With the release of FIX 5.0 I thought id reflect upon the part of my presentation at Tradetech Fixed Income 2004 where I asked whether FIX will become the defacto standard protocol throughout fixed income, while showing this picture (when hell freezes over)
Back at that Tradetech I was waving the flag about a new service we’d launched that utilised FIX for fixed income, which made a significant process improvement to allocations on electronic trade flow. We’d plumbed the Bloomberg allocation screens (which customers use to send allocation instructions for their AutoEx trades) into our allocation engine via a FIX pipe, so the customer could actually interface direct with our engine rather than just msging a sales contact who’d then need to do something to manually book the shape and confirm when done. Two years on our service is still there, still used, but despite Bloomberg offering same functionality to all other dealers at no extra cost, I am told we are still the only dealer who is actually using this FIX connection rather than relying on their salesguys or middle office getting a msg and manually booking the allocations. This leads me to question whether FIX is as ubiquitous outside of OMGEO etc as is made out for fixed income allocations?
Since that 2004 Tradetech we have established FIX connections in which our customers conduct their fixed income price discovery and electronic order flow. Yes, customers trading Fixed Income with us via FIX. Many sellside have said for some time “were ready for fixed income FIX as soon as customers want to use it”, but I’m referring here to real customers, real fixed income FIX connections, with real fixed income flow via FIX each and every day. But its generally just those customers using FIX to trade other products with us already and who have the organisational structure allowing them to leverage this existing connection – as the majority of FI customers remain more than happy to continue in the usual multidealer platforms at this point. If you read my blog then you already know I expect FIX to become the de-facto standard for connectivity on the customer side in the cross-asset space, but as this is driven by uptake of OMS, and specifically for etrading within the OMS, this is still some way off in terms of reach critical mass.
As the FPL “process” to agree requirements and establish releases means FIX is only ever deployed in OTC space to replace whatever the original connectivity solution was, would moving to an open source development model mean product initiators could “do it in FIX ” from the outset rather than building their proprietary connectivity which will at some point in the fullness of time be supported in FIX? Surely once the guys at the coal face of developing electronic trading for totally new capital markets initiatives have the ability to deploy a full lifecycle FIX-based solution from day1, FIX would really work its way into the hearts and minds of the movers and shakers who sponsor these development projects? …at which point uptake would surely snowball.
The OTC FX and FI electronic markets .. and so the current generation of pricing/trading systems .. were built upon the premise of using proprietary APIs, and these APIs have each evolved in-line as each liquidity pool has evolved over the years. So for existing market participants FIX will clearly need to do that connectivity better in order to be considered a candidate to replace any of this existing connectivity. Of course one “better” might be in the aggregation of disparate feeds and venues; e.g. the pitch of Currenex’s “OpenSTP/FX” (FIX-based STP protocol for FX, MM and precious metals), though this clearly only works as a standard if more than one of the liquidity pools offers it :-s (and even then this particular Currenex example is only really of interest to those who don’t have an existing STP capture in place).
It’s not all doom and gloom. FPL has addressed some of its major issues in FIX though; with FAST aiming to shut up everyone who says FIX can’t cope with streaming data, and 5.0’s transport independence addressing the complaint that there are too many versions of FIX in use for it to really be a “standard” (- though is the irony that we still need everyone to update to 5.0 in order to actually use this??), the hurdle to using FIX appears now to be just a matter of “implementation” – shifting the problem from being a technical issue into a “FIX implementation” resourcing issue instead.
It’s the new venues and market participants (and for examples of these see the “2015” category) that will drive the uptake of FIX in the OTC space. So in terms of that slide I was talking about at the start of this… while it’s definitely not freezing over in Hell just yet, I think it has probably got colder there in the last couple of years.