CHAPTER 19 – INDEX NUMBERS AND TIME SERIES – UNIT 1 – INDEX NUMBERS

Chapter 19 – Index Numbers and Time

Series

Unit 1 – Index Numbers

Introduction
An index number is a ratio of two or more time periods, one of which is the base time period. The
value at the base time period serves as the standard point of comparison.
The base time period is that time period from which the comparisons are to be made. For example,
in 2009 the price of a McAloo Tikki burger was ₹20; in 2020, it’s ₹40. Now, if I need to compare
the price of 2020 with the price of 2009, 2009 will be the base time period, and 2020 will be current
time period. The price in the base time period is denoted as P0 . The price in the current time period
is denoted as P1. The ratio of the price of the current period (2020, i.e., P1 ) to the price of the base
period (or reference period, i.e., 2009, i.e., P0 ), is known as the Price Relative, and is denoted as
P1
P01 . Therefore, P01 = .
P0

Pn
Therefore, Price Relative = . It is expressed as a percentage as follows: Price Relative
P0
Pn
=  100.
P0

Questions to be solved from Scanner

1. Page 3.970 – Question 10
2. Page 3.986 – Question 40
3. Page 3.989 – Question 46
4. Page 3.984 – Question 37 – Homework
5. Page 3.993 – Question 56 – Homework

Simple Aggregative Method

Simple Aggregative Price Index = 

Pn
100
P 0

Questions to be solved from Scanner

1. Question 38

P a g e 19.1 | 19.18
 CHAPTER 19 – INDEX NUMBERS AND TIME SERIES – UNIT 1 – INDEX NUMBERS

Simple Average of Price Relatives

 P1 
  P 100 
Index =  0 
N
Questions to be solved from Scanner
1. Page 3.990 – Question 49
2. Page 3.976 – Question 19

Weighted Average Method

In this method, we assign a weight to the prices of the commodities. Thereafter, the average is
calculated as follows:
𝑆𝑢𝑚 𝑜𝑓 𝑃𝑟𝑜𝑑𝑢𝑐𝑡𝑠
𝐺𝑒𝑛𝑒𝑟𝑎𝑙 𝐼𝑛𝑑𝑒𝑥 =
𝑆𝑢𝑚 𝑜𝑓 𝑊𝑒𝑖𝑔ℎ𝑡𝑠

Questions to be solved from Scanner

1. Question 8
2. Question 14 – Homework
The weights are usually the quantities of the commodities. These indices can be classified into two
broad groups:
1. Weighted Aggregative Index
2. Weighted Average of Relatives

Weighted Aggregative Index

In this method, weights are assigned to the prices of the commodities. The weights are usually
either the quantities or the value of goods, sold either during the base year, or the given year, or an
average of some years. Various alternative formulae used are as follows:
1. Laspeyres’ Index: In this Index, base year quantities are used as weights:

Laspeyres Index =  n 0 100

PQ
 P0Q0
Questions to be solved from Scanner:
a. Question 93
b. Question 2 – Homework
c. Question 62
d. Question 80 – Homework
e. Question 88 – Homework
2. Paasche’s Index: In this Index current year quantities are used as weights:

Passche's Index =  n n 100

PQ
 P0Qn
Questions to be solved from Scanner:
a. Question 65
P a g e 19.2 | 19.18
 CHAPTER 19 – INDEX NUMBERS AND TIME SERIES – UNIT 1 – INDEX NUMBERS

b. Question 22 – Homework
c. Question 32 – Homework
d. Question 42 – Homework
3. Methods based on some typical Period:

Index =  n t 100 , where t stands for some typical period of years, the quantities of
PQ
 P0Qt
which are used as weights.
The Marshall-Edgeworth index uses this method by taking the average of the base year and
the current year.

Marshall-Edgeworth Index =  n 0
P ( Q + Qn )
100
 P0 ( Q0 + Qn )
Questions to be solved from Scanner:
a. Question 11
4. Bowley’s Price Index: This index is the arithmetic mean of Laspeyres’ and Paasche’s.
Laspeyres' + Paasche's
Bowley’s Index =
2
Questions to be solved from Scanner:
a. Question 47
b. Question 30
5. Fisher’s ideal Price Index: This index is the geometric mean of Laspeyres’ and Paasche’s.

