Karnataka J. Agric. Sci.

,27 (1): (36-39) 2014

Stability analysis of mango export markets of India: Markov Chain approach*

D. K. KUSUMA AND H. BASAVARAJA
Department of Agricultural Economics
University of Agricultural Sciences Dharwad-580 005, Karnataka, India
E-mail : kusuma3836dk@gmail.com
( Received: February 2013 ;

Accepted: March 2014)

Abstract: India is native to mango and is also the largest producer of mangoes with 44.14 per cent of the total world
production. The export of fresh mangoes has increased from ` 35.2 crores in 1991-92 to `162 crores in 2010-11. The paper
attempts to quantify the changing structure of Indian mango exports. Data for analysis was taken for a period of 10 years
from 2001-02 to 2010-11. Compound growth rate (CGR) was used for analyzing the growth in mango production, area,
yield, export quantity and export value over the years. The markov chain analysis was attempted through linear programming
method to assess the transition probabilities for the major mango markets. The major Indian mango export markets were
categorized as stable market (Bangladesh, U.K, U.A.E) and unstable markets (Nepal, Saudi Arabia) based on the magnitude
of transition probabilities. The major export markets for Indian mangoes are Bangladesh (46.22 %), U.A.E (33.26 %), Nepal
(6.06 %), Saudi Arabia (3.63 %) and UK (3.06 %). The increasing share of other countries clearly shows the need to explore
and exploit the market potential of other countries. Efforts are also needed to improve the efficiency of production and
quality in order to stabilize the markets and also to make the product acceptable and price competitive in other importing
countries.
Key words: Compound growth rate, Direction of trade, Markov chain, Structural change

Introduction

Material and methods

India is the fruit and vegetable basket of the world. It is the

home of wide variety of fruit and vegetables and holds a unique
position in production. India is the largest producer of fruits in
the world with an annual production of 74.87 million tonnes
from an area of 63.83 million hectares and contributes more than
12 per cent to the world fruit production. Mango, grapes and
banana are the major export earning commodities. Mango
(Mangifera indica) is the main fruit of Asia and has developed
its own importance all over the world. Being an useful and
delicious fruit, it was the part of culture and religion since long
time. Besides a fine taste, it has many good qualities thus making
it to be called as King of Fruits. Mango continues to dominate
the Indian fruit basket contributing 36 per cent to total fruit area
and 20.3 per cent to total fruit production. It is grown over an
area of 2.2 million ha with a production of 15.5 million tonnes in
the country (Anon., 2011). India is the largest producer of
mangoes with 44.14 per cent of the total world production. Thus
India has immense potential to assert a strong presence in the
global mango market, which is still under-developed.
The export value of fresh mangoes has increased from ` 35.2
crores in 1991-92 to ` 162 crores in 2010-11. The major export
markets for Indian mangoes are Bangladesh (46.22 %), U.A.E
(33.26 %), Nepal (6.06 %), Saudi Arabia (3.63 %) and UK
(3.06 %). This is made possible due to increased export and due
to diversification of geographical concentration. Even though
export of mangoes is increasing, the share of quantity exported
out of total production is meagre (0.38 %) of total production is
exported. Therefore, an attempt was made to quantify the
changing structure of Indian mango exports. The main objective
of the paper was to study the direction of export and structural
change in mango exports.

The study is based on time series data on area, production,

productivity and exports of mango from India obtained from
various published issues of APEDA for a period of 2001-02
to 2010-11.
The compound growth rate analysis was carried out to
ascertain the growth in area, production, productivity and
exports of mango from India. The structural change and direction
of change in the export of mango was analysed using markov
chain analysis
Yt = abt u t
Where,
Yt = area, yield, production or export of mango in year t
t = year which takes value 1, 2 .n
ut = disturbance term
a and b are regression parameters to be estimated. Logarithmic
transformation of provided the estimating equation was
ln Yt= ln a + t ln b + ln ut
The equation was estimated by ordinary least square
technique (OLS). Compound growth rate (g) was then estimated
by the identity given in equation
g = (anti log of b-1) 100
Where,
g = estimated compound growth rate in per cent per year and
b = anti log of b
Annual export data for period 2001-02 to 2010-11 were used
to analyze the direction of trade and changing pattern of Indian
mango export. The major Indian mango importing countries
considered were U.A.E, Bangladesh, UK, Nepal and Saudi Arabia.
Estimation of the exports was done for the study period using
Markov chain analysis.

