In the world of tech, there’s this push-and-pull dynamic whereby firms scale up via acquisition but do so at the risk of growing too quickly and creating a disjointed offering. In the best instances, a company acquires a piece of tech that seamlessly integrates with the acquirer’s other functionalities. But when these integrations go wrong, the result is a technological quagmire.

I was recently speaking with a senior technologist at a very large trading technology provider that has grown through acquisitions. His company’s greatest challenge—which has become their main focus—is creating a common taxonomy in which all the different platforms the company has acquired could interoperate and talk to one another. Despite years of deals, it’s something the vendor only seriously began to address about two years ago.

Acquiring is the easy part—merging is easy only to put off for another day.

Earlier this year, the London Stock Exchange Group built on its acquisitive streak with the addition of buy-side order and execution management system and portfolio management system provider Tora for $325. The appeal for the exchange operator is clear: LSEG doesn’t currently have an OEMS with a strong presence in Asia’s growth markets. Tora helps LSEG expand its coverage of fixed income, foreign exchange, derivatives, and, perhaps most importantly, cryptocurrencies. The exchange doesn’t currently have any products that enable digital asset execution.

When LSEG acquired Refinitiv in the summer of 2019, we wrote that the deal was about much more than just market data. While it’s the second largest market data provider, Refinitiv’s portfolio of platforms includes the Eikon trading terminal, Redi EMS, FXall trading platform, and buy-side OMS AlphaDesk. The process of integrating Refinitiv has been an arduous one.

“Geographical expansion makes sense, but how do you offer a product? Clients don’t want three products for three geographies. Just because you have Tora and Redi, that doesn’t mean that your clients can magically trade globally,” a former Refinitiv senior executive told me.

A former LSEG executive raised concerns about support capabilities. “By putting Tora support into a central department, where you have one customer service rep who’s supposed to service Eikon and Redi and FXall—they all have vastly different client engagements,” they said. “What they do is take it down to the lowest common denominator—the same guy is supporting six things because they want to save costs. They’re not Bloomberg, where everything is in one box. They’ve never been able to achieve that, and they never will because they’ve already abandoned Eikon.”

Dean Berry, group head of trading and banking solutions at LSEG, strongly pushed back on these assertions that integration would be problematic for the exchange operator. As is true of any integration project, time will tell. But it’s worth noting that two former LSEG execs say that the exchange is not done acquiring. They say that after losing out on TradingScreen (to Francisco Partners), Tora become the next target, but not the last.

[Editor’s note: A day after this column was published, LSEG announced that it is acquiring Global Data Consortium, which provides identity verification data to support clients with KYC requirements. While not in the same vein as those aforementioned trading platforms, it’s still another integration project.]

When I asked one of the former LSEG execs—who seemed largely positive toward the deal—if there was a risk in adding more platforms when the Refinitiv integration was still ongoing (both with LSEG, and even integrating those other Refinitiv properties to one another), they paused for a long time.

“I think there’s something to be said that there’s not many assets left in the market right now. So if you look at what’s going on in fintech, with the amount of M&A that’s going on, they feel that they need to be participating in some way in that.”

Doing it and doing it well

A few weeks ago, Max Bowie examined a handful of market data platform integration projects after some recent M&A—specifically, Exegy-Vela, Options-Activ, and Iress-QuantHouse.

Looking at that last one, Iress has spent the last 18 months integrating the connectivity provider’s tech. Arthur Tricoire, a general manager at the Aussie market data and trading workstation provider, told Max that they now have “a solid roadmap of tasks for the next 34 months.” Integrations, after all, are never quick and easy if they’re to be done well.

With the deal, Iress was able to add a proprietary market data collection infrastructure. Previously, it relied on data sourced from other third-party distributors. But the devil lies in the niggling details of an integration project.

“The heavy lifting part is… that the more you look at a new data model, you realize there are differences, and that other things need to be updated, and you need to make changes upstream—i.e., one small difference in data formats can lead to a lot of changes,” he told Max.

Additionally, a fair amount of regression testing is necessary.

“A big part of the QA and engineering tasks is regression testing. Most of the platform footprint is display end-users, so we had to make software tags and API layers match one-to-one, so we didn’t need to involve clients. We would not have been able to achieve this migration roadmap if we had to stop pieces to involve clients. Clients have roadmaps and timeframes of their own” for other, internal projects.

The benefits to proper integration are obvious: greater coverage, a better user experience, and efficiency gains. Most consequentially, though, is that the acquiring vendor can focus all its efforts on supporting a single platform, rather than having to spread its resources across multiple code bases.

If having more data is vitally important to a market data vendor’s revenue, having control over that data is paramount. Otherwise, you run into thorny issues around data licensing and distribution, data overlap (and data spend increases), and data cannibalization.

Additionally, when integration is done incorrectly, it creates something of a Hunger Games effect. As an executive at a broker-dealer put it to me, when you have a string of acquisitions that results in bottlenecks and delays, talent bolts out the door.

“There are overlaps clearly, but there are so many complimentary pieces, too,” the exec says of the LSEG-Tora deal. “The problem is that integrating like this—through repeated acquisition—is very hard. There were already rumblings with the Refinitiv integration, so adding another layer to that will only add complexity. As a result, there will probably be a lot of intra-company competition now because you’re fighting over resources and turf. And that’s when talent gets antsy, and they start to look at their options.”

Acquisition is a necessity—in this competitive landscape, it’s easy to get squeezed out in the middle. But it needs to be done responsibly. There’s a reason why interoperability companies like OpenFin, Cosaic and Glue42 exist—systems have been built closed off and getting them to talk to one another is painstaking. The complexity only gets worse for end-users, who are using numerous vendor platforms in addition to their internally built systems. The alphas will be those that can harvest the most insights from the data flowing through their systems. That begins with proper integration.

Think I’m missing something? Let me know: anthony.malakian@infopro-digital.com.

The image accompanying this column is “Construction of an Elevated Railway: Bridge over the Cours de Vincennes” by Paul Désiré Trouillebert, courtesy of the Cleveland Museum of Art’s open-access program.