Fisher's Index =
PQ  PQ
n 0 n n
100
PQ PQ
0 0 0 n

Questions to be solved from Scanner:

a. Question 86
b. Question 34 – Homework
c. Question 43
d. Question 55
e. Question 41 – Homework
f. Question 6
g. Question 16

Weighted Average of Relatives

In this method, weighted arithmetic mean is used to calculate the index.

 Pn 
  P  ( P Q )
0 0

Index =  0   100
PQ 0 0

The Chain Index Numbers

Till now, we have been taking a fixed base; however, when conditions change rapidly, the fixed
base does not suit the required needs. In such a case, changing base is more suitable. For example,
the base for the year 1999 could be 1998; the base for the year 2000 could be 1999 (not 1998), the

P a g e 19.3 | 19.18
 CHAPTER 19 – INDEX NUMBERS AND TIME SERIES – UNIT 1 – INDEX NUMBERS

base for the year 2001 could be 2000 (neither 1998, nor 1999), and so on. If it is desired to associate
these relatives to a common base, the results are chained. Thus, under this method the relatives of
each year are first related to the preceding year, called the link relatives, and then they are chained
together by successive multiplication to form a chain index.
𝐿𝑖𝑛𝑘 𝑅𝑒𝑙𝑎𝑡𝑖𝑣𝑒 𝑜𝑓 𝑡ℎ𝑒 𝐶𝑢𝑟𝑟𝑒𝑛𝑡 𝑌𝑒𝑎𝑟 × 𝐶ℎ𝑎𝑖𝑛 𝐼𝑛𝑑𝑒𝑥 𝑜𝑓 𝑡ℎ𝑒 𝑃𝑟𝑒𝑣𝑖𝑜𝑢𝑠 𝑌𝑒𝑎𝑟
𝐶ℎ𝑎𝑖𝑛 𝐼𝑛𝑑𝑒𝑥 =
100
For example,
Year Price Link Relatives Chain Indices
(Taking Previous Year as Base Year) (Taking 1991 as Base Year)
1991 50 100.00 100.00
60 120
1992 60  100 = 120.00  100 = 120.00
50 100
62 103.33
1993 62  100 = 103.33 120 = 124.00
60 100
65 104.84
1994 65  100 = 104.84 124 = 129.90
62 100

Quantity Index Numbers

1. Simple Aggregate of Quantities

Index =  n 100
Q
 Q0
2. Simple Average of Quantity Relatives:
 Qn
Index =  0 100
Q
n
3. Weighted Aggregate Quantity Indices:
a. With base year weight (Laspeyre’s index)

Index =  n 0 100
QP
 Q0 P0
b. With current year weight (Paasche’s index)

Index =  n n 100
QP
 Q0 Pn
c. Fisher’s Ideal (Geometric mean of the above)

Index =
Q P  Q P
n 0 n n
100
Q P Q P
0 0 0 n

4. Base-year weighted average of quantity relatives

Q 
  Qn  ( P0Q0 )
Index =  0   100
 P0Q0
P a g e 19.4 | 19.18
 CHAPTER 19 – INDEX NUMBERS AND TIME SERIES – UNIT 1 – INDEX NUMBERS

Value Indices
Value = Price × Quantity

Value Index =  PQ

Vn
= n n

V 0 PQ0 0

Limitations and Usefulness of Index Numbers

Limitations
1. As the indices are constructed mostly from deliberate samples, chances of errors creeping
in cannot be always avoided.
2. Since index numbers are based on some selected items, they simply depict the broad trend
and not the real picture.
3. Since many methods are employed for constructing index numbers, the result gives
different values and this at times creates confusion.
Usefulness
1. Framing suitable policies in economics and business: They provide guidelines to make
decisions in measuring intelligence quotients, research etc.
2. They reveal trends and tendencies in making important conclusions in cyclical forces,
irregular forces, etc.
3. They are important in forecasting future economic activity. They are used in time series
analysis to study long-term trend, seasonal variations and cyclical developments.
4. Index numbers are very useful in deflating i.e., they are used to adjust the original data for
price changes and thus transform nominal wages into real wages.
5. Cost of living index numbers measure changes in the cost of living over a given period.

