Ever since MiFID ended the stock exchange concentration rule, competition and confusion have reigned. "The Fidessa fragmentation index for liquid stocks continues to grow in Europe from around 1 in Spain to 3 in the UK compared to 5 in the US, so the problems of a consolidated tape are likely to grow as well," says André Kelekis, head of market data strategy at BNP Paribas. "Most illiquid stocks are still just traded on the primary exchange." He argues that while trading fees have clearly fallen under MiFID, market data fees have risen hugely, perhaps 70% since 2000, according to Kelekis, encouraged by the shift from mutuals to for-profit exchanges. "This offsets most of the benefit," he says. "Part of this of course is the increase in transaction volumes due to technology, but we need to address the cost issues."
"In the beginning every train company had their own gauge of track. Then we had national standards and finally European standards. As transport costs came down commerce soared. That's what we should see with standard market data in Europe. The benefits are enormous."
"While MiFID clearly mandated transparency for equities trading, there was less clarity around just what needed to be reported," notes Andrew Allwright, business manager for MiFID solutions at Thomson Reuters, "So people reported everything, but there was no way for the market to classify what they saw around OTC, thus effectively overstating liquidity." He describes how CESR has now provided technical advice to the Commission confirming and addressing the problems with duplication and poor quality data, and documenting the lack of any clear mechanisms for errors to be reported or resolved.
"Some publish their data in 3 seconds, others in 3 minutes," adds Kelekis. "Some firms confide to have introduced artificial buffers into the reporting to avoid leakage. It's very unsatisfactory. With the poor data quality there is very little real transparency."
"Each exchange has different standards to distinguish instruments in floor trading from the electronic books for example or lit markets from dark pools," explains Chris Pickles, head of marketing for financial markets and wholesale banking at BT Global Services. "The OTC categories are even more diverse, and we're still struggling to come up with a common symbology and reference data. How do we identify platforms for example?"
"None of the existing commercial solutions for post trade can be comprehensive, accurate, consistent, or timely because the data they get is just so poor and diverse," says Denzil Jenkins, director of regulation at Chi-X Europe.
"Currently there are no standards around how firms publish their trades, only that they are obliged to do so," says Allwright at Thomson Reuters. "CESR is now proposing a system of Approved Publishing Arrangements or APAs for all OTC prints similar to the FSA's Trade Data Monitor regime.
This would apply rigorous quality controls." As a result, simply publishing to the web site would no longer be acceptable. "They are also tightening up normal reporting times to within 1 minute of the trade instead of 3, and block trades would also be reported more quickly," he continues. "Finally a more granular classification of trades should eliminate duplicates by identifying, for example, end-of-day volume weighted average trades or give-up facilitation trades that don't represent new liquidity. Currently OTC prints just have too much noise to be useful." With these enhancements, he asks, "Would a consolidated tape per se really add very much?"
"The buy side doesn't want to connect to every platform, but they do want to know the prices, so we have to get this right," insists Kelekis at BNP Paribas. "In Europe market data is 7 to 10 times more expensive than in the US."
"In the US they say that the consolidated tape is the 'truth', and this is what's lacking in Europe," says Bradley Duke, managing director of Knight Direct Europe. "At Knight, we believe a consolidated tape is an essential component of efficient markets. Out of necessity, we've had to create our own consolidated view, but if we do a VWAP or participation in-line execution for a client, we cannot escape the fact that there is no unified and universally accepted reference point used by the industry at large with common standards."
Not everyone, however, sees the US model as ideal. "The US has had a consolidated tape since long before Reg NMS, but the jury is still out whether it serves a useful purpose," says Herbie Skeete, co-founder and managing director of Mondo Visione, the global exchange and trading intelligence provider. "Like most utilities it was launched with great enthusiasm, but has been left to wither with ramshackle hardware and software due to underinvestment. No one wants to pay for it. The exchanges would rather bypass it to make more money."
"People are generally satisfied with the tape in the US," Duke argues. "It's an SEC mandated, not-for-profit utility, and, while it is not entirely without flaws, it does provide a fairly clear reference point of what traded where and how much."
Building the market railways
"A consolidated tape is desirable for all segments of the market: high frequency trading, retail, institutional buy side etc.," says Peter Randall, CEO of Equiduct Systems. "However, even more important is a 'consolidat-able tape', a set of agreed standards that everyone can follow. Without standards covering everything from content to format to standard categories, timestamps and clock arrangements, we can't ensure the quality of the tape." He then uses a striking metaphor. "We have to think of it like the railways," says Randall. "Before Brunel's Great Western Railway in the 1830s, there were no synchronized clocks across the UK. Now we take synchronization for granted. It's the same with exchanges. We need to synchronise them across Europe within our technical limits. There are huge benefits to be had, but everyone just thinks of the headaches to get there. Once it's done there will be no looking back."
