Lecture 1

TIME SERIES AND THEIR COMPONENTS

A time series is a set of numerical data that is obtained at regular periods

over time.

For example, the daily closing prices of a particular stock on Karachi

Stock Exchange, the monthly total number of passengers carried by rail
constitutes a time series.
 Objectives

The basic assumption underlying time-series analysis is that the factors

that have influenced patterns of activity in the past and present will
continue to do so in more or less the same manner in future. Thus, the
major goals of time-series analysis are to identify and isolate these
influencing factors for predictive (forecasting) purposes as well as for
managerial planning and control.

One of the most time-consuming and difficult parts of forecasting is the

collection of valid and reliable data. Four criteria can be applied to the
determination of whether data will be useful.

1. Data should be reliable and accurate.

2. Data should be relevant.
3. Data should be consistent.
4. Data should be timely.
 Components of time series

There are four basic types of movements for time series and are known as
Components of time series.

1. Trend Component or Secular trend (T)

2. Seasonal component or seasonal variation (S)
3. Cyclical component or cyclical fluctuation (C)
4. Irregular component or random variation. (I).
Trend component (T): A secular trend is a long term upward or
downward pattern of movement that persists for many years; not less than
10 years, e.g. sales, industrial production, capital formation, etc.
Reason for Influence: Changes in technology, population, wealth, value,
etc.
Seasonal component (S): Fairly regular short term periodic fluctuations
that occur within each 12-month period year after year (monthly or
quarterly data), e.g. the sales of soft drinks which are high in summer but
low in the winter.

Reason for Influence: weather conditions, social customs, religious

customs, etc.
 Cyclical component
(C): Repeating up-and-
down swings or movements
through four phases:
from peak (prosperity) to
contraction (recession)
to trough (depression) to
expansion (recovery or
growth), within 2 – 10 years with differing intensity for a complete cycle.
Reason for Influence: Interaction of numerous combinations of factors
influencing the economy.

A Typical Product Life Cycle

Similar products and markets often display similar growth patterns or life
cycles. The S-shaped product life cycle is a typical example. It is generally
divided into four stages:

 Introduction - a period of slow growth while the product is

introduced onto the market.
 Growth - sales rapidly increase at an increasing growth rate as
the market accepts the product.
 Maturity - sales increase slowly but with a decreasing growth
rate. The product has now been accepted by the majority of the
people that are likely to buy it.
  Decline - a decline in sales caused by changes in tastes,
increased competition or a shift away from your product towards
a new or improved product.

As stated earlier, the ups and downs in business activities are the
effects of cyclical variation. A business cycle showing these
oscillatory movements has to pass through four phases-prosperity,
recession, depression and recovery. In a business, these four phases
are completed by passing one to another in this order.
Irregular Component (I): The erratic or residual fluctuations in a time
series that exist after taking into account the systematic effects -- trend,
seasonal and cyclical. These variations occur in completely unpredictable
manner for short duration and non-repeating and also called accidental
variations. Irregular variation may be divided into two components:
  Episodic fluctuations
 Residual fluctuations.

Monthly Value of Building Approvals, Australian Capital Territory (ACT)

Episodic fluctuations: Unpredictable variation in a time series that is due

to unusual causes that can be identified such as strikes, tornado damage, or
fire.

Episodic variations
are unpredictable, but they can be identified. Major floods, hurricanes, or
strikes are examples of unpredictable events but it is possible to identify
the time period in which they happened.
Residual fluctuations: Unpredictable variation in a time series that cannot
be identified.

Reasons for Influence: Random variation in data due to unforeseen events

such as strikes, hurricanes, war, political assassinations, etc.

To study the components of a time series, the analyst must consider

how the components relate to the original series. This task is accomplished
by specifying a model (mathematical relationship) that expresses the time
series variable Y in terms of the components T, C, S, and I.
 A model that treats the time series values as a sum of the components
is called an additive components model. A model that treats the time series
values as the product of the components is called a multiplicative
components model. Both models are sometimes referred to as unobserved
components models, since, in practice, although we observe the values of
the time series, the values of the components are not observed. The
approach to time series analysis is: given the observed series, estimate the
values of the components. These estimates can be used in the process what
is called seasonal adjustment. It is difficult to deal with the cyclical
component of a time series. To the extent that cycles can be determined
from historical data, both their lengths (measured in years) and their
magnitudes (differences between highs and lows) are far from constant.
This lack of a consistent wavelike pattern makes distinguishing cycles
from smoothly evolving trends difficult. Consequently, to keep things
relatively simple, we will assume any cycle in the data is part of the trend.
Initially, then, we consider only three components, T, S, and I.

