GRA6515 Quantitative Methods for Finance

Practice Exercises: Sampling and Estimation

1. Identify each of the following groups as a population or a sample. If the group is a sample,
identify the population to which the sample is related.
a. The S&P MidCap 400 Index viewed as representing US stocks with market capital-
ization falling within a stated range.
The S&P Mid Cap 400 Index represents a sample of all US stocks in the mid-cap or
medium capitalization range. The related population is “all US mid-cap stocks.”
b. UK shares that traded on 11 August 2003 and that also closed above L100/share as
of the close of the London Stock Exchange on that day.
The statement tells us to enumerate all members of a group and is sufficiently precise to
allow us to do that. The statement defines a population.
c. Marsh & McLennan Companies, Inc. (NYSE: MMC) and AON Corporation (NYSE:
AOC). This group is part of Standard & Poor’s Insurance Brokers Index.
The two companies constitute a sample of US insurance brokers. The related population
is “US insurance brokers.”
d. The set of 31 estimates for Microsoft EPS for fiscal year 2003, as of the 4 June 2003
date of a First Call/Thomson Financial report.
The statement defines a population. The 31 estimates for Microsoft EPS are the population of
publicly available US analyst estimates of Microsoft’s FY2003 EPS, as of the report’s date.
2. As reported by Liang (1999), US equity funds in three style categories had the following
mean monthly returns, standard deviations of return, and Sharpe ratios during the period
January 1994 to December 1996:

January 1994 to December 1996

Strategy Mean Return (%) Standard Deviation (%) Sharpe Ratio
Large-cap growth 1.15 2.89 0.26
Large-cap value 1.08 2.20 0.31
Large-cap blend 1.07 2.38 0.28

Assume that fund returns are normally distributed. Basing your estimate of future-period
monthly return parameters on the sample mean and standard deviation for the period
January 1994 to December 1996,
a. construct a 95 percent confidence interval for the monthly return on each of the funds.
Hint: you are looking for an interval that would contain 95% of all observations;
b. for each fund, calculate the probability that a fund will earn a monthly return of 0
percent or less.
c. What is the assumed risk-free rate?

a. Using the properties of the normal distribution: 95% of observations lie within the interval
µ  1.96σ.



b. P pR ¤ 0q  P

R µR
¤
µσ  N
µσ
R R
σR R R

95% confidence interval P pR¤ 0q

Large-cap growth 1.15  1.96  2.89  r
4.51; 6.81s N
2.89
1.15
  0.3453
Large-cap value 1.08  1.96  2.20  r
3.23; 5.39s N
2.20  0.31174
1.08

Large-cap blend 1.07  1.96  2.38  r
3.59; 5.73s N
1.07

2.38  0.3265

1
 c. Using one of the Sharpe ratios, compute the risk free rate: 0.26 
or r  1.15

1.15 r
2.89 ,
0.26  2.89  0.4%.

3. The table below presents the annual total return on the MSCI Germany Index and the
MSCI World Index. Assume that during the period 2003-2016 the reference risk free rate
had a mean annual return of 4.1 percent.

Year MSCI Germany MSCI World

2003 63.80 33.11
2004 16.17 14.72
2005 9.92 9.49
2006 35.99 20.07
2007 35.21 9.04
2008 -45.87 -40.71
2009 25.15 29.99
2010 8.44 11.76
2011 -18.08 -5.54
2012 30.90 15.83
2013 31.37 26.68
2014 -10.36 4.94
2015 -1.89 -0.87
2016 2.75 7.51

a. Create a frequency distribution for MSCI Germany Index with five equally spaced
classes (round up at the second decimal place in computing the width of class inter-
vals). Then calculate the absolute and the relative frequency of the data.

Questions b.-f. to be answered for both indexes.

b. To describe the central tendency of the distributions, calculate the sample mean and
the median return.
c. To describe the compound rate of growth of the indexes, calculate the geometric mean
returns.
d. Calculate the 30th percentiles.
e. Calculate the sample standard deviation and the semideviation. Explain the difference
between the two measures.
f. Calculate the skewness and excess kurtosis. How do these values characterize the
distribution of the indexes?

g. Calculate the coefficient of variation for both indexes, as well as for the 60/40 portfolio
(60 percent invested in MSCI Germany and 40 percent invested in MSCI World
Index).
i. Calculate the Sharpe ratio for both indexes, as well as for the 60/40 portfolio.

a. Given the maximum value of 63.80% and the minimum value of
45.87, we can compute
the range and then the length of each interval 63.80
p

5
45.87q
 21.93.