Deflating Time Series Using Index Numbers

𝐶𝑢𝑟𝑟𝑒𝑛𝑡 𝑉𝑎𝑙𝑢𝑒
𝐷𝑒𝑓𝑙𝑎𝑡𝑒𝑑 𝑉𝑎𝑙𝑢𝑒 =
𝑃𝑟𝑖𝑐𝑒 𝐼𝑛𝑑𝑒𝑥 𝑜𝑓 𝑡ℎ𝑒 𝐶𝑢𝑟𝑟𝑒𝑛𝑡 𝑌𝑒𝑎𝑟
or
𝐵𝑎𝑠𝑒 𝑃𝑟𝑖𝑐𝑒 (𝑃0 )
𝐶𝑢𝑟𝑟𝑒𝑛𝑡 𝑉𝑎𝑙𝑢𝑒 =
𝐶𝑢𝑟𝑟𝑒𝑛𝑡 𝑃𝑟𝑖𝑐𝑒 (𝑃𝑛 )
Year Wholesale Price Gross National Product at Real Gross National
Index Current Prices Product
7499
1970 113.1 7499  100 = 6630
113.1
7935
1971 116.3 7935  100 = 6823
116.3
8657
1972 121.2 8657 100 = 7143
121.2

P a g e 19.5 | 19.18
 CHAPTER 19 – INDEX NUMBERS AND TIME SERIES – UNIT 1 – INDEX NUMBERS

9323
1973 127.7 9323  100 = 7301
127.7
Questions to be solved from Scanner
1. Page 3.995 – Question 60
2. Page 3.977 – Question 21
3. Page 3.994 – Question 58 – Homework
4. Page 3.1006 – Question 94
5. Page 3.965 – Question 3
6. Page 3.998 – Question 68 – Homework
7. Page 3.971 – Question 12 – Homework
8. Page 3.988 – Question 44 – Homework
9. Page 3.992 – Question 54 – Homework
10. Page 3.966 – Question 63 - Homework

Shifting and Splicing of Index Numbers

Shifting of Index Numbers
𝑂𝑟𝑖𝑔𝑖𝑛𝑎𝑙 𝑃𝑟𝑖𝑐𝑒 𝐼𝑛𝑑𝑒𝑥
𝑆ℎ𝑖𝑓𝑡𝑒𝑑 𝑃𝑟𝑖𝑐𝑒 𝐼𝑛𝑑𝑒𝑥 = × 100
𝑃𝑟𝑖𝑐𝑒 𝐼𝑛𝑑𝑒𝑥 𝑜𝑓 𝑡ℎ𝑒 𝑌𝑒𝑎𝑟 𝑜𝑛 𝑤ℎ𝑖𝑐ℎ 𝑖𝑡 ℎ𝑎𝑠 𝑡𝑜 𝑏𝑒 𝑠ℎ𝑖𝑓𝑡𝑒𝑑
Year Original Price Index Shifted Price Index to Base 1990
125
1988 125 100 = 89.3
140
131
1989 131 100 = 93.6
140
140
1990 140 100 = 100.0
140
147
1991 147 100 = 105.0
140
Splicing of Index Numbers
Splicing means combining two index covering different bases into a single series. Splicing two
sets of price index numbers covering different periods of time is usually required when there is a
major change in quantity weights. It may also be necessary on account of a new method of
calculation or the inclusion of new commodity in the index.
Year Old Price Index Revised Price Index Spliced Price Index
[1900 = 100] [1995 = 100] [1995 = 100]
100
1990 100.0 100 = 87.6
114.2
102.3
1991 102.3 100 = 89.6
114.2
105.3
1992 105.3 100 = 92.2
114.2
107.6
1993 107.6 100 = 94.2
114.2
P a g e 19.6 | 19.18
 CHAPTER 19 – INDEX NUMBERS AND TIME SERIES – UNIT 1 – INDEX NUMBERS

111.9
1994 111.9 100 = 98.0
114.2
1995 114.2 100.0 100.0
1996 102.5 102.5
1997 106.4 106.4
1998 108.3 108.3
1999 111.7 111.7
2000 117.8 117.8

