Markowitz Portfolio Selection Model

The Investment Decision

• Top-down process with 3 steps:

1. Capital allocation between the risky portfolio and
risk-free asset
2. Asset allocation across broad asset classes
3. Security selection of individual assets within each
asset class
 Diversification and Portfolio Risk

• Market risk
• Risk attributable to marketwide risk sources and
remains even after extensive diversification
• Also call systematic or nondiversifiable
• Firm-specific risk
• Risk that can be eliminated by diversification
• Also called diversifiable or nonsystematic
 Figure 7.1 Portfolio Risk and
the Number of Stocks in the Portfolio

Panel A: All risk is firm specific. Panel B: Some risk is systematic or marketwide.
 Portfolios of Two Risky Assets:
Return
• Portfolio return: rp = wDrD + wErE
– wD = Bond weight
– rD = Bond return
– wE = Equity weight
– rE = Equity return

E(rp) = wD E(rD) + wEE(rE)

 Portfolios of Two Risky Assets:
Risk
• Portfolio variance:
 p = wD D + wE E + 2wD wE Cov ( rD , rE )
2 2 2 2 2

–  D2 = Bond variance

–  2
E
= Equity variance

–Cov ( rD , rE ) = Covariance of returns for bond

and equity
 Portfolios of Two Risky Assets:
Covariance
• Covariance of returns on bond and equity:
Cov(rD,rE) = DEDE

– D,E = Correlation coefficient of returns

– D = Standard deviation of bond returns
– E = Standard deviation of equity returns
 Portfolios of Two Risky Assets:
Correlation Coefficients
• Range of values for 1,2
- 1.0 >  > +1.0
– If  = 1.0, the securities are perfectly positively
correlated
– If  = - 1.0, the securities are perfectly negatively
correlated
 Portfolios of Two Risky Assets:
Correlation Coefficients
• When ρDE = 1, there is no diversification

 P = wE E + wD D
• When ρDE = -1, a perfect hedge is possible

D
wE = = 1 − wD
 D + E
Minimum Variance Portfolio
Weight
Optimal Risky Portfolio Weight
Figure 7.3 Portfolio Expected Return as a Function of
Investment Proportions
Table 7.3 Expected Return and Standard Deviation with
Various Correlation Coefficients
Portfolio Expected Return as a function of Standard
Deviation
Markowitz Portfolio Optimization Model

• Capital Allocation and the Separation Property

• Portfolio choice problem may be separated into
two independent tasks
• Determination of the optimal risky portfolio is
purely technical
• Allocation of the complete portfolio to risk-free
versus the risky portfolio depends on personal
preference
 Markowitz Portfolio Selection Model

• Many risky securities + Risk-free assets

Three steps:
1. Identify the risk-return combinations available
from the risky assets (investment opportunity
set),
2. Find the optimal risky portfolio (i.e. the risk-
return combination that produces the steepest
CAL),
3. Find the optimal complete portfolio (Utility
Maximization).
17
 Step 1. Minimum-Variance Frontier
– Plot the different portfolio combinations of
assets.
– Pick a target return and find the portfolio
minimum variance that
E{Rp}

Standard deviation
 Step 1. Minimum-Variance Frontier
Points below the efficient frontier are “dominated”
– R is dominated by S; Y is dominated by Z

E{Rp}
Efficient Frontier
Z
S

Standard deviation
Minimum-Variance Frontier: Summary

20
Markowitz Portfolio Optimization Model

• Search for the CAL with the highest reward-to-

variability ratio
• Everyone invests in P, regardless of their
degree of risk aversion
• More risk averse investors put more in the risk-
free asset
• Less risk averse investors put more in P
Figure 7.11 The Efficient Frontier of
Risky Assets with the Optimal CAL
 Step 2: Find the optimal risky portfolio

• The aim is to reach the steepest capital

allocation line (CAL), i.e. the maximum Sharpe
Ratio.

23
Step 2: Find the optimal risky portfolio

24
Step 3: Find the optimal complete portfolio

Optimal Allocation Formula:

25
 Summary – Capital Allocation Decision
1. Choose the risky assets that will form your risky portfolio
2. Find all feasible combinations of portfolio risk and portfolio return with
all-risky assets
3. Eliminate all that are "dominated" or, equivalently, mean-variance
inefficient
• Find "Efficient Frontier" (which involves only risky assets)
4. Find all feasible combinations of portfolios on efficient frontier with
risk-free asset
• Eliminate all that are "dominated” ➔Capital Allocation Line
with steepest slope
• Along the line: Different combinations of risk-free asset with
one unique portfolio of risky assets
5. Given this CAL, agents choose optimal allocation between risky portfolio
and risk-free asset by maximizing utility (i.e. Reach the highest
indifference curve).
26
Chapter 3, Sections 3.6, 3.7

MARGIN TRADING
SHORT SALES

27
 Buying Stocks on Margin
• "Margin" is borrowing money from your
broker to buy a stock and using your
investment as collateral.
– Investors generally use margin to increase their
purchasing power so that they can own more
stock without fully paying for it.
– But margin exposes investors to the potential for
higher losses.