2
 Frequency distribution:
Frequency
absolute relative
Interval r
45.87;
23.94q
1 1 0.07
Interval r
23.94;
2.01q
2 2 0.14
Interval 3r
2.01; 19.92q 5 0.36
Interval 4r19.92; 41.85q 5 0.36
Interval 5r41.85; 63.80s 1 0.36

X  13.11%, X  9.72%. For even number of observations: M edianG  9.92% 216.17% 

G W
b.
13.045%; M edianW  9.49% 211.76%  10.625%
d
±
 p1 Ri q
1  9.57%, X G  7.96%
G n W
c. XG n


i 1

d. First compute the location of the 30th percentile: L30  p14 1q 100
30
 4.5. Hence, the
required value is in the middle between values Xp4q and Xp5q :
G
P30 
1.89 2.75
 0.43%
  6.225%
2
W 4.94 7.51
P30 2
d
°

i1pXi
X q  
n
1
e. The sample standard deviations are: sG n 1
2 27.39%, sW

18.26%. c °
The sample semideviaiton: sG
 1
n 1
pXi
X q2  20.3%, sW
 14.98%.
Xi X¤
Important: when we compute the semideviation we divide by the total number of obser-
vations minus one. While the standard deviation measures the dispersion of all outcomes,
the semideviation focuses on outcomes below the mean, that is, on downside risk.
°
f. The sample skewness: skewG  pn
1qpn n
2q pXs
X q 
0.378, skewW 
1.556.
i
3

° pX
X q
3

npn 1q

pn3
pn2qp
n1q
3q  0.67,
4

n 2
i
The sample excess kurtosis: KurtG
E pn
1qpn
2qpn
3q s
i 1
4

KurtWE  3.97.
The distributions of both indexes are negatively skewed, that is, they are subject to
extreme negative realizations. The positive kurtosis of both indexes indicates that the
distributions are more peaked than normal. To compare, the distribution of MSCI Germany
index is closer to the normal compared to MSCI World.
g./i. Compute the mean return and the standard deviation of the 60/40-portfolio:

µptf  0.6  0.1311 0.4  0.0972  0.1175

σptf  0.232

Coefficient Sharpe ratio

of variation
MSCI Germany 2.09

0.1311 0.041
 0.33
MSCI World 1.88

0.2739
0.0972 0.041
 0.31
60/40-portfolio 1.97

0.1826
0.1175 0.041
 0.33
0.232

4. Ten analysts have given the following fiscal-year earnings forecasts for a stock:

3
 Forecast (Xi ) Number of Analysts (ni )
1.40 1
1.43 1
1.44 3
1.45 2
1.47 1
1.48 1
1.50 1

Assume that the analyst forecasts are normally distributed.

° nand
a. What are the mean forecast standard deviation of forecasts?
The mean forecast is X  ° ni  1.45.
i Xi

2
The variance of the forecast s2
?  ° n1
1 ° ni Xi
X
 0.000778, then the standard
°
deviation is s  0.000778  0.02789. Let denote n  ni  10.
i

b.
?
Provide a 95%-confidence interval for the population mean of the forecasts.
The confidence interval can be computed as X  tα{2 pn
1qs{ n, where we use the
t-distribution with n
1 degrees of freedom for the reliability factor, since the variance
is unknown and the sample size is small. Plugging in the values, we get
1.45  2.262  0.02789
?  r1.43005; 1.46995s.
10

5. The table below gives the deviations of a hypothetical portfolio’s annual total returns
(gross of fees) from its benchmark’s annual returns, for a 12-year period ending in 2003.

