UNIT-2

TIME SERIES AND INDEX NUMBER

TIME SERIES ANALYSIS

Time seriesanalysis refers to such a series in which statistical data are presented on the basis
of time of occurrence or in a chronological order. The measurement of time may be year,
month, week, day, hour or even minutes or seconds.
Definition—
1) “A set of data depending on the time is called a time series.”- Kenny &Kieping
2) “A time series is a sequence of values of the same variate corresponding to successive
point in time.’- Werner Z. Hirsch

COMPONENTS OF A TIME SERIES

The main components of time series may be classified as shown below-
Original Data (O or Y)
A) Long term Movement or Trend or Secular Trend (T)
B) Short- Time Fluctuations
1) Regular short time fluctuation( S+C)
a) Seasonal Variation (S)
b) Cyclical Fluctuation (C)
2) Irregular or Random Fluctuations (I)
A) Long term Movement or Trend or Secular Trend (T)-Trend refers to that tendency
which indicates the general direction of fluctuation in a long period. According to
Simpson and Kafka, “Trend also called secular or long term trend, is the basic
tendency of production, sales, income, employment, or the like to grow or decline
over a period of time. The concept of trend does not include short range oscillations
but, rather, steady movements over a long time.”
B) 1) Regular short time fluctuation( S+C)- The variations arising out on account of
such regular or periodical repetitions are called regular short time oscillation, which
may of two type-
a) Seasonal Variation- Seasonal variations refer to such movements which are
regular and repetitive and which operate in a regular and periodic manner over a
span of less than a year. This span may be a day, week, month, half-year, etc.
b) Cyclical Fluctuation-It also occurs periodically like seasonal variations, but the
period of their reoccurrence is more than a year. It is called cyclical because they
occur in cyclical nature and in this cycle there are four stages- (i) Prosperity, (ii)
Recession, (iii) Depression, (iv) Recovery.
3) Irregular or Random Fluctuations (I)- Irregular or random fluctuations occur
accidently in time series. For instance, decline in profits due to break of fire in the
factory in a particular year, decrease in production due to sudden strike or scarcity
of petroleum product due to war, etc.
 MEASUREMENT OF TREND
The following are the four important method which are used in estimating the trend-
1) Free hand curve method
2) Semi-average method
3) Moving average method
4) Method of Least square
5) 1) Free hand curve method- This method is also called as curve fitting by
inspection. The procedure for knowing the trend by this method is as follows-
a) First of all the original value of a time series are plotted on a graph paper and a
historigram is obtained by joining these points.
b) After this a smooth curve is drawn through these points keeping in view the
direction of fluctuations is such a way that the curve represent the general
tendency of the data.
2) Semi-average method- The procedure of finding out trend by semi-average method is as
follows-
a) Division of time series in to two equal parts.
b) Calculation of average in both the part.
c) Plotting of original data on the graph.
d) Plotting of both semi averages.
e) Trend line- It is a straight line is drawn by joining the two points of semi-averages and this
line represent the trend of data.

3 ) Moving Average Method- It consists in obtaining a series of moving averages of

successive overlapping groups or sections of the time series. From the point of view of
calculation of moving averages, the question can be divided in two categories-
a) When Period is Odd- 3, 5, 7, 9 year moving average method
b) When Period is Even- 2, 4, 6 year moving average method

4) Method of Least Squares- This method is called the method of least squares because the
sum of squares of deviations of various points of trend line from original data would be the
least as compared to the sums of squares of the deviations obtained by using any other line.
Fitting a straight line trend by least Square Method-
Equation is YC= a+ b X