28
 Before You Trade – Minimum Margin

• Before trading on margin, the NYSE and

NASDAQ, for example, require you to deposit
with your brokerage firm a minimum of
$2,000 or 100 percent of the purchase price,
whichever is less.
• This is known as the "minimum margin."
• Some firms may require you to deposit more
than $2,000.

29
 Margin Calls

• If your account falls below the firm's

maintenance requirement, your firm generally
will make a margin call to ask you to deposit
more cash or securities into your account.
• If you are unable to meet the margin call, your
firm will sell your securities to increase the
equity in your account up to or above the
firm's maintenance requirement.

30
 Example
• Suppose an investor initially pays $8,000 toward the
purchase of $16,000 stock.
– 100 shares at $160 per share.
• Borrows the remaining $8,000 from her broker.
Assets
Value of Stock $16,000
Liabilities and Owner's Equity
Loan from broker $8,000
Equity $8,000

• The initial percentage margin is: 50%

31
 Example
• If the price drops to $120 per share, the account
balance becomes:
Assets
Value of Stock $12,000
Liabilities and Owner's Equity
Loan from broker $8,000
Equity $4,000
• The percentage margin becomes: 33.33%
• Suppose the maintenance margin is 25%.
– You need to have $3,000 (=25% * $12,000) in equity.
• There is no margin call….

32
 Example
• If the price drops to $120 per share, the account
balance becomes:
Assets
Value of Stock $12,000
Liabilities and Owner's Equity
Loan from broker $8,000
Equity $4,000
• The percentage margin becomes: 33.33%
• Suppose the maintenance margin is 40%.
– You need to have $4,800 (=40% * $12,000) in equity.
• Margin call!!!
– Up to, or above the maintenance requirement.
33
 Why to buy on Margin?

• To achieve greater upside potential.

• Example:
– You have $10,000, buy 100 shares of IBM stock at
$100 per share
– Expect that the price of IBM stock will increase by
30%, expected return is
– If borrow $10,000 at 9%:
• The total investment in the IBM stock is
• If the stock price increases by 30%, the rate of return
will be
34
 Downside Risk from Margin Trading?

• Example:
– You have $10,000, buy 100 shares of IBM stock at
$100 per share
– Expect that the price of IBM stock will increase by
30%, expected return is
– If borrow $10,000 at 9%:
• The total investment in the IBM stock is
• If the stock price decreases by 30%, the rate of return
will be

35
 Example 3.1
Margin Trading: Initial Conditions
Share price $100
60% Initial Margin
40% Maintenance Margin
100 Shares Purchased
Initial Position
Stock $10,000 Borrowed $4,000
Equity $6,000
 Example 3.1
Margin Trading: Margin Call
Stock price falls to $70 per share
New Position
Stock $7,000 Borrowed $4,000
Equity $3,000

Margin% = $3,000/$7,000 = 43%

 Example 3.2
Margin Trading: Maintenance Margin
How far can the stock price fall before a
margin call? Let maintenance margin = 30%
Equity = 100P - $4000
Percentage margin = (100P - $4,000)/100P
(100P - $4,000)/100P = 0.30
Solve to find:
P = $57.14
 Example 2
• You purchased 100 shares of IBM common
stock on margin at $70 per share. Assume the
initial margin is 50% and the maintenance
margin is 30%. Below what stock price level
would you get a margin call? Assume the stock
pays no dividend; ignore interest on margin.

39
 Short Sales
• The sale of a stock you do not own.
• Investors who sell short believe the price of
the stock will fall.
– If the price drops, you can buy the stock at the
lower price and make a profit.
– If the price of the stock rises and you buy it back
later at the higher price, you will incur a loss.
– Old Rule: Exchange rules permit short sales only
when the last recorded change in stock price is
positive (The uptick rule was removed in July
2007). 40
 Short Sales
• When you sell short, your brokerage firm
loans you the stock.
– The stock you borrow comes from either the
firm’s own inventory, the margin account of
another of the firm’s clients, or another brokerage
firm.
• Short-seller is subject to the margin rules, and
other fees and charges may apply.
• If the stock you borrow pays a dividend, you
must pay the dividend to the person or firm
making the loan. 41
 Cash Flows from Purchasing versus Short-
selling of Shares

Purchase of Stock
Time Action Cash Flow
0 Buy share - Initial Price
1 Receive dividend, sell share + Ending Price + Dividend

Profit =

Short Sale of Stock

Time Action Cash Flow
0 Borrow share, sell it + Initial price
1 Repay dividend and buy share to replace -Ending price - Dividend
the share originally borrowed

Profit =
42
 Example 3.3
Short Sale: Initial Conditions
Dot Bomb 1000 Shares
50% Initial Margin
30% Maintenance Margin
$100 Initial Price

Sale Proceeds $100,000

Margin & Equity $50,000
Stock Owed 1000 shares
 Example 3.3
Short Sale: Dot Bomb falls to $70 per share

Assets Liabilities
$100,000 (sale proceeds) $70,000 (buy shares)
$50,000 (initial margin)
Equity
$80,000

Profit = Ending equity – Beginning equity

= $80,000 - $50,000 = $30,000
= Decline in share price x Number of shares sold short
 Example 3.3
Short Sale: Margin Call
How much can the stock price rise before a
margin call?