Portfolio’s Deviations from Benchmark Return,

1992–2003 (%)
1992 -7.14
1993 1.62
1994 2.48
1995 -2.59
1996 9.37
1997 -0.55
1998 -0.89
1999 -9.19
2000 -5.11
2001 -0.49
2002 6.84
2003 3.04

a. Calculate the frequency, cumulative frequency, relative frequency, and cumulative

relative frequency for the portfolio’s deviations from benchmark return, by grouping
the data into 4 equal intervals.

Cumulative Relative Cumulative Relative

Return Interval Frequency Frequency Frequency (%) Frequency (%)
-9.19% ¤ A -4.55% 3 3 25.00 25.00
-4.55% ¤ B 0.09% 4 7 33.33 58.33
0.09% ¤ C 4.73% 3 10 25.00 83.33
4.73% ¤ D ¤ 9.37% 2 12 16.67 100.00

b. Construct a histogram using the data.

4
 c. Tracking risk is the standard deviation of the deviation of a portfolio’s gross-of-fees
total returns from benchmark return. Calculate the tracking risk of the portfolio,
stated in percent.
The portfolio’s tracking risk was 5.41 percent a year which corresponds to the sample
standard deviation of the data in the table.

6. The table below gives the annual total returns on the MSCI Germany Index from 1993 to
2002. The returns are in the local currency.

MSCI Germany Index Total Returns, 1993–2002

Year Return (%)
1993 46.21
1994 -6.18
1995 8.04
1996 22.87
1997 45.90
1998 20.32
1999 41.20
2000 -9.53
2001 -17.75
2002 -43.06

a. To describe the distribution of observations, create a frequency distribution with five

equally spaced classes (round up at the second decimal place in computing the width
of class intervals).
b. Calculate the cumulative frequency of the data. Calculate the relative frequency and
cumulative relative frequency of the data.

Cumulative Relative Cumulative Relative

Return Interval Frequency Frequency Frequency (%) Frequency (%)
-43.06 ¤ R -25.20 1 1 10 10
-25.20 ¤ R -7.34 2 3 20 30
-7.34 ¤ R 10.52 2 5 20 50
10.52 ¤ R 28.38.52 2 7 20 70
28.38 ¤ R ¤ 46.24 3 10 30 100

c. State whether the frequency distribution is symmetric or asymmetric. If the distri-

bution is asymmetric, characterize the nature of the asymmetry.
The last interval in the frequency distribution contains 30 percent of the observations,
whereas the first interval contains 10 percent of the observations. The middle three
intervals each contain 20 percent of the observations. The last two intervals thus contain
50 percent of the observations. The distribution is asymmetric. With most observations
somewhat concentrated to the right but with one extreme negative observation (in the
first interval), we conclude that the distribution is negatively skewed.
d. To describe the central tendency of the distribution, calculate the sample mean return.
10.802
e. Calculate the median return. 14.18
f. Identify the modal interval (or intervals) of the grouped returns. The modal return
interval: from 28.38 percent to 46.24 percent.

5
 g. To describe the compound rate of growth of the MSCI Germany Index, calculate the
geometric mean return. 6.70 percent a year.
h. To describe the values at which certain returns fall, calculate the 30th percentile.
The estimate of the 30th percentile is P30 
9.53 0.3r
6.18
p
9.53qs 
9.53
0.3p3.35q 
9.53 1.005 
8.53.
i. To describe the dispersion of the distribution, calculate the range, the mean absolute
deviation (MAD), the sample variance, the sample standard deviation, the semivari-
ance, and the semideviation. 89.27, 24.50, 896.844711, 29.95, 491.6943, 22.17.
j. To describe the degree to which the distribution may depart from normality, calculate
the skewness. Explain the finding for skewness in terms of the location of the median
and mean returns. -0.39. In the sample period, the returns on the MSCI Germany Index
were slightly negatively skewed. For a negatively skewed distribution, the median is greater
than the arithmetic mean. In our sample, the median return of 14.18 percent is greater
than the mean return of 10.80.
k. Calculate excess kurtosis. -0.73
l. Contrast the distribution of annual returns on the MSCI Germany Index to a normal
distribution model for returns. In contrast to a normal distribution, the distribution of
returns on the MSCI Germany Index is somewhat asymmetric in direction of negative
skew and is somewhat platykurtic (less peaked).

