Online FX a quick history lesson From http://www.amazines.com/article_detail.cfm/376305?articleid=376305

Electronic Foreign Exchange trading has been with us for over 15 years. Telex machines were one of the first electronic communication tools used between banks, then Reuters Direct Dealing allowed dealer to dealer communication. With the advent of desk top Pcs and networks, plus a desire from the banks to execute currencies quickly and efficiently, EBS was developed in the mid 90s and has been one of the most successful and sustaining ECNs for FX in existence. However its early days only catered to a ‘club’ of banks, those that were in control of the rates and liquidity, no buy-side client could access this market place.

The banks themselves were beginning to develop their own internet based platforms, UBS in particular targeting regional tier 2 and 3 banks but also some active currency managers. Deutsche Bank had begun the first stages of its FX-Trades tool, another RFQ product designed for regional banks but not particularly targeted at the buyside.

So what of buyside clients? During the 90s they had FX Connect, at that time only really accessed by the Real Money community. Owned by State Street, it naturally targeted its large Global Custody clients, many of whom had 100s of tickets a day to execute and had to do it with their custodian.

As more of them looked to trade out-with their custodian it put pressure on State Street to add more banks but their technology was not sufficient to support the growing demand for multiple bank requests, ticket processing and quick pricing. Indeed much of the actual processing and pricing was done manually by the banks.

Three ECNs were now in development going into the year 2000. Currenex, Fxall and Atriax, all looking to exploit the failure of FX Connect to move quickly.

These platforms were going to revolutionise online FX Trading. Accessible over the internet and deployed via a desktop, they also allowed their target clients (Corporates and Real Money Funds) to trade multiple tickets at once, compare pricing among a select number of banks and STP into their Risk Management or Treasury Management systems. All done under the new Request for Quote process, where by a client would input trade details and send a request to a bank for a rate. At this time relationship was still very much important to the clients and they would tell the banks what side. Most banks at this time still had a sales person manually pricing these tickets.

Atriax owned by Deutsche Bank, Citibank and JPM plus some large client shareholders had first mover advantage, but its lack of support from other banks, lack of support internally and clients still uncertain which platform or indeed, if, internet trading was the way forward, led to Citibank leaving for Fxall and eventually the collapse of Atriax in 2002. After this failure, there was a sense in the market that others would follow. Currenex, late to the party. was finding the support from the same sell side banks on Fxall difficult to gain. Its funding by TH Lee, a venture capital firm was rumoured to be waning and Fxall had managed to gain traction with the very clients Currenex was targeting.

During this period the banks (sellside) were improving technology and their ability to deliver rates, clients were becoming more frustrated with the RFQ process, and indeed within some banks they were still priced by a sales person sitting on the other end of the request. However, the first sign of change was afoot, banks now looking to streamline their pricing capabilities began to develop pricing engines for RFQ requests. Now an RFQ within the parameters defined by the bank, would automatically be priced by a programme, a robot trader if you will. As clients were quite transparent online, this allowed the banks to tailor pricing and spread accordingly. The API/FIX revolution was beginning.

However, one sector of the market had been left out in the cold: another ECN had slipped under the radar avoiding much publicity in the Real Money and Corporate sector. It was the place to be if you were a hedge fund. Hotspot FXI marketed itself as an Anonymous FX Exchange, with the new pricing engines now able to stream constant spot prices, Prime Brokerage now readily accepted in FX they cornered the market. The idea was simple, client to client trading and only ever settle in the name of their PB, allowed for the first time, clients to make prices into an FX market place without showing a name, being read on a side and not being controlled by a banks view on risk.

The sell side banks had been left behind. This new technology, was supposed to bring efficiencies to their clients and yet allow them to retain their position as the leaders in providing rates. What it had done was increase transparency of rates, thus tightening spreads, expose inefficiencies in their pricing engines (latency arbitrage) and dramatically reduce their revenue. They were unsure of their next step and a phrase that makes everyone groan was born, ‘liquidity mirage’. Many more ECN based around streaming pricing were created, Lava, E-Speed and Currenex had begun its revolution with ESP. The banks were claiming that due to the speed of the internet and the ability for clients to deal on the growing number of venues the viewable liquidity across these venues was more than was actually available, thus creating a false sense of volume. At the time everyone nodded and agreed and the Bank of England Committee took it very seriously. The statement in fact was wrong, it was scaremongering. There was no liquidity mirage, there was an inability in certain banks to control their exposure as they had not put in place the necessary risk tools to manage it, but nothing had actually changed. Go back to the early 90s and spot desks were making prices on hundreds of mios of dollars at one time but their only risk control was the spread, they had no real access other than Reuters and voice brokers to the liquidity available.

By 2004 Currenex had redesigned itself as a technology vendor. Rather than trying to be a market place like Hotspot, it tapped into new sources of liquidity via a white label platform with Refco. This actually gave the banks their first taste of retail FX but also gave Currenex that much required niche. This was the beginning of their turnaround. Lava had now appeared with the same concept as Hotspot (indeed under ex Hotspot founder) but with more advanced order management and other platforms were to follow each offering something slightly different.

The banks had begun to get to grips with the new technology, credit was gradually under control, and actually utilising some of the hedge fund API technology had been able to remove the latency arbitrage risk from their pricing engines. A better understanding of their clients was developed by actively monitoring their trades, knowing if abc client buys and is right 25% allows you to either cover the trade or indeed narrow your price to encourage wrong trades, these tools were the edge the banks needed.

Much has now changed since the early days in the 90s, the supposed demand for an FX Exchange has led to yet another platform FX Market Space a combination between Reuters and the CME has appeared, its volumes suggest it has some work to do. EBS opened its doors to the hedge fund world and the banks took profit, selling to Icap almost immediately. Currenex was purchased by State Street for an astronomical figure of $564 mio, Hotspot was purchased by Knight Trading for over $70 mio and Citibank took control of Lava.

The market has fully realised the importance of FX in the institutional world there are no doubters now, online FX is growing year on year, there are now over 30 platforms available to choose if you are an institutional trader and 100s more for the new market, retail FX. Understanding where to go, how to get credit and who is the best is not for the faint hearted. Making the right choice shouldn’t be left up to chance.

John Farrell JFX Consulting FX E-Commerce Consultants www.jfxconsulting.com jfarrell@jfxconsulting.com FX E-Commerce Consultancy

ECN: Electronic Currency Network

ESP: