Price Determination and Demand Flexibilities in the Ex-Vessel Market for Tuna in the Republic of Maldives

Abstract

This article assesses the relationship between export and ex-vessel prices for tuna fish in the Republic of Maldives. The economic welfare of fishermen depends to a great extent on the price received for fish. The price of fish is set by external factors exogenous to fishermen. It is important in understanding the welfare of fishermen to understand the price links in the fish supply chain and the factors that impact the ex-vessel price of fish. This article uses three models to investigate price determination in the ex-vessel market; ARMAX, inverse demand equation and margin equation. The results provide statistically important price relationships and price flexibilities on asymmetric export price effects.

Keywords:

• asymmetric export price effects
• price flexibilities
• supply chain management
• tuna

ACKNOWLEDGMENT

We thank Trond Bjørndal and two anonymous referees for comments and suggestions that improved the style and content of the article.

Notes

Statistical Yearbook Maldives (2009).

The information in this section is taken from ‘Maldives Background Report’ by Hussain Sinan, FAO Rome, 2011. Also, see Bjørndal and Gordon (2010) for a statistical evaluation of the fishery.

The vessel crew take a portion of the catch for personal consumption.

Studies suggest that bigeye (Thunnus obsesus) tuna may account for up to 5% of the total yellowfin catch (Anderson & Hafiz, 1991).

Keep in mind that the canneries do produce somewhat different products. They both produce frozen skipjack and yellowfin tuna, and salted dried or canned skipjack, however, cannery 1 produces steamed skipjack loins.

Although we are interested in a short-run forecasting model a Johannson cointegration test shows strong statistical support for a long-run relationship for ex-vessel and export price. Based on this result an alternative modelling approach would be to use an error-correction model.

Autoregressive Integrated Moving Average Model.

For an excellent review of applied time series econometrics, see Enders (2010).

For an interesting discussion of the first serious price forecasting model, see Gordon & Kerr (1997).

The restriction on the exogenous variables requires that there be no feedback effect to the dependent variable (Enders, 2010).

Seasonal and trend variables were also included in specification but did not statistically improve the forecast.

Estimation is carried out using STATA 12 software.

Bayesian Information Criteria.

The null for the Dickey–Fuller statistic is fist-difference stationary, whereas the null for the KPSS statistic is level stationary.

Different specifications of the AR and MA components were estimated but all specifications were dominated by the equations reported in Table 3.

Confidence limits (not shown in Figure 8) fail to include the original series during these periods.

We test for cointegration and price leadership between ex-vessel prices offered by companies 1 and 2. We find both cointegration and price leadership by company 1.

In estimation we will include seasonal dummies in the harvest equation but not in the price equation. As seems reasonable, initial testing showed that the price relationship does not vary with the season.

A 2SLS estimator is used to ensure correct standard errors.

The margin could be defined with respect to ex-vessel prices lagged one or two periods, but this makes little difference in the estimation. The correlation between the current margin and a one- and two-period lagged margin effect is 0.997 and 0.995, respectively.