Portfolio Theory

Dr. Vinod Kumar

 Correlation and Covariance
• Random variables: Whose possible values are numerical outcomes of a random
phenomenon.
• Mean
• Without probability distribution
• With probability distribution
• Variance
• Without probability distribution
• With probability distribution
• Finite sample correction
• Covariance of two variables
• Correlation coefficients:
• To measure how strong a relationship is between two variables.
• Between two variables
• Statistical measure and not a mathematical relationship
• Variance of sums of Random Variables
 Mean
• is probability function

• Mean: Discrete variable

• for discrete case
• E (x) = for continuous case (is just symbolic, chose the limit
from domain of x)
• Example
f (x) = X2/9 0≤x≤3
Ans. E(X) = 2.25
• Property of mean
• E(aX) = aE(X) a and b are constants,
• E(aX + b) = aE(X) + b a and b are constants,
• E(aX + bY) = aE(X) + bE(Y) a and b are constants,
• E(XY) = E(X)*E(Y) X and Y are independent random variables
• if g is function of X, discrete case
• E () = if g is function of X, continuous case
 Variance
• Variance: how closely or widely the individual X values are spread around their mean value.
• = E(X2) − [E(X)]2
Var if X is discrete
Var (x) = if X is continuous
• Example
x −2 1 2
f (x) 5/8 1/8 2/8
Variance = (29/8) − (−5/8)2 = 207/64 = 3.23.
• Properties of Variance
1. E(X − μ)2 = E(X2) − μ2, as noted before.
2. The variance of a constant is zero.
3. If a and b are constants, then var (aX + b) = a2 var (X)
4. If X and Y are independent random variables, then
var (X + Y) = var (X) + var (Y)
var (X − Y) = var (X) + var (Y)
5. If X and Y are independent and a and b are constants, then
var (aX + bY) = a2 var (X) + b2 var (Y)
 Covariance:
• Let X and Y be two rv’s with means μx and μy, respectively.
• Covariance (X, Y) = E{(X − μx)(Y − μy)} = E(XY) − μxμy
•
= for discrete variables
• Covar (X, Y) =
= for continuous variables

Properties of covariance:
• If X and Y are independent, their covariance is zero,
• cov (a + bX, c + dY) = bd cov (X, Y) where a, b, c, and d are constants.
 Variance and Covariances of Correlated Variables
• Correlation Coefficient (ρ) = cov (X, Y)/(σxσy) -1≤ ρ ≤+1
Example: Let Cov (X, Y) = 2.24
σx = 2.05 and σy = 1.50.
Therefore, ρ = 2.24/[(2.05)(1.50)] = 0.73.

• Let X and Y be two random variables. Then

var (X + Y) = var (X) + var (Y) + 2 cov (X, Y)
= var (X) + var (Y) + 2ρσxσy
var (X − Y) = var (X) + var (Y) − 2 cov (X, Y)
= var (X) + var (Y) − 2ρσxσy
 Characteristics of Individual Securities
• The characteristics of individual securities that are of interest are the:
1. Expected Return
2. Variance and Standard Deviation
3. Covariance and Correlation (to another security or index)
 Portfolio
• Collection of financial investments
• E.g. stocks, bonds, commodities, cash, cash equivalents, etc.
including closed-end finds, exchange traded funds (ETFs), real
estate, art, and private investments.

• We are focusing here on stock and bonds

 Portfolio returns

• Suppose there are n assets in a portfolio where expected return on ith asset is ȓi
• Further, the weight (% value) of ith asset is wi
• Expected portfolio return on a portfolio ȓp
ȓp = w1ȓ1 +w2ȓ2 +w3ȓ3…….+wnȓn
Share ȓi wi ȓi**wi
IRCTC 18% 40% 7.2%
Infosys 12% 50% 6.0%
HLL 10% 30% 3.0%
Borrowing 8% -20% -1.6%
ȓp 14.6%
 Portfolio Risk
• Risk is measured by standard deviation, variance, coefficient of variation, covariance
• Stand alone risk vs portfolio risk σX σY

• Correlation ρ -1 ≤ ρ ≤ 1

• or, = ρ

• New variance of portfolio after adding a stock:

= +2
= +2 ρ

• If ρ is negative, the variance in portfolio will decrease, even if expected return

increased.
• If ρ <1, the variance in portfolio will increase, but at decreasing rate.
 Portfolio Risk: Perfect Correlation
• If ρ = -1, it is called perfectly negative correlation (risk can be completely eliminated)
• All the risk are diversified away
• If ρ = +1, it is called perfectly positive correlation (no benefit of diversification)
• No risk is diversified away
ρ= -1   ρ= +1
Year
Stock X Stock Y Portfolio XY Stock X Stock Y Portfolio XY
2013 40% -10% 15% 40% 40% 40%
2014 -10% 40% 15% -10% -10% -10%
2015 35% -5% 15% 35% 35% 35%
2016 -5% 35% 15% -5% -5% -5%
2017 15% 15% 15% 15% 15% 15%
Average 15% 15% 15% 15% 15% 15%
Std. dev. 22.64% 22.64% 0.00% 22.64% 22.64% 22.64%
 Correlation and reduction in portfolio risk
• If ρ <1 diversification reduce risk of the portfolio (normally ρ is between +.5 to +0.7)
• = +2 ρ

• How to chose weight so that is minimised?