7. Which of the following is a random variable?

a. The population mean

b. The population size, N
c. The sample size, n
d. The sample mean
e. The variance of the sample mean
f. The largest value in the sample
g. The population variance
h. The estimated variance of the sample mean

d,f,h

8. How would you respond to a friend who asks you, “How can we say that the sample mean
is a random variable when it is just a number, like the population mean? For example, if
a simple random sample of size 50 produced X = 938.5; how can the number 938.5 be a
random variable?” It’s random because it is computed from a random sample.

9. Two populations are surveyed with simple random samples. A sample of size n1 is used
for population I, which has a population standard deviation σ1 ; a sample of size n2  2n1
is used for population II, which has a population standard deviation σ2  2σ1 . Ignoring
finite population corrections, in which of the two samples would you expect the estimate
of the population mean to be more accurate?
Hint: use formula for the standard error of the sample mean.

10. (*) Suppose that a simple random sample is used to estimate the proportion of families
in a certain area that are living below the poverty level. If this proportion is roughly 0.15,
what sample size is necessary so that the standard error of the estimate is 0.02? n = 319

6
11. This problem introduces the concept of a one-sided confidence interval. Using the central
limit theorem, how should the constant k be chosen so that the interval p
8, X ksX q
is a 90% confidence interval for µ — i.e., so that Ppµ ¤ X ksX q  0.9? This is called
a one-sided confidence interval. How should k be chosen so that pX
ksX , 8q is 95%
one-sided confidence interval?
1.28, 1.645
12. An investigator quantifies her uncertainty about the estimate of a population mean by
reporting X  sX . What size confidence interval is this? N
1 p1
α{2q  1, α  0.3176,
68% confidence interval
13. A population consists of three strata with N1  N2  1000 and N3  500. A stratified
random sample with 10 observations in each stratum yields the following data:

Stratum 1 94 99 106 106 101 102 122 104 97 97

Stratum 2 183 183 179 211 178 179 192 192 201 177
Stratum 3 343 302 286 317 289 284 357 288 314 276

Estimate the population mean and give a 90% confidence interval.

14. (*) Using the notation from the lecture the optimal allocation of stratified sampling is
defined as follows. If resources of a survey allow only a total of n units to be sampled, the
optimal allocation specifies how to choose n1 , . . . , nL to minimize V arpXsq subject to the
constraint n1    nL  n. It can be shown that

nl  n °LWl σl , l  1, . . . , L (1)
Wk σk
k 1
where Wl  Nl {N . The following table (Cochran 1977) shows the stratification of all farms
in a county by farm size and the mean and standard deviation of the number of acres of
corn in each stratum.
Farm Size Nl µl σl
0–40 394 5.4 8.3
41–80 461 16.3 13.3
81–120 391 24.3 15.1
121–160 334 34.5 19.8
161–200 169 42.1 24.5
201–240 113 50.1 26.0
241+ 148 63.8 35.2

a. For a sample size of 100 farms, compute the sample sizes from each stratum for
proportional and optimal allocation, and compare them.
b. Calculate the variances of the sample mean for each allocation and compare them to
each other and to the variance of an estimate formed from simple random sampling.
c. What are the population mean and variance?

a. For optimal allocation, the sample sizes are 10, 18, 17, 19, 12, 9, 15. For proportional
allocation they are 20, 23, 19, 17, 8, 6, 7.
b. V arpX opt q  2.90, V arpX prop q  3.4, V arpXsimple q  6.2
15. (*) The designer of a sample survey stratifies a population into two strata, H and L. H
contains 100,000 people, and L contains 500,000. He decides to allocate 100 samples to
stratum H and 200 to stratum L, taking a simple random sample in each stratum.

7
 a. How should the designer estimate the population mean?
b. Suppose that the population standard deviation in stratum H is 20 and the standard
deviation in stratum L is 10. What will be the standard error of his estimate?
c. Would it be better to allocate 200 samples to stratum H and 100 to stratum L?
d. Would it be better to use proportional allocation?

a. 16 X H 5
6 XL
b. 0.68
c. No, the standard error would be 0.87.
d. No, the standard error would be 0.71.