TIME SERIES
Definition
This is a sequence of a variable values that change over a uniform set of time. The
variable values represent statistical data while time can be in seconds, hours. days, weeks
etc. Many business and economic studies are based on time series data.
Examples
1. Monthly production level for a company over several years
2. Weekly sales for a chain of supermarkets over a couple of months etc.
Time series components
All-time series contain at least one of the following four components:
1. Secular trend
2. Seasonal variations
3. Cyclical variations
4. Random/ irregular erratic variations
1. Secular trend (T)
This is the general underlying tendency of the time series data to increase, decrease or
remain constant for a long period of time.
The importance of the trend includes the following:
It permits to project past patterns or trend into the future.
It is used to describe a historical pattern in the given data. This may be used to evaluate
the success or failure of a given action.
Identifying the secular trend enables its elimination in the trend component and thus
makes it easier to study other components of the time series.
2. seasonal variations/variations (S)
Are periodic movements of the data where the duration is less than a year. The factors
that mainly cause these variations are: -
a) climatic changes
b) the customs and habits that people follow at different times
The main objective of measuring the seasonal variations is to isolate them so that their
effect can be understood and used for future extrapolation.
3. Cyclical variations/ fluctuations (C)
Are periodic movements within the time series data where the duration is more than a
year. They are not as regular as the seasonal variations but their sequence of change is the
same. The causes of the QUANTITATIVE ANALYSIS
Contact: 0724398053 Page 210
cyclical variations are the four phases of an economic cycle which include: the
boom/peak, decline/downturn, depression/trough and recovery/upswing.
4. random/residual/irregular erratic occurrences (R)
These are completely unpredictable variations within the data caused by unpredictable
events like sickness, machine breakdown, weather conditions, strikes etc. They are non-
recurring influences which cannot be mathematically captured yet they have profound
consequences on a time series.
Time series (decomposition)
This analysis provides techniques that may be used to isolate the four components of a
time series. Decomposition may be used to measure the degree of impact each component
has on the direction of time series itself i.e the influence each component has on the
movement of the time series. In this analysis a standard line diagram representing the
time series data is also plotted. The diagram is known as histogram or a time series plot.
This is a plot of the variable values on the y axis against time points on the x axis
ILLUSTRATION
The data Tue Wed Thur Friday Sat Sun
below
represent
the daily
sales
(sh000)
for
business
is a
week’s
period.
Mon
12 9 11 14 13 10 15
