Time Series Analysis

MAT 3103: Computational Statistics and Probability

Chapter 8: Time Series Analysis
Time series is a sequence of data points recorded in time order, often taken at successive equally
paced points in time. Data are taken yearly, monthly, weekly, hourly or even by the minute.
Stock prices, Sales demand, website traffic, daily temperatures, quarterly sales etc. are some
examples.

Terms and concepts:

Dependence: Dependence refers to the association of two observations with the same variable,
at prior time points.
Stationarity: Shows the mean value of the series that remains constant over a time period; if
past effects accumulate and the values increase toward infinity, then stationarity is not met.
Differencing: Used to make the series stationary, to De-trend, and to control the auto
correlations; however, some time series analyses do not require differencing and over-
differenced series can produce inaccurate estimates.

Components of a Time Series

1. Trend: is a general direction in which something is developing or changing. A trend can
be upward (uptrend) or downward (downtrend). It is not always necessary that the
increase or decrease is consistently in the same direction in a given period.
2. Seasonality: Predictable pattern that recurs or repeats over regular intervals. Seasonality
is often observed within a year or less.
3. Cycles: Occur when a series follows an up and down pattern that is not seasonal. Cyclical
variations are periodic in nature and repeat themselves like business cycle, which has
four phases (1) Peak, (2) Recession, (3) Trough/Depression, and (4) Expansion.
Seasonality is different from cycles, as seasonal cycles are observed within a calendar
year, while cyclical effects can span duration shorter or longer than a calendar year.
4. Irregular fluctuation: Variations that occur due to sudden causes and are unpredictable.
For example, the rise in prices of food due to war, flood, earthquakes, farmers striking
etc.

1
 Time Series Analysis

Time Series and Forecasting

The long-term tendency of the data to move in an upward or downward direction can be
measured by the following methods:
1. Graphical Method
The values of a time series are plotted on a graph paper by taking time variable on the X-axis and
the values variable on the Y-axis. After this, a smooth curve is drawn with free hand through the
plotted points. The trend line drawn above can be extended to forecast the values.
Advantages
• It is very easy and simple.
• No mathematical calculations are needed.
• It can be used even if the trend is not linear.
Disadvantages
• It is a subjective method.
• Trend values obtained by different statisticians would be different and hence not reliable.

Example 8.1
GDP per capita (in USD) in Bangladesh is listed below:
Year 2011 2012 2013 2014 2015 2016 2017 2018 2019
GDP 836 857 952 1085 1211 1359 1517 1625 1827

Fit a trend line applying graphical method to forecast future GDPs.

2000
1800
1600
1400
1200
1000
800
600
400
200
0
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

2
 Time Series Analysis

2. Semi-Average Method
The series is divided into two equal parts and the average of each part is plotted at the mid-point
of their time duration.
▪ In case the series consists of an even number of years, the series is divisible into two
halves. Find the average of the two parts of the series and place these values in the mid-
year of each of the respective durations.
▪ In case the series consists of odd number of years, it is not possible to divide the series
into two equal halves. The middle year will be omitted. After dividing the data into two
parts, find the arithmetic mean of each part. Thus, we get semi-averages.
▪ The trend values for other years can be computed by successive addition or subtraction
for each year ahead or behind any year.
Advantages
• This method is very simple and easy to understand.
• It does not require many calculations.
Disadvantages
• This method is used only when the trend is linear.
• It is used for calculation of averages and they are affected by extreme values.
Example 8.2
Calculate the trend values using semi-averages methods for the income from the forest
department.
Year 2008 2009 2010 2011 2012 2013
Income (in crores) 46.17 51.65 63.81 70.99 84.91 91.64
Solution:
Year Income 3-year semi total 3-year semi average Trend values
2008 46.17 53.877 - 9.545 = 44.332
2009 51.65 161.63 53.877 44.332 + 9.545 = 53.877
2010 63.81 53.877 + 9.545 = 63.422
2011 70.99 63.422 + 9.545 = 72.967
2012 84.91 247.54 82.513 72.967 + 9.545 = 82.512
2013 91.64 82.512 + 9.545 = 92.057

3
 Time Series Analysis

Difference between the semi-averages = 82.513 – 53.877 = 28.636

Increase in trend value for one year = 28.636 /3 = 9.545
Example 8.3
Calculate the trend values using semi-averages methods for the income from the forest
department.
Year 1951 1961 1971 1981 1991 2001 2011
Income (in crores) 301.2 336.9 412 484.1 558.6 624.1 721.4

