS
 Optimal Risky Portfolio: An Illustration
 Efficiently diversified global portfolio using stock market
indices of six countries
 Standard deviation and correlation estimated from
historical data
 Risk premium forecast generated from fundamental
analysis (not for historical averages)

31
FIGURE 6.11 EFFICIENT FRONTIERS/CAL: TABLE 6.1

0.09
Efficient Frontier
Capital Allocation Line
0.08 Eff Front - No Short
Min-Var with short sales
Min-Var no short sales
0.07 Optimum with short sales
Optimum no short sales
0.06 Individual countries
Risk Premium

0.05
0.04
0.03
0.02
0.01
0.00
0.00 0.05 0.10 0.15 0.20 0.25 0.30 32
Standard Deviation

Click the chart to access Excel data

WHAT WOULD YOU CHOOSE?

What factors would you consider if you want to derive a formula to

explain stock price movement/stock return?

What if you can only choose one factor?

33
6.5 A SINGLE-INDEX MODEL
 Index model
 Single-index model assumes that there is only one
macroeconomic factor affecting all stock returns and
this macroeconomic factor can be represented by the rate of return on
a market index, such as the Hang Seng index, FTSE 100, S&P 500,
etc.

 It relates stock returns to returns on broad market index/firm-

specific factors

34
6.5 A SINGLE-INDEX MODEL

 Ri = αi + βiRM + ei
 ri- rf = αi + βi( rM –rf) + ei

• Ri : excess return, rate of return in excess of risk-free rate

• RM :market excess return
• βi: sensitivity of security’s returns to market factor
• ei: residual or random return, assume the mean value=zero
component of return independent of/unexplained by market factor
• αi : expected return beyond that induced by market index, abnormal return
• r : return
• rf : risk free return
35
FIGURE 6.12 SCATTER DIAGRAM FOR DELL
Graphical Representation of Single-Index Model

Security Characteristic Line (SCL). Plot of security’s excess return from excess return of
market 36

You can easily get Beta of the stock you want. How do we make use of this in practice?
 6.5 A SINGLE-INDEX MODEL
 Statistical Representation of Single-Index Model
 How well does the model fit the data?
 R-square = ρ2

Total variance = Systematic variance + Residual variance

= β2Dσ2M + σ2(eD)
37
Residual variance measures the dispersion of the scatter
of actual returns about the regression line
FIGURE 6.13 VARIOUS SCATTER
DIAGRAMS

38

https://www.investopedia.com/terms/r/regression.asp
EXAMPLE

The characteristic line is:

R = 4.05 +1.32RM

What does R-square mean?

The model can explain 12% of the return variance,

39
in other words, how well does market changes, rM, explain the stock return
QUESTIONS
We know how to calculate the expected return and the risk of a portfolio
based Markowitz portfolio theory. How about their calculations based on
index model?
-refer to the word file: Portfolio risk_index model_ch6
What are the advantages of using index model over Markowitz model?
Which one is better?
- Let me show you two studies that I can download easily from the
internet (they are not representative). If you are interested, there are
Journal articles available from library for you to find out.
ComparisonOfTheMarkowitzAndSingleIndexModelBasedOnMVCriterionInOptimal
PortfolioFormation.pdf

Comparison among Different Models in Determining Optimal.pdf

40