Forums › Ask ACCA Tutor Forums › Ask the Tutor ACCA MA – FIA FMA › Time series analysis
- This topic has 7 replies, 2 voices, and was last updated 4 years ago by John Moffat.
- AuthorPosts
- April 16, 2019 at 3:13 pm #513095
Question: Unemployment numbers actually recorded in a town for the second quarter of the year 2000 were 4700. The underlying trend at this point was 4300 people and the seasonal factor is 0.92. Using the multiplicative model for seasonal adjustment, what is the seasonally-adjusted figure (in whole numbers) for the quarter?
April 16, 2019 at 6:13 pm #513132Please do not simply set test questions and expect an answer.
You must have an answer in the same book in which you found the question, so ask about whatever it is in the answer that you are not clear about and then I will explain.
The actual unemployment figure is 4,700. However had it is lower than it would have been expected because of the seasonality – it is only 0.92 of what it would have been without seasonality.
Therefore the seasonally adjusted figure is 4,700 / 0.92.April 17, 2019 at 8:57 am #513224@czxy said:
Question: Unemployment numbers actually recorded in a town for the second quarter of the year 2000 were 4700. The underlying trend at this point was 4300 people and the seasonal factor is 0.92. Using the multiplicative model for seasonal adjustment, what is the seasonally-adjusted figure (in whole numbers) for the quarter?In the book the answer is simply given as 4700/0.92, but why is actual amount 4700 not multiplied by 0.92 to get the forecast trend — why is the process in reverse?
Also why the seasonal factor is 0.92? It should be 4700/4300 and then multiplied by 100 which is equal to 109.3%. And so the forecast trend should have been calculated as 4700/1.09. Why this is not the case and what is the relationship between 4700 and 4300?April 17, 2019 at 3:33 pm #513259Firstly, 0.92 is the same as 92%.
If the question had asked you to forecast the second quarter of the year 2,000, then you would have taken 0.92 x the trend of 4,300.
However, this is not what was asked. As I wrote in my previous reply, the actual unemployment figure is 4,700. However had it is lower than it would have been expected because of the seasonality – it is only 0.92 of what it would have been without seasonality.
Therefore the seasonally adjusted figure is 4,700 / 0.92.April 17, 2019 at 4:46 pm #513287The seasonally adjusted figure is also called the trend or centered average and this figure of 4300 is also stated as trend in the question. So what is the difference?
April 18, 2019 at 6:05 am #513336No – the seasonally adjusted figure is not the same as the trend.
The trend is the ‘average’ per period in the past (which we can use to forecast for the future).
The seasonally adjusted figure is the actual result with the seasonality adjusted for.
See my previous explanation.April 18, 2019 at 7:19 am #513355So this trend ( the average per period) is the same thing as the moving average?
April 18, 2019 at 2:51 pm #513398The trend is the centred average (if there are an even number of periods then the moving average needs centering, if there are an odd number of periods, then it is already centred.
- AuthorPosts
- You must be logged in to reply to this topic.