Académique Documents
Professionnel Documents
Culture Documents
Introduction
The reason for portfolio theory mathematics:
Linear Combinations
A portfolios performance is the result of the performance of its components. When
an investor distributes money across a handful of securities:
Return
The expected return of a portfolio is a weighted average of the expected returns of
the components:
E ( R p )
i 1
x i E ( R i )
(1 .1 )
xi 1
i 1
return of 0.
Example
Assume the following statistics for Stock A and Stock B:
Stock A
Stock B
Expected return
.015
.020
Variance
.050
.060
Standard deviation
.224
.245
Weight
40%
60%
Correlation coefficient
.50
Table 1
What is the expected return of this two-security portfolio?
E ( Rp ) xi E ( Ri )
i 1
x A E ( RA ) xB E ( RB )
0.4(0.015) 0.6(0.020)
0.018 1.80%
Variance
Understanding portfolio variance is the essence of understanding the mathematics of
diversification. The variance of a linear combination of random variables is not a
weighted average of the component variances.
2p xi xj iji j
(1.2)
i 1 j 1
2p x A2 A2 xB2 B2 2 xA xB AB A B
(.4) 2 (.05) (.6) 2 (.06) 2(.4)(.6)(.5)(.224)(.245)
.0080 .0216 .0132
.0428
Minimum Variance Portfolio
The minimum variance portfolio is the particular combination of securities that will
result in the least possible variance. Solving for the minimum variance portfolio
requires basic calculus
B2 A B AB
xA 2
A B2 2 A B AB
(1.4)
xB 1 x A
3
Example (contd)
Assume the same statistics for Stocks A and B as in the previous example. What are
the weights of the minimum variance portfolio in this case?
Solution: The weights of the minimum variance portfolios in this case are:
xA
B2 A B AB
.06 (.224)(.245)(.5)
59.07%
A2 B2 2 A B AB .05 .06 2(.224)(.245)(.5)
xB 1 xA 1 .5907 40.93%
reduction.
p2 xi x j ij i j
(1.2)
i 1 j 1
From a portfolio construction perspective, the important feature of this equation lies
in the fact that it includes the correlation coefficient between all pairs of securities in
the portfolio. Because ij i j COV (ij ) , equation (1.2) can be expressed in either
covariance or correlation terms, and it is common to see it either way.
n 2.
For a five-security
portfolio, (25 5)/2 = 10 covariances are necessary, plus the five variances. You
can convert a covariance matrix into a correlation matrix by dividing each covariance
by the product of the two standard deviations. Any portfolio construction technique
using the full covariance matrix is called a Markowitz model
Single-Index Model
While the Markowitz model makes use of the full covariance matrix with all its
pairwise statistics, the single-index model simplifies matters by comparing all
securities to a single benchmark.
Computational Advantages
A fifty-security portfolio requires the calculation of 1,225 covariances to forecast
portfolio variance. Thousands of stocks exist, and a pairwise comparison of them all
would be a very unwieldy task. Instead of comparing a security with each of the
others, why not compare each of the securities with a single benchmark?
By
observing how two independent securities behave relative to a third value, we learn
something about how the securities are likely to behave relative to each other.
5
This is one of the merits of beta. Beta is a measure of how a security moves relative
to overall market movements and that linear regression is a convenient way to
estimate this statistic. If beta is greater than one, the security tends to show price
swings greater than the market average. A beta less than one means that the
security moves less than average.
A securitys beta is a function of the market returns and the securitys covariance
with them. See equation (1.5):
COV ( Ri , Rm )
i
(1.5)
m2
where R return on the market index
m
n
n
xi i m2 xi2 ei2
i 1
i 1
2 2
2
p m ep
Variance of a Portfolio:
(1.7)
2
p
p2 m2
Variance of a Portfolio Component: i2 i2 m2 ei2 (1.8)
Equation (1.7) shows that if we know the variance of return on the market index and
the betas of the portfolio components, we can calculate the portfolio variance. The
last term in equation (1.7) reflects the unsystematic risk in the portfolio; this term
approaches zero as the number of portfolio components increases. Unsystematic
risk is the variance of the error term in the market model. For all practical purposes,
this term is small after the portfolio size reaches about twenty securities.
The last term of the equation indicates that some of the risk of a
security is company specific and unrelated to market movements. This is the type of
risk that diversification seeks to reduce. The first term on the right-hand side of
equation (1.8) measures systematic risk, whereas the second term measures
unsystematic risk.
Multi-Index Model
Although beta is a very useful statistic in the construction of portfolios and in security
analysis, it does not completely explain why security prices change.
In fact,
Ri ai im Im i1I1 i 2 I2 ... in In
(1.10)
where ai constant
Im return on the market index
I return on an industry index
j
Although the idea of a multi-index model employing industry (or other) effects is
appealing, financial research has not uncovered much evidence that such models
are particularly useful in forecasting security price behaviour. There is ongoing work
in this area, but the traditional market model, with its single index, continues to be
the method of choice for the bulk of current portfolio management purposes.