  = 0.16% = 1% Port. Variance for
WA WB RA RB RP ρ=1 ρ=0 ρ=-1
1 0 5% 8% 5% 4.00% 4.00% 4.00%
0.75 0.25 5% 8% 6% 5.50% 3.91% 0.50%
0.5 0.5 5% 8% 7% 7.00% 5.39% 3.00%
0.25 0.75 5% 8% 7% 8.50% 7.57% 6.50%
0 1 5% 8% 8% 10.00% 10.00% 10.00%

• What kind of curve do you see in Return - Std. Dev. Relationship?

 Portfolios with Various Correlations

return
100%
 = -1.0 stocks

 = 1.0
100%  = 0.2
bonds

• Relationship depends on correlation coefficient.
-1.0 < r < +1.0
• If r = +1.0, no risk reduction is possible.
• If r = –1.0, complete risk reduction is possible.
 Portfolio Risk: Diversification and Correlation

• Within the two limits of correlation, there are partial diversification benefits

• In absence of perfect negative correlation, with increasing diversification the

risk reduces, but up to a certain limit.
• Some risk can be eliminated through diversification: diversifiable risk
• Some risk can not be diversified away: non diversifiable risk or Market risk

• A diversified portfolio may have substantial amount of risk but not as much as
in standalone investment

• If we include all the assets in the portfolio then it is called market portfolio
• The observed variance in market portfolio is about 20%.
Portfolio Risk and Number of Stocks
In a large portfolio the variance terms are effectively
diversified away, but the covariance terms are not.

Diversifiable Risk;
Nonsystematic Risk;
Firm Specific Risk;
 Unique Risk
Portfolio risk
Nondiversifiable risk;
Systematic Risk;
Market Risk
n
Total Risk
• Total risk = systematic risk + unsystematic risk

• Standard deviation of returns is a measure of total risk.

• For well-diversified portfolios, unsystematic risk is very small.

• Consequently, the total risk for a diversified portfolio is essentially

equivalent to the systematic risk.
 More about risk

• Diversifiable risk is idiosyncratic risk (company specific) such as lawsuit,

strikes, successful and unsuccessful marketing programs, the winning or
losing of a major contract, etc.

• Market risk is risk related to common economywide events, that

systematically affects most of the firms in the economy, such as war,
inflation, recessions, high interest rate etc.
• Since most stocks are negatively affected by these events, diversification can
not eliminate this risk.

• Relevant risk of an asset is its contribution to the risk of a well diversified

portfolio.
 “Minimum risk” Portfolio: 2 asset case

= +2 ρ

For minimum differentiating against and setting equal to 0, we have

 Return and Risk: N asset case

Matrix Form:
 Optimal Portfolio with risk free asset

= +2 ρ

Let us have B as risk free asset, then For, = 0

ȓp = (1-wA) rf +wA rA

ȓp = rf + wA(rA - rF )
 Market Portfolio: Efficient Frontier
The Efficient Set for Many Securities

return Individual
Assets

P
Consider a world with many risky assets; we can still identify the
opportunity set of risk-return combinations of various portfolios.
 The Efficient Set for Many Securities

return
tie r
fr o n
nt
ffi cie
e
minimum
variance
portfolio

Individual Assets

P

The section of the opportunity set above the minimum

variance portfolio is the efficient frontier.
 Optimal Portfolio with a Risk-Free Asset

return
100%
stocks

rf
100%
bonds

Now consider a world that also has risk-free securities like T-bills in
addition to stocks and bonds,
 Riskless Borrowing and Lending – I
=
L ȓp = (1-wA) rf +wA rA

return
CM 100% ȓp = rf + wA(rA - rF )
stocks
Balanced  ȓp is an equation of line with
fund
intercept rf and slope (rA - rF )

rf  and ȓp both increase linearly

100% with
bonds


Now investors can allocate their money across the T-bills and an equity portfolio on
efficient frontier.
Any point on the CML Line is achievable
Remember =
Riskless Borrowing and Lending – II
L

return
CM efficient frontier

rf
P

With a risk-free asset available and the efficient frontier identified, we

choose the capital allocation line with the steepest slope.
Market Equilibrium – I

return
L
CM efficient frontier

rf

P

With the capital allocation line identified, all investors choose a point along the
line—some combination of the risk-free asset and the market portfolio M.
In a world with homogeneous expectations, M is the same for all investors.
Market Equilibrium – II

return
CM 100%
stocks
Balanced
fund

rf
100%
bonds

Where the investor chooses along the Capital Market Line depends on her risk
tolerance. The big point is that all investors have the same CML.
 Measuring market risk in context of Market Portfolio

• Researchers have shown that the best measure of the risk of a security
in a large portfolio is the beta (b) of the security.
• Beta measures the responsiveness of a security to movements in the
market portfolio (i.e., systematic risk).

Cov(Ri,RM )
i  2
 (RM )