Periodic auctions came to Europe as an answer to the speed race. They are now routine. Several operators run them, and randomised auction lengths and speed-neutral matching have become standard rather than special. A venue arriving today has to say what it does that the others do not.
OneChronos, which began production trading on its UK and EU multilateral trading facilities on 29 July, puts its answer in the matching engine.
“We specialise in optimisation techniques, leveraging mathematical modelling based upon combinatorial auction theory,” says Scott Bradley, CEO of OneChronos Markets UK Limited. “Instead of price-time priority, and ‘what’s the most number of shares I can cross, and then I’ll tell you the price we’ve uncrossed at’, we actually turn it round. It’s not a price-time priority. It’s how do we find the greatest notional price improvement, and then share it equitably amongst all the participants. We’re democratising access to liquidity by removing speed from the equation. Periodic auctions do that naturally, but we are the only periodic auction mechanism where we control the timer. We’re not building a book. There’s no queue priority.”Rethinking the Auction Timer and Cadence
Bradley counts six incumbent operators in Europe, and most trigger auctions on demand, stock by stock, when an order arrives that could match. OneChronos sets the clock itself. A Poisson distribution decides when the collection phase ends. The book then locks, a data buffer runs, and the optimisation follows. Auctions run about 10 times a second.
“It’s a very binary outcome: you’ve either made it into the auction during the collection phase before we randomly lock the book, or you’ve missed it,” Bradley says. “When auction n is locked, auction n+1 then starts collecting. So you get that feeling of a continuum of periodic auctions, but you have time as a discrete function.”
“When we run our auction timers, we run them for the whole universe of securities we cover at the same time, on the same cadence,” says Bradley. “We’re not running a set of discrete auctions on an individual security basis. In Europe we cover over 4,000 securities on a pan-European basis. When we put out a message saying the auction is running, we mean for all securities.”
His example is pairs trading.
“It creates the perfect atomic pair opportunity. I will trade A and I’ll trade B if the ‘and/or’ condition between the two securities is satisfied, and it’s happening in the same auction… you can almost think of OneChronos as providing the platform to have that optimised layer of trading constraints taking place at the venue level, rather than at the algorithm level.”
Adam Sherlock, CEO of OneChronos Markets NL B.V., starts from what a firm can say when it sends an order today.
“When we look at and study European and global markets, trading firms really have only a handful of constraints to play with,” Sherlock says. “You have your size, you have your minimum quantities, you have certain other bespoke constraints, but there’s a handful. A couple of handfuls at most.”
The 4,000 figure covers both entities. The UK venue trades UK and Swiss stocks. The EU venue trades EEA stocks. Between them they reach 13 primary markets. Each runs its own cadence, so a constraint tying a London line to a Paris line falls across two auctions on two separate MTFs.Bradley is clear about why there are two. Brexit, MiFID II and the share trading obligation oblige the firm to run two legal entities, each answering to its own regulator, to serve clients across Europe. It operates them as one European business.
Integration and Market Traction
The firm confirms that standard FIX tags carry order constraints, alongside some custom tags it has made available. “We operate on a FIX 4.2 protocol, so very standard FIX messaging, and it’s nothing more than applying certain FIX tags with specific instructions,” says Bradley.
Members of OneChronos are generally banks, brokers and market makers, with the buy side trading through them. Bradley argues that low-latency firms have been moving into periodic auctions as they take on mid-duration strategies, and that a book where speed buys nothing lets market makers trade with each other at lower risk of being picked off. Sherlock’s adds that the book is a level playing field, and each firm can judge the fit for itself.
Preparing for Agentic Trading Systems
The company describes it as building “the bazaar for agentic systems”. Sherlock explains it as a matter of complexity. As AI helps firms think across asset classes, they will want to say more in an order, and the constraint language is where that lands. Bradley takes it further, to the same model running power and compute markets outside capital markets.
Asked what guardrails a venue needs if agentic systems are sending the orders, Sherlock says: “We only take direct MiFID investment firms or credit institutions as direct users, and we can only accept those as the members, effectively, of our venues. What they are utilising to power that trading is their conversation with their regulators. What is driving those underlying decisions comes one level removed from where we operate.”
Regulatory Landscape and Outlook
ESMA published a call for evidence on 30 April covering dark trading, periodic auctions and systematic internalisers. Responses closed at the end of June, with a feedback statement due in the second half of this year.
Ultimately, OneChronos is betting that the future of institutional execution relies on venue-level optimisation rather than raw speed. By shifting mathematical constraints into the matching engine and preparing for AI-driven order flow, the platform offers a distinct alternative to conventional auction structures. Whether market participants fully embrace this model remains to be seen, but OneChronos has successfully established the infrastructure to support that shift if the market is ready.
Subscribe to our newsletter
