By Nolan Schiff, Chief Solutions Officer, Eventus.
It’s a question nearly as old as enterprise software itself: should you pay a vendor to deliver your technology, or dedicate internal resources to building and maintaining it yourself? For trade surveillance, it’s a decision that requires weighing cost against control, performance against customization and the speed of a proven platform against the fit of something purpose-built.
The quandary isn’t new. But the pace and scale of everything around it is changing rapidly – and the stakes of getting the decision right are rising in kind.
24-hour trading is moving into the mainstream, bringing thinner markets, wider spreads and a risk profile that daytime procedures weren’t built for – spoofing, pump-and-dump schemes and outright price-setting all get easier and cheaper to attempt in off-hours sessions.
Prediction markets raise a different set of questions, many of them around information asymmetry, influence and resolution integrity. Did a user know something material before the market moved? Can the venue reconstruct what happened around a given event? Surveilling these markets means watching for prohibited persons and insiders, oracle and settlement risk in how outcomes get resolved and coordinated or social-media-driven trading, categories that didn’t exist in most surveillance programs even two years ago.
We’re seeing these shifts arrive at a moment when the allure of utilizing AI to stand up surveillance systems is stronger than ever. Building detection logic for spoofing in an illiquid overnight session, or spotting a coordinated insider play in a prediction market, appears much more solvable with a model doing the pattern matching. While these systems can certainly make the first pass, the ultimate judgment calls still require experts to verify and validate their decisions, particularly in new, untested categories of misconduct.Amid all of this, firms are pushing into new markets and geographies more aggressively than ever. The line between asset classes is blurring, with several leading platforms now offering crypto trading, event contracts and overnight equities under one roof – an example of the “super app” model that is quickly becoming commonplace. In a 2025 Datos Insights survey, respondents ranked multi-asset and global coverage as their top-priority consideration in developing a surveillance strategy, with cross-asset correlated alerts cited by 45% as the functional area most in need of improvement.
These developments are exciting for firms – and they create a lot more surface area for a surveillance program to cover. Against that backdrop, the classic buy-or-build considerations matter more than ever.
The Case for Buying
Scale, continuity, independence and coverage are the core principles underpinning the case for buying. A vendor’s surveillance logic is shaped by every organization on its platform – an enhancement built for one firm rolls out to the entire client base, and the underlying models improve as they ingest a wider, more diverse pool of alerts than any single institution could generate alone. Vendors also have teams whose entire role is to keep improving the platform, whereas internal teams will always be pulled in different directions – often functions that provide more direct value for clients. In practice, in-house platforms tend to live or die with the engineers who built them.
There’s also the matter of coverage. A single, normalized view across every asset class, venue and data format is one of the hardest things to build well internally, and vendors that specialize in this problem are built to solve it holistically – delivering a level of pattern recognition, from wash trade detection to cross-product surveillance, that an internal team would otherwise have to develop from scratch.
Lastly, partnering with a vendor makes it easier to stay nimble as your business evolves. Whether it’s a new asset class, new venue, a merger or expansion into a new region, a vendor will be faster to market due to its pre-existing infrastructure, ability to map to new data formats and regulatory expertise. By contrast, a firm building in-house has to develop that capability from a standing start every time your business changes shape.
The Case for Building
This isn’t to say building in-house is always wrong. There’s a legitimate case for keeping full responsibility for your surveillance roadmap via a bespoke solution shaped around your own workflows. But two assumptions are worth pressure-testing.
The first is control: the belief that a vendor can’t fully understand your data or environment. Ingesting and normalizing raw data without forcing a rigid schema is a problem that mature vendors have solved repeatedly.
The second is cost. Building can look cheaper on paper, but it rarely accounts for the ongoing burden of tracking new obligations, retraining staff and maintaining the platform as markets evolve. Do that work properly, in perpetuity, and the savings tend to shrink. Cut corners, and the regulatory and reputational cost of missing something can be severe.
Finding the Right Fit
“Buy versus build” is never going to be settled once and for all, especially for a function as complex and mission-critical as trade surveillance. The right answer still depends on a firm’s internal expertise, resources and risk appetite.
This pace isn’t letting up anytime soon. Firms looking to start trading overnight, adopt perpetual futures or move into prediction markets need a surveillance platform that will help them do so with speed and confidence. Rather than asking your technology team to build a program from scratch, working with a trusted partner helps ensure you’ll have access to a dynamic solution that scales with your business and a dedicated support team that gets your analysts up to speed quickly.
Now in its 3rd year, TradingTech Buy AND Build: The Future of Capital Markets Technology takes place on September 29th at America Square Conference Centre, London.
Subscribe to our newsletter
