J.K.SHAH CLASSES S.Y.J.C.

- STATISTICS

CHAPTER 4 TIME SERIES

Introduction
 Time series is a sequence of observations made on a variable at regular time
intervals over a specified period of time.
 Time series Analysis help us in monitoring & forecasting data with help of
appropriate statistical models.
 Analysis of time of time series data requires maintaining records of values of the
variable overtime.
 Time series is statistical data that are arranged and presented in chronological
order i.e. over a period of time.

Examples:
1. Monthly, Quarterly OR yearly production of an industrial product.
2. Monthly sales in a departmental store
3. Yearly GDP of a country
4. Daily closing price of a share at a stock exchange

Use of Time series Analysis:

 Used for studying past behaviour of a variable
 Used for forecasting future behaviour of a variable
 Used in evaluating the performance.
 Used in making a comparative study.

Components of Time Series

Systematic Part Unsystematic Part

- Secular Trend Irregular variation

- Seasonal variation
- Cyclical variation
-
Note:
1. Every time series has some or all of these components.
2. Only the systematic components of a time series are useful in forecasting its
future values.

: 316 :
J.K.SHAH CLASSES S.Y.J.C. - STATISTICS
A] Secular Trend or Simple Trend (T)
 It is the long term pattern of time series to move in upward or downward
direction.
 Sector trend shows smooth & regular movement of time series.
 Does not include short term fluctuation but only consist of steady
movement over a long period of time.
 General tendency of a variable to increase, decrease or remain constant in
long term called trend of a variable

Eg.
1) Population of a country has increasing trend over a year.
2) Due to modern technology, agricultural & industrial production is
increasing.

B] Seasonal variation (S)

 Over a span of one year, seasonal variation taken place due to rhythmic
forces which operates in a regular & periodic manner.
These forces have the same or almost similar pattern year after year.
 Seasonal variation could be seen & calculated if the data are recorded
quarterly, monthly, weekly, Daily or hourly basis.
 Seasonal variation is measured with the help of seasonal indices, which are
useful for short term forecasting.
 A bank manager can we such short term forecast in managing cash flow on
diff. days of a week or month.
 In time series with only annual data no seasonal variation.

Eg.
1) Sale of cold drinks rise in summer & fall in winter
2) Sale of clothes might go up in festival like Diwali, Christmas, Navratri etc. as
compared to other days.

C] Cyclical Variation
 Cyclical variation is a long term oscillatory movement in values of a time
series.
 Cyclical variation occurs over a long period.
 One complete round of Oscillation is called a cycle.
 Cyclical variation are also termed an business cycle or trade cycle

: 317 :
J.K.SHAH CLASSES S.Y.J.C. - STATISTICS
Typical business cycle consists of following 4 phases:
1) Prosperity
2) Recession
3) Recovery
4) Depression

D] Irregular Variation
Irregular Variation are unexpected variation in times series caused by
unforeseen events include natural disasters like floods, or Famines, political
events like strike, international events like wars or other conflicts.
 Irregular variation do not follow any patterns & it cannot be predicted in
advance.
 Irregular variation are also known as unexplained variation or unaccounted
variation.
 Mathematical models of Time series
Based on secular trend (T), Seasonal variation (S), Cyclical variation (C) &
Irregular variation (I)

Two standard Mathematical models of Time series

Additive model Multiplicative model

Yt = T + S + C + I Yt = T .S. C. I

: 318 :
J.K.SHAH CLASSES S.Y.J.C. - STATISTICS
Additive Model Multiplicative model
1. Assumes that four component are 1. Does not assume independence of four
independent components.
2. All four components must be 2. Trend (T) expressed as unit of
measured in same unit of measurement and others are expressed as
measurement percentage or relative values.
Hence, are free from unit of measurement.
3. Magnitude of seasonal variation does no 3. Magnitude of seasonal variation
change as the series go up or down. increases as the data value increases &
decreases as data value decreases

 Measurement of secular trend

I) Method of free hand curve (graphical Method)
 Quit flexible to use but its risky
 Involves min. amt. of work.
 Helps to know trend not trend value.

II) Method of moving Averages.

 3 yearly moving Avg.
 4 yearly centered moving Avg.
 5 Yearly moving Avg.

