Look ahead · decide & act

foresight

The decision engine. Signals, stress tests, VaR and hedging rules that turn what the desk sees into what it does next, with every recommendation traceable to the flow behind it.

What it does
  • Signals and account recommendations
  • Stress tests and value at risk
  • Automated hedging and A/B-book routing
foresight · Recommendations

Recommendations, not just dashboards

Detectors run on a schedule and classify accounts (toxic, fast-hedge, A-book, B-book, scalper, martingale, news trader, copy trader and more) with a confidence score and the evidence behind every match. Accept or reject, and the outcome feeds straight back into hindsight.

Reads the whole decay curveThe toxic detector measures the sub-30-second markout curve against the spread the client actually paid.
Hedge ConfigNOP, symbol VaR, portfolio VaR, time-based, speed and blackout rules, assigned to books.
The Client Flags Account Activity Monitor: a table of active flags awaiting action, each with the account, a critical, warning or info severity, the rule that fired (news trading, machine-gun dealing, shared connection, profitable large trades), the current state, when it became active and when it was last checked.
Inside foresight

Know what the next move costs, before it happens

Stress the live book, size risk against a budget and test a signal on your own flow, all without placing a trade.

Stress Test

Replay a hypothetical market move through the live book, account by account. See which orders and stops it triggers, which clients are closed out, what happens to your LP positions, and what is owed. Nothing is saved.

  • ScenariosOne symbol, every symbol, or a lead symbol with correlated moves implied by the risk model.
  • Two pathsAn instantaneous gap and a steady walk, side by side, to bracket the outcome.
  • Shock tablePreview and override every symbol's move before running.
  • OutcomeFirm, hedge and client P&L, orders triggered, close-outs and LP stop-outs.
  • Event logEvery order, stop and close-out, in sequence.
Analytics › Risk & Context

Signals

Per-symbol buy, sell or neutral signals built from client deal imbalance, RSI, MACD, moving-average stacks and news gates, then combined into composites. Backtest any signal against your own flow to see whether the book dealt with it, and what skewing on it would have made.

  • Signal gridSymbol by signal, coloured by direction, refreshing every second.
  • CompositesAgree, majority, weighted and gate, neutral when an input goes stale.
  • AnalysisDeals classed with, against or off the signal, and the skew earned or given back.
  • Signal skewRules that skew price while a signal is on, shadow mode first, with a killswitch.
Live › Risk Management · Signal Analysis in Analytics

Value at Risk

VaR of the live book for any slice of it (the whole firm, a book, client groups or individual accounts) over any horizon from an hour to a year. Offsetting clients net out, and every contribution sums to the portfolio total.

  • MethodParametric variance-covariance from EOD price history, scaled by trading days.
  • Confidence90% to 99.5%, with expected shortfall alongside.
  • DiversificationWhat netting and correlation remove versus standalone VaR.
  • ContributionsBy symbol and by account, via Euler decomposition.
  • Size to a budgetThe positions whose VaR exactly matches a budget you set.
Analytics › Risk & Context
One platform, three lenses

Part of a single loop

foresight closes the loop. Every recommendation it makes feeds straight back into hindsight as something to measure.

Get started

See foresight against your own book.

Book a 30-minute walkthrough and we'll map it to your own venues, books and clients.

Request a demo →