Exegy and MarketsIO Target the Market Data Migration Deadlock

For years, financial institutions have faced an uncomfortable trade-off around enterprise market data infrastructure. Legacy platforms may be expensive, complex and increasingly ill-suited to modern trading environments, but replacing technology embedded across hundreds or thousands of applications can introduce even greater cost and operational risk.

The partnership, announced earlier this month, gives clients access to normalised data from more than 300 global trading venues through Exegy, alongside MarketsIO’s enterprise distribution and control layer. Crucially, existing application interfaces can initially be retained, allowing the underlying infrastructure to change without forcing simultaneous changes across the application estate.

For David Taylor, CEO of Exegy, the difficulty of replacing legacy infrastructure has historically been largely economic and operational.

“Until recently the cost did not justify the benefit,” he tells TradingTech Insight. “In this case, the ‘cost’ is the effort, technical and operational risk of migrating hundreds or thousands of applications to a new enterprise market data platform.”

Previous alternatives have also faced the problem of having to integrate with incumbent data models, interfaces and entitlement systems, he says, leaving firms carrying much of their existing technical debt into supposedly modernised environments.

“Decades of incremental evolution encumbered these legacy solutions with loads of technical debt that often nullified the technical benefits that the new alternatives offered,” says Taylor.

Changing the migration equation

Taylor believes that financial institutions are now becoming more willing to challenge established market data arrangements, while alternative technologies have developed sufficiently to provide a more practical migration path.

“Two things have changed: clients are committed to change and there now exists alternative technology and commercial models that address the former impediments to change,” he says. “Fed up with paying significantly more for no additional benefit, many of the clients have made the strategic – and emotional – decision to change. The cost of doing nothing is now appreciably higher than the cost of change.”

The Exegy-MarketsIO model is designed to avoid a large-scale cutover. The initial step is to introduce MarketsIO as an abstraction layer between applications and their underlying data sources, separating the technology used to distribute and manage data from the source of the content itself.

Once that layer is in place, firms can introduce alternative sources incrementally. Exegy’s contribution can include Nexus or Ticker Plant for low-latency direct feeds in colocation environments, either deployed on-premises or hosted by Exegy. Firms can also migrate individual markets from incumbent consolidated feeds onto Exegy’s Axiom service.

Taylor says the objective is to break the historical coupling between market data content and the infrastructure through which firms distribute and consume it.

“The MarketsIO platform is essentially an off-the-shelf vendor abstraction layer that allows firms to select and manage best-of-breed data sources,” he says. “This breaks the historical coupling of content with the technology used to distribute, consume, and manage it.”

That separation could also allow institutions already consuming direct feeds for latency-sensitive trading to make greater use of those feeds elsewhere rather than purchasing the same underlying content through multiple channels.

Separating content from infrastructure

For MarketsIO CEO Terry Roche, separating the layers of the market data stack has implications extending beyond the migration itself.

Many institutions continue to obtain data, distribution technology, entitlements and application connectivity through tightly integrated vendor environments. Changing the data source can therefore involve changing infrastructure, applications and commercial arrangements at the same time.

“The practical difference is independence,” says Roche. “Separating those layers lets firms choose or combine the best data sources for each requirement without redesigning the enterprise architecture or unpicking a bundled contract.”

MarketsIO’s EventStream distribution technology provides instrument-level resilience, allowing problems with individual securities to be handled without requiring an entire feed or service to fail over. Its Autonomy entitlements platform, meanwhile, is designed to provide more granular information about who or what is consuming data, including the application, process or service involved and the authority under which it is being used.

Roche argues that decoupling these functions also introduces greater competitive pressure into a technology layer that has often remained closely tied to market data provision.

“Some widely deployed enterprise market-data infrastructure has architectural roots going back decades and has evolved largely through successive upgrades,” he says. “When technology is insulated from direct competition because it is bundled with other services, the pressure to invest aggressively in development, modernisation and support can diminish. Separating the layers means the technology has to keep earning its place.”

The potential economic effect, he argues, is competition at individual layers of the stack, giving institutions greater flexibility over suppliers as well as additional negotiating leverage.

Compatibility without preserving the past

One potential tension in an incremental approach is that compatibility with legacy applications could itself perpetuate older technology. MarketsIO’s answer is to treat backward compatibility as a transitional capability rather than an end state.

“We see backward compatibility as a migration mechanism, not an architectural destination,” says Roche. “A large financial institution may have dozens or hundreds of applications connected to existing market-data infrastructure. Requiring all of them to change before the underlying platform can be modernised creates significant cost and risk.”

Instead, institutions can replace the underlying infrastructure while applications continue to operate against familiar interfaces, then migrate individual applications when appropriate. Roche says moving an application onto MarketsIO’s Fusion API Suite typically takes one or two days.

MarketsIO also provides clients with the Fusion API Suite source code, which Roche says is intended to prevent the migration simply replacing one dependency with another.

“The result is a staged approach to modernisation,” he says. “Upgrade the underlying platform first, preserve continuity for existing applications, then move applications to newer interfaces and consumption models where it makes business sense.”

A broader role for Exegy

The partnership also reflects Exegy’s expansion beyond the ultra-low-latency market data technology for which it has traditionally been known.

Taylor describes Exegy’s ambition as becoming a platform for “mission-critical trading”, spanning enterprise access to global market data through to nanosecond-level optimisation for latency-sensitive strategies. For large institutions operating multiple trading businesses, that means addressing different performance requirements through a smaller number of technology relationships.

While deterministic performance has traditionally been associated most closely with high-frequency trading, Taylor argues that resilience and predictable performance under stressed market conditions have become increasingly relevant to enterprise applications as data volumes and trading demands increase.

The Exegy-MarketsIO partnership therefore addresses two related aspects of market data modernisation: changing the technology itself and changing the way institutions can get there. Rather than asking firms to justify another wholesale infrastructure replacement, the proposition is that data sourcing, distribution and applications can be progressively disentangled, allowing each layer to evolve on its own timetable.