CONCLUSIONS
We have assessed the behaviour of the spot foreign exchange
market quotations in terms of volatility, average
spread, and the number of quotations within half-hour
intervals, as well as certain informational aspects of these
processes. It seems that a log-linear relationship among
these three processes is a considerably better approximation
to the true data generating process functional form, than the
linear one; however, it is by far worse than the functional
form presented here.
A new variable was introduced: the number of observations
within a speciÞc time interval. This variable plays an
important role in the determination of volatility and average
spread, either directly or through the error terms. The
contemporaneous correlation of the number of quotations
and volatility leads us to hypothesize that the former process
could be a proxy for the volume of trade, or for the
number of transactions in the spot FOREX market, for
which data are unavailable. This is in line with studies in
stock market volume and volatility data [see Gallant, Rossi,
and Tauchen (1990), and Lamoureux and Lastrapes (1990)].
It turns out that informational theories can only partially
explain the facts documented here. Although, high trading
and volatility at the opening of markets can be explained
along the lines of the Admati and Pßeiderer (1988) theory,3
the di¤erent behaviour of the two currencies in di¤erent
markets at the same (and di¤erent) time periods points
towards the need to take into account local and currencyspeci
Þc behaviour. The same can be said for the models of
Foster and Viswanathan (1990), Subrahmanyan (1989), and
Son (1991).
An important result of this paper is that the inclusion of
half-hourly dummies, and taking account of simultaneity
between volatility, average spread, and number of quotations,
considerably reduces the GARCH type e¤ects in the
conditional variance of these two exchange rates. What
remains of such GARCH e¤ects can then probably be
attributed to private information and the uncertainty associated
with it.
Finally, having Þtted weekly, daily and half-hour dummies,
we can identify inter- and intra-day patterns of activity,
volatility and average spread. Some of these, for
example, the impact of the Tokyo lunch hour, we have
previously documented. Others are already well known in
markets, for example, the rise in spreads and decline in
activity on Fridays. But we were surprised by the Þnding of
the continuing high volatility, in both currencies, throughout
the period of US market opening, despite steadily falling
activity, which we had expected. Much of the public information
on economic news in the US is released at, or
before, the market opening, so exactly what keeps volatility
so high during the afternoons in the US is a mystery to us