Statistics FC301: IME Eries

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 24

Statistics

FC301

TIME SERIES
Time Series Data
• Cross Sectional data: collected at one point in time

• Time series data is a sequence of observations which


are ordered in time (space)

• Examples in Business and Economics


- Daily: Stock market, currency prices
- Weekly: Money supply
- Quarterly: Balance of payments
- Annually: Budget

2
Time Series Data
• Basic assumption of time-series: factors that
have influenced activities in the past and
present will continue to do so in more or less
the same way in the future

• Major goal of time series is to identify and


isolate these influencing factors

3
Factors Influencing Time Series Data
Component Definition Reason for Influence Duration

Trend Overall long-term Changes in Several Years


upward or downward technology,
pattern of movement population or wealth

Seasonal Regular periodic Weather conditions, Within 12 months (or


influences that occur social customs, school monthly or quarterly
within each 12 month schedules etc data)
period
Cyclical Up and down swings Interactions of Usually 2-10 years
through four periods: numerous with differing intensity
peak, contraction, combinations of for a complete cycle
trough, expansion factors influencing the
economy
Irregular Other movements Unforeseen events Short duration and
after accounting for such as strikes, wars non repeating
other three factors and disasters
4
Example 1: Moving Averages (MA)

• With an even number of values in each


period, centred Moving Averages (CMA) must
be calculated
Step 1
• The next slide shows a 4 period Moving
Average for: Quarterly Productivity 1992-
1995

5
Year Sales Calculation of MA MA Calculation of CMA
CMA
1992 Q1 81
Q2 84 (Q1+Q2+Q3+Q4)/4 82
Q3 78 (Q2+Q3+Q4+Q1 (1993) )/4 81.5 (82+81.5)/2= 81.75
Q4 85 (Q3+Q4+Q1 +Q2)/4 82.25 (81.5+82.25)/2= 81.875
1993 Q1 79 (Q4+Q1 +Q2+Q3)/4 83.25 (82.25+83.25)/2= 82.75
Q2 87 (Q1 + Q2+ Q3+ Q4)/4 84.25 (83.25+84.25)/2= 83.75
Q3 82 (Q2 +Q3+Q4+Q1 (1994) )/4 85.25 (84.25+85.25)/2= 84.75
Q4 89 (Q3 + Q4 + Q1 + Q2)/4 86 (85.25+86)/2= 85.625
1994 Q1 83 (Q4 + Q1+ Q2+Q3)/4 85.75 (86+85.75)/2= 85.875
Q2 90 (Q1 + Q2 + Q3 +Q4)/4 86.75 (85.75+86.75)/2= 86.25
Q3 81 (Q2 +Q3+Q4+Q1 (1995) )/4 87 (86.75+87)/2= 86.875
Q4 93 (Q3+Q4+Q1+Q2)/4 87.75 (87+87.75)/2= 87.375
1995 Q1 84 (Q4 + Q1+Q2+Q3)/4 88.25 (87.75+88.25)/2= 88
Q2 93 (Q1+Q2+Q3+Q4)/4 89.25 (88.25+89.25)/2= 88.75
Q3 83
Q4 97 6
Step 2: Plotting the graph
We need to plot two things on the same graph:

1. The sales (or whatever the variable is in your


data) against the quarters
2. The CMAs against their appropriate quarters
in order to get a trend

Let’s do this for the data on previous slide:


7
Sales over time
100

90
Sales

80 Sales
CMA

70
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
1992 1993 1994 1995
Quarters
You must know how this graph is constructed from the data on slide
6 and be able to construct it yourself. Notice that the first CMA is in
Q3 of the first year!
8
Step 3: Drawing a Trend line and making a
Prediction
• The next step once you have the graph on the
previous slide is to use it to draw a trend line
(a line of best fit through the CMAs) and
using it to make a prediction.

Let’s say we want to predict the Sales in Q2 of


1996. We must extend our horizontal axis and
see what the trend line tells us. See next slide:
9
Our trend line
predicts: 92 sales in
Q2 in 1996 Sales over time
100
Trend line

92
90
Sales

Sales
80
CMA

70
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2
1992 1993 1994 1995 1996
Quarters
10
Step 4: Analyse your time series graph

You will always be asked to comment on time


series graphs. Your comments should pick out
things such as:
• What is the highest and lowest value for your
variable?
• What is the overall trend?
• Is there a quarter that is the highest or lowest
each year?
11
Summary
For time series you need to be able to perform the
following steps:
1. Calculate the moving and central moving
averages (MAs and CMAs) given some data.
2. Plot a graph with the original values plotted and
your CMAs.
3. Draw a trend line through your CMAs and know
how to use it to predict future outcomes
4. Analyse your graph
12
Example 2: Moving Averages (MA)
Mr Jones says that his Gas bill is getting more expensive.
Mrs Jones says it is actually getting cheaper. Below are
Mr Jones’ gas bills for a 2 year period. Use Time series
data to show:
who do you think is right? Why?
Hint: As the data shows 4 separate periods in each year,
calculate a four-point moving average.
Jan 10 – Apr 10 – Jul 10 – Oct 10 – Jan 11 – Apr 11 – Jul 11 – Oct 11 –
Quarter
Mar 10 Jun 10 Sep 10 Dec 10 Mar 11 Jun 11 Sep 11 Dec 11
Bill
122.24 87.08 73.04 118.46 136.80 96.34 80.06 135.96
(£’s)
Example 2: Answer- Moving Averages (MA)

