Deseasonalizing a Time Series
The Ratio to Average Method allows us to identify the component of the seasonal variation in time series data and the indices themselves help us to nullify the effects of seasonality on the time series. The use of indices to nullify the seasonal effects in the common parlance is referred to as Deseasonalizing the time series. Deseasonalizing a time series involves dividing the original data points with the relevant seasonal index expressed as a percentage.
In our example, the deseasonalization process is carried out as follows.
|
Year
|
Quarter
|
Actual Data
|
Seasonal index/100
|
Deseasonalized data
|
|
|
(1)
|
(2)
|
(3)
|
(4)
|
(5) = (3)/(4)
|
|
|
1998
|
I
|
1
|
112.73/100
|
0.887
|
|
|
II
|
2
|
98.41/100
|
2.032
|
|
|
III
|
2
|
118.10/100
|
1.693
|
|
|
IV
|
1
|
70.77/100
|
1.413
|
|
The removal of the seasonal component helps us to analyze the components of secular, cyclical and irregular variations. The secular trend then obtained can be utilized for projecting the trend into the future.