Less noise.
More certainty.
Your Risk team doesn’t need another system that flags more accounts. It needs to know which accounts really trade as one, and why.
Behavioural Relationship Intelligence (BRI) finds them from trading behaviour alone and gives Risk a short, ranked list, with the trades behind every finding. Fewer false leads. Less time chasing alerts. More time on the accounts that matter.
23 of 24 groups were right.
In a blind test on 841 public strategies, 23 of the 24 groups BRI found belonged to the same owner.
0 false links.
Copiers planted into a real book of 250 accounts: BRI found 74 of 78, without linking a single account that did not belong.
Evidence, not a score.
Every finding comes with the matching trades and a timing control that rules out coincidence.
Accounts that trade as one
cost money in three ways.
Payouts to accounts that work together.
Several accounts copy or hedge each other to pass more challenges. Each one passes your rules. You pay out to all of them.
Exposure that looks small per account.
A group of linked accounts that win together changes your book. Seen one by one, every account stays under your limits.
Investigations that stop at one account.
Your team investigates the account that stood out, and never sees the others that were trading with it.
Your controls link accounts by identity.
BRI links them by behaviour.
Accounts can use different names, devices and payment methods and still trade as one. Each looks normal on its own. Most monitoring asks: does this account break a rule? or do these accounts share an identity? BRI asks: which accounts trade as one, whatever their identity? It runs beside your existing controls, from trading data you already have. No IP, device, KYC or payment data. Nothing to replace.
Two “independent” strategies.
One owner.
Two public trading strategies, presented as separate, with nothing in common on paper. BRI looked only at their trades. It found the same trade, in the same instrument and direction, again and again, about half a minute apart. Move one of them a few weeks in time and the matches disappear: not coincidence.
Same accounts. Same numbers.
A different decision.
On an anonymised production book, BRI formed groups of accounts from trading behaviour alone. Realised client P&L was attached only afterwards. Seen as one group, the combined client-side result is a material exposure for the broker. Seen per account, the gains are concentrated in one account, while the rest of the group shows a different financial picture.
Tested blind
against the truth.
841 public strategies. BRI looked only at their trades and was never told who owned what. Afterwards the public owner records were checked: 23 of the 24 groups it found belonged to the same owner, about 52 times more often than chance.
Verified on a real
production trading book.
Between July and August 2026, Yevhen Mykuliak, Founder of OB Group, reviewed the engine on a real production trading book, on anonymised account data, in his own environment and with his own tooling. He verified reproducibility, determinism, the falsification tests and reporting integrity, and signed his findings.
“I re-ran the method independently and reproduced the published output exactly, including the intermediate values. I found no fitting to the result.”
“What I measured was the behavioural analytics: whether the engine identifies the account relationships it claims to identify, and whether it does so consistently and reproducibly. On that question the answer is yes…”
“In this market that combination is not common, and I say so without reservation.”
Put BRI to the test
on your own trading book.
What do you have to lose? Nothing. You send anonymised files under NDA, we run BRI, you get the results. No integration, no access to your systems, no personal data. Before the run, we agree in writing on one financial metric: BRI’s selection of accounts against a simple baseline selection of the same size.
You send anonymised CSV files.
Your standard trading exports, with anonymised account IDs, under NDA. No names, no personal data, no access to your systems.
We run BRI.
Your book is analysed in isolation, and the results are locked with a fingerprint before anything is compared. No access to your investigation lists is needed.
You compare and measure.
The groups BRI found, ranked by strength, each with its evidence and combined P&L. Your Risk team puts them next to its own findings: what was known, what was missed, what is new. The agreed metric shows whether BRI’s selection beats the baseline.
BRI shows where to look.
Risk decides.
A link between accounts is not, by itself, proof of fraud, collusion, common ownership or intent. BRI does not replace Risk, compliance, KYC, AML or surveillance. No black-box scores, no automatic verdicts: just evidence your team can examine.
What are your controls missing?
Ask for a walkthrough, or a pilot on your own book. The address is in the footer below.
Less noise. More certainty. Let Risk decide.