The two simplest models relating the observed value (Y ) of a time series
t
 to the trend (T ), seasonal ( S ), and irregular ( I ) components are the
t t t

additive components model:

Y t =T t +S t + I t

and the multiplicative components model:

Y t =T t × S t × I t

The additive components model works best when the time series being
analyzed has roughly the same variability throughout the length of the
series. That is, all the values of the series fall essentially within a band of
constant width centered on the trend.
The multiplicative components model works best when the variability of
the time series increases with the level. That is, the values of the series
spread out as the trend increases, and the set of observations has the
appearance of a megaphone or funnel.
 1,000
Time Series with Constant Variability (Top) and a Time
Series with Variability Increasing with Level (Bottom)
900
Milk Production

800

700

600

Now, we briefly talk about linear trend.

Later, we will discuss nonlinear trend. There
50 100 150
Month
900

800
are various methods to compute the trend.
700 Among them two well-known methods are
600
Monthly Sales

500

400 i. k-Moving averages in which trend is

computed by
300

200
100

0 Y t +Y t−1+ …+Y t −k+1

10 20 30 40 50 60 70
T^ t=
k
Month

where T^ = forecast value for next period ,

t
Yt = actual value at period t,
k = number of terms in moving average.
ii. Fitting least square regression line by
T^ t=a+bt

where
a=Ý −b t́

and
b=
n ∑ tY −∑Y ∑t
n ∑t 2 −¿ ¿ ¿
.

Since the independent variable, time, is equally spaced, it can be coded as

i = 1, 2, 3, . . . . Thus calculation can be made easier by using the
following formulas:
n
n (n+1)
∑ i= 2
i=1

and
 n
n(n+1)(2n+1)
∑ i 2= 6
i

Method of moving averages

To smooth out large variations

To take away the short-term seasonal and irregular variations
Computed by averaging the observations over a certain
number of time periods. The same number of time periods
is used in each successive average
It is the estimate of the long-run average of the variable.

Example 1: Cedar Fair operates seven amusement parks and five

separately gated water parks. Its combined attendance (in thousands) for
the last 8 years is given in the following table. A partner asks you to
study the trend in attendance. Compute a three-year moving average
and a three-year weighted moving average with weights of 0.2, 0.3, and
0.5 for successive years.
 Year Attendance Moving Weighted
(000) Averages Moving
Averages
2000 11,703 * *
2001 11,890 11991.0
12097.6
2002 12,380 12150.3
12182.5
2003 12,181 12372.7
12408.8
2004 12,557 12479.3
12553.3
2005 12,700 14852.3
15971.4
2006 19,300 18033.3
19380
2007 22,100 * *

Example 2: The table below shows the quarterly sales for Toys
International for the years 2001 through 2006. The sales are reported in
millions of dollars. Determine quarterly moving averages.
 Year Quarter Sales Four-Quarter Centered
($ moving average Moving
millions)
Average
2001 Winter 6.7 * *
Spring 4.6 * *
8.500
Summer 10.0 8.4750
8.450
Fall 12.7 8.4500
8.450
2002 Winter 6.5 8.4250
8.400
Spring 4.6 8.5125
8.625
Summer 9.8 8.6750
8.725
Fall 13.6 8.7750
8.825
2003 Winter 6.9 8.9000
8.975
 Spring 5.0 9.0375
9.100
Summer 10.4 9.1125
9.125
Fall 14.1 9.1875
9.250
2004 Winter 7.0 9.3000
9.350
Spring 5.5 9.4625
9.575
Summer 10.8 9.5875
9.600
Fall 15.0 9.6250
9.650
2005 Winter 7.1 9.6875
9.725
Spring 5.7 9.6625
9.600
Summer 11.1 9.7125
9.825
Fall 14.5 9.8875
9.950
2006 Winter 8.0 9.9875
10.025
Spring 6.2 10.0750
10.125
Summer 11.4 * *
Fall 14.9 * *
 Sales (millions)

Year Coded variable Yt t Yt

t = Year-2000

2001 1 125

2002 2 113

2003 3 189

2004 4 201

2005 5 206
 2006 6 241

2007 7 230

2008 8 245

2009 9 271

2010 10 291

Total

Need:∑ t , ∑ t2 , ∑ tY ∧∑ Y