Sei sulla pagina 1di 22

SOME STATISTICAL CONCEPTS

Chapter 3

Distributions of Data
Probability Distribution

Expected Rate of Return


Variance of Returns
Standard Deviation
Covariance
Correlation Coefficient

Coefficient of Determination

Historical Distributions
Various Statistics

Relationship Between a Stock and the Market


Portfolio
The Characteristic Line
Residual Variance

DISTRIBUTIONS OF DATA

When evaluating security and portfolio returns, the


analyst may be confronted with:
1. possible returns in some future time period
(probability distributions of possible future
returns), or
2. past returns over some historical time period
(sample distribution of past returns).
The same statistics may be used to describe both
types of distributions (probability and sample). For
each type of distribution, however, the procedures
for calculating the various statistics vary
somewhat.
In the following examples, statistics are discussed
first with respect to probability distributions, and
then with respect to sample distributions of
historical returns.

PROBABILITY DISTRIBUTION
(Evaluating Possible Future Returns)
Probability
(hi)
_________
.05
.10
.20
.30
.20
.10
.05

Possible Return (%)


(ri)
_________
-20
-10
5
30
55
70
80

PROBABILITY DISTRIBUTION
(Continued)

Probability
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
-20

-10

30

55

Possible Return (%)

70

80

Expected Rate of Return (Best Guess)


n

E(r) h i r i
i 1

E(r) = .05(-20) + .10(-10) + .20(5) + .30(30)


+ .20(55) + .10(70) + .05(80)
= 30%

Variance of Returns (Potential for deviation of the


return from its expected value)

(r) h i [ri E(r)]


2

i 1

2(r) = .05(-20 -30)2 + .10(-10 -30)2 + .20(5 -30)2


+ .30(30 -30)2 + .20(55 -30)2 + .10(70 -30)2
+ .05(80 -30)2
= 820
Standard Deviation

(r)

(r)

(r)

820 28.64%

Covariance (A measure of the interrelationship


between securities)
A positive number indicates positive correlation.
A negative number indicates negative
correlation. A value of zero indicates zero
correlation.

Covariance - An Example:

Joint
Probability (hi)
_____________
.10
.20
.40
.20
.10

Possible Returns (%)


Stock A (rA,i)
Stock B (rB,i)
__________
__________
5
10
10
20
20
40
40
50
70
60
n

Cov(rA , rB ) h i [rA,i E(rA )][rB,i E(rB )]


i 1

Covariance - An Example (Continued)


E(rA) = .10(5) + .20(10) + .40(20) + .20(40) + .10(70)
= 25.5%
E(rB) = .10(10) + .20(20) + .40(40) + .20(50) + .10(60)
= 37.0%
Cov(rA,rB) = .10(5 - 25.5)(10 - 37)
+ .20(10 - 25.5)(20 - 37)
+ .40(20 - 25.5)(40 - 37)
+ .20(40 - 25.5)(50 - 37)
+ .10(70 - 25.5)(60 - 37)
= 241.50 (Positive Covariance)

Graphic Illustration of Positive Covariance


Return on Stock A
80
70
60
50
40
30
20
10
0
10

20

40

50

Return on Stock B

60

Correlation Coefficient [Ranges between +1.0


(perfect positive correlation) and -1.0 (perfect
negative correlation)].
Cov(r , r )
A B

A, B (r ) (r )
A
B
2 (r ) .10(5 25.5) 2 .20(10 25.5) 2 .40(20 25.5)2
A
+ .20(40 25.5) 2 .10(70 25.5) 2 342.25
(r ) 342.25 18.5%
A
2 (r ) .10(10 37) 2 .20(20 37)2 .40(40 37) 2
B
+ .20(50 37) 2 .10(60 37)2 221.0
(r )
B

221.0 14.87%

Cov(r , r )
241.5
A B

.878
A, B (r ) (r ) (18.5)(14.87)
A
B

Coefficient of Determination
Percentage of the variability in returns
on one investment that can be
associated with the returns on another
investment