To Find the value of a and b we used following two equations-

𝚺Y= Na + b𝚺X
𝚺XY= a𝚺X + b 𝚺X2

Example-
Find trend by three yearly moving average method from the given data-
Years: 1960,61,62, 63, 64,65,66,67,68,69,70,71,72,73,74
Sales: 5, 7 , 9, 12, 11, 10 ,8, 12,13,17,19,14,13,12,15
Solution-
year sales 3-yearly moving sum Moving average
1960 5 - -
61 7 5+7+9=21 7
62 9 7+9+12=28 9.33
 63 12 9+12+11=32 10.67
64 11 12+11+10=33 10
65 10 11+10+8=29 9.67
66 8 10+8+12=30 10
67 12 8+12+13=33 11
68 13 12+13+17=42 14
69 17 13+17+19=49 16.33
70 19 17+19+14=50 16.67
71 14 19+14+13=46 15.33
72 13 14+13+12=39 13
73 12 13+12+15=40 13.33
74 15 - -

20

18

16

14

12

10 sales
8 trend value

0
1960 61 62 63 64 65 66 67 68 69 70 71 72 73 74

Example-2 Fit a straight line trend by the method of least square to the following time
series data-

Year - 1951 1952 1953 1954 1955

Price - 107 110 114 114 115
Show the above series and trend line on the graph paper.

Solution-
year Price,y x X2 xy Trend value- y = a+bx
1951 107 0 0 0 108+2*0 = 108
1952 110 1 1 110 108+2*1 = 110
1953 114 2 4 228 108 +2*2 =112
1954 114 3 9 342 108 +2*3 = 114
 1955 115 4 16 460 108 +2*4 = 116
2
Total N=5 ∑ y = 560 ∑x =10 ∑ x =30 ∑ xy = 1140

∑y=Na+b∑x ∑x y = a∑x + b ∑ x 2
560 =5a + 10 b ….. (1) 1140 = 10a + 30 b…….(2)

Multiply equation (1) by 2 and subtract it from equation (2)

1140 = 10a + 30 b
- 1120 = - 10 a +-20 b

20 = 10 b

b=2
substitute the value of b in equation -1
560 = 5a + 20
5a = 540
a = 108

118

116

114

112
price
110
trend value
108
Column1
106

104

102
1951 1952 1953 1954 1955

INDEX NUMBER
Index number is a specialized average through which charges in relative to time or
comparative form.
Definition-
1) Index number are used to measure the changes in some quantity which we cannot
observe directly.- A. L. Bowley
2) Index numbers are a series of numbers by which changes in the magnitude from time
to time or from place to place.- Secrist
3) Index number is a single ratio which measure the combined change of several
variables between two different time, place, situation.- A. M. Tuttle
Characteristics of Index Numbers-
1) Relative or comparative measures
2) Specialized average
 3) Measurement of common characteristics of a group of items.
4) Measurement of changes not capable of direct measurement.
5) Comparison on the basis on time or place.
6) Universal use.
Points to be considered in the construction of Index Number-
1) Purpose of Index Number
2) Selection of base year
3) Selection of representative items of commodities.
4) Selection of representative price
5) Problem of weighting
6) Choice of suitable average
7) Selection of appropriate formula
Method of base year-
A) Single year fixed base
B) Multi year average base
Single year fixed base- In this method any normal year is selected as base year. The price of
base year is denoted as P0 and the price of other year as P1 and index number or price relative
(PR).
Index Number or (PR) = P0/ P1
Multi year average base- When there is difficulty in selecting a particular year as a base
year, the average price of a few years is taken as base price and this average price expressed
as P0.
Chain base method- In this method price relative for every current year is calculated on the
basis of price of the immediately preceding year.

Method of constructing Index Number

A) Unweighted index number- It is also known as simple aggregative method. In this

