The large country effect, contagion and spillover effects in the GCC

Abstract

This article examines contagion and crisis propagation (spillovers) in the Gulf Cooperation Council (GCC) economies over the period 1960 to 2002. It also examines whether contagion occurred in Saudi Arabia (large country) after the 1987 US stock market crash and the 1997 Thai exchange rate devaluation and whether these contagion shocks spillover to smaller countries of the region. Spillovers are likely to occur among interdependent countries within the same geographical region. Tests based on correlation coefficients, ARCH/GARCH estimates and direct change (generalized least squares regression) propagation effects indicate that contagion from the US stock market crash and the Thai devaluation occurred in Saudi Arabia, and these external shocks were propagated to smaller GCC countries. This suggests that GCC countries are likely to mitigate such propagations through economic integration. Thus, the idea of GCC formation may help insulate Gulf economies against crisis propagation.

Acknowledgements

The author thanks Dr. Kirsten Madden for helpful comments and Ms. Beth Colvin for her assistance.

Notes

¹Saudi Arabia was a prime actor in setting up the GCC in 1981. Other members are Bahrain, Kuwait, Oman, Qatar and the United Arab Emirates. Progress towards economic integration has been slow because of the down turn in the 1980s and the Gulf Wars. The GCC seeks to strengthen cooperation in areas such as agriculture, security and trade. The Gulf Standards Organization was established in 1982 and the Gulf Investment Corporation in 1984.

²Another type of test for contagion studies changes in the long-run relationship between markets instead of short-run changes after a shock. Such tests add to the other methods, testing for changes in the cointegrating vector between asset markets instead of in the variance/covariance matrix. This approach (Longuin and Slonik 1995) is flawed if contagion is defined as a significant increase in market comovement after a shock. It assumes that real linkages between markets or noncrisis contingent theories such as trade flows remain constant over the sample period. By focusing on long time periods, cointegration tests (Granger et al., 2000) could miss brief periods of contagion. Also, if tests show that the cointegrating relationship increased over time, this could be a permanent shift in cross-market linkages instead of contagion.

⁴The entire sample (1960–2002) results (available upon request) are not included in all estimations of this article. The emphasis is to compare the stable period (before the 1987 US stock market crash and the 1997 Thai devaluation) with the unstable one (after the crises).

³The ARCH/GARCH results affirm the previous correlation findings of international reserves, showing pure contagion to Saudi Arabia, but insignificant (weak) contagion to smaller GCC countries. However, changes in the other GCC countries' international reserves are almost perfectly explained by changes in Saudi reserves, indicating that Saudi Arabia transfers external shocks to smaller countries in the region. For brevity the results will not be discussed.

⁶The external factors are introduced in separate regressions to focus on the contagion effects of each crisis in isolation.

⁵Different authors have constructed different indexes for measuring the pressure on the foreign exchange market. Hernandez et al. (2001) employed a weighted average of the depreciation rate of domestic currency, the increase in domestic interest rates and the losses in international reserves. Sachs et al. (1996) used the drop in stock prices to measure the severity and extent of the crisis.

⁷Results for these tests are not reported to save space.