BizAims

The foreign exchange market is primarily an over-the-counter (OTC) market, ie one where contracts are agreed bilaterally between participants, rather than on-exchange. The market consists of different agents, trading for various reasons. ‘End-users’—such as corporates, investors and governments—may enter into foreign exchange trades with market intermediaries (usually banks) in order, for example, to facilitate the purchase of foreign currency bonds, or to exchange foreign currency proceeds from exports into their domestic currency. There is a large professional interbank market that enables intermediaries to manage the risks arising from this activity—at the simplest level, that exchange rate moves change the value, in domestic currency terms, of an asset denominated in foreign currency—by trading to transfer risk between themselves.

Participants in the foreign exchange (FX) market have been executing transactions across electronic messaging or broking systems such as Reuters and EBS, which match buyers and sellers, for many years. But these are proprietary, closed systems, and largely restricted to the interbank market. In contrast, the market between end-users and banks was for many years based on telephone contact. But in recent years internet-based trading platforms have appeared, and are being used by a much broader range of market participants.

There are two main types.

" First, ‘proprietary’ or ‘single-bank’ systems. Here a bank allows its customers to trade with it, on its own internet-based platform, essentially as an alternative to the telephone. There are advantages for both parties: time is saved in processing trades, especially small ones; the system can be linked electronically to each party’s in-house systems for recording, settling, accounting and risk-managing trades and therefore reduces the need for re-keying and aids straight through processing; and it simplifies complex cross-product transactions (eg some systems can automatically calculate the FX implications of a string of cross-currency securities trades). All these factors should reduce costs.

" Second, ‘multi-bank’ systems or ‘portals’. Several of these have been set up, some by consortia of banks and others independently. The key difference between a multi-bank and a single-bank system is that in the case of the former, a number of different banks offer prices—that is quote exchange rates or ‘provide liquidity’—on these platforms in competition with one another. In addition to the advantages of single-bank platforms, there is an argument that the multi-bank portals may provide ‘finer’ prices (that is, narrower spreads between the prices that the bank quotes for buying and selling a currency pair—known as ‘bid’ and ‘offer’ prices respectively). They also allow customers to demonstrate, for example to their auditors, that they achieved the best price available.

The newest types of platform involve end-users disintermediating by matching transactions between themselves.

Developments in electronic platforms

During 2002, as described in the subgroup’s previous report, the market’s attention was clearly on multi-bank portals, in part reflecting the closure of Atriax, one of the major multi-bank portals. But in 2003, proprietary bank systems have been back in focus again. The most interesting e-commerce developments have been in ‘prime brokerage’, ‘white labelling’ and ‘liquidity-exchange’(1) models, to which proprietary systems are central. Further, banks that have aggressively marketed their proprietary platforms have reportedly seen much higher volumes across these platforms than through their participation in the multi-bank portals. A clear trend across all banks is that e-commerce volumes have continued to grow.

Multi-bank portals

Market participants suggest that the market leaders are widely perceived to be FXall, Currenex and FXConnect, as was the case in 2002. The ownership structure of these portals varies. A consortium of banks owns FXall; Currenex is independently owned; and FXConnect is owned by a single bank, State Street (although it is a multi-bank system in that other banks are able to offer prices). FXall and Currenex have tended to attract corporate customers whereas FXConnect has tended to attract fund managers. All systems are reportedly looking to expand their customer base into other sectors, further increasing competition between the portals.

According to one survey the daily volumes through these portals are estimated to have risen rapidly, from $7 billion per day in May 2002 to $14 billion per day by October 2002. Anecdotal evidence suggests that volumes have continued to grow into 2003.

·   " FXall reported that its average daily trading volume in April 2003 was $7.5 billion.

·   " FXConnect reported that its average daily trading volume in April 2003 was $10 billion. "

·   Currenex has not released turnover data.

However, these volumes are small in the context of the overall foreign exchange market. The survey quoted above estimated that trading over multi-bank portals accounted for around 7% of wholesale foreign exchange market turnover. There are geographical differences in foreign exchange e-trading take-up. In Europe and North America, 35% of larger organisations are estimated to trade electronically, compared with 25% of such organisations in Japan.

A new development is end-user to end-user matching systems, such as Hotspot FXi. These enable participants to post bid and offer prices anonymously, and to accept market prices posted by others. Banks can provide liquidity by posting bid and offer prices but are not permitted to accept prices placed by end-users. This model is attractive primarily because it offers end-users, such as institutional funds, hedge funds and corporates, the opportunity to trade with each other rather than via an intermediary such as a bank, which should therefore be cheaper for them. The multi-bank portals are considered to have strong brand names, and market participants believe that they may in due course expand into other products, such as money market instruments. Some market participants expect there to be further consolidation among the multi-bank portal businesses at some point. Such consolidation, should it occur, is considered unlikely to affect the broad trends described above.