Time Series Components Lecture 02

 Components of Time Series

Lecture 02

Introduction

The factors that influence the data gathered over time are called components of time series. These components are helpful to study the pattern and behaviour of the time series, which is very useful in forecasting and informing the decision-making process.

The components that can be distinguished from the observed data are considered to have an impact on a time series. Following are the four main components:

Secular Trend

Cyclical Fluctuation

Seasonal Variations 

Random or Accidental Variations.

Secular Trend

The secular trend indicates the general pattern or direction of change in the observed data of an activity; it is a long-term movement or line that has persisted for at least 12 years. It also shows a slow or gradual increase or decrease in the data. The secular trend is represented by a straight line or any handy curve. Therefore, a trend in a time series can be defined as a long-term movement. For instance, there is an upward tendency in the population growth and demand for housing in Pakistan, where a downward trend can be observed in series relating to epidemic mortality rates.

Graphically, the secular trend is represented as

Advantages of Secular Trend

i. It indicates the basic tendency of growth or decline.

ii. It makes it possible to compare two or more time series at a time and draw useful conclusions.

iii. It is helpful in analyzing short-term variation.

iv. It is helpful in forecasting.

Cyclical Variation

The cyclical variations are medium-term changes in time series data caused by circumstances that repeat in cycles. Most of the economic and financial time series show some kind of cyclical variation. e.g., a business cycle consists of four phases, viz. i) Prosperity, ii) Decline, iii) Depression, and iv) Recovery.

A cycle is an economic or commercial cycle that oscillates around a trendline-related long-period oscillation. Cycle fluctuation indicates medium-range expansion and contraction in an activity.

A time of prosperity is followed by a decline in economic activity known as a recession. If things stay as they are, a period of depression will set in during which economic activity will have reached its lowest point. After some time, the forces of recovery come into play, and economic activity once again picks up momentum and eventually transitions into a prosperous period.

 

 Main Causes of Cyclical Fluctuation

i. Price increases or decreases.
ii. A rise in the pay rates
iii. Supply and money concentration.
vi. High bank interest rate.
v. Introduction of novel and efficient products, etc.

Advantages

i. It aids in keeping the prosperity phase stable.

ii. Fashion designers and policymakers benefit from it.

iii. It makes audacious choices for the future and can be utilised to identify turning points.

Seasonal Variation

The seasonal variations are short-term variations that take place on a regular basis throughout the course of a year. These variations are brought on by yearly recurring, predictable events like seasonal weather shifts and religious events.

These variations may occur due to:

Climatic and Natural Causes: The weather and natural forces affect the different economic activities differently.

For example, wool and woollen products have a good demand during winter, while fans and A.Cs have good demand in summer.

Religious and Social Causes: The religious and social events like Hajj, Eid, New Year’s ceremonies, etc. also affect the different economic activities differently.

Advantages 

The study of seasonal variation is very useful to the sales managers.

The consumer purchases certain articles in the off-season at discount prices.

The main objective of seasonal variation is to study their effect and isolate the trend.

 Accidental or Random Variations

The accidental variation is short-term, irregular, and unsystematic in nature and affects the economic activity “Yt.” Random variation is caused by some unusual events like flood, strikes, earthquakes, droughts,  and blasts, etc.

There is no formally defined statistical method for the measurement of accidental variation.

Examples

i. An abrupt decline in sales as a result of a nearby blackout of electricity.

ii. An instantaneous increase in website traffic once a social media post goes viral.

iii. The effect on a local economy of a natural calamity.

iv. Stock prices may fluctuate temporarily as a result of a significant news event.

• Read More: Time Series Decomposition