2
A,B

(.878) .77 77%


2

HISTORICAL DISTRIBUTIONS
(Evaluating Past Returns)

Time Period
(e.g., month)
(t)
________
1
2
3
4
5

Percent Returns
Stock A
Stock B
(rA,t)
(rB,t)
________
________
5
10
10
5
5
15
20
20
40
5

Graph of Past Returns


Return on Stock A
45
40
35
30
25
20
15
10
5
0
10

15

20

Return on Stock B

Mean Return
r

r
t 1

n
rA (5 10 5 20 40) / 5 16%
rB (10 5 15 20 5) / 5 11%

Variance and Standard Deviation


n

2 (r)

2
(r

r
)
t
t 1

n 1
2 (rA ) [(5 16) 2 (10 16) 2 (5 16) 2
+ (20 16) 2 (40 16) 2 ] / 4 217.5
(rA )

217.5 14.75%

2 (rB ) [(10 11) 2 (5 11) 2 (15 11) 2


+ (20 11) 2 (5 11) 2 ] / 4 42.5
(rB )

42.5 6.52%

Covariance
n

Cov(rA , rB )

[(r

t 1

Cov(rA , rB )

rA )(rB,t rB )

n 1
(5 16)(10 11)
(10 16)(5 11)
(5 16)(15 11) / 4 26.25

(20 16)(20 11)


(40 16)(5 11)

Correlation Coefficient

A,B

A,t

Cov(rA , rB )
26.25

.27
(rA ) (rB ) (14.75)(6.52)

Coefficient of Determination

2A,B ( .27) 2 .073 7.3%

Relationship Between a Stock


and the Market Portfolio
Time Period
(e.g., month)
(t)
________
1
2
3
4
5

Percent Returns
Stock j
Market
(rj,t)
(rM,t)
________
________
-7
-10
6
5
15
25
9
15
22
30

Mean Returns
rj ( 7 6 15 9 22) / 5 9%

rM ( 10 5 25 15 30) / 5 13%

Variance and Standard Deviation


2 (rj ) [( 7 9) 2 (6 9) 2 (15 9) 2
+ (9 9) 2 (22 9) 2 ] / 4 117.5
(rj )

117.5 10.84%

2 (rM ) [( 10 13) 2 (5 13) 2 (25 13) 2


+ (15 13) 2 (30 13) 2 ] / 4 257.5
(rM )

257.5 16.05%

Covariance
( 7 9)( 10 13)
(6 9)(5 13)

Cov(rj , rM ) (15 9)(25 13) / 4 171.25

(9

9)(15

13)

(22 9)(30 13)

Correlation Coefficient

j,M

Cov(rj , rM )

171.25

.984
(rj ) (rM ) (10.84)(16.05)

The Characteristic Line


rj j jrM
where :
Cov(rj , rM ) 171.25 .665
j
2 (rM )
257.5
j rj jrM 9 (.665)(13) .355

The Characteristic Line for


Stock (j) and the Market (m)
Return on the Stock
25

Line passes through


The means of both
variables

20
15
10

When the Markets return is zero,


the stocks return is .355

5
0
-20

-10

0
-5
-10

10

20

30

40

Return on the Market

Residuals
Deviations from the characteristic line:

r
j
j,t rj,t (
j M, t )
1.
2.
3.
4.
5.

-7 - [.355 + .665(-10)] = - .705


6 - [.355 + .665( 5)] = + 2.32
15 - [.355 + .665(25)] = - 1.98
9 - [.355 + .665(15)] = - 1.33
22 - [.355 + .665(30)] = + 1.695

Residual Variance
Propensity to deviate from the line:
n

( j )
2

t 1

2
j,t

n 2

( j ) [( .705) (2.32) ( 1.98)


2

+ ( 1.33) (1.695) ] / 3
= 4.814
2

Potrebbero piacerti anche