*Part of Ph. D. thesis submitted by the first author to the University of Agricultural Sciences, Dharwad - 580 005, India
36

Karnataka J. Agric. Sci.,27 (1): 2014

Markov chain analysis was employed to analyze the
structural change in any system whose progress through time
can be measured in terms of single outcome variable
(Dent, 1967). In the present study, the dynamic nature of trade
patterns that is the gains and losses in export of Indian mango
in major importing countries was examined using the Markov
chain model. Markov chain analysis involves developing a
transitional probability matrix P, whose elements, Pij indicate
the probability of exports switching from country i to country
j over time. The diagonal element Pij where i=j, measures the
probability of a country retaining its market share or in other
words, the loyalty of an importing country to a particular
countrys exports.
In the context of current application, structural change was
treated as a random process with seven importing countries for
mango the assumption was that the average export of mango
from India amongst importing countries in any period depends
only on the export in the previous period and this dependence
was same among all the periods. This was algebraically
expressed as

e is the vector of absolute errors

Y is the proportion of exports to each country.
X is a block diagonal matrix of lagged values of Y
V is the vector of errors
G is a grouping matrix to add the row elements of P arranged in
P* to unity.
Prediction of quantity of fresh mango export were made by using
the Transitional Probability Matrix.
Bt = B0 * T
Bt+i = Bt+i-1 * T
Where,
B0 = Quantity exported in Base years
Bt+i = Quantity exported in next year (prediction)
T
= Transitional probability matrix
Results and discussion
Compound growth rates for area, production, productivity
and export of mango were analysed for a period of ten years i.e
from 2001-02 to 2010-11 and have been presented Table 1. The
results revealed that both area (6.77 %) and production (3.70 %)
exhibits positive growth, while productivity (-0.72 %) of mango
found to be negatively growing. This negative growth in
productivity of mango may be due to poor management practices
by the producers. The study also depicted the growth in export
quantity and value of export where in growth of value of export
was increasing at 9.98 per cent per annum and quantity of export

Where,
Ejt = exports from India to the jth country in the year t
Eit-1 = exports of ith country during the year t-1
Pi j = the probability that exports will shift from ith
country to jth country
e jt = the error term which is statistically independent
of Eit-1
n
= the number of importing countries

Table 1. Compound growth rates of export of fresh mangoes from

India (2001-02 to 2010-11)
Particulars

CGR (% Per annum)

Area
Production
Productivity
Quantity of Export
Value of Export

The transitional probabilities Pij, which can be arranged in a

(c x n) matrix, have the following properties.

6.77
3.70
-0.72
5.65
9.98

Table 2. Destination -wise growth rates in export of fresh mangoes

from India during 2001-02 to 2010-11
Destinations

Thus, the expected export share of each country during

period t is obtained by multiplying the exports to these countries
in the previous period (t-1) with the transitional probability matrix.
The probability matrix was estimated for the period 2000-01 to
2010-11.
Thus transitional probability matrix (T) was estimated using
linear programming (LP) framework by a method referred to as
minimization of Mean Absolute Deviation (MAD).
Min, OP* + I e
Subject to
X P* + V = Y
GP* = 1
P* > 0
Where,
P* is a vector of the probabilities P i j
O is the vector of zeros
i is an appropriately dimensional vectors of areas

UAE
Bangladesh
Nepal
Saudi Arabia
UK
Others
Total

Quantity

Growth rate (% Pa)

Value
Unit value

8.46
5.47
33.82
-4.18
11.35
-5.10
5.66

15.13
3.18
32.03
1.31
16.92
3.66
9.98

6.15
-2.17
-1.34
5.73
5.01
9.22
4.09

Table 3. Transitional Probability Matrix for fresh mangoes export

from India (2001-02 to 2010-11)

U.A.E
Bangladesh
Nepal
Saudi Arabia
UK
Others
37

U.A.E

Bangla Nepal
desh

Saudi
Arabia

0.7152
0.0000
0.4544
0.0000
0.0000
0.5316

0.0237
0.0000
0.5456
0.0000
0.0000
0.1689

0.0203
0.0408
0.0000
1.0000
0.0000
0.0000

0.0766
0.0000
0.0000
0.0000
0.0000
0.2034

UK
0.0285
0.0000
0.0000
0.0000
0.5162
0.0018

Others
0.1358
0.9592
0.0000
0.0000
0.4838
0.0943

38
6157.21
(10.4)

3589.98

(4.04)

2393.93

(4.95)

3687.61

(4.73)

3956.75

(9.55)

5191.94

(6.83)

5400.42

(5.05)

3511.92

(5.07)

2712.76

(5.53)

3348.18

(4.52)

1719.07

(3.15)

(51.1)

7200.95

(38.92)

23049.70

(45.06)

33549.90

(53.89)

45104.50

(31.40)

17063.60

(54.25)

42887.52

(47.08)

32770.90

(60.78)

32503.22

(39.30)

