Long-run income and price elasticities of demand for Colombian nontraditional exports: a multivariate cointegration framework

Abstract

A long run income and price elasticities of demand is estimated for Colombian nontraditional exports through a multivariate cointegration analysis. Based on the combination of cointegration and exogeneity concepts and the inclusion of the complete dynamic system, the paper shows the existence of a long-run relationship among nontraditional exports, relative price and foreign demand, and higher long-run elasticities than those provided by the long-run cointegration vector coefficients that are usually reported in the trade literature.

Acknowledgements

We are indebted to Augustine C. Arize, Enrique Lopez, Hugo Oliveros, Carlos E. Posada and Luisa Silva for helpful comments on earlier drafts of this paper. The views expressed are those of the authors and not of the Central Bank of Colombia or of its Board of Directors. The usual disclaimer applies to any remaining errors or omissions.

Notes

¹ For details see Misas et al. (2001).

² For example, Villar (1984 and 1992), Botero and Meisel (1988), Alonso (1993), Steiner and Wüllner (1994) and Mesa et al. (1999), among others.

³ For instance see Villar (1984), Botero and Meisel (1988), and Alonso (1993).

⁴ One exception is Villar (1992).

⁵ See for instance Chowdhury (1993), Reinhart (1995), In and Sgro (1998) and Arize et al. (1999).

⁶ For details, see Misas et al. (2001).

⁷ In average, between 1996 and 2000, nontraditional exports oriented to Venezuela represent 18% of that kind of exports; see Misas et al. (2001).

⁸ Following this scheme, a volatility measure was constructed using m = 4, since quarterly data is used.

⁹ See Bera and Higgins (1993), Engle (1995) and Franses and van Dijk (2000).

¹⁰ It is worthwhile to mention that the GARCH models were estimated using monthly data; the quarterly measure was obtained by Drost and Nijman's (1995) procedure.

¹¹ The results of these tests are available upon request from the authors.

¹² As Hendry and Doornik (1994) and Arize (1999), among others, the critical values used in the cointegration analysis do not take this into account.

¹³ The results are available from the authors on request.

¹⁴ In September 1999, the exchange rate target zone regime was changed for the exchange rate floating system.

¹⁵ Only the selected models are reported.

¹⁶ It is worth mentioning that, in the case of developing countries, including Colombia, Reinhart (1995) and Arize et al. (1999) also use a multivariate cointegration analysis. However, the error correction model set up could be inadequate because they did not analyse the exogeneity degree among the variables of the system.

¹⁷ Johansen (1995), Chapter 4.

¹⁸ It is worthwhile to mention, that the small differences between the results from Table 6, C matrix Johansen's approach and from Fig. 2 (IRF) Lütkepohl's approach are explained by the fact that in the Equation 7, the first approach only include the long-run behaviour, while the second one, takes into account also the transitory component.

¹⁹ See Cuthbertson et al.(1992).

²⁰ See Bardsen (1994).