Analysis of Marketing Function, Marketing Efficiency and Spatial Co-Integration of Rohu (Labeo rohita) Fish in Some Selected Areas of Bangladesh

The study was undertaken to examine the value chain, value addition, marketing cost & margin, marketing efficiency and market integration of rohu fish in selected areas of Bangladesh during the month of July-August 2013. The objectives of the study were to estimate costs and margins, seasonal price variation and to test market integration of rohu fish. Primary and secondary data were used for this study. The higher marketing cost was incurred by paiker and the lowest by retailer. On the other hand, retailers earned the highest net marketing margins. Chain II was found the most efficient chain. Analysis of market integration shows that rohu fish market in Bangladesh was well integrated. The findings of the study revealed that the marketing of rohu fish was a profitable business and some recommendations were provided for the improvement of rohu fish marketing in the country.


INTRODUCTION
In the agro-based economy of Bangladesh, the fisheries sector contributes near about 58% of animal protein to the daily diets of the population, about 3.74% to GDP, 4.04% in export earnings and 20.87% to agriculture in A higher value of ME denotes higher level of efficiency and vice versa

Market integration
Market integration was measured by co-integration method.The bulk of econometric theories have been based on the assumption that the underlying data process is stationary. Stochastic process is said to be stationary if its mean and variance are constant over time and the value of co-variance between two time periods depends only on the distance or gap or lag between the two time periods and not the actual time at which the co-variance is computed (Gujarati, 2003, p.797). In practice, most economic time series are non-stationary.An applying regression models to non-stationary data may arise the problem of "spurious or nonsense" correlation (Gujarati, 2003, p. 792).To overcome such problems, the concept of co-integration was used because it offers a means of identifying and hence avoiding the spurious. The underlying principle of co-integration analysis is that although trend of many economic series show upward or downwards over time in a non-stationary fashion, group of

Unit Root and Co-integrationTest:
The individual price series were tested for the order of integration to determine whether or not they are stationary which is known as the unit root test (Gujarati, 2003, p.799). A number of tests for stationary are available in the literature; these include the Dickey-Fuller (DF) test (Dickey and Fuller, 1979), the Augmented Dickey-Fuller (ADF) test (Dickey and Fuller, 1981) and the Philips-Perron (PP) test (Perron, 1988). For theoretical and practical reasons, the Dickey-Fuller test is applied to regressions run in the following forms: Y t is a random walk or without constant: ∆Y t = δ Y t-1 + e t ………………….
(1) Y t is a random walk with drift or constant: ∆Y t = β 1 +δ Y t-1 + e t .. In each case the null hypothesis is δ = 0(ρ =1); that is, there is a unit root in the time series i.e. the series is non-stationary. The alternative hypothesis is that δ is less than zero; that is, the time series is stationary. Under the null hypothesis, the conventionally computed t statistics is known as the τ (tau) statistic, whose critical values have been tabulated by Dickey and Fuller. If the null hypothesis is rejected, it means that Y t is a stationary time series with zero mean in the case of (1), that Y t is stationary with a non-zero mean [ = β 1 / (1 _ ρ) ] in the case of (2), and that Y t is a stationary around a deterministic trend in equation (3).
It is extremely important to note that the critical values of the tau test to test the hypothesis that δ = 0, are different for each of the preceding three specifications of the DF test. If the computed absolute value of the tau statistics (τ) exceeds the DF or MacKinnon critical tau values, we reject the hypothesis that δ = 0, in which case the time series is stationary. On the other hand, if the computed (τ) does not exceed the critical tau value, we do not reject the null hypothesis, were the time series is non-stationary.
In conducting the DF test as in (1), (2), or (3), it was assumed that the error term e t was uncorrelated. But in case the e t are correlated, Dickey and Fuller have developed a test known as the augmented Dickey-Fuller (ADF) test. This test is conducted by "augmenting" the preceding equation by adding the lagged values of the dependent variable ∆Y t .
The ADF test here consists of estimating if the error term e t is auto correlated, one modifies (4) as follows: where ε t is a pure white noise error term and where, ∆ Y t-1 = (Y t-1 _ Y t-2 ), ∆ Y t-2 = (Y t-2 _ Y t-3 ), etc., that is, one uses lagged difference terms. The number of lagged difference terms to include is often determined empirically, the idea being to include enough terms so that the error term in (4) is serially uncorrelated. The null hypothesis is still that δ = 0 or ρ = 1, that is, a unit root exists in Y (i.e., Y is non-stationary).

