MARKET DELINEATION OF THE SPANISH CLAM MARKET

Abstract

This article analyzes the long-term relationship among prices of the most consumed clams in Spain. The clams considered are: the striped venus (Chamelea gallina) from Spain and Italy, the Japanese carpet shell (Ruditapes philippinarum) and the grooved carpet shell (Ruditapes decussatus). The striped venus and grooved carpet shell are wild-caught whereas the Japanese carpet shell is from aquaculture farms. Unit root tests were used to check for stationarity and cointegration and the Law of One Price (LOP) were analyzed. Also short-run and long-run causality among prices is studied. The empirical results show that the markets for the Italian and Spanish striped venus and the Japanese carpet shell are interrelated and to some extent these clams can be considered to be substitutes. However, only the Japanese carpet shell and the Spanish striped venus are perfect substitutes.

Keywords:

• causality
• cointegration
• market interaction
• Spanish clam markets

ACKNOWLEDGMENTS

The authors would like to thank the Marketing Director of Mercamadrid, Antonio Roldán Bonilla, for having provided us with all the statistical data needed to carry out this study. The authors are also indebted to two anonymous referees for their helpful comments. They are also grateful to an anonymous English editor and Jeanette Corbett for their editing work on the article. The study has been partly funded by the EU project OCIPESCA of the Interreg IIIB Atlantic Area Program and the IFAPA (Junta de Andalucía) project called “Análisis del Mercado Español de Productos Pesqueros y Acuícolas. Consecuencias para los Productores Andaluces.”

Notes

Notes: (1) The auxiliary test regression contains a constant and seasonal dummy variables.

(2) The auxiliary test regression contains a constant, seasonal dummy variables and a trend.

*Significant at the 10% level. **Significant at the 5% level.

Maximum number of lags = 12.

Only the lags that are statistically significant have been included in the auxiliary regressions (Franses, 1991).

^a)ϕ*(L) = (1 − ϕ₁ L ⁷ − ϕ₂ L ¹²).

^b)ϕ*(L) = 1.

^c)ϕ*(L) = (1 − ϕ₁ L ¹⁰).

Notes: *Model and cointegration rank which have been chosen.

Number of lags = 6. Significant at the 10% level.

Note: *Significant at the 10% level.

**Significant at the 5% level.

Notes: *Model and cointegration rank which have been chosen.

Number of lags = 6. Significant at the 10% level.

Note: *Significant at the 10% level.

**Significant at the 5% level.

Note: *Significant at the 10%.

**Significant at the 5%.

Note: *Significant at the 10% level.

**Significant at the 5% level.

Note: *Significant at the 10% level.

**Significant at the 5% level.

To study the BMF seasonal unit root test in more detail, see Ghysels & Osborn (2001).

Centred seasonal dummy variables were included in the models in order to consider seasonal components which have not been removed by filtering seasonal unit roots out. Some mis-specification tests have been performed to study whether the models are well specified or not. The results of these tests can be provided upon request.

Clams are usually cooked in Andalusia by frying them lightly with some olive oil, parsley and salt and are then served as a course. In contrast, in the centre of Spain they are included in stews. This culinary tradition comes from the times where the taste of fish and shellfish was hidden between many other ingredients.

Other reasons which go beyond the scope of this study can be related to the environment. This is due to the fact that aquaculture farms devoted to the Japanese carpet shell are introducing a foreign species into an environment where the grooved carpet shell has been the dominant species to date.