III) Method of least square :

Equation of Trend line are

= (U)

Where,
t − middle ′t′ v lue
h If n odd

Or

(t − me n of middle ′t′v lue)

If n Even
t time (yr) h

Two normal Equation are

: 319 :
J.K.SHAH CLASSES S.Y.J.C. - STATISTICS

OBJECTIVES

I) Choose the correct alternative.

1. Which of the following can’t be a component of a time series?
(a)Seasonality (b) Cyclical
(c) Trend (d) Mean

2. The first step in time series analysis is to

(a) Perform regression calculations
(b) Calculate a moving average
(c) Plot the data on a graph
(d) Identify seasonal variation

3. Time-series analysis is based on the assumption that

(a) Random error terms are normally distributed.
(b) The variable to be forecast and other independent variables are
correlated.
(c) Past patterns in the variable to be forecast will continue unchanged
into the future.
(d) The data do not exhibit a trend.

4. Moving averages are useful in identifying

(a) Seasonal component
(b) Irregular component
(c) Trend component
(d) Cyclical component

5. We can use regression line for past data to forecast future data. We then
use the line which
(a) Minimizes the sum of squared deviations of past data from the line
(b) Minimizes the sum of deviations of past data from the line.
(c) Maximizes the sum of squared deviations of past data from the line
(d) Maximizes the sum of deviations of past data from the line.

6. Which of the following is a major problem for forecasting, especially when

using the method of least squares?
(a) The past cannot be known
(b) The future is not entirely certain
(c) The future exactly follows the patterns of the past
(d) The future may not follow the patterns of the past

: 320 :
J.K.SHAH CLASSES S.Y.J.C. - STATISTICS
7. An overall upward or downward pattern in an annual time series would be
contained in which component of the times series
(a) Trend (b) Cyclical (c) Irregular (d) Seasonal

8. The following trend line equation was developed for annual sales from
1984 to 1990 with 1984 as base or zero year.
Y1 = 500 + 60X (in 1000 Rs). The estimated sales for 1984 (in 1000 Rs) is:
(a) Rs 500 (b) Rs 560 (c) Rs 1,040 (d) Rs 1,100

9. What is a disadvantage of the graphical method of determining a trend

line?
(a) Provides quick approximations
(b) Is subject to human error
(c) Provides accurate forecasts
(d) Is too difficult to calculate

10. Which component of time series refers to erratic time series movements
that follow no recognizable or regular pattern.
(a) Trend (b) Seasonal (c) Cyclical (d) Irregular

II) Fill in the blanks

1. ________components of time series is indicated by a smooth line.
2. component of time series is indicated by periodic variation year after year.
3. ________component of time series is indicated by a long wave spanning
two or more years.
4. ___________component of time series is indicated by up and down
movements without any pattern.
5. Addictive models of time series______ independence of its components.
6. Multiplicative models of time series______ independence of its
components.
7. The simplest method of measuring trend of time series is______
8. The method of measuring trend of time series using only averages is_____.
9. The complicated but efficient method of measuring trend of time series
is______.
10. The graph of time series clearly shows of_____ it is monotone.

: 321 :
J.K.SHAH CLASSES S.Y.J.C. - STATISTICS
III) State whether each of the following is True or False.
1. The secular trend component of time series represents irregular variations.
2. Seasonal variation can be observed over several years.
3. Cyclical variation can occur several times in a year.
4. Irregular variation is not a random component of time series.
5. Additive model of time series does not require the assumption of
independence of its components.
6. Multiplicative model of time series does not require the assumption of
independence of its components.
7. Graphical method of finding trend is very complicated and involves
several calculations.
8. Moving average method of finding trend is very complicated and involves
several calculations.
9. Least squares method of finding trend is very simple and does not
involve any calculations.
10. All the three methods of measuring trend will always give the same results.

Answers:
I) 1. (d)
2. (c)
3. (c)
4. (c)
5. (a)
6. (d)
7. (a)
8. (a)
9. (b)
10. (a)

II) 1. Trend
2. Seasonal
3. Cyclical
4. Irregular
: 322 :
J.K.SHAH CLASSES S.Y.J.C. - STATISTICS
5. Assume
6. Does not assume
7. Graphical
8. Moving average
9. Least square
10. Trend

III) 1. False
2. True
3. False
4. False
5. False
6. True
7. False
8. False
9. False
10. False

: 323 :