Test of Adequacy
There are four tests:
1. Unit Test –
a. This test requires that the formula should be independent of the unit in which (or,
for which) prices and quantities are quoted.
b. All the formulae satisfy this test, except for the simple (unweighted) aggregative
index.
2. Time Reversal Test –
a. It is a test to determine whether a given method will work both ways in time,
forward and backward.
b. The test provides that the formula for calculating the index number should be such
that two ratios, the current on the base and the base on the current should multiply
into unity.
c. In other words, the two indices should be reciprocals of each other. Symbolically,
P P
P01  P10 = 1 , where, P01 = 1 , and P10 = 0 .
P0 P1
d. Check of Different Methods
i. Laspeyres’ method

P01 =
 PQ
1 0
, P10 =  0 0
PQ
 P0Q0  PQ
1 0

P01  P10 =
 PQ   P Q
1 0 0 0
1
 P Q  PQ
0 0 1 0

Therefore, Laspeyres’ Method does not satisfy this test.

ii. Paasche’s method

P01 =
 PQ
1 n
, P10 =  0 n
PQ
 P0Qn  PQ
1 n

P01  P10 =
 PQ   P Q
1 n 0 n
1
 P Q  PQ
0 n 1 n

Therefore, Paasche’s Method does not satisfy this test.

iii. Fisher’s Ideal

P01 =
 PQ   PQ
1 0 1 1
, P10 =
PQ  PQ
0 1 0 0

PQ PQ
0 0 0 1  PQ  PQ
1 1 1 0

P a g e 19.7 | 19.18
 CHAPTER 19 – INDEX NUMBERS AND TIME SERIES – UNIT 1 – INDEX NUMBERS

P01  P10 =
 PQ   PQ 1 0 1 1

PQ  PQ
0 1 0 0
=1
PQ PQ 0 0 0 1  PQ  PQ
1 1 1 0

Therefore, Fisher’s Idea does satisfy this test.

3. Factor Reversal Test –
a. This states that the product of price index and the quantity index should be equal to

the corresponding value index, i.e.,  1 1 .

PQ
 P0Q0
b. Symbolically, P01  Q01 = V01 .
c. Check for Fisher’s Method

P01 =
 PQ   PQ , Q =  Q P   Q P
1 0 1 1 1 0 1 1

PQ PQ Q P Q P
01
0 0 0 1 0 0 0 1

P01 Q =
 PQ   PQ   Q P   Q P
1 0 1 1 1 0 1 1

 P Q  P Q Q P Q P
01
0 0 0 1 0 0 0 1

P01 Q =
 PQ   PQ   Q P   Q P
1 0 1 1 1 0 1 1

 P Q  P Q Q P Q P
01
0 0 0 1 0 0 0 1

(  PQ )
2

Q =
1 1
P01
( P Q )
01 2
0 0

P01 Q =
 PQ 1 1

PQ
01
0 0

Therefore, Fisher’s Method satisfies this test as well.

d. While selecting an appropriate index formula, the Time Reversal Test and the
Factor Reversal test are considered necessary in testing the consistency.
e. Since Fisher’s Index number satisfies both the tests (Time Reversal, as well as
Factor Reversal), it is called an Ideal Index Number.
4. Circular Test –
a. As per this test, P01  P12  P20 = 1 .
For example,
Question
If the 1970 index with base 1965 is 200, and 1965 index with base 1960 is 150,
what will be the index of 1970 on base 1960?
(May, 2018)
Solution
Let the year 1960 be P0 , the year 1965 be P1 , and the year 1970 be P2 .
We need to find out the index of 1970 ( P2 ) , on base 1960 ( P0 ) . Therefore, we need
to find P02 .
As per the question,
i. the 1970 index with base 1965 is 200. This means that P12 = 200.