Spatial Price Relationship:
To test the market integration, the following co-integration regression was run for each pair of price series: Y it = α 0 + α 1 Y jt + ε t ……………………… (5) Where, Y i and Y j are price series of a specific commodity in two markets i and j, and ε t is the residual term assumed to be distributed identically and independently. The test of market integration is straightforward if Y i and Y j are stationary variables but if the price series proved as non-stationary then we have to done another test (Engle-Granger test) Testing whether the variables are co-integrated is merely another unit root test on the residual in equation (5).
However, since the Y i and Y j are individually non-stationary, there is the possibility that the regression is spurious. The DF and ADF tests in the present context are known as Engle-Granger (EG) test whose critical values was provided by Engle-Granger (Ramakumar, 1998). The test involved regression the first-difference of the residual lagged level and lagged dependent variables (Engle-Granger test) is as follows: For Engle-Granger (EG) test, ∆ε t = β ε t-1 …………………. (6) If the computed value of 't' of regression coefficient β is higher (in absolute term) than tabulated value, our conclusion is that the residuals from the regression are I (0), that is they are stationary and the regression is not spurious even though individually two variables are non-stationary.

RESULTS AND DISCUSSION
Buying and selling An efficient marketing system is essential for earning fair profit for the fish farmers and traders. Marketing functions may be defined as major specialized activities performed in accomplishing the marketing process of concentration, equalization and dispersion (Kohls, 2005). The activities involved in the transfer of goods are completed through buying and selling functions. Aratdars do the functions of negotiation between buyers and sellers of fish and help them at their own business premises on receipt of commission. They do not take the ownership of the products. Rohu farmers sell 81% of their fish to paiker through aratdar, 15% to paiker directly and the final 4% to retailer. Paikers sell 74% of their fishes to retailers and 26% to retailers through aratdars. Retailers sell the entire fish to ultimate consumers. Paiker of Rohu purchases 88% from farmers through aratdar and 12% directly from farmers. Retailer purchases 87% from farmers through aratdar and 13% from farmers. Consumer purchases 100% of rohu from the retailers in the study area (Table 1).

Grading
Grading is the basic function of sales transactions and is defined as the classification of products according to some standards or measures (Kohls and Uhl, 2005; p. 314). Grading is the sorting of different market quality which facilitates exchange by simplifying buying and selling as it makes the sale by showing sample and description possible. It also simplifies the concentration process and makes easier and less costly the movement of goods through the marketing channel. Grading facilitates sale since different sizes of fish have different prices. In Bangladesh, all intermediaries grade fish on the basis of weight (Table 2).

Table 2. Grading practices of rohu fish on the basis of size and weight Species Size Weight
Rohu fish Large 2.5 kg above Medium 1.0 kg to 2.5 kg Small Less than 1 kg Source: Field survey, 2013.

Storage
The storage facilities help buyers and sellers to reduce the wide fluctuation of prices between peak and lean seasons. The storage function is primarily concerned with making goods available at the desired time and enables traders to receive better prices for their products. Because of high perishability, fish requires extremely specialized storage facilities matching the seasonal demand. Other intermediaries use only ice to transport fishes from one place to another. Surprisingly, no refrigerated vans are used in Bangladesh to transport fish. All intermediaries use ice during marketing, but their ice to fish ratio is not appropriate. So the quality of fish is deteriorated. In retail selling, some retailers use ice.

Transportation
Transportation is a basic function of making goods available at proper place and it creates place utility. Perishable goods must be moved as early as possible from the producing centre to the consumer centre. So transportation is essential for highly perishable commodities like fish. Adequate and efficient transportation is a cornerstone for the modern marketing system (Kohls and Uhl, 2005, p.319). In the study areas, the fish farmers and intermediaries use various modes of transports such as van, rickshaw, truck, passenger bus, pickup, nasimon (locally made pick-up type van for transporting passengers and goods), head load etc, to transfer product from the producing areas to the consumption centre. Table 3, show different modes of transport used by the intermediaries to transport fish from one place to another.

Financing
The financing function is the advancing of money by someone to carry on the business. For effective operation, financing is of crucial importance in the whole marketing system of rohu. Table 4 shows that most of the fish farmers, aratdars, paikers and retailers of rohu, are self-financed. Other sources of finance for farmers are banks, friends and relatives, and dadon. A minor portion of Aratdar's sources of finance are banks and friends and relatives. Paikers take loan from banks, NGO and friends and relatives. In addition to the use of their own fund, retailers also borrow from NGOs and friends and relatives. Packaging Packaging may be defined as the general group of activities in product planning which involves designing and producing the container or wrapper for a product (Stanton, 1991). Packaging is essential for proper transportation of fish. 'Basket' made of bamboo, rope and polythene is used by farmers, paikers and retailers of major carps, pangas and tilapia fish. Plastic drums are usually used when fish is transported in live form. Currently, 'plastic crate' is commonly used by all types of intermediaries in Bangladesh.

