«Volume Title: The Microstructure of Foreign Exchange Markets Volume Author/Editor: Jeffrey A. Frankel, Giampaolo Galli, Alberto Giovannini, editors ...»
A dealer who has a given interpretation of the significance of publicly available information on the exchange rate during the remainder of the day may take a position based on this view by asking a price from another dealer and transacting accordingly, but he does not make a large adjustment to his own price immediately if he is not absolutely sure that all other marketmakers have already made a similar price adjustment. If he did, he might receive unwanted orders because his quote might be inconsistent with prices quoted by others.
Rather, he is likely to widen the spread in order to reduce the likelihood of such an occurrence.
It is plausible to assume that dealers opening their screens in the morning have different outlooks for the remainder of the day even in the absence of any private inside information. Dealers are cautious not to start the day with a price far out of line with other quotes, which explains the wide spread. But with a wide spread there is little or no trading and hence little or no trading profit.
Therefore, the dealers are eager to start trading as soon as possible, which explains the rapid convergence of the spreads to the level that makes active trading possible. Once established, the "feel" of the market is maintained because the prices and the spreads can be constantly tested in active trading.
The short period during which the "feel" of the market is established in an opening time zone need not affect prices quoted in a time zone where the markets are soon to close, provided there is sufficient overlapping. Dealers in the opening zone may get attractive quotes from the soon-to-close zone, where some dealers may get rid of their remaining positions at an attractive price.
In addition, sufficient overlapping provides dealers in the soon-to-close zone sufficient time to wait until the "feel" is established in the opening zone.
Additional Remarks The paper marks an opening to a new area of research, that is, the intermarket connections across overlapping time zones. It would be interesting to have comparable data on the overlapping Asian and European markets. Data on the 71 Bid-Ask Spreads in Foreign Exchange Markets spread and volatility at closing time in Continental Europe might also be illuminating.
The data are based on indicative quotations from the Reuters' screen (FXFX), which do not give the actual transaction prices or trade volume. As the authors point out, it would be useful to check the results using alternative and more accurate sources of information.
As the data reveal, the changeover from the opening phase, when dealers are getting the "feel" of the market, to active trading takes place in a relatively short span of time. It is more or less over by the time corporate treasurers and other nondealer customers step in and begin to request prices. Similarly, most of the nonbank customers have probably already disappeared when, shortly before closing, the spread widens and the volatility increases.
This raises the question of the role of nonbank customers in the foreign exchange market. Despite the high proportion of interdealer trade, nondealer customers may be needed after all to maintain the depth of the market. Otherwise, how would dealers in the aggregate earn their trading income over the long run?