Business Mathematics and Statistics · Ch 9 — Applied Statistics (Time Series, Index Numbers, Statistical Quality Control)
Time Series: Meaning and Components
Time Series: Meaning and Components
A time series is a set of numerical values (sales, prices, production, rainfall, population, and so on) recorded at successive, usually equal, intervals of time — daily, weekly, monthly, quarterly, or yearly. Writing the values against time and looking at how they move over the years is one of the most widely used tools in business forecasting, and this syllabus's treatment of time series draws on the same statistical principles taught across Indian commerce and statistics curricula.
A time series is usually thought of as the combined effect of four components:
| Component | What it captures | Typical example |
|---|---|---|
| Secular Trend () | The smooth, long-term direction (upward, downward, or flat) of the series over many years, ignoring short-term ups and downs | Steady rise in a country's per-capita income over 30 years |
| Seasonal Variation () | A regular, short-term pattern that repeats within a fixed period — usually a year, but sometimes a week or a day | Umbrella sales rising every monsoon quarter |
| Cyclical Variation () | Wave-like swings that repeat over a longer, less regular period than a year — usually tied to the business cycle (boom, recession, depression, recovery) | A 6-8 year swing in industrial output |
| Irregular / Random Variation () | Sudden, unpredictable movements with no regular pattern, caused by one-off events | A factory fire, a pandemic lockdown, a sudden strike |
The classical (multiplicative) model expresses an observed value as
so that once trend is separated out (the focus of this chapter), the remaining components can, in principle, be studied on their own. The practical value of decomposing a series this way is forecasting: a business that knows its secular trend and its seasonal pattern can plan production, staffing and inventory well ahead, while treating irregular shocks honestly as the part of the series that cannot be predicted.
A set of observations of a variable recorded at successive points or intervals of time, arranged in chronological order.
The general, smooth long-term movement (increase, decrease, or stagnation) shown by a time series over a fairly long period, after short-term fluctuations are ignored.
A regular, periodic fluctuation in a time series that completes itself within a period of one year (or shorter) and repeats year after year, usually caused by climate, festivals or social custom.
A wave-like oscillation in a time series with a period usually longer than a year, generally associated with alternating phases of the business cycle.
An unpredictable, non-recurring movement in a time series caused by unforeseen events such as strikes, wars, floods, or accidents.