Intra-Market Price Discovery in an Emerging Stock Market: Vector Fractionally-Integrated Error Correction Model and Toda-Yamamoto Level VAR Approaches

Abstract

The study investigates the price discovery process by which
markets attempt to find equilibrium prices among a system of
disaggregate daily spot share price indices relating to the Malaysian
Stock Exchange using a VAR. Specifically: (i) a vector fractionallycointegrated
error-correction model is proposed and estimated to
investigate the short-run dynamics accounting for the long-run
information via a fractionally-integrated error-correction term; and
(ii) the Toda-Yamamoto (1995) [k + d(max)]th-order VAR procedure
to specify a ‘level’ VAR containing integrated and cointegrated
processes of arbitrary orders is adopted to uncover the long-run
driving forces behind stock market linkages. The results are
interpreted in the context of the price discovery process among
spot stock prices. Our findings indicate consistently that the price
discovery process was focused on the palm oil market in Malaysia in
the sense that this market played, relatively, the leading role (both in
the short- and long-term) being the most exogenous of all. In
particular, we demonstrate that previous research, by using ordinary
difference VARs either by intent or lack of support of cointegration
in the standard sense, ignored an important component of linkages
displayed over the long run.
Keywords: Intra-Market Price Discovery; Granger Causality;
Mean-Reversion; Fractional Cointegration.
JEL classification: G15, C52