23797.13

(35.24)

13392.85

(47.34)

21033.74

P
1400.13

Bangladesh
A

30262.93

6654.96

(36.31)

31175.58

(43.44)

(30.74)

21503.97

(34.39)

25725.00

22890.98

(26.50)

25608.20

22180.85

(29.35)

(39.01)

(41.34)

24570.90

21204.13

22469.62

(26.50)

(27.88)

(32.84)

20951.06

22045.51

(38.12)

(20.32)

22857.99

(19.33)

26533.76

10868.46

(33.58)

10338.61

20330.53

(34.77)

(36.49)

21056.16

13865.03

(36.93)

(28.11)

14033.56

12486.98

(28.83)

12809.55

U.A.E

1770.12

(3.36)

1991.30

(5.45)

4058.20

(5.69)

4765.00

(13.89)

7550.89

(10.19)

8055.73

(5.91)

4116.01

(6.36)

3400.94

(4.84)

2930.11

(1.12)

426.19

(0.23)

101.40

(7.1)

4183.40

8829.19

(4.75)

2812.11

(4.04)

3006.42

(3.36)

2816.24

(4.36)

2372.27

(2.87)

2271.25

(4.02)

2801.40

(2.86)

1529.72

(5.15)

3119.83

(6.49)

2465.25

(5.03)

2235.67

Nepal

(52.64)

(2.69)

1592.20

(4.23)

3147.10

(2.56)

2141.30

(2.74)

1488.95

(1.67)

1323.56

(2.25)

1564.15

(4.30)

2300.53

(6.35)

3845.72

(5.49)

2085.02

(6.62)

2942.88

(7.6)

4494.95

(11.24)

(5.16)

3055.38

(6.76)

5036.34

(5.35)

4480.92

(4.86)

2641.73

(4.45)

3521.41

(4.94)

3440.41

(7.17)

3836.81

(8.66)

5244.53

(7.67)

2916.62

(9.14)

4061.38

Saudi Arabia
A

2013-14
30086.61
5076.27
4820.33
5361.07
(50.8)
(8.6)
(8.1)
(9.1)
(3.1)
Note. A-Actual exports in tonnes. P- Predicted exports in tonnes. Figures in parentheses indicate exports share in per cent

2012-13

2011-12

2010-11

2009-10

2008-09

2007-08

2006-07

2005-06

2004-05

2003-04

2002-03

2001-02

Table 4. Actual and Predicted quantity of fresh mangoes export from India to selected countries (Qty in tonnes)

(12.16)

(4.60)

2723.50

(3.97)

2958.70

(3.02)

2527.40

(4.74)

2575.37

(2.38)

1883.19

(1.21)

839.97

(2.45)

1308.56

(2.50)

1511.63

(3.23)

1227.57

(3.09)

1372.87

1822.34
(20.4)

(3.1)

1817.47

(6.06)

(3.62)

2145.80

(3.04)

2265.72

(2.40)

2012.56

(3.63)

1974.94

(2.03)

1604.94

(1.72)

1195.90

(1.83)

976.47

(2.30)

1393.36

(2.75)

1045.76

(2.44)

1084.68

UK

(2.99)

(6.99)

4139.07

(6.90)

5138.51

(5.49)

4594.07

(5.89)

3202.37

(3.62)

2865.37

(5.43)

3781.81

(6.78)

3628.16

(12.24)

7410.57

(17.99)

6836.74

(13.88)

12054.15

(20.8)

12304.81

(14.91)

(46.11)

27309.59

(50.46)

37573.54

(57.65)

48255.85

(38.57)

20965.79

(57.31)

45311.79

(51.43)

35798.98

(62.74)

33555.80

(44.78)

27114.90

(42.08)

15990.21

(52.13)

23160.48

Others
6168.88

Stability analysis of mango ......