P a g e 19.8 | 19.18
 CHAPTER 19 – INDEX NUMBERS AND TIME SERIES – UNIT 1 – INDEX NUMBERS

ii. the 1965 index with base 1960 is 150. This means that P01 = 150.
As per the circular test, we know that P01  P12  P20 = 1 .
1 1
Therefore, 150  200  P20 = 1  P20 = 100 = .
150  200 300
1
Therefore, P02 = = 300.
P20
b. Therefore, this property enables us to adjust the index values from period to period
without referring to the original base every time.
c. The test of this shiftability of base is called the circular test.
d. This test is not met by Laspeyres, or Paasche’s or the Fisher’s ideal index.
e. The simple geometric mean of price relatives and the weighted aggregative with
fixed weights meet this test.

P a g e 19.9 | 19.18
 CHAPTER 19 – INDEX NUMBERS AND TIME SERIES – UNIT 2 – TIME SERIES

Unit 2 – Time Series

Introduction
A Time Series is a set of observations taken at specified times, usually at equal intervals.

Components of Time Series

There are various forces that affect the values of a phenomenon in a time series; these may be
broadly divided into the following four categories, commonly known as the components of a time
series.
1. Long term movement or Secular Trend
2. Seasonal variations
3. Cyclical variations
4. Random or irregular variations

Long term movement or Secular Trend or Simple Trend

1. Secular trend is the long-term tendency of the time series to move in an upward or
downward direction.
2. It indicates how it has behaved over the entire period under reference.
3. A general tendency of a variable to increase, decrease or remain constant in long term is
called trend of a variable. However, in a small interval of time, the variable may increase
or decrease.
4. Some examples are:
a. Population of a country has an increasing trend over the years.
b. Due to modern technology, agricultural and industrial production is increasing.
c. Due to modern technology and health facilities, death rate is decreasing and life
expectancy is increasing.
5. These are results of long-term forces that gradually operate on the time series variable.
6. A few examples of theses long term forces (which make a time series to move in any
direction over long period of the time) are:
a. long term changes per capita income,
b. technological improvements of growth of population,
c. changes in Social norms etc.
7. Most of the time series relating to Economic, Business and Commerce might show
a. an upward tendency in case of
i. population,
ii. production & sales of products,
iii. incomes,
iv. prices; or
b. a downward tendency in case of
i. share prices,

P a g e 19.10 | 19.18
 CHAPTER 19 – INDEX NUMBERS AND TIME SERIES – UNIT 2 – TIME SERIES

ii. death,
iii. birth rate etc.
due to global meltdown, or improvement in medical facilities etc.

Seasonal Variations
1. Over a span of one year, seasonal variation takes place due to the rhythmic forces which
operate in a regular and periodic manner.
2. These forces have the same or almost similar pattern year after year.
3. It is common knowledge that the value of many variables depends in part on the time of
year.
4. For Example, Seasonal variations could be seen and calculated if the data are recorded
quarterly, monthly, weekly, daily or hourly basis.
5. So, if in a time series data, only annual figures are given, there will be no seasonal
variations.
6. The seasonal variations may be due to various seasons or weather conditions; for example,
sale of cold drink would go up in summers & go down in winters.
7. These variations may also be due to
a. man-made conventions,
b. habits,
c. customs, or
d. traditions.
For example, sales might go up during Diwali & Christmas or sales of restaurants &
eateries might go down during Navratri’s.
8. The methods of seasonal variations are:
a. Simple Average Method
b. Ratio to Trend Method
c. Ratio to Moving Average Method
d. Link Relatives Method

Cyclical variations
1. Cyclical variations are the periodic movements. These are also generally termed as
business cycles.
2. These variations in a time series are due to ups & downs recurring after a period from
Season to Season.
3. Though they are more or less regular, they may not be uniformly periodic.
4. These are oscillatory movements which are present in any business activity, and is termed
as business cycle.
5. It has got four phases consisting of
a. prosperity (boom),
b. recession,
c. depression, and
d. recovery.
6. All these phases together may last from 7 to 9 years may be less or more.

Random or Irregular Variations

1. These are irregular variations which occur on account of random external events.

P a g e 19.11 | 19.18
 CHAPTER 19 – INDEX NUMBERS AND TIME SERIES – UNIT 2 – TIME SERIES

2. These variations either go very deep downward or too high upward to attain peaks abruptly.
3. These fluctuations are a result of unforeseen and unpredictable forces which operate in
absolutely random or erratic manner.
4. They do not have any definite pattern and it cannot be predicted in advance.
5. These variations are due to floods, wars, famines, earthquakes, strikes, lockouts, epidemics
etc.