Characteristics of Market Participants
In the chain of fish marketing of the study areas, the product moves from farmers to consumers through market intermediaries such as Fish farmers, aratdar paiker, Faria and retailer.
Fish farmers and fishermen are the first link in the fish marketing channels. The fish Farmers (producers) of rohu fish usually sell their fish to the local aratdar.
The aratdars are at the centre of the entire marketing system and their role goes far beyond what one would normally expect of a commission agent, including financing of suppliers and buyers, and often dealing on their own account (Coulter and Disney, 1987). When fish arrives at the wholesale markets, aratdars take the responsibility and control of each sale. They sell the fish through an auctioning system and get a commission of 3% to 4% depending on fish species. Most of the time aratdars recruit koyal (person who organizes auction by uttering and offering different prices for buyers for sale). Koyals have a significant role on pricing the fish. Generally, the aratdars are self-financed. They hire necessary salaried persons or labourers depending upon their volume of business.
Paiker or bepari is conceptually same but used interchangeably in different fish marketing system in Bangladesh who transacts large volume of product. Another type of paiker is seen in hilsha marketing system called L/C paiker. They purchase fish from fishermen through aratdar and sell (export) their entire product to overseas market, especially the Indian markets. Some paikers/beparis receive money in advance from the aratdar on condition that they would sell their fish through them.
Faria is another type of intermediary in the marketing system. They purchase a small quantity of fish form distant fishermen far away from the market and carry it to the terminal point and sell it to aratdar or retailer in the study areas.
Retailer the last intermediaries of fish marketing channel, do not have any permanent establishment but they have fixed places in the market centre or are wandering with hari (aluminium pot) on head from door to door. Usually retailers buy fish from aratdar and sell directly to ultimate consumers. Mostly they purchase fish on cash. Sometimes they also purchase on credit for short term periods. If the size of fish is too large then buyers want the fish to cut into pieces as cutters have sufficient instruments to cut the large fish. Retailers may cut the whole fish for consumers or uses the services of cutters to remove scales and cut into pieces. Depending on the convenience, extra money is charged for removing scales or cut into pieces. In spite of being self-financed, the retailers often borrow money from non-institutional sources at the time of need.

Marketing Chains
Marketing chains are the alternative routes of product flows from producers to consumers (Kohls and Uhl, 2005; p. 501). Value chain may be long or short for a particular commodity depending on the qualities of products, size and nature of consumers and producers and the prevailing social and physical environment.

Value Addition Costs by Different Actors
The cost incurred to transport the product from producers to consumers is ordinarily known as marketing cost. In other words, the cost of marketing represents the cost of performing various marketing functions (Kohls and Uhl, 2005; p.96). Marketing costs are incurred when commodities are shipped from the farm to the final market. Intermediary-wise marketing costs are discussed below:

Fig 2. Total marketing cost involved in rohu fish marketing (Taka per maund)
Total marketing cost of fish includes all costs incurred by different intermediaries like inter district paikers, local paikers, aratdars, retailers and farmers who perform some marketing functions in the study areas. Products get value added during their movement across items. Share of transportation cost is the highest (40.54%), followed by aratdar's commission (26.92%), icing (8.23%), wages and salaries (4.81%) and tips & donations (4.32%) for rohu marketing (Table 7). Total value added cost per maund is Taka 953.13 from production point to consumption point. Amongst them, Paiker's value added cost is the highest while aratdar's value added cost is the lowest. Aratdars negotiate between buyers and sellers of fish and assist them in buying and selling at their own business premises on receipt of commission.

Marketing Margin
A marketing margin is the percentage of the final weighted average selling Price taken by each stage of the marketing chain. The margin must cover the costs involved in transferring produce from one stage to the next and provide a reasonable return to those doing the marketing activities. (Crawford, 1997). It is also termed as Price spread as it represents the difference between the buying and selling Price. Total marketing margin is the difference between the Price received by the fish Farmers and the Price paid by the final consumers. Marketing margins of fish are calculated separately for different intermediaries. Gross marketing margin of each type of intermediaries is calculated by deducting the purchase Price of fish from their sale Prices while net margin or profit component is calculated by deducting the marketing cost from gross marketing margins. Average net marketing margins of all intermediaries for rohu fish are presented in Table 3. Farmer average marketing cost is Taka 135.00 per maund for all fishes. Amongst all intermediaries, profit of retailers is the highest of Taka 624.29 per maund of fish. Profit of intermediaries varies due to variation in their costs; purchase Price and sales Price (Table 8).