Karnataka J. Agric. Sci.,27 (1): 2014

was increasing at 5.65 per cent per annum. Compared to
production, export of mangoes was growing at a positive and
significant growth. Growth in value of export was found to be
very high indicating good potential and higher profit for Indian
mangoes. These results are in line with findings of
Ramachandra (2006).
Production has increased from 10020.20 thousand metric
tonne in 2001-02 to 15188.00 thousand metric tonne with a growth
rate of 6.77 per cent per annum. Quantity of mangoes exported
from India has been increased from 44429.32 metric tonne to
59220.77 metric tonne with a growth rate of 5.65 per cent per
annum. But the share of export out of the total quantity of mango
produced is very less, this is ranging between only 0.30 per cent
to 0.70 per cent. Though India is the largest producer of mango
in the world, the quantity exported is less than one per cent
during the study period.
The major importing countries of Indian mangoes were UAE,
Bangladesh, Nepal, Saudi Arabia, UK and others have been
presented in Table 2. Export of mango to Nepal recorded a
significant and positive growth rate of 33.82 per cent and
32.03 per cent in terms of quantum and value of export but it
recorded a negative growth rate of -1.34 per cent in unit value of
export. Saudi (-4.18 per cent) and other countries (-5.10 per cent)
exhibited negative growth in quantum of export but positive
growth in value and unit value of export. UAE, Bangladesh, UK
export were having moderate positive growth. These results are
in line with findings of Prakash Mokashi (2012).
The Transitional Probability Matrix presented in Table 3
provides a broad indication of changes in the direction of export
of fresh mango from India for the study period (2001-02 to
2010-11). The major Indian fresh mango importing countries were
UAE, Bangladesh, Nepal, Saudi Arabia, UK and all other
importing countries were grouped under the category of the
other countries. The row elements in the transitional probability
matrix provide the information on the extent of loss in trade, on
account of competing countries. The columns element indicates
the probability of gains in volume of trade from other competing
countries and the diagonal element indicates probability of
retention of the previous years trade volume by the respective
country.
It is evident from Table 3, that Saudi Arabia was the most
stable market among the major importers of Indian mango as

reflected by the probability of retention at 100 per cent. The

most unstable markets among the importing countries were
Bangladesh and Nepal with the zero per cent retention. U.A.E
retained 71 per cent and UK with 51.62 per cent of total export
from India. These results are in line with findings Pramod Kumar
et al. (2007) and Yeledhalli et al. (2012.)
The market share projections of Indian fresh mango exports
to the major importing countries were computed up to 2013-14
using the transitional probability matrix. Table 4 presents the
actual and predicted values of Indian mango exports to major
importers from 2001-02 to 2010-11 and also projections up to
2013-14. The actual share of U.A.E in fresh mango export had
shown fluctuation over the study period (2001-02to 2010-11) on
the whole it had increased from 28.83 per cent to 43.44 per cent.
Similar picture was in prediction of export share too, where the
increase was from 28.11 per cent to 50.8 per cent. The estimation
for 2013-14 suggested a increase from 36.31 per cent to
50.8 per cent.
With regard to Bangladesh the actual and predicted export
share showed fluctuations from 47.34 per cent to 38.92 per cent
and 3.15 per cent to8.6 per cent respectively from 2001-02 to
2010-11. But, the estimation for 2013-14 suggested a increase
from 4.04 per cent in 2010-11 to 8.6 per cent. The actual proportion
of Nepal market share of imports from India showed a increasing
trend from 0.23 per cent to 3.36 per cent. The predicted export
share also increased from 5.03 per cent to 8.1 per cent during the
study and the estimation for 2013-14 suggested a increase from
4.75 per cent in 2010-11 to 8.1 per cent. With respect to the Saudi
Arabia, the actual proportion of exports showed a decreasing
trend i.e., decreased from 6.62 per cent to 2.69 per cent. The
predicted proportion showed a constant share 9 per cent. With
regard to UK, the actual and predicted market share of Indian
fresh mango exports showed a slight increase from 3.09 per cent
to 4.60 per cent and 2.44 per cent to 3.1 per cent. The actual and
predicted proportion of exports share of Indias mango exports
to other countries showed a decreasing trend from
13.88 per cent to 6.99 per cent and 52.13 per cent to 20.4 per cent.
Though India is the largest producer of mango in the world,
it exports less than one per cent of its total production. India
has a good potential for export of mango and it needs to strive
hard to improve its export by improving upon the quality of
mango exports and also by improving the yield levels.

References
Dent, W. T., 1967. Application of Markov Analysis to International
Wool Flows, Review of Economics and Statistics,
49(2): 613-616.

Ramachandra, V. A., 2006, Production and marketing of Sapota in

Northern Karnataka - An economic analysis. M. Sc. (Agri.)
Thesis, Univ. Agric Sci, Dharwad, India.

Prakash Mokashi, 2012, An economic analysis of export potential of

grapes from Northern Karnataka. M. Sc. (Agri.) Thesis, Univ.
Agri. Sci., Dharwad (India).

Yeledhalli, R. A., Praveen H. Patil, Chidanand Patil and Vinod R. Naik.

(2012). Changing direction and magnitude of Indias major fruit
export to middle east countries. Int. J. Agricult. Stat. Sci.,
8(2): 651-658.

Pramod Kumar, Badal, P. S. and Lalith Achoth., 2007, Changing direction

of Indian mango exports. Indian J. Agril. Mktg.,
21(1):130-137.

Annonymons, 2011, Indian Horticulture database 2011.

www.apeda.com

39