Models of Time Series

There are two models which are generally used for decomposition of time series into its four
components. The objective is to estimate and separate the four types of variations and to bring out
the relative impact of each on the overall behaviour of the time series.
1. Additive model
2. Multiplicative model

Additive Model
In additive model, it is assumed that the four components are independent of one another. Under
this assumption, the four components are arithmetically additive, i.e., magnitude of time series is
the sum of the separate influences of its four components, i.e., Yt = T + C + S + I , where,

Yt is time series;
T is trend variation;
C is cyclical variation;
S is seasonal variation;
I is random or irregular variation.

Multiplicative Model
In this model, it is assumed that the forces that give rise to four types of variations are
interdependent, so that the overall pattern of variations in the time series is a combined result of
the interaction of all the forces operating on the time series. Therefore, time series is the product
of its four components, i.e., Yt = T  C  S  I .

Multiplication model which is used more frequently.

Measurement of Secular Trend

The following are the methods most commonly used for studying & measuring the trend
component in a time series:
1. Graphic or a Freehand Curve Method
2. Method of Semi Averages
3. Method of Moving Averages
4. Method of Least Squares

Graphic or a Freehand Curve Method

• The data of a given time series is plotted on a graph and all the points are joined together
with a straight line.
P a g e 19.12 | 19.18
 CHAPTER 19 – INDEX NUMBERS AND TIME SERIES – UNIT 2 – TIME SERIES

• This curve would be irregular as it includes short run oscillations.

• These irregularities are smoothened out by drawing a freehand curve or line along with the
curve previously drawn.
• This curve would eliminate the short run oscillations and would show the long period
general tendency of the data.
• While drawing this curve, it should be kept in mind that the curve should be smooth and
the number of points above the trend curve should be more or less equal to the number of
points below it.
• Merits:
o It is very simple and easy to understand.
o It does not require any mathematical calculations.
• Disadvantages:
o This is a subjective concept. Hence different persons may draw freehand lines at
different positions and with different slopes.
o If the length of period for which the curve is drawn is very small, it might give
totally erroneous results.
Question 1
The following are figures of a Sale for the last nine years. Determine the trend by line by the
freehand method.
Year 2000 2001 2002 2003 2004 2005 2006 2007 2008
Sale in lac units 75 95 115 65 120 100 150 135 175

Solution

Sale in lac units

Sale in lac units Linear (Sale in lac units)
200 175
150
Sale in Lac Units

150 135
115 120
95 100
100 75 65
50

0
2000 2002 2004 2006 2008
Years

Method of Semi Averages

• Under this method, the whole time series data is classified into two equal parts and the
averages for each half are calculated.

P a g e 19.13 | 19.18
 CHAPTER 19 – INDEX NUMBERS AND TIME SERIES – UNIT 2 – TIME SERIES

• If the data is for even number of years, it is easily divided into two. If the data is for odd
number of years, then the middle year of the time series is left and the two halves are
constituted with the period on each side of the middle year.
• The arithmetic mean for a half is taken to be representative of the value corresponding to
the midpoint of the time interval of that half.
• Thus, we get two points. These two points are plotted on a graph and then are joined by
straight line which is our required trend line.
Question 2
Fit a trend line to the following data by the method of Semi-averages.
Year 2000 2001 2002 2003 2004 2005 2006
Sale in lac units 100 105 115 110 120 105 115

Solution
Here, since there are 7 years (odd), we’ll leave out the middle one, i.e., the year 2003, and take the
average of the first three years and the last three years.
100 + 105 + 115
Average for the first three years = = 106.67 . This is taken to be representative of
3
the value corresponding to the midpoint of the time interval of the first half, i.e., the year 2001.
120 + 105 + 115
Average for the last three years = = 113.33 . This is taken to be representative of
3
the value corresponding to the midpoint of the time interval of the second half, i.e., the year 2005.

Sale in Lac Units

114 113.33
113
Sale in Lac Units

112
111
110
109
108
106.67
107
106
2000 2002 2004 2006 2008
Year

Method of Moving Averages

• A moving average is an average (Arithmetic mean) of fixed number of items (known as
periods) which moves through a series by dropping the first item of the previously averaged
group and adding the next item in each successive average.