"Before Brunel's Great Western Railway in the 1830s, there were no synchronized clocks across the UK. Now we take synchronization for granted. It's the same with exchanges."
Kelekis at BNP Paribas draws a similar comparison. "It's like the spread of the railways in Europe," says Kelekis. "In the beginning every train company had their own gauge of track. Then we had national standards and finally European standards. As transport costs came down commerce soared. That's what we should see with standard market data in Europe. The benefits are enormous."
Standards are apparently key, but visions differ. "There appears to be a need for standards to allow people to build a consolidated tape of record for post trade TCA and best execution purposes," says Allwright at Thomson Reuters, "but there's much less agreement on the need for a real-time consolidated tape as such. It's likely just to add costs without significantly adding value and there are too many conflicting interests."
Paul O'Donnell, COO for BATS Europe envisages a somewhat broader role. "People need a consolidated tape to benchmark their trades, smart route their flow and work out which venues they need to access," says O'Donnell. "The APAs are for OTC trades and are similar to the Trade Reporting Facilities run by exchanges in the US except here they're not limited to an exchange."
Meanwhile, regulators may have even greater ambitions. "For some a consolidated tape is linked to a European trade through rule, like the US," says Allwright. "Personally I can't see that happening in Europe. MiFID's best execution rules are very different, but some regulators have so lost faith in the market that they may refuse to listen."
"There's even debate over the scope of a consolidated tape," says Pickles at BT, "all EU listed stocks or just blue chips or just liquid stocks or the underlyings of designated indices etc. There are various views." Indeed, he argues that the data will need to be tagged in different ways so people can slice and dice it to meet their needs. "We need more of a tool kit rather than a bundled data feed which might only increase costs," he adds.
"There's a broad consensus that some of the MiFID principles need to be tightened," says Brian Schwieger, head of EMEA algorithmic trading at Merrill Lynch. "We're very supportive of the introduction of the APA regime given that it closely mirrors the existing UK regime adopted by Markit BOAT." He agrees that post trade reporting does need attention because of potential duplication and uncertainties over what the print means and whether a trade contributes to price formation or not. "That needs to be fixed," he says.
Behind it all, however, is the ever-present debate over costs. "Firms can't cope with maintaining all the different proprietary feeds," says Pickles at BT. "With 40 or 50 feeds at least one of them is changing every week. We need to simplify things and reduce costs."
But who pays?
"Naturally there are many open questions," says Duke at Knight. "Who owns the data and who pays? Is it just a set of standards or a central utility? How will the tape be policed? Ultimately it's in everyone's interest to have a clean tape that's as accurate as possible."
"Whether we need a central body to create the tape is open to debate," says Schwieger at Merrill Lynch. "The data vendors are already doing it; they just need the standardized inputs."
In its position paper to the European Commission in May, the London Stock Exchange clearly set out its opposition to the development of a mandatory European consolidated tape, arguing that such a development would be too costly and inflexible, and unable to meet the needs of market participants in the post MiFID world, citing the US experience. Consequently the Exchange argued that participants would be obliged to make additional alternative arrangements, adding further to costs. In the Exchange's view, the market is best served by commercial-led providers, capable of responding nimbly to meet the changing needs of investors and market participants.
"At this point no one is mandating a US style Consolidated Tape Association," says O'Donnell at BATS. "The priority is sorting out what gets reported and how, which will take some time, and we need a system that can evolve."
"A centralised authority to produce the tape is definitely on the table, even if decentralised solutions are also possible," says Jenkins at Chi-X. "We think the industry has perhaps 18 months to come up with a solution or face a mandated central authority. For CESR the consolidated tape is the number one problem and the political will to address this issue is there."
"In Europe with MiFID, I don't think the regulators really ever considered the commercial impact of fragmentation," says Skeete at Mondo Visione. "Real time market data costs are huge. At the moment most MTFs are not charging for their data, but how long can that last? Even if we get a consolidated tape for post trade data, how will it be funded if everyone can sell their data? The regulators may want a consolidated tape but I can't see how it will work. Investors just won't wear the fees and who then owns the data? The exchange shareholders certainly don't want to see their data revenues go down."
"Unfortunately the price of data is linked to the revenues generated by the main market indices like FTSE, CAC or DAX so there are many entrenched interests," says Jenkins. "The consolidated tape immediately spotlights these issues."