method all commodities in base year and current year are added separately and they
are denoted as 𝚺P0 and 𝚺P1
Index Number= ∗ 100
B) Weighted index number- In this method appropriate weights are be assigned to
various commodities to reflect their relative importance. If these weights are base on
the actual quantity, the symbol q is used for it. The formula for the construction of the
index number.
Index Number= ∗ 100
There are some of the important methods to constructed Index number-
1) Laspeyre’s Method: In this method, weights are assigned by the quantities (q0) in
the base year. The formula is:
Index Number (P01) = ∗ 100
2) Paasche’s Method: In this method, weights are assigned by the quantities (q1) in
the current year. The formula is:
Index Number (P01) = ∗ 100
3) Marshall-Edge worth’s Method: In this formula, weights are given on the basis
of average of quantity of base year and that of current year. The formula is:
Index Number (P01) = ∗ 100
 4) Dorbish Bowley’s Method- This method is a combination of Laspeyre’s and
Paasche’s Method and we find out the arithmetic average of both these index
numbers. The formula is:
Index Number (P01) = + ∗ 100/2
5) Kelly’s Method: This formula, named after Truman L. Kelly, requires the
weights to be fixed for all periods and is also sometimes known as aggregative
index with fixed weights. These weights may be assigned on the basis of
quantities of base year or current year or average of both these years or of any
other year assumed as standardized year. The formula is :
Index Number (P01) = ∗ 100
6) Fisher’s Ideal Index Number: It was developed by Prof. Irvin Fisher. It is a
geometric mean of the Laspeyre’s and Paasche’s method. The formula is :
Index Number (P01) = ∗ * 100

Reversibility Tests-For an ideal index number it is necessary that it should satisfy

reversal test, which are as follows:

1) Time Reversibility test- Time reversal test provides that if the index number of
current year (P01) is constructed on the basis of base year and then the index number
of base year (P10) on the basis of current year, both should be reciprocal to each other.
P01*P10 = 1
Fisher’s ideal index number satisfies this test as explained below:
(P01) = ∗ * 100 or (P10) = ∗ * 100

Thus, P01 x P10 = ∗ ∗ ∗

P01 x P10 = 1

2) Factor Reversibility Test- It is provides that if quantity index number (Q01) is

constructed by substituting ‘quantity’ in place of ‘price’ and ‘price’ in place of
‘quantity’ and index number is multiplied by current year’s price index number,
product should be in ratio of total expenditure of current year (𝚺p1q1) and total
expenditure of base year (𝚺p0q0).
𝜮𝑷𝟏𝐪𝟏
P01*Q01 = 𝜮𝑷𝒐𝒒𝟎
Fisher’s formula satisfies also this test as explained below:

(P01) = ∗ * 100 Q01 = ∗ * 100

P01*Q01 = ∗ ∗ ∗
𝜮𝑷𝟏𝐪𝟏
P01*Q01 =
𝜮𝑷𝒐𝒒𝟎
Example-
Constructlaspeyre, paasche, marshall- edgeworth, dorbish-bowley and fisher price index
numbers from the following data:

item 1987 1997

price quantity price quantity
A 4 20 6 10
B 3 15 5 20
C 2 25 3 15
D 5 10 4 40

Solution-

item 1987 1997 P0q0 P0 q1 P1q0 P1q1

price quantity price quantity
A 4 20 6 10 80 40 120 60
B 3 15 5 20 45 60 75 100
C 2 25 3 15 50 30 75 45
D 5 10 4 40 50 200 40 160
total 225 330 310 365

Laspeyre index no - = ∗ 100

= (310/225) *100 = 137.78

paasche Index Number (P01) = ∗ 100

=(365/330)*100 = 110.61

marshall- edgeworth Index Number (P01) = ∗ 100

= [(310+365)/(225+330)]*100 = 121.62

dorbish-bowle Index Number (P01) = + ∗ 100/2

 = [310/225 + 365/330]*100 /2
= (1.37 78+ 1.1061) *100 /2 =124.195

fisher price Index Number (P01) = ∗ * 100

= √ 137.78 *110.61
= 123.45
Exercise-
1- Calculate three yearly, five yearly moving average from the given data-
year import
1966 20
67 24
68 19
69 18
70 24
71 25
72 21
73 23
74 28
75 22
76 26

2- Given the following data calculate-

 laspeyre price index no.
 paasche price index no.
 fisher price index no.
 and also satisfy time reversal test and factor reversal test.

item 1990 2007

price quantity price quantity
A 5 50 10 56
B 3 100 4 120
C 4 60 6 60
D 11 30 14 24
E 7 40 10 36