This document discusses the ratio to trend method for time series analysis.
It involves obtaining trend values through least squares fitting, expressing original data as percentages of the trend values to isolate seasonal components, averaging the percentages for each month/quarter to obtain seasonal indices, and adjusting the indices to total 1200 for monthly data or 400 for quarterly data.
While it attempts to remove cyclical variations, the seasonal indices may be biased if the series exhibits strong cycles. Its advantage over moving averages is that seasonal indices can be obtained for each data point without losing early or late observations.
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
100%(1)100% found this document useful (1 vote)
1K views1 page
Ratio To Trend Method
This document discusses the ratio to trend method for time series analysis.
It involves obtaining trend values through least squares fitting, expressing original data as percentages of the trend values to isolate seasonal components, averaging the percentages for each month/quarter to obtain seasonal indices, and adjusting the indices to total 1200 for monthly data or 400 for quarterly data.
While it attempts to remove cyclical variations, the seasonal indices may be biased if the series exhibits strong cycles. Its advantage over moving averages is that seasonal indices can be obtained for each data point without losing early or late observations.
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 1
TIME S E R I E S A N A L Y S I S 2-47
.6-2, Ratio to Trend Method. This method is an improvement over the
simple nle averages method and is based on the assumption that seasonal variation averages meti San for any given month is constant factor of the trend. The mecasurement of asonal variation by this method consists on the followingmethod seasonal steps (i) Obtain the trend values by the least square by fitting a nathematical curve, straight line or 2nd degree polynomial, etc. data as the percentage of the trend values. i ) Express the original the seasonal, Assiuming the multiplicative model, these percentages will contain components. cyclic and irregular are then wiped out by averaging (ii) The cyclic and irregular components for different months (quarters) if the data are monthly (quarterly), the percentages with indices of seasonal variations. Either arithmetic mean or thus leaving us be used for averaging, but median is preferred to arithmetic mean median can extreme values which are not primarily since the latter gives undue weightage to there are few abnormal values, modified mean (which due to seasonal swings. If after dropping out the extreme or consists of calculating arithmetic mean abnormal values) may be used with advantage. these indices, obtained in step (iii), are adjusted to a total of (iv) Finally, data by multiplying them throughout 1200 for monthly data or 400 for quarterly by a constant k given by 1200 400 and k Total of the indices RTatal of the indices for monthly and quarterly data respectively. at ironing out the Merits and Demerits. Since this method attempts the of averaging, the purpose will be cyclical or irregular components by process or they are are known to be absent accomplished only if the cyclical variations if the series exhibits not so pronounced if present. On the other hand, even obtained by the least square method pronounced cyclical swings, the trend values 12-month moving average and as can never follow the actual data as closely as method are liable to be 'ratio to trend' ucn the seasonal indices obtained by method discussed biased than those obtained by 'ratio to moving average' more in $ 2.5.3. over the moving average method lies The obvious advantage of this method which the data obtained for each month for e lact that 'ratio to trend' can be there is unlike the 'ratio to moving average method, 4vailable and as such, no loss of data. first fitting simplificd to a great extent by a Kemark. The calculations are lso ahtainina the manthlv