Thevertex ingests real-time market data and applies predictive modelling to surface risk-adjusted opportunities, then executes according to a defined mandate while you remain offline or in transit.
Thevertex continuously ingests pricing, volume and macroeconomic data streams, then applies statistical and machine-learning models to identify pricing asymmetries that have historically preceded favourable risk-adjusted outcomes.
The objective is not to forecast every market move. It is to quantify probability and size exposure accordingly, so that capital allocation reflects measured risk rather than conviction alone.
Every signal the model generates is logged alongside its confidence score and the data inputs that produced it, giving a transparent audit trail rather than a black-box recommendation.
The platform follows a fixed sequence from raw data to executed decision. You define the mandate once; the system handles the rest and reports back for your review.
Market feeds, order book depth and macro indicators are pulled continuously and normalised into a single dataset the model can evaluate.
The engine re-weights exposure against your stated risk mandate, factoring correlation and drawdown history before any change is proposed.
Approved logic triggers execution within pre-set limits, and a plain-language summary of the change is sent for your review, whenever you next connect.
Strategies are validated against historical data before they are made available within the platform. The figures below describe the structure of that testing, not a forecast of future results.
Methodology note: strategies are tested against historical price data using walk-forward validation, with out-of-sample periods excluded from model training.
Past performance, whether backtested or live, does not guarantee future results. Historical testing is subject to data availability bias and does not account for every market condition that may occur going forward.
Location-independent investors rarely have a fixed desk or a stable connection. The platform is structured so that oversight, not continuous management, is what's required.
A consolidated activity log shows every executed decision, the model confidence behind it and the resulting exposure change, so a short daily check from a café connection is enough to stay informed.
Global instruments across major exchanges and sessions, with data normalised to a single time-zone reference for consistent reporting regardless of where you are.
Rather than requiring you to track separate market hours, the platform aggregates coverage into a single interface, so a decision made overnight in one region is reflected in the same report you review the next morning.
Notifications are only sent when an action exceeds a threshold you define, limiting interruptions to decisions that genuinely require your attention.
Routine rebalancing happens without a prompt. You are alerted when exposure moves outside your defined mandate, or when the model's confidence in a position drops below an agreed level.