"Besides the technical issues, we need to have a grown-up conversation around the commercials," insists Randall at Equiduct. "The exchanges don't want to lose their data revenues, while the MTFs don't want to pick up chunky data fees. It won't be easy but we must grasp the nettle. The aim should be low cost access to all, without paying monopoly rents, but also without putting people out of business either."
"People talk about huge costs to build a central utility for a consolidated tape," says Kelekis at BNP Paribas. "Figures like €500 million are mentioned. This is ridiculous." He notes that technology costs like CPU, storage, or bandwith continue to fall. "Once we have standards a utility gathering and providing a real-time and historical service should only cost 10 to 100 times less or around €10s of millions with current technology," says Kelekis, arguing that data vendors could then enrich that, but the feed should be relatively inexpensive and unbundled, so not subject to any other subscription. "It would hugely reduce costs for the end user on the buy side or retail," concludes Kelekis. He notes that the Commission is also looking to revise the Prospectus Directive to oblige firms to publish their securities data in standard electronic format into a central repository. "Managing quality at the source level will further reduce costs by eliminating multiple data sourcing and facilitate trading," says Kelekis. "The answers are not that complicated. We just need the will to act."
Meanwhile, Jenkins at Chi-X sees further issues looming. "More important perhaps is the debate over dark pools itself and the grander debate over the desirable size of OTC," he argues. Here there are apparently strong feelings between those who thought MiFID permits OTC and those who thought MiFID would push nearly everything onto exchange. "There's a lack of knowledge and ill informed speculation," says Jenkins. "Some in the Commission claim it's grown, while the banks insist it has diminished with the rise of high frequency trading. Everyone is now agreed that we need proper data to judge such issues, and improved standards and reporting granularity will give us that."
Regarding the size of OTC, Jenkins points out that in the US true OTC liquidity is only 17% of the market once you remove facilitation trades. "So we think the data will be similar in Europe and not the 30-40% people are currently getting so alarmed over," says Jenkins. "One of the main aims of the consolidated tape initiative will be to give both traders and regulators better data on which to base their decisions."
More unanswered questions
"Pre-trade consolidation is a less pressing issue because market participants and vendors have managed to handle this well enough," says Duke. "At Knight, we build our own view of the full market depth in the Knight Direct EMS for traders and as a reference for our liquidity sourcing and aggregation algorithms. It's crucial to smart order execution." Duke argues that given the multi-factor approach to best execution under MiFID - price, size, likelihood of execution, latency etc. - there can be a lot of noise, people posting and canceling liquidity etc., so it is a very challenging problem to solve.
"For consolidated quotes," says Allwright at Thomson Reuters, "the vendors are currently offering a wide range of solutions, whereby users might select the venues they wish to monitor and then see a customized, real-time aggregation, as with our Reuters 3000 XTRA, for example. Buy side firms we've talked to don't see a need for further regulation in relation to pre-trade data."
"The consolidated quote is less important for the large institutional players," says Jenkins at Chi-X, "but for the retail end of the market it's more problematic. The consolidated tape has the greater priority because it makes best execution decisions more feasible and will increase competition. However, the debate over a consolidated quote is unlikely to go away even then."
On the pre trade side, Randall believes that Equiduct's virtual best bid or offer (VBBO) product already offers does a great job with price consolidation. "We calculate VBBO 100s of millions of times per day based on feeds from all the major platforms," says Randall. "It's being used in production to price retail trades for market makers so that investors know they are always getting best execution, with the security of printing on a recognized exchange, namely, Boerse Berlin." Being focused more on the retail segment, he admits that Equiduct volumes are still relatively low but notes they are growing quickly. "We recently surpassed those of the Irish Stock Exchange, for example," says Randall, "and have just acquired another big retail broker, so there's clear demand for a pre-trade consolidated tape as well."
This retail end of the market is important. "Some 12% of the market today is not achieving best execution," says Randall, "and much of that disproportionately represents retail investors who are just transacting on their primary exchange. That has to be addressed and I think the European Commission will do it."
Nevertheless, a consolidated quote feed would face many challenges. "As for an officially mandated EBBO, issues around latency arbitrage and increased risk of 'flash crash' scenarios should be considered," says Schwieger at Merrill Lynch.
The other big issue is who should take the lead on agreeing the standards. "I think it should be up to an industry organization like FIX to sort out the standards," says Randall at Equiduct, "which inevitably will have to evolve, not be chiseled in stone." He argues that FPL has wide industry participation from all the stake holders, exchanges, MTFs, investment banks, buyside, technology firms, data vendors, and managed service providers. It has no commercial interests and is at the heart of current state-of-the-art developments."