Kestrel Vermira calibrates predictive models on historical and live order-book data, giving UK-based traders a transparent, evidence-led view of entry and exit timing — without promises of guaranteed returns.
Kestrel Vermira was built around a simple constraint: any strategