Statistics II: Prof. Christos Agiakloglou

Statistics II
Prof. Christos Agiakloglou

The Weighted Mean
• The concept of Weighted Mean arises from portfolio
analysis.
• The weighted mean assigns different weights to each
observation, whereas the arithmetic mean assigns equal
weights (1/N or 1/n) to each observation.
• The weighted mean ( X W ) is computed as:
n
X W   wi X i
i 1
n
where w
i 1
i 1
• Note that if wi = 1/n for all i, then the weighted mean is

equal to the arithmetic mean.
The Weights
• Can be given, i.e.,
– 20% is invested in Government bonds,
– 50% is invested in the stock market and
– 30% in invested in foreign markets.
• Can be obtained from the data, i.e., if we are dealing

with grouped data.
– Relative frequency.
– Sampling distribution of the variable.
• Can be obtained from a well know distribution.

Population – Probabilities.
Example of given weights
• A portfolio manager allocates 60% of the portfolio to
Canadian stocks and 40% to Canadian bonds. The returns
for last year were (-1.8%) from equities and (6.2%) from
bonds.
• The weighted mean portfolio return is 1.4%, i.e.,:
(0.60)(-1.8) + (0.40)(6.2) = -1.08 + 2.48 =
1.4
• If the weights were equal, then the arithmetic mean
portfolio return would have been 2.3%, i.e.,
(0.50)(-1.8) + (0.50)(6.2) = -0.9 + 3.1 = 2.2
Summarizing Data – Grouped Data
• Sometimes, especially when we are dealing with large data sets,
it is better to summarize the data and make the computations
based on the new data set which is called grouped data.
• Data which have been arranged in groups or classes rather than
showing all the original figures.
• For example, think of monthly stock returns, say for two years.
Of course in this case the sample is too small but still it is a
good example.
• We define returns as:
Pt  Pt 1
Rt 
Pt 1
• Note Pt and Pt-1 are the closing stock price at month t and t-1.
We assume no dividend.
Example -- Monthly Stock Returns
Jan-04 -4.5% Jan-05 0.3
Feb-04 -3.1 Feb-05 0.7
Mar-04 1.8 Mar-05 0.9
Apr-04 2.2 Apr-05 1.4
May-04 2.6 May-05 -0.8
Jun-04 -0.4 Jun-05 -1.6
Jul-04 -0.6 Jul-05 1.1
Aug-04 0.8 Aug-05 2.4
Sept-04 1.5 Sept-05 2.9
Oct-04 -2.4 Oct-05 3.6
Nov-04 -3.4 Nov-05 4.7
Dec-04 -1.4 Dec-05 2.3
Summarizing Data
• Short the data in an ascending order and define min and max
value, i.e., min = -4.5 and max = 4.7.
• Compute the range, i.e., Range = 4.7 – (-4.5) = 9.2
• Decide on the number of equally spaced intervals, i.e., k = 5.
• Determine the interval width and define the intervals, i.e., (-
5, -3), (-3, -1), (-1, 1), (1, 3), (3, 5)
• Count the number of observations in each interval and
construct the frequency distribution.
• Then, the number of observations in each interval is called the
absolute frequency or simply frequency or (fi).
• Note that the sum of frequencies equals to the sample size, i.e,
n
f
i 1
i n
Frequency Distribution
Interval Frequency
(fi)
-5 to -3 3
-3 to -1 3
-1 to 1 7
1 to 3 9
3 to 5 2
Total 24
Cumulative Frequency, Relative Frequency,
Cumulative Relative Frequency
• Cumulative frequency (cfi) is the sum of frequencies in that

and previous intervals.
• Relative frequency (rfi) is the absolute frequency of each

interval divided by the total number of observations, i.e.,
and n
• rf i  f i n  rf i  1
Cumulative relative frequency (crf ) is the sum of relative
i 1
i
frequencies in that and previous intervals.
Relative Frequency, Cumulative
Frequency, Cumulative Relative Frequency
Cumulative Relative Cum. Rel.
Interval Frequency Frequency Frequency Frequency
fi cfi rfi crfi
-5 to -3 3 3 0.125 0.125
-3 to -1 3 6 0.125 0.250
-1 to 1 7 13 0.292 0.542
1 to 3 9 22 0.375 0.917
3 to 5 2 24 0.083 1.000
Total 24 1
The Mean of Grouped Data
• The mean of grouped data is the weighted mean, where the
weights (wi) are the relative frequencies (rfi).
• For population
1 N N

N
 f m   rf m
i 1
i i
i 1
i i
• For Sample 1 n n
X   f i mi   rf i mi
n i 1 i 1
where mi is the midpoint of each interval.

The Mean Return
Relative
Interval Midpoint Frequency Frequency
mi fi rfi rfimi
-5 to -3 -4 3 0.125 -0.50
-3 to -1 -2 3 0.125 -0.25
-1 to 1 0 7 0.292 0
1 to 3 2 9 0.375 0.75
3 to 5 4 2 0.083 0.332
Total 24 1 0.332
Mean return =
0.332%
The Variance of Grouped Data
• For population
1 N N
2 
N
 f (m
i 1
i i   ) 2   rf i (mi   ) 2
i 1
• For Sample 1 n
s2  
n  1 i 1
f i ( mi  X ) 2
where mi is the midpoint of each interval.

The Variance of Returns
Interval Midpoint Frequency

mi fi (mi – mean) fi (mi –mean)2
-5 to -3 -4 3 -4.332 56.30
-3 to -1 -2 3 -2.332 16.32
-1 to 1 0 7 -0.332 0.77
1 to 3 2 9 1.668 25.04
3 to 5 4 2 3.668 26.91
Total 24 125.34
Variance = 125.34/23 = 5.45 and standard deviation = 2.33%

Histogram
A Histogram is a bar chart graphical presentation of the data.
Histogram of Monthy Returns
10
8
F req u en cy
0
(-5, -3) (-3, -1) (-1, 1) (1, 3) (3, 5)
Return intervals

Statistics II: Prof. Christos Agiakloglou

Uploaded by

Copyright:

Available Formats

Statistics II: Prof. Christos Agiakloglou

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Statistics II: Prof. Christos Agiakloglou

Uploaded by

Copyright:

Available Formats

Statistics II

Prof. Christos Agiakloglou

• Note that if wi = 1/n for all i, then the weighted mean is

• Can be obtained from the data, i.e., if we are dealing

• Can be obtained from a well know distribution.

• Cumulative frequency (cfi) is the sum of frequencies in that

• Relative frequency (rfi) is the absolute frequency of each

where mi is the midpoint of each interval.

where mi is the midpoint of each interval.

Interval Midpoint Frequency

Variance = 125.34/23 = 5.45 and standard deviation = 2.33%

Histogram of Monthy Returns

You might also like