ANALYSIS OF TIME SERIES

MEANING OF TIME SERIES

A TIME SERIES is a collection of observations of well-defined data items obtained through repeated
measurements over time.

For example, measuring the value of retail sales each month of the year would comprise a time series. This is
because sales revenue is well defined, and consistently measured at equally spaced intervals. Data collected
irregularly or only once are not time series.

An observed time series can be decomposed into four components (pattern):

• The Secular Trend (long term direction),

• The Seasonal Variation(systematic, calendar related movements)
• The Cyclic Variation (regular and/or recurring variation)
• The Irregular Variation (unsystematic, short term fluctuations).
 USAGE OF TIME SERIES

• Time series is used to predict future values based on previously observed values.

• Time series analysis is used to identify the fluctuation in economics and business.

• It helps in the evaluation of current achievements.

• Time series is used in pattern recognition, weather forecasting and earthquake prediction.

It can be said that time series analysis is a big tool in the hands of business executives to plan
their sales, prices, policies and production.
 COMPONENTS OF TIME SERIES

1. Secular Trend (T) - The concept of trend does not include short range oscillations but
rather steady variations or movement over a long period of time. Secular trend
movements are attributable to factors such as population change, technical progress and
large-scale shift in consumer tastes.
 COMPONENTS OF TIME SERIES

2. Seasonal Variation (S) – Seasonal Variations are those periodic movements in business activity which
occur regularly every year. Since these variations repeat during a period of 12 month they can be predicted fairly
accurately. Nearly every type of business activity is susceptible to seasonal influence to a greater or lesser
degree and as such these variations are regarded as normal phenomenon recurring every year.
- Climate & weather Conditions.
- Customs Traditions and habit.
 COMPONENTS OF TIME SERIES

3. Cyclic Variation (C) - Most of the time series relating to economics and business show some kind of
cyclical or oscillatory variation. Cyclical fluctuations are long-term movements that
represent consistently recurring rises and decline in activity. A business cycle consists of the recurrence of the
up down movement of business activity from some sort of statistical trend or normal.

There are four well defined periods or phases in the business cycle-
I. Prosperity II. Decline (Recession)
III. Depression IV. Improvement (Recovery)
 COMPONENTS OF TIME SERIES
4. Irregular Variation (I) - Irregular variation, also called 'Erratic', ‘Accidental’, ‘Random’ refer to such

variation in business activity which do not repeat in a definite pattern. Irregular variation are caused by such

isolated special occurrence as floods, earthquake, strikes, wars and pandemics like Corona. Sudden change in

demand or very rapid technological progress may also be included in this category. (very irregular and

unpredictable).
ADDITIVE & MULTIPLICATIVE MODEL OF TIME SERIES
 MEASUREMENT OF TREND

The various methods that can be used for determining trend are:

• Freehand or graphic method.

• Semi-average method.

• Moving average method.

• Method of least squares.

 Moving average method
▸ When a trend is to be determined by the method of moving averages, the average value for a number of years (or

months/ week) is secured, and this average is taken as the normal or trend value for the unit of time falling at the middle of

the period covered in the calculation of the average.

▸ The effect of averaging is to give a smoother curve, lessening the influence of the fluctuation that pull the annual figure

away from the general trend.

▸ While applying this method, it is necessary to select a period for moving average such as 3-yearly moving average,

5-yearly moving average, 7-yearly moving average etc.

Data—If the data points are: a, b, c, d, e, f, g, h

𝑎 + 𝑏 + 𝑐 𝑏 + 𝑐 + 𝑑 𝑐 + 𝑑 +𝑒
The 3-yearly moving average shall be computed as follows: 3
, 3 , 3 ,…

𝑎+𝑏+𝑐+𝑑+𝑒 𝑏 + 𝑐 + 𝑑+ 𝑒 + 𝑓 𝑐 + 𝑑 +𝑒+𝑓+ℎ
The 5-yearly moving average shall be computed as follows: , , ,…
5 5 5
Example: Calculate the 5-yearly and 7-yearly Moving Averages for the following data of the numbers of commercial and
industrial failures in a country during 1988 to 2003
Year No. of Failures
1988 23
1989 26
1990 28
1991 32
1992 20
1993 12
1994 12
1995 10
1996 9
1997 13
1998 11
1999 14
2000 12
2001 9
2002 3
2003 1
 Moving average method : Fitting of Trend
No. of Failures
35

30

25
No. of Failures

20

15

10

0
1986 1988 1990 1992 1994 1996 1998 2000 2002 2004
Year
 Method of Least Square
This method is most widely used in practice. It is a mathematical method and with its help a trend

line is fitted to the data in such a manner that the following two conditions are satisfied:

• Σ (Y - 𝑌𝐶 ) = 0; i.e., the sum of deviations of the actual values of Y and the computed values of Y is

zero.