($150,000* - 1000P)/(1000P) = 30%

P = $115.38

* Initial margin plus sale proceeds

 Example
• You sold short 300 shares of common stock at
$55 per share. The initial margin is 60%. At
what stock price would you receive a margin
call if the maintenance margin is 35%?

46
Chapter Eight
Index Models
 Inputs for Markowitz Portfolio Selection
Suppose your security analysts can thoroughly analyze 50 stocks. This
means that your input list will include the following:

– This is a formidable task, particularly in light of the fact that a 50-

security portfolio is relatively small. Doubling n to 100 will nearly
quadruple the number of estimates to 5,150.
– If n 3,000, roughly the number of NYSE stocks, we need more than
4.5 million estimates.

48
 A Single-Factor Market

• Advantages
• Reduces the number of inputs for diversification
• Easier for security analysts to specialize
• Model
ri = E(ri ) + i m + ei
• βi = response of an individual security’s return to
the common factor, m; measure of systematic risk
• m = a common macroeconomic factor
• ei = firm-specific surprises
The Regression Equation of the Single
Index Model
• The regression equation is:

(ri - rf) = α i + ßi(rm - rf) + ei

Ri = α i + ßiRm + ei

• Regress the excess return of a security on the

excess return of the index.

50
 Single-Index Model

• Regression equation:
Ri (t ) = i + i RM (t ) + ei (t )

• Expected return-beta relationship:

E(Ri ) = i + i E(RM )
 Single-Index Model

• Variance = Systematic risk + Firm-specific risk:

 =   +  (ei )
i
2
i
2 2
M
2

• Covariance = Product of betas × Market index

risk:
Cov ( ri , rj ) = i  j 2
M

Cov (ri , rj ) =  i  j M2
 Single-Index Model

• Correlation = Product of correlations with the

market index

i  j M2 i M2  j M2
Corr ( ri , rj ) = =
 i j  i M  j M
= Corr ( ri , rM )  Corr ( rj , rM )
 Index Model and Diversification

• Variance of the equally-weighted portfolio of

firm-specific components:
2

 (e p ) =     (ei ) =  (e)
2
n
1 2 1 2
i =1  n  n

• When n gets large, σ2(ep) becomes negligible

and firm specific risk is diversified away
 Figure 8.1 The Variance of an Equally
Weighted Portfolio with Risk Coefficient βp
 Estimating the Single Index Model
(ri - rf) = α i + ßi(rm - rf) + ei

• Let’s put the model in practice…

• Pick a period to build the model.
– How many observations should the model have?
• Data inputs
– Stock prices and dividend distributions for N
securities.
– T-bill rates

56
 Estimating the Single Index Model
• Let’s go over an estimation example for Hewlett
Packard Stock
(rHP - rf) = α HP + ßi(rS&P500 - rf) + eHP

(1) Download stock prices from

– WRDS – CRSP Database
– Monthly from March, 2000 to March, 2006
– We will have 61 monthly observations.
– Use the prices adjusted for dividends and splits
(2) Calculate returns using the prices
- We will have 60 monthly returns 57
Estimating the Single Index Model
(rHP - rf) = α HP + ßi(rS&P500 - rf) + eHP

(3) Get the T-bill rates for the same period

http://research.stlouisfed.org/fred2/categories/115
(4) Calculate the excess returns.
(5) Define the market and calculate market returns.
– Most common practice is to use the S&P 500
Index Portfolio.
– http://finance.yahoo.com/q/hp?s=%5EGSPC

58
 Estimating the Single Index Model
(6) Calculate the market excess returns.
Now we have all the data we need to estimate the
model (i.e. run the regression)

(rHP - rf) = α HP + ßi(rS&P500 - rf) + eHP

(7) Use a statistical software to estimate the
equation
– Excel will do the job….

59
Figure 8.2 Excess Returns on
HP and S&P 500
Figure 8.3 Scatter Diagram of HP, the S&P 500,
and HP’s SCL

RHP (t ) =  HP +  HP RS &P500 (t ) + eHP (t )

Table 8.1 Excel Output: Regression Statistics
for the SCL of Hewlett-Packard
 Table 8.1 Interpreting the Output

• Correlation of HP with the S&P 500 is 0.7238

• The model explains about 52% of the variation
in HP
• HP’s alpha is 0.86% per month (10.32%
annually) but it is not statistically significant
• HP’s beta is 2.0348, but the 95% confidence
interval is 1.43 to 2.53
 Example

The index model has been estimated for stocks A and B with the following
results:

RA = 0.03 + 0.7RM + eA.

RB = 0.01 + 0.9RM + eB.
σM = 0.35; σ(eA) = 0.20; σ(eB) = 0.10.

The covariance between the returns on stocks A and B is

64