P a g e 19.14 | 19.18
 CHAPTER 19 – INDEX NUMBERS AND TIME SERIES – UNIT 2 – TIME SERIES

• The value so computed is considered the trend value for the unit of time falling at the centre
of the period used in the calculation of the average.
• 3 Year Moving Average:
For computing 3 yearly moving average, the value of 1st, 2nd and 3rd years are added up,
and arithmetic mean is found out and the answer is placed against the 2nd year; then value
of 2nd, 3rd and 4th years are added up, and arithmetic mean is derived and this average is
placed against 3rd year (i.e. the middle of 2nd, 3rd and 4th) and so on.
• 4 Year Moving Average:
For computing 4 yearly moving average, the values of 1st, 2nd, 3rd and 4th years are added
up, and the total is written between the second and the third year in the third column.
Thereafter, the values of the 2nd, 3rd, 4th, and 5th years are added up, and the total is written
between the 3rd and the 4th year in the third column. Thereafter, a fourth column is prepared,
which contains the totals of groups of two values in the third column. A fifth column is
then prepared to calculate the average, which is given by dividing the figure in the fourth
column by the total number of years.
• 5 Year Moving Average:
For computing 5 yearly moving average, the values of the 1st, 2nd, 3rd, 4th, and 5th years are
added up, and arithmetic mean is found out and answer is placed against the 3rd year; then
value of 2nd, 3rd, 4th, 5th, and 6th years are added up, and the arithmetic mean is derived and
this average is placed against 4th year, and so on.
• This technique is called centring & the corresponding moving averages are called moving
average centred.
Question 3
The wages of certain factory workers are given as below. Using 3 yearly moving average, indicate
the trend in wages.
Year 2004 2005 2006 2007 2008 2009 2010 2011 2012
Wages 1200 1500 1400 1750 1800 1700 1600 1500 1750

Solution
Year Wages 3 Yearly Moving Totals 3 Yearly Moving Average, i.e., Trend
2004 1200
2005 1500 (1200 + 1500 + 1400) = 4100 4100 ÷ 3 = 1366.67
2006 1400 (1500 + 1400 + 1750) = 4650 4650 ÷ 3 = 1550.00
2007 1750 (1400 + 1750 + 1800) = 4950 4950 ÷ 3 = 1650.00
2008 1800 (1750 + 1800 + 1700) = 5250 5250 ÷ 3 = 1750.00
2009 1700 (1800 + 1700 + 1600) = 5100 5100 ÷ 3 = 1700.00
2010 1600 (1700 + 1600 + 1500) = 4800 4800 ÷ 3 = 1600.00
2011 1500 (1600 + 1500 + 1750) = 4850 4850 ÷ 3 = 1616.67
2012 1750

Question 4
Calculate 4 yearly moving average of the following data.

P a g e 19.15 | 19.18
 CHAPTER 19 – INDEX NUMBERS AND TIME SERIES – UNIT 2 – TIME SERIES

Year 2005 2006 2007 2008 2009 2010 2011 2012

Wages 1150 1250 1320 1400 1300 1320 1500 1700

Solution
Calculation of 4 Year Centred Moving Average
2 Year Moving Total of 4 Year Moving
Year Wages 4 Year Moving Total
Column 3 (Centred) Average (Centred)
(1) (2) (3)
(4) (5) = (4) ÷ 8
2005 1,150 – – –
2006 1,250 – – –
1,150 + 1,250 + 1,320
+ 1,400 = 5,120
2007 1,320 5,120 + 5,270 = 10,390 10,390 ÷ 8 = 1,298.75
1,250 + 1,320 + 1,400
+ 1,300 = 5,270
2008 1,400 5,270 + 5,340 = 10,610 10,610 ÷ 8 = 1,326.25
5,340
2009 1,300 10,860 1,357.50
5,520
2010 1,320 11,340 1,417.50
5,820
2011 1,500
2012 1,700

Question 5
Calculate five yearly moving averages for the following data.
Year 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Value 123 140 110 98 104 133 95 105 150 135

