The appetite for the fastest possible interface will always be present and FPL’s continuous, collaborative work with the exchange community, evidenced with a number of working groups, has resulted in improved latency, making FIX messages more suitable for high speed trading. Through FIX 5.0, for example, exchange clients will find it easier to implement a more flexible, faster connection to the exchange. Exchanges are increasingly taking advantage of the latest advancements in FIX and using it to establish points-of-presence in major liquidity venues worldwide — thereby providing local connectivity for local customers, which in itself significantly reduces cross-regional connectivity costs.
The cost of defining a new protocol is considerable and these costs continue for the lifetime of the product. Every new software release must include additional regression tests based on very demanding performance tests. This will ensure performance gained is real and constant. Benefits are intrinsically about the additional flow bringing revenue that the exchange attracts, either from competitors as a result of the speed offered or from new flow that is created due to this speed. This cost calculation for customers is also important. The customer will be looking for measureable benefits and/or lower costs. If the binary interface uses FIX the cost of developing and using the interface may be lower. If the data types are common with FIX, the integration with existing OMS systems may be easier.
The Nordic Growth Market (NGM) is a good example of an exchange who has applied this approach. Choosing a proprietary interface over a standard like FIX has consequences beyond just order data. Latency data for instance are now scrutinized across the entire trading cycle and are used by some innovative trading strategies – take for example Royal Bank of Canada’s new system Thor, which bases order slicing and routing on latency data for the link to each liquidity venue. To promote transparency and fuel innovation, the industry is nearing developing standards to handle and distribute such data, such as the work of FPL’s Inter-Party Latency Working Group. New standards will thus emerge which will likely leverage and extend existing ones. As a result, an exchange that continues to use their own proprietary API over standards like FIX may be left out in the cold; isolated in the new domain and thereby excluded from those new trading strategies, or having to incur additional costs in retrofitting support for the new standards in their proprietary API.
If we examine arguments against a generic standard these are mostly around it lacking the latest functionality, and therefore market requirements. FIX versus proprietary exchange API’s varies in functionality, with degrees of variation determined by those driving them. Exchanges must weigh in competitive pressures when it comes to making technology decisions. Introducing FIX standardization and the potential resulting reduction in costs it brings, for it and also its members, has to be weighted with the market differentiation opportunity in the area of speed. Important to note, time and time again FIX has provided a greater degree of flexibility compared to proprietary exchange technology.
The FIX Protocol organization actively listens and responds to the ever evolving requirements of the financial community. If you examine both the ECNs in the US and new entrants in the European markets, virtually all of these use FIX as their primary interface. It is also no coincidence that many exchanges have launched spin-off technology businesses to
generate additional revenues but more importantly, propagate their technology choices in the hope they become the industry standard.
Today’s state of play sees the healthy, exponential uptake and adoption of FIX by a broad cross section of trading venues encompassing all sizes and shapes globally. In an increasingly cross-asset, cross-border driven execution environment, this is a good thing and most pundits would agree vastly superior to any alternate outcome.