Tse (2000) argues that different data sets (quotes versus trades), rather than models or
periods, matter when measuring relative contributions to price discovery. Thus, market
‘‘A’’ trades may provide information to market ‘‘B’’ traders even if the economic meaning
of market ‘‘A’’ quotes is negligible. Therefore, a methodology intended to study the price
discovery process of a dually listed stock should take into account at least both quotes and
trades. Current methodologies do not openly model the trading process and so cannot
differentiate between trade-related and trade-unrelated sources of information. As a result,
they provide an incomplete characterization of the dominant-satellite relationships
(Garbade and Silber, 1979).