Marketing Efficiency
Marketing efficiency is essential for measuring the degree of marketing performance. On the basis of three methods for measuring marketing efficiency, chain 2 is more efficient (Table 9). In chain 2 the consumer paid lower price per maund rohu compared to chain 1though in practice chain 1 is mainly used for selling rohu fish in the study area. Considering two value chain on the basis of product flow it was found that value chain-2 is more efficient than value chain-1. Consumer paid lower price in value chain-1 compared to value chain-2.   Table 10. Shows the percentages of total value addition cost and total net profit by different intermediaries for different fish marketing system in Bangladesh. For rohu fish, major cost is borne by paikers (32.03% of total cost) and major net profit is earned by retailers (51.98% of total net profit). For rohu, major cost is borne by the inter district beparis, paikers and fishermen but major net profit is earned by retailers and processing plant owners. Farmers, in rohu marketing, bear the major marketing cost (23.70% of total cost) because they have to pay aratdar's commission which ultimately increases their marketing cost.

Market integration of rohu
To test the stationary of the data, the ADF (Augmented Dickey-Fuller) test with 13 lags of rohu Prices for Dhaka, Chittagong, Khulna, Sherpur, Comilla, Bogra, Rangpur, Rajshahi, Mymensingh, Sylhet, Gazipur and Noakhali were performed over 2000 to 2012 period and the estimated tau) (τ statistics and P values in their level and first difference are presented in Table 11 Real Price data were used and data were transformed into natural log. Real Price was calculated by multiplying nominal Price with the corresponding consumer price index (CPI) and dividing by the last CPI of the series. The tau (τ) statistics which were compared with p values indicate that all the rohu. Price series data were non-stationary at level. This set of regression was run once more after differencing all the markets. The tau(τ) statistics on the lagged first-difference terms are significantly negative indication that the series are stationary after first differencing. The study revealed that the rohu prices are stationary after differencing once that is they are all (11) processe To examine whether bivariate co-integration exists between different Prices of rohu market, Dhaka wholesale market was considered as reference market, Dhaka is a capital market and the largest city and it would appeared to be the dominant influences on inter districts rohu markets in Bangladesh. The reference market is a dominant market serving as a hub in a sort of "radial market structure" where different feeder (local) markets are at the rim. The reference market dominates the Price formation in the feeder markets. Every individual feeder market was affected by the reference market Price, though it alone cannot affect the reference markets Price. Normally, the reference markets has a high turnover so that supply and demand shocks originating in the individual feeder markets are absorbed without creating much effect on the Price prevailing in the reference markets 63 (Ravallion, 1986). As there will be different combinations of the given 12 wholesale rohu Price markets, all combinations in a systems of bivariate relationships. The Engle-Granger (EG) and Augmented Engle-Granger (AEG) tests of residual equation confirm the stationarity of the residual series. Thus ADF results of unit root equation indicate that the real rohu Price series are I(1), while Engle-Granger (EG) and Augmented Engle-Granger (AEG) results of residual equation indicate that the residual series (which are linear combination of above rohu real Price series ) are I(0).Thus above fact that the Price series being I(1) and their linear combination being I(0) point out that the series are cointegrated without any exception. According to the Engle -Granger (EG) and Augmented Engle and Granger (AEG) test rohu markets of Bangladesh are statistically significance at 1% level (Table 12). An important finding of the study is that Dhaka market is significantly integrated to all regional markets of rohu markets in Bangladesh due to having the facility of information technology, which closely connected the markets to each other. This study strongly supports marketing efficiency in the selected rohu markets. Price move in the unison in all the markets together. Central Price policy making will be effective in these markets.

Strong forms of market integration
For testing strong form of rohu markets integration, the null hypotheses were applied to find rohu market integration against alternate hypotheses where rohu markets might not be integrated. The result of strong form of market integration in selected rohu markets is given in Table 13. It is seen from Table 13 that strong form of market integration was observed all the rohu markets in Bangladesh due to congenial atmosphere existed in these markets. This empirical finding strongly supports marketing efficiency in the markets. Price move in the unison in all the markets together. Central Price policy would be effective for these markets. For the estimation of strong form of market Integration we used the following restriction.  To calculate pair wise regression of the selected 12 domestic rohu markets prices, the time period was considered from January 2000 to December 2012 and the results are presented in Table 14. Monthly wholesale market price and Log linear model used for the study.