Solution:
Year Income 3-year semi total 3-year semi average Trend values
1951 301.2 350.03 – 71.17 = 278.86
1961 336.9 1050.1 350.03 278.86 + 71.17 = 350.03
1971 412 350.03 + 71.17 = 421.2
1981 484.1 421.2 + 71.17 = 492.37
1991 558.6 492.37 + 71.17 = 563.54
2001 624.1 1904.1 634.7 563.54 + 71.17 = 634.71
2011 721.4 634.71 + 71.17 = 705.88

Difference between the semi-averages = 634.7 – 350.03 = 284.67

Increase in trend value for one year = 284.67 / 4 = 71.17
3. Moving Averages Method
Moving averages is a series of arithmetic means of variate values of a sequence. To find the
trend values by the method of 3-yearly moving averages, the following steps have to be
considered:
▪ Add up the values of the first 3 years and place the yearly sum against the median year.
▪ Leave the first-year value, add up the values of the next three years and place it against its
median year.
▪ This process must be continued till all the values of the data are taken for calculation.
▪ Each 3-yearly moving total must be divided by 3 to get the 3-year moving averages,
which are our required trend values.

4
 Time Series Analysis

Advantages
• It can be easily applied.
• It is useful in case of series with periodic fluctuations.
• It does not show different results when used by different persons
• It can be used to find the figures on either extreme; that is, for the past and future years.
Disadvantages
• In non-periodic data this method is less effective.
• Selection of proper ‘period’ or ‘time interval’ for computing moving average is difficult.
• Values for the first few years and as well as for the last few years cannot be found.

Example 8.4
Calculate the 3-years moving averages for the loans issued by co-operative banks for non-farm
sector/small scale industries based on the values given below:
Year 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014
Loan 41.82 40.05 39.12 24.72 26.69 59.66 23.65 28.36 33.31 31.60 36.48

Solution:
Year Income 3-year semi total 3-year semi
average
2004 41.82 --- ---
2005 40.05 120.99 40.33
2006 39.12 103.89 34.63
2007 24.72 90.53 30.18
2008 26.69 111.07 37.02
2009 59.66 110 36.67
2010 23.65 111.67 37.22
2011 28.36 85.32 28.44
2012 33.31 93.27 31.09
2013 31.60 101.39 33.80
2014 36.48 --- ---

5
 Time Series Analysis

4. Method of least squares

One way of finding the trend values with the help of mathematical technique is the method of
least squares. This method is most widely used in practice and in this method the sum of squares
of deviations of the actual and computed values is least and hence the line obtained by this
method is known as the line of best fit. It helps for forecasting the future values. It plays an
important role in finding the trend values of economic and business time series data.
Advantages
• The method of least squares completely eliminates personal bias.
• Trend values for all the given time periods can be obtained
• This method enables us to forecast future values.
Disadvantages
• The calculations for this method are difficult compared to the other methods.
• Addition of new observations requires recalculations.
• It ignores cyclical, seasonal and irregular fluctuations.
• The trend can be estimated only for immediate future and not for distant future.
We will apply this least square technique in regression analysis later in the chapter
“Correlation and Regression”. Example will be illustrated there.
Exercise 8
8.1 Show the principle of Moving average for the given data (Using 3 years).
Month Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Revenue 5.0 8.0 7.0 8.0 8.0 9.0 7.0 9.0 5.0 7.0 5.0 8.0

8.2 Fit trend lines by the methods of i) graphical ii) semi-averages for the given data.
Year 2003 2005 2007 2009 2011 2013
Demand 650 700 800 810 700 900

8.3 Fit a trend line by the method of semi-averages for the given data.
Year 1990 1991 1992 1993 1994 1995 1996
Sales 15 11 20 10 15 25 35

6
 Time Series Analysis

8.4 Measure the trend by the method of semi-averages by using the table given below. Estimate
the value for the year 1994-1995.
Years Value in Million
1984 – 85 18.6
1985 – 86 22.6
1986 – 87 38.1
1987 – 88 40.9
1988 – 89 41.4
1989 – 90 40.1
1990 – 91 46.6
1991 – 92 60.7
1992 – 93 57.2
1993 – 94 53.4