Jan 10 – Apr 10 – Jul 10 – Oct 10 – Jan 11 – Apr 11 – Jul 11 – Oct 11 –


Quarter
Mar 10 Jun 10 Sep 10 Dec 10 Mar 11 Jun 11 Sep 11 Dec 11
Bill
122.24 87.08 73.04 118.46 136.80 96.34 80.06 135.96
(£’s)

1st four-point Moving average: (122.24 + 87.08 + 73.04 + 118.46) ÷ 4 = 100.21

2nd four-point Moving average: (87.08 + 73.04 + 118.46 + 136.80) ÷ 4 = 103.85

3rd four-point Moving average: (73.04 + 118.46 + 136.80 + 96.34) ÷ 4 = 106.16

4th four-point Moving average: (118.46 + 136.80 + 96.34 + 80.06 ) ÷ 4 = 107.92

5th four-point Moving average: (136.80 + 96.34 + 80.06 + 135.96) ÷ 4 = 112.29

Recommend: show your results on a Time Series Table:


Example 2: Answer- Moving Averages (MA)
Jan 10 – Apr 10 – Jul 10 – Oct 10 – Jan 11 – Apr 11 – Jul 11 – Oct 11 –
Quarter
Mar 10 Jun 10 Sep 10 Dec 10 Mar 11 Jun 11 Sep 11 Dec 11
Bill
122.24 87.08 73.04 118.46 136.80 96.34 80.06 135.96
(£’s)

Year Gas bill MA Calculation of CMA CMA

2010 Q1 122.24
Q2 87.08 100.21
Q3 73.04 103.85 (100.21+103.85)/2= 102.03
Q4 118.46 106.16 (103.85+106.16)/2= 105.005
2011 Q1 136.80 107.92 (106.16+107.92)/2= 107.04
Q2 96.34 112.29 (107.92+112.29)/2= 110.105
Q3 80.06
Q4 135.96

Recommend: show the results from the table on a Time Series Graph
and draw a trend line:
Example 2: Answer- Moving Averages (MA)
Time Series Graph for Gas Bills
160

140

120

100
Cost (£s)

80

60

40

20

0
Example 2: Answer- Moving Averages (MA)

Remember to interpret the graph to answer the question:


who do you think is right? Why?
By using the trend line, we see the gas bill is getting slightly more expensive
during this period of time. Therefore Mr Jones is correct.
Example 3: Moving Averages (MA)
Mrs Smith is the Manager of ‘Smith’s Toy Shop’. She is
keen to see whether her sales are improving. Every 6
months she does a review of the shop’s sales. The results
are shown in the table below:

Oct 07 – Apr 08 – Oct 08 – Apr 09 – Oct 09 – Apr 10 – Oct 10 – Apr 11 –


Period
Mar 08 Sept 08 Mar 09 Sept 09 Mar 10 Sept 10 Mar 11 Sept 11
Sales
120.3 57.2 134.8 60.7 146.9 68.5 157.1 74.2
(£1,000’s)

Mrs Smith has got an appointment with her accountant


tomorrow. Should he be worried?
Example 3: Answer- Moving Averages (MA)
As there are 2 separate periods in each year we would
calculate a two-point moving average.

Oct 07 – Apr 08 – Oct 08 – Apr 09 – Oct 09 – Apr 10 – Oct 10 – Apr 11 –


Period
Mar 08 Sept 08 Mar 09 Sept 09 Mar 10 Sept 10 Mar 11 Sept 11
Sales
120.3 57.2 134.8 60.7 146.9 68.5 157.1 74.2
(£1,000’s)

1st two-point Moving average: (120.3 + 57.2) ÷ 2 = 88.75


2nd two-point Moving average: (57.2 + 134.8) ÷ 2 = 96
3rd two-point Moving average: (134.8 + 60.7) ÷ 2 = 97.75
4th two-point Moving average: (60.7 + 146.9) ÷ 2 = 103.8
5th two-point Moving average: (146.9 + 68.5) ÷ 2 = 107.7
6th two-point Moving average: (68.5 + 157.1) ÷ 2 = 112.8
7th two-point Moving average: (157.1 + 74.2) ÷ 2 = 115.65

Recommend: show your results on a Time Series Table:


Example 3: Answer- Moving Averages (MA)
Time Series Graph for Toy Sales
180

160

140

120
Sales (£100's)

100

80

60

40

20

Draw a trend line to analyse (interpret) the graph :


Example 3: Answer- Moving Averages (MA)
Time Series Graph for Toy Sales
180

160

140

120
Sales (£100's)

100

80

60

40

20

0
Example 3: Answer- Moving Averages (MA)

Remember to interpret the graph to answer the question:


Should he be worried?
By using the trend line, we see the sale is increasing during this period of
time. Therefore, Mrs Smith or her accountant should be fine.
End of Session 1

23
In session 2
COMPLETE WORKSHEET 1

6.9.2- FC301- WORKSHEET 1- Time Series

mini test

24

You might also like