There’s always plenty of talk in the press of new technological developments and market movements shifting the balance of FX trading away from the traditional centres of New York, London, Singapore and Tokyo. Reuters published an article recently about new fibre optic connections which threaten to remove the perceived advantage these cities have by being located close to high-capacity undersea cables. The gist of the story is that high-speed, high-capacity fibre optic cables can move data inland at rates which would negate the advantage gained by being close to an undersea cable, which means that cities further from the coast could conceivably become centres for FX trading.
Reports of the demise of these cities as global FX trading centres have been doing the rounds for some time. The reasons cited include the proliferation of electronic trading (no need to be next to the exchange), competition from emerging economies and the global nature of FX trading. But as yet none of these predictions have come to pass. In fact London has recently become the largest clearer of Renminbi outside of China. According to the Bank of England, London is the world’s largest single FX trading centre with USD 2.15tn traded daily, giving it 41% of the market. New York has 19% and Singapore and Tokyo each have around 6%.
So why, when many commentators are sounding the death knell, are these FX trading centres still thriving, and actually growing?
Location, Location, Location?
To understand what keeps these cities in the positions they are in, it’s essential to understand how modern FX trading operates. According to Greenwich Associates, 74% of FX trading is electronic and (because it’s electronic) it requires vast amounts of hardware, processing power, servers, storage and connectivity. All this needs to be up and running all the time and kept secure, cool and clean. Given the huge amount of space required to house it all, it’s impossible to do in a crowded metropolis. Thus, single and multi-bank FX platforms are hosted in specialised data centres outside the city and as e-trading is largely a latency game (the closer you physically are to the market the better) most participants will co-locate their trading platforms on servers at the same datacentre.
Indeed, co-location brings benefits other than latency reduction. Most players in the FX market need to connect to a large number of markets and clients. Because everyone is located in the same data centre it’s much more efficient to do this through cross connects than via long distance telecoms links
Thus, trading on a new venue or accessing a new data feed isn’t always about putting in new cables or digging up streets any more. Clearly it’s going to take a lot to shift all this infrastructure to a new city when it becomes the next London, New York, Tokyo or Singapore.
It’s this web of cross connects rather than new infrastructure or market forces that will maintain the status quo. The FX market is global and trading occurs where the liquidity is. The web of cross connects, built up over many years in traditional FX-trading centres, is now so interdependent and complex that it is impractical to untangle it and then recreate it at another location. The addition of a high speed connection to a new city is not a strong enough driver for market participants to decamp, so liquidity is also likely to stay where it currently is unless something forces it to move. Because of the infrastructure already in place, the current dominant centres will continue to dominate.
Plus ca Change
So does that mean that infrastructure improvements are pointless? In short, no. Clearly high speed connections to these major cities are essential. If, for example, you’re a FX trading desk sitting in Geneva wanting to trade using your eFX trading platform co-located in a datacentre just outside London, you will still require a fast and reliable connection to your trading engines from your trading desk in Geneva for command and control.
Many infrastructure providers can now offer access to multiple markets and market participants through extranet services co-located at these datacentres. Simply connecting to the extranet service gives you low-latency access to whichever trading venue you need to trade on. This is why the reports of the demise of London, New York, Tokyo and Singapore are being exaggerated.
Before the proliferation of low latency, high capacity connections, location was everything. However this is no longer the case. The huge investment in infrastructure means FX trading now occurs in a virtual world where the actual location of the trading venue or the trader hardly matters. Through new extranet services, the ability to locate trading platforms in the same facility as FX platforms gives any trading firm the proximity to a platform that was historically only available to those in the city. London, New York, Tokyo and Singapore will continue to be dominant FX-trading centres because in (virtual) reality, there are no FX-trading centres.
This blog was first published on Tabb Forum