• Σ (Y - 𝑌𝐶 )² is least; i.e., the sum of the squares of the deviations of the actual and computed values

is least from this line and hence the name method of least squares. The line obtained by this

method is known as the line of best fit.

 Method of Least Square
This method is most widely used in practice. It is a mathematical method and with its help a trend

line is fitted to the data in such a manner that the following two conditions are satisfied:

• Σ (Y - 𝑌𝐶 ) = 0; i.e., the sum of deviations of the actual values of Y and the computed values of Y is

zero.

• Σ (Y - 𝑌𝐶 )² is least; i.e., the sum of the squares of the deviations of the actual and computed values

is least from this line and hence the name method of least squares. The line obtained by this

method is known as the line of best fit.

 Method of Least Square
The straight line trend is represented by the equation 𝑌𝐶 = a + b X,

Where 𝑌𝐶 is used to designate the trend values to distinguish them from the actual Y values,

‘a’ is the Y-intercept or the computed trend figure of the Y variable when X = 0.

‘b’ represents the slope of the trend line or amount of change in Y variable that is associated with a

change of one unit of X variable in time series analysis represents time.

In order to determine the values of the constants ‘a’ and ‘b’ the following two normal equations are

to be solved:
ΣY = an + b ΣX
ΣXY = a ΣX + b Σ𝑋 2
Where N represents number of years for which data are given.
 Method of Least Square : VALUE OF ‘a’ & ‘b’
ΣX = 0. In other words, the time variable is measured as a deviation from its mean. Since ΣX = 0 the

above two normal equations would take the form.

ΣY = an + b ΣX--------------------------> ΣY = an
ΣXY = a ΣX + b Σ𝑋 2 −−−−−−−−−−−−−−−−−−−−> ΣXY = b Σ𝑋 2

The values of ‘a’ and ‘b’ can now be determined easily.

ΣY
Since ΣY = an gives a = 𝒐𝒓 𝒎𝒆𝒂𝒏 𝒐𝒇 𝒀
𝒏

ΣXY
And ΣXY = b Σ𝑿𝟐 gives b =
Σ𝑿𝟐
 Method of Least Square
Example: Below are given the figures of production (in thousand quintals) of a sugar factory:

Year 1997 1998 1999 2000 2001 2002 2003

Production 80 90 92 83 94 99 92
(‘000 Quintals)

(i) Fit a straight line trend to these figures.

(ii) Plot these figures on a graph and show the trend line.
Solution:
Production
Year
(‘000 Quintals)
1997 80
1998 90
1999 92
2000 83
2001 94
2002 99
2003 92
MEASUREMENT OF SEASONAL VARIATION
 MEASUREMENT OF SEASONAL VARIATION

The effects of trend cycles and irregular fluctuations must be eliminated from the original time series data to

obtain an estimate of seasonal variation. Once they are eliminated, we get measures of seasonal variations in

the behavior of any variable. These measures are called seasonal indexes. By using such indexes, we can

deseasonalised the time series. Such deseasonalised data are known as seasonally adjusted data.

There are four methods of constructing seasonal index:

1. Method of Simple Averages (Weekly, Monthly, or Quarterly)

2. Ratio-to-Trend Method

3. Moving Average Method

4. Link Relatives Method,

 MEASUREMENT OF SEASONAL VARIATION : Method of Simple Averages
This is the simplest method of obtaining a seasonal index. The following steps are required to calculate index:

(i) Arrange the unadjusted data by years and months (or quarters if quarterly data are given).

(ii) Find the totals of January, February, etc. (or quarters)

(iii) Divide each total by the number of years for which data are given. For example, if we are given monthly

data for five years then total each month for five years and divide each total by 5 to obtain an average.

(iv) Obtain an average of monthly/quarterly averages.

(v) Compute the seasonal index as follows:

𝑄𝑢𝑎𝑟𝑡𝑒𝑟𝑙𝑦 𝐴𝑣𝑒𝑟𝑎𝑔𝑒 𝑓𝑜𝑟 𝑄𝑢𝑎𝑟𝑡𝑒𝑟 𝐼 𝑄𝑢𝑎𝑟𝑡𝑒𝑟𝑙𝑦 𝐴𝑣𝑒𝑟𝑎𝑔𝑒

Seasonal Index for Quarter I = × 100 = × 100
𝐴𝑣𝑒𝑟𝑎𝑔𝑒𝑠 𝑜𝑓 𝑄𝑢𝑎𝑟𝑡𝑒𝑟𝑠 𝐺𝑒𝑛𝑒𝑟𝑎𝑙 𝐴𝑣𝑒𝑟𝑎𝑔𝑒

If instead of the average of each month, the totals of each month are obtained, we will get the same result.
 MEASUREMENT OF SEASONAL VARIATION : Method of Simple Averages
EXAMPLE: Compute the average seasonal movement for the following series:

Quarterly Production
Year
I II III IV
1994 3.5 3.9 3.4 3.6
1995 3.5 4.1 3.7 4.0
1996 3.5 3.9 3.7 4.2
1997 4.0 4.6 3.8 4.5
1998 4.1 4.4 4.2 4.5
 MEASUREMENT OF SEASONAL VARIATION : Method of Simple Averages
Solution:

Quarterly Production
Year
I II III IV
1994 3.5 3.9 3.4 3.6
1995 3.5 4.1 3.7 4.0
1996 3.5 3.9 3.7 4.2
1997 4.0 4.6 3.8 4.5
1998 4.1 4.4 4.2 4.5
TOTAL
Seasonal
Average
 MEASUREMENT OF SEASONAL VARIATION : Ratio to Trend Method
This method isolates the seasonal factor after eliminating the trend from time series. Trend is eliminated by
computing the ratios, and random elements disappear when the ratios are averaged. The steps involved in
computation of index are:

1. Determine trend values by the method of least squares.

2. Divide the original data month by month (quarter by quarter) by the corresponding trend values and
express them as percentages.
3. Average the different values for a month/quarter.
4. Adjust all these averages to a total of 1200.
5. If the data is distributed in quarters instead of months, adjust the respective averages to a total of 400.

It may be noted here that step (2) eliminates the trend and the values so obtained include cyclical and irregular
variations. Step (3) frees the values from cyclical and irregular variations.
 MEASUREMENT OF SEASONAL VARIATION : Ratio to Trend Method
EXAMPLE: Find seasonal variations by the Ratio-to-Trend Method from the data given below:

Quarterly Data
Year
I II III IV
1990 30 40 36 34
1991 34 52 50 44
1992 40 58 54 48
1993 54 76 68 62
1994 80 92 86 82
Solution: Table : 1/3
Deviations Quarterly
Year Yearly from 1992 Averages XY 𝑿𝟐 𝒀𝒄
Totals (X) (Y)
1990
1991
1992
1993
1994

Total
Solution: Table : 2/3

Trend Values
Year
Quarter I Quarter II Quarter III Quarter IV
1990
1991
1992
1993
1994
 Quarterly Data (Observed Values) Quarterly Data (Trend Values)
Year Year
I II III IV I II III IV
1990 30 40 36 34 1990 27.5 30.5 33.5 36.5
1991 34 52 50 44 1991 39.5 42.5 45.5 48.5
1992 40 58 54 48 1992 51.5 54.5 57.5 60.5
1993 54 76 68 62 1993 63.5 66.5 69.5 72.5
1994 80 92 86 82 1994 75.5 78.5 81.5 84.5

Ratio to Trend (Percentage of observed values on trend values)

Year
Quarter I Quarter II Quarter III Quarter IV
1990
1991
1992
Table : 3/3 1993
1994
TOTAL
Average
Adjusted Seasonal Index
 MEASUREMENT OF SEASONAL VARIATION : Ratio to Moving Average Method
Under this method, firstly the trend is eliminated from the data, and the resulting data representing the short-

term oscillations is expressed as percentages of the original figures. These percentages are then averaged to

obtain the seasonal index. The total of percentages of the monthly averages should be 1200. If the computed

total of these averages is more or less than 1200, the figures are so adjusted as to make the total equal to 1200

(if quarter then adjusted sum of percentage = 400).

Note that under this method, the trend of the data set is calculated by the moving averages. If data is monthly,

12 months centred moving averages are calculated, and if the data is given in quarters, 4 quarterly centred

moving average is calculated. The resultant moving averages are trend figures of the data set.
 MEASUREMENT OF SEASONAL VARIATION : Ratio to Moving Average Method
EXAMPLE: Find seasonal variations by the Ratio-to-Moving Average Method from the data given below:

Quarterly Data
Year
I II III IV

2005 68 62 61 63

2006 65 58 66 61

2007 68 63 63 67
Year Quarter Data 4-Quarter Moving 2-figure Total 4-year moving Percentage of
Total average moving average

I 68

II 62
2005
III 61

IV 63

I 65

II 58
2006
III 66

IV 61

I 68

II 63
2007
III 63

IV 67
 MEASUREMENT OF SEASONAL VARIATION : Ratio to Moving Average Method
EXAMPLE: Find seasonal variations by the Ratio-to-Moving Average Method from the data given below:

Ratio to Moving Average (Percentage of Moving Average on Observed Values)

Year
I II III IV

2005

2006

2007

TOTAL

AVERAGE

ADJUSTED
SEASONAL INDEX