Time Series Analysis – A First Look
Imagine you track your monthly pocket money for two years. You get ₹500 in January, ₹600 in February, ₹550 in March, and so on. If you plot these numbers on a graph with months on the horizontal axis and rupees on the vertical axis, you get a time series – a sequence of data points recorded at regular intervals over time.
That is the everyday intuition. Now the precise meaning.
What Exactly Is a Time Series?
A time series is a set of observations on a variable (or several variables) collected at successive points in time, usually at equal intervals. In economics, those intervals are typically daily, monthly, quarterly, or yearly.
The key idea is that the order of the data matters. If you shuffle the months, you destroy the pattern. Time series analysis is the set of tools economists use to study such data – to understand what drives the ups and downs, to detect long-term trends, and to forecast what might happen next.
Why Does It Matter in Economics?
Economic data almost always comes as time series: GDP every quarter, inflation every month, unemployment every month, stock prices every day. A government deciding the next budget needs to know whether the economy is growing or slowing. A business planning production needs to know whether demand is rising or falling. Time series analysis gives them a systematic way to answer those questions.
The Four Components of a Time Series
When you look at any economic time series, you can usually break it into four parts:
- Trend (T) – The long-term direction. Is GDP generally rising over decades? That is the trend.
- Seasonal variation (S) – Regular patterns that repeat every year. Ice cream sales peak every summer. Tax collections peak every March. These are seasonal effects.
- Cyclical variation (C) – Ups and downs that last longer than a year, typically tied to the business cycle. A recession followed by a boom is a cycle.
- Irregular or random variation (I) – One-off shocks: a flood, a war, a sudden policy change. These cannot be predicted.
In most Class 12 Economics syllabi, you are expected to recognise these four components and understand that the observed value of a variable at time t can be thought of as:
Yt=Tt+St+Ct+It
(additive model) or sometimes as a product (multiplicative model). The additive model works when the seasonal swings are roughly constant in size; the multiplicative model works when they grow with the trend.
How Is It Used? A Simple Example
Suppose you have monthly sales data for a shop for three years. You want to know whether sales are genuinely growing or just fluctuating seasonally. …