# Time Series analysis

#### Time series analysis: please note that this lecture relates to Chapter 12 of the Course Notes (and not Chapter 11 as stated in the lecture)

• says

I have absolutely no idea – I do not write the exam!!

You have presumably read the first paragraph of the chapter in the free Lecture Notes (because there is no point in watching the lectures without the lecture notes in front of you) and you will therefore know that the examiner has said that she will not ask calculations. The most she can expect is an understanding of the idea (it is assumed knowledge from Paper F2).

1. says

This december will be the sixth time i will sit for F5 and it is the first time I understand Time Series. Thanks John

2. says

I am currently using BPPs new study text for dec… time series is not in the text.. will time serious still be tested for dec sitting?

• says

If you read at the introductory paragraph of the chapter in our Course Notes on Quanititative Techniques (which are to be used with this lecture) it explains that Time Series (and Regression Analysis) are assumed knowledge from Paper F2.
The examiner has said that she will not ask calculations, but you are expected simply to be aware of the idea.
We show the calculations just to make sense of the idea, but again you cannot be asked calculations in the exam.

• says

Sorry for the trouble, I had not looked at the class notes as yet… will download ASAP… thank you for the Clarification

3. says

Please make me understand the logic here when calculating seasonal variation,are subtracting actual from the trend or the trend from the actual?

• says

It is actual minus trend.

However, as it states clearly in the notes, you will not be asked arithmetic on time series analysis.

4. says

Hi John,

I hope you are well… It’s been a long time since studying this sort of material at school. I just wanted you to clarify if you may for me please.

You said partway through the lecture that you would need to calculate the moving average and then the centered average/trend if its a 4 quarter period. However, if it was a three part period, the trend would already be calculated.

I am somewhat confused. No doubt it is very basic but your help would be much appreciated.

I am now going back to look at F2 – Regression Analysis but as mentioned, your help would be great.

Thank you.

Chris.

• says

The reason is the the middle of 1, 2, 3, 4 is in between 2 and 4. However the middle of 1, 2, 3 is 3.

I am happy to explain more, but I do not want to waste your time. As it says in the Course Notes, in Paper F5 you cannot be expected to do calculations. All that is expected is that you are aware of the idea (and even asking that is not very likely in Paper F5).

• says

Hi John, it’s me again 😀

I still don’t understand this… Are you able to clarify again… Sorry

• says

Think about it this way. Quarter 1 is Jan, Feb, Mar. Quarter 2 is Apr, May, Jun, and so on. Assume the sales each month are rising steadily.
When we calculate the average quarterly sales, we add up the four quarters and divide by 4. This tells us what the sales would be for 3 months in the middle of the year (earlier in the year will be lower, and later in the year will be higher). However the very middle three months of the year will be June, July and August (5 months before and 5 months after – these 3 are in the middle). So……we cannot compare these three months with the actual sales for any of the individual quarters – none of the quarters I listed in the first sentence covers Jun, July, August. That is why we need to take a second average.

• says

Thanks John… I Appreciate it’s not asked but better to understand than just simply nod head and quickly turn over the page

Best wishes.

Chris.

5. says

What do you mean by:

“Forecast the Trend, Adjust by reevant average seasonal variation”?

• says

The trend is the basic pattern (without the seasonal variation).
You can forecast this, but the actual forecast will be higher or lower than the trend because of the seasonal variation.

So you are being asked to forecast the trend, and then to adjust it (higher or lower) to account for the seasonal variation.