Long memory linkages amongst Latin American stock markets. A fractional cointegration approach

ABSTRACT

This paper examines long-run relationships amongst six Latin American stock markets as possible evidence for their economic development, by using fractional cointegration which is applied to monthly observations for the period September 2002 to November 2019. Additionally, a novel summary table is proposed in an exhaustive way, which attempts to help to understand the puzzle of market integration around the different economic regions of the world. Hereby, the analysis suggests that there are four cointegrating vectors among the six equity markets, suggesting that Latin American stock markets are not fully nor perfectly integrated.

KEYWORDS:

• Fractional cointegration
• Latin America
• financial integration
• financial market cointegration

JEL classification:

• C32
• F41
• G15

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. For its Spanish initials.

2. The hypotheses ${\mathrm{H}}_{\mathrm{i}}^{\beta }$are used to test whether a given stock market is part of a cointegrating relationship and existing long-run equilibrium. If these hypotheses are rejected, we can conclude that a long-run equilibrium relationship does not exist.

3. The hypotheses ${\mathrm{H}}_{\mathrm{i}}^{\alpha }$ is used to test whether each variable is individually weakly exogenous. If a row of α is zero, the variable does not respond to disequilibrium in the relationship. A rejection of the null hypothesis implies that a market index adjusts towards the long-run equilibrium after a shock.

4. To test the presence of unit roots, the estimates were obtained using first-differenced data, because the original series might be above 0.5 and this test requires that the results are limited to the interval −0.5 < d < 0.5, then adding 1 to obtain the proper estimates of d. According to the literature, the bandwidth size spans from 0.25 to 0.8. In our study, the three bandwidths selected are 0.4, 0.5, and 0.6.

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.