Dutch sandbox aims to broker EU consolidated tape

The European Commission has moved to create the legislative conditions that will foster the emergence of EU consolidated tapes (CTs) of market data. Late last year, the bloc published a proposal that looks to amend the Markets in Financial Instruments Regulation (Mifir) to allow a set of tapes to materialize. As WatersTechnology reported last year, data quality issues and some regulatory uncertainty over data are giving potential consolidated tape vendors pause for thought, as a lack of standards and gaps in the data have impeded efforts to develop workable prototypes for a consolidated tape.

Dutch markets regulator the Authority for the Financial Markets (AFM) has been working with financial technology vendors to understand these data quality issues, develop viable prototype tapes, and feed its findings back to the EC, pan-European regulator the European Securities Markets Authority (Esma), and industry participants. The AFM last year opened its tech sandbox to vendors that were interested in developing proofs of concept for a fixed-income and derivatives tape, including fixed-income connectivity provider TransFicc and investment management software vendor Finbourne.

The regulator has also set up a working group that is in the early stages of discussing obstacles standing in the way of the tape.

The AFM runs its sandbox, Innovation Hub, in conjunction with the Dutch central bank as an environment where innovators can test solutions and check that they don’t infringe on Dutch regulations. Early last year, when it became clear that the EC was looking to review Mifir to create the tape, the AFM began to think about opening the sandbox to consolidated tape hopefuls, which it did in the summer.

By November, the project consisted of five participants—TransFicc; Finbourne; analytics provider Ediphy Analytics, which has announced its own CT pilot; and two others that could not be named—focusing on fixed-income instruments.

“[This exercise] was a very good opportunity for us to see what was already possible regarding the CTs from a technical and organizational standpoint. A number of these fintechs were existing data providers that were already working with current Mifid data to offer CT-like structures. Most of the work of the fintechs that are currently in the sandbox was based on existing solutions,” says Matthijs Geneste, a senior supervisor in the AFM’s capital markets team, and the head of the consolidated tape regulatory sandbox project.

“We wanted to understand what the barriers were to the CTs in terms of data quality, data consistency and data formats, as well as the current barriers on the regulatory side that stand in the way of consolidated tapes emerging at some point in Europe,” Geneste says.

The AFM has also launched the working group—which Geneste says the regulator tried to make representative of the various industry participants, including venues and data vendors—though he declines to say who exactly is participating. “We’re working with that group now to find these barriers and these market data and data quality issues, mostly from a technical perspective, to provide useful, valuable representative feedback to the EC, and to Esma,” he says.

European politicians have been concerned that a lack of access to consolidated market data may cause Europe to lose market share to rival financial centers in the UK or US. The UK government said two weeks ago that it would be reforming capital markets regulation, including giving the Financial Conduct Authority (FCA) the tools to develop a consolidated tape. The US is pressing on with plans to improve its equity tape. Its fixed-income tape, the Trade Reporting and Compliance Engine (Trace), is a cheap source of real-time and historic data for corporate bonds.

The EU proposed to introduce a consolidated tape as part of its 2018 update to Mifir/Mifid II. Since Mifid II came into force, though, no consolidated tape provider (CTP) has emerged. EU lawmakers have pointed to several reasons for this—a lack of financial incentives, insufficient data quality, data vendor competition, and restrictive regulations. The EU’s November proposals sought to address these issues and put forward the model of one CT provider per asset class (derivatives, equities, fixed income, and ETFs).

Market participants say one of the challenges of building a consolidated tape under the current legislative framework is the cost of cleaning and aggregating unstandardized reported data, while ensuring the data covers enough of the market to make it valuable for users to purchase the data.

At the root of these issues is the lack of standards from execution venues and Approved Publication Arrangements (APAs), which publish trade reports on behalf of investment firms and must make this information publicly available for free 15 minutes after publication. Mifir doesn’t mandate any particular formats or standards for how these APAs must publish the data. APAs make the trade information available as CSV or JSON files, for example; or market participants interpret post-trade reporting flags differently. A consolidator must then normalize these into one view.

Another issue is that firms trading non-equity instruments can choose to defer the publication of their post-trade data for weeks under certain scenarios, such as in large-scale trades. This leads to gaps in the reported data.

The idea of a consolidated tape, especially one that disseminates real-time data, has not found favor among all stakeholders. Trading venues see price controls on market data, or mandating their contributions to a tape, as existential threats to their businesses.

Geneste says AFM’s role is to be an “honest broker,” adding that the regulator steers away the working group from the most intractable topics so as to facilitate a productive dialogue. “We want to make the discussion about the tech practical and concrete, rather than the principled, high-level discussions that we otherwise end up in. So we try to avoid politically sensitive topics around costs and fees, fee distribution, etc.,” he says.

The working group began with the AFM soliciting written contributions on various issues; it is currently following up on these with individual conversations, ahead of a roundtable discussion in April. At this meeting, “we will hopefully agree with the group on some high-level principles to achieve a corporate bond CT, and lay the foundations for a derivatives CT. We want to facilitate those technical and practical discussions, and for what we agree with trading venues, sell sides, buy sides, and infrastructure firms could be the basis for these fintechs to start developing actual proofs of concept,” Geneste says.

“At the same time, we can feed those insights back into the policy discussions that are currently ongoing in the Mifir review on the Level 1 text, and later in the Level 2 discussions in Esma.”

Proving the concept

Geneste says the AFM’s first question when opening its sandbox for the CT was a basic one: Is a consolidated tape even technically possible, and is developing one prohibitively expensive? The sandbox showed them that there weren’t technical barriers to a tape, says Coen de Putter, supervision officer of trading venues at the AFM. And the costs weren’t remarkable, as the vendors, while distinct in their business models, were all analytics providers already working with regulatory data.