Solution
Value 5 Year Moving Totals 5 Year Moving Average
Year
(’000 ₹) (’000 ₹) (’000 ₹)
2003 123 – –
2004 140 – –
2005 110 123 + 140 + 110 + 98 + 104 = 575 575 ÷ 5 = 115.0
2006 98 140 + 110 + 98 + 104 + 133 = 585 585 ÷ 2 = 117.0
2007 104 110 + 98 + 104 + 133 + 95 = 540 540 ÷ 2 = 108.0
2008 133 98 + 104 + 133 + 95 + 105 = 535 535 ÷ 2 = 107.0
2009 95 104 + 133 + 95 + 105 + 150 = 587 587 ÷ 2 = 117.4
2010 105 133 + 95 + 105 + 150 + 135 = 618 618 ÷ 2 = 123.6
2011 150 – –
2012 135 – –

P a g e 19.16 | 19.18
 CHAPTER 19 – INDEX NUMBERS AND TIME SERIES – UNIT 2 – TIME SERIES

Method of Least Squares

This method is used for finding a straight-line equation which represents the given data. The
straight-line equation is given by Yc = a + bX .

The values of a and b can be found as follows:

a=
 Y , b =  XY
N X 2

Here,

Y = Sum of actual values of Y variable, i.e. Sales, Profit, etc.

N = No. of years or months or any other period
 X = Sum of values of X.
Note: When the number of years is odd, then X is the deviation of every year from the central year;
however, if the number of years is even, then, first the deviations are taken from the average of the
two middle most years, and then these deviations are multiplied by 2. This gives us the value of X.
 XY = Sum of the products X and Y
X 2
= Sum of squares of deviations from X

Question 6
Fit a straight-line trend to the following data by Least Square Method and estimate the sale for the
year 2019.
Year 2013 2014 2015 2016 2017
Sales (₹ in lakhs) 10 20 30 50 40

Solution
Year Sales (Y) Deviations of the Years from 2015 (X) XY X2
2013 10 2013 – 2015 = –2 –20 4
2014 20 2014 – 2015 = –1 –20 1
2015 30 2015 – 2015 = 0 0 0
2016 50 2016 – 2015 = 1 50 1
2017 40 2017 – 2015 = 2 80 4
Total 150 0 90 10
Now, we’ll find out the values of a and b as follows:

a=
 Y  a = 150 = 30
N 5

b=
 XY  b = 90 = 9
X 2
10

Therefore, the best fit line is Yc = 30 + 9 X .

P a g e 19.17 | 19.18
 CHAPTER 19 – INDEX NUMBERS AND TIME SERIES – UNIT 2 – TIME SERIES

Now, 2019 means that X is 2019 – 2015 = 4. Therefore, estimated sales of 2019 is given by
Y2019 = 30 + ( 9  4 ) = 30 + 36 = 66.

Question 7
Fit a straight-line trend to the following data by Least Square Method and estimate the sale for the
year 2012.
Year 2005 2006 2007 2008 2009 2010
Sale (in ’000s) 70 80 96 100 95 114

Solution
Year Sale (in ’000s) (Y) Difference from 2007.5 X = Deviations × 2 XY X2
2005 70.00 2005 – 2007.5 = –2.50 –5.00 –350.00 25.00
2006 80.00 2006 – 2007.5 = –1.50 –3.00 –240.00 9.00
2007 96.00 –0.50 –1.00 –96.00 1.00
2008 100.00 0.50 1.00 100.00 1.00
2009 95.00 1.50 3.00 285.00 9.00
2010 114.00 2.50 5.00 570.00 25.00
Total 555.00 0.00 0.00 269.00 70.00
Now, we’ll find out the values of a and b as follows:

a=
 Y  a = 555 = 92.5
N 6

b=
 XY  b = 269 = 3.843
X 2
70

Therefore, the best fit line is Yc = 92.5 + 3.843 X .

Now, 2012 means that the deviation from 2007.5 is 2012 – 2007.5 = 4.5, and therefore, X is 4.5 ×
2 = 9. Therefore, estimated sales of 2012 is given by Y2012 = 92.5 + (3.843 × 9) = 127.09.

P a g e 19.18 | 19.18