AGEC 386 P 2 Solution

Experiment no - 02: Analysis of time series data.
Problem: The agricultural wholesale price index of Potato from 1980 to 2005 is given as bellow:
Year Code Agricultural Export Year Code Agricultural Export earnings
year wholesale earnings year wholesale from agriculture (
price index from price index million $)
agriculture
(million $)
1980 1 510 397 1993 14 1437 356
1981 2 580 340 1994 15 1519 411
1982 3 596 411 1995 16 1609 442
1983 4 727 456 1996 17 1611 463
1984 5 883 550 1997 18 1701 428
1985 6 919 438 1998 19 1848 398
1986 7 1023 429 1999 20 1847 356
1987 8 1087 402 2000 21 1802 315
1988 9 1175 392 2001 22 1810 328
1989 10 1276 467 2002 23 1922 364
1990 11 1297 402 2003 24 1994 367
1991 12 1333 396 2004 25 2060 485
1992 13 1353 381 2005 26 2204 614
Question:
a. Fit a linear (Power of parameter is always 1) regression model and exponential model, where
agricultural wholesale price index is dependent variable & time (t) is independent variable.
b. Find the value of agricultural wholesale price index of 2008, 2010 & 2012 on the basis of both
models.
c. Estimate the growth rate of agricultural wholesale price index from the exponential model.
d. Examine the year to year fluctuation of export earnings from agriculture.
Note:
Arrangement of statistical data according to change of time is called time series.
Component of time series:

1. Trend: Upward/downward pattern.
2. Seasonal Variation: Short term variation of data (Less than one year, may be
weekly/monthly/quarterly/half-yearly)
3. Cyclical Movement/fluctuation: (3-10 years duration)
4. Irregular Movement
Wholesale price index: Relative change (between base year wholesale price and present year wholesale
price) in the level of phenomenon (Price, quantity, value) with respect to variable characteristics (time,
location, production, income, expenditure, import, export).
There are some methods of measuring this-

1. Simple aggregate method
2. Weighted aggregate method
Characteristics:
1. Index number has no unit and Expressed in percentage form.
2. Express the change of phenomenon in respect to variable
Solution:
a. Let,
The linear regression model, y = a + b t
Where, t = independent variable (code year)
y = agricultural wholesale price index
Using data from the table, after calculation, we find the value of a= 531.71 and b = 63.531
Therefore, y = a + b t or, y = 531.71 + 63.531 t
So, the linear regression model is, y = 531.71 + 63.531 t
Again, the exponential model, y = aebt
Or, ln y = ln aebt
Or, ln y = ln a + ln ebt
Or, ln y = ln a + bt ln e
Or, ln y = ln a + b t [Since, ln e = 1] 1, ln510 M+
After calculation we find, ln a = 6.46 or, a = 0.05
Therefore, The exponential model, y = 6.46 + 0.05 t
b. For the year 2008, t = 29

On the basis of linear regression model, y = (531.71 + 63.531* 29) = 2374.08
On the basis of exponential model, y = (6.46 + 0.05*29) = 7.91
For the year 2010, t=3131

On the basis of linear regression model, y = 531.71 + 63.
On the basis of linear regression model, y = 6.46 + 0.05 X 31

For the year 2012, t =33 33
c. The estimated exponential model is, y = 6.46 + 0.05 t

In any exponential model, the value of b is considered as growth rate.
So, the growth rate is 0.05
d. Table: calculation of year to year fluctuation

Year Export earnings Year to year Year Export earnings Year to year
from agriculture fluctuations from agriculture fluctuations
(million $) (%) (million $) (%)
1980 397 - 1993 356 -6.56
1981 340 -14.36 1994 411 15.45
1982 411 20.88 1995 442 7.54
1983 456 10.95 1996 463 4.75
1984 550 20.61 1997 428 -7.56
1985 438 -20.36 1998 398 -7.00
1986 429 -2.055 1999 356 -10.55
1987 402 -6.294 2000 315 -11.52
1988 392 -2.490 2001 328 4.127
1989 467 19.133 2002 364 10.98
1990 402 -13.92 2003 367 0.824
1991 396 -1.493 2004 485 32.153
1992 381 -3.788 2005 614 26.598
We know, change in export earnings

Fluctuation = x 100
Export earnings of previous year
For example, 340 - 397

Fluctuation in 1981 = x 100
397
= -14.36%
411 - 340
Fluctuation in 1982 = x 100
340
= 20.88%

AGEC 386 P 2 Solution

Uploaded by

Copyright:

Available Formats

AGEC 386 P 2 Solution

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

AGEC 386 P 2 Solution

Uploaded by

Copyright:

Available Formats

Experiment no - 02: Analysis of time series data.

Component of time series:

There are some methods of measuring this-

Again, the exponential model, y = aebt

Or, ln y = ln a + b t [Since, ln e = 1] 1, ln510 M+

After calculation we find, ln a = 6.46 or, a = 0.05

Therefore, The exponential model, y = 6.46 + 0.05 t

b. For the year 2008, t = 29

On the basis of exponential model, y = (6.46 + 0.05*29) = 7.91

For the year 2010, t=3131

On the basis of linear regression model, y = 6.46 + 0.05 X 31

On the basis of linear regression model, y = 531.71 + 63.531 X 33

On the basis of linear regression model, y = 6.46 + 0.05 X 33

c. The estimated exponential model is, y = 6.46 + 0.05 t

d. Table: calculation of year to year fluctuation

We know, change in export earnings

For example, 340 - 397

You might also like