“The discussions the AFM had in the sandbox with the tech providers showed us they had already built consolidated tapes with the data they already had. And that’s the kind of feedback that we want to produce and show actual evidence for, so the Commission can understand exactly what it’s going to cost. And these models would be feasible from a technical perspective,” Geneste says.

Exchanges say building a tape would be too expensive, and the market would be better off with a tape of 15-minute delayed data. The Federation of European Stock Exchanges (Fese) estimates that a close-to-real-time consolidated tape would cost €77 million ($84.5 million) to build, and €35 million ($38.4 million) to run. The EC’s consolidated tape proposal’s estimates are far lower, with a total annual cost of €1.46 million to €5.4million ($1.6 million to $6 million) for a bonds or derivatives tape provider, assuming a five-year payback period.

Once the AFM was satisfied that a tape was a realistic proposition, questions centered on the data quality issues.

Geneste says the output of the tapes will be shaped by market demand—it’s unlikely to be difficult to work out what the market will need. The inputs, however—the raw material data—are where the difficulties lie.

“That was one of the main insights we gained from working with the fintechs: that getting that data is quite difficult. The European Commission’s proposals are trying to streamline that”—considering the harmonization of mandatory data reporting standards—“which makes sense. But they need to think carefully about the standards for getting the data from the data contributors into the CT, with how they’re going to prescribe the preferred formats,” Geneste says.

If the data inputs were cleaner and the work of consolidation easier and cheaper, vendors could use the tape to build value-added products on top of it.

“In the US, you have this whole ecosystem that has emerged around Trace, with different types of data vendors basing their offerings on that. That is a situation that we would like to see in Europe, where we have a level playing field, an equal starting point with the data and access to the data, and various types of solutions can be developed around it, be it for transaction cost analysis, historical data, various forms of analytics—whatever you can think of,” Geneste says.

Regulators like the AFM could also use the tape for fixed-income market surveillance, de Putter adds. “If we are looking at the same picture, it’s going to make life so much easier. You will see all these use cases evolve that we haven’t thought of yet,” he says.

Launching pilot

TransFicc said in February that it had produced a pilot for a fixed-income consolidated tape in the AFM’s sandbox. The company says its consolidated tape pilot uses “key components from TransFicc’s normalized post-trade feed,” and is inviting clients and regulators to test the pilot for performance, resilience and ease of integration.

Clients testing the pilot can access TransFicc’s client library API, which supports pre-trade data, central limit order book, request-for-quote trading, and post-trade. They will be subscribed to the pilot solution’s feed, which publishes 30 fixed-income messages per second via low-latency messaging protocol Aeron.

TransFicc’s offering that it markets to clients is a normalization service that brings together APIs from various fixed-income, currency, and commodity execution venues, and provides one point of contact for banks to interact with these venues, whether for streaming prices, execution, or post-trade.

Tim Whipman, head of business development at TransFicc, says the company started seriously considering its bid as a CT provider when it realized that it fits most of the EU’s resilience and tech attributes for a tape, as laid out in the consolidated tape proposal.

“We connect to pretty much every major venue in Europe. We don’t take in aggregated venue data, but we do take in that data with the same fields for banks individually, after they have traded. So we have the connectivity and relationships with venues in place, and we can leverage that for a consolidated tape because it’s essentially the same thing,” Whipman says.

A CT should be able to disseminate data at high speed. “We allow banks to stream pre-trade pricing into these venues. That is a lot of data that needs to be pushed into the likes of Bloomberg and Tradeweb at high speed. We already do that today. Also, the servers we use to host these services are efficient, and the way our infrastructure is constructed makes it low cost for firms to use us for this service. And that is another reason why we believe that using this type of infrastructure could allow a consolidated type to be built for a sensible amount of money, as opposed to the sums we’ve seen in the press over the last few months,” Whipman says.

TransFicc hosts multiple regions and availability zones in the cloud and physically in Equinix, Whipman adds, so it has no single point of failure.

TransFicc has also built an API aimed at allowing smaller investment firms and banks to contribute data should the EU and Esma decide that reporting doesn’t just to have to come from the APAs. “That’s not been defined yet—who must report? Are we aggregating APAs? Are we aggregating APAs plus venues? Or are we saying that people can direct report? We think the pilot can accommodate all those groups, and that’s really important,” Whipman says.

By design

Finbourne has been looking at how to turn its flagship data aggregation platform Lusid into a consolidated tape. The vendor also launched a forum last year that it calls its Design Council, which brought together some 18 representatives from industry groups, buy- and sell-side firms, and consultants to discuss issues around the practical implementation of a tape. (The AFM is an advisor to the Design Council.)

The vendor also conducted research into a dataset of about 70 million transactions, where it tried to get granular detail on the inconsistencies within the transaction data that will feed into a CT. Finbourne published a series of whitepapers on the back of this work, dealing with the consolidation, consistency and cohesion of data for the consolidated tape.

It announced late last year that it was joining the AFM’s sandbox, saying that it would share this research there also.

“We’ve gone down to trade level data to see what’s in the fields. And a lot of what’s in the fields is incomprehensible stuff, where people put prices in, and the prices are the inverse of what they mean. Or they put yields in, and you have no idea what the yields are benchmarked against,” says Neil Ryan, CTP lead at Finbourne. “We’ve asked the Design Council to pick three or four of the core fields, and identify solutions.”

The Design Council has heard concerns from members that regulators could end up imposing strict data standards on the industry, rather than leaving it to the industry to oversee data quality, Ryan adds.

“Let’s have an industry-led solution to an industry-wide problem. That is the approach we are taking, and from that, we’ll look to build an infrastructure that makes sense, and build it with partners we’ve involved in the Design Council,” he says.