It’s been quite a week for me thanks to my digital life being turned upside down (I have never had to deal with so many help desks at one time) and the passing of the founder of my favourite restaurants in London, Dick Barfoot of Sweetings. So with a “vale” to Dick and a gesture of a different kind to the various (un)help desks (once again I am reminded that a customer’s perception of an organisation is largely created through the experience when things go wrong), let’s get things back on track. I sense there is a real pushback against fragmentation in FX markets.
Several years ago, I was told a rather amusing story about a call received to the EBS help desk during an outage on that platform. It was from another platform asking when they were going to be up again. Clearly that platform’s liquidity providers were recycling prices from the primary venues and creating the liquidity mirage that was so often spoken out in the middle of the first decade of this century, but at the time we didn’t think much of it beyond it was funny.
Last week at Forex Network Chicago I found several people for whom the joke has gone sour and who wanted to reverse the course of fragmentation. These were not, I should stress, people with vested interests, but people I would consider neutral in the debate. Of course, if one speaks with someone in the connectivity business they would argue a different corner, but as the bedrock of their service it could be argued that their opinion is skewed by commercial interests, just as it can be for some LPs struggling with the technology cost of managing so many channels.
The earlier story does have elements of farce about it, but there is a serious side, not least how so many venues rely upon – yes I am going to say it – last look to support their business. Is anything actually being added to market quality through recycled liquidity and last look? Could it not be argued that these practices are actually damaging market quality?
I ask this because a key part of the debate around fragmentation is another pushback against last look. It may be the impending three-year review of the FX Global Code, but the noise around last look has risen a notch or two and critics of the practice are not only coming from the buy side; several LPs are now getting on board – bank and non-bank.
Last year, the BIS Markets Group released a paper discussing how price discovery still relied upon the primary venues, even though volumes on those platforms were dropping. Fragmentation has undoubtedly hurt the primary venues thanks to the newer venues being nimbler and having more modern technology – and there also seems to be a sense in the industry that the primary venues are no longer critical to the health of the FX market.
I would tend to disagree with that viewpoint due to the aforementioned story, which is clearly still relevant, and the industry’s reliance on the primary venues for data. It could also be argued, however, that is really just a plumbing and wiring issue – if another venue was hitting the $40-50 billion per day mark with as clean market data then if market participants replumbed their infrastructure that venue could easily take over. As to whether firms will be willing to go to that expense just to change the name of their primary venue is for individuals to decide, this individual thinks they will not because it is too far down the “to-do” list.
The view that the primary venues are becoming irrelevant is, without that rewiring of the system, a concern to me because it reflects yet another facet of the “age of entitlement” in which we live – not least that people maybe don’t think having a primary venue is important. They expect the primary venues to deliver the data they require to price, but they are less concerned about actually using those venues to trade. This inevitably has to degrade the quality of the data coming from those venues because volumes will continue to drop, but is anyone noticing?
Of course it can be argued that these players are paying (handsomely) for this data and it is perhaps a function of platforms separating the data and trading business lines, but the bottom line is, if we think the primary venues are important to the health of the FX ecosystem then we have to consider how they are actually used. If they are considered unimportant then it’s simply a case of rewiring the system to use different (probably more) data sources – even if a fraction of that data is actually original (one senior player at a major LP told me last week in Chicago that in terms of genuinely original liquidity they estimated that more than 70% of traded volume on the public ECNs was on recycled liquidity).
I should stress this is not a call to arms for EBS Market or Matching, more that as an industry there seems to be a reassessment taking place over the value of so many trading channels. The identity of the primary venues going forward is for the market to decide and incumbency should not be relied upon too heavily to protect one’s position.
The fact is though, the cost of maintaining so many channels, with so many different protocols, is not inconsiderable and as the buy side becomes more engaged in the markets on a direct basis that segment too will face the same challenges as the LPs have – and it is this change that I sense is driving the pushback against fragmentation. There are buy side firms out there who have the data to support the supposition that there are actually a pretty limited number of genuine LPs out there and therefore they are asking themselves the question ‘why bother with the cost of connecting to so many venues when all that changes is the name on the trade and the rate of rejections?’ Likewise, the genuine LPs are getting better at watermarking their liquidity and identifying the more egregious venues when it comes to recycling.
The upshot of all this could be that we have hit the peak in terms of fragmentation in the public market. I fully expect the disclosed, bespoke liquidity landscape to fragment further but it will be around the same protocols, pipes and commercial relationships. Clients will want to access more bespoke liquidity pools because it is easier to control the environment on both sides of the trade. If the execution methodology is too aggressive it will be highlighted – just as it will if an LP is externalising too quickly and polluting the quality of the flow. That can be managed by a few firms – certainly fewer than currently exist in the public sphere – indeed it could be argued that this evolution will take place away from the public platforms entirely as buy side firms build their own pools. The technology is available almost off-the-shelf, what’s to stop them?
If this does play out as I suspect it will, we will see a modern-day version of an old market structure emerge – one in which clients execute with a limited number of LPs in a limited number of private pools, and those LPs go to the public market to clear out any risk. Those LPs will want to be able to shift risk in an opaque environment and that does not mean pricing across multiple venues and signalling the trade to everyone. It may not mean they use the existing primary venues, as I noted who emerges as the key venues could change, but it will mean that less channels are used and for those of you that like circularity, that should mean better data from the venues of choice.
White noise is a feature of all markets, however I sense an effort to reduce it in FX markets, especially in the areas of data and liquidity. There is a genuine body of opinion that believes the FX market structure has over-developed and it currently out of control thanks to excessive fragmentation. The solution is not that hard – fewer venues – but the real challenge will be how the industry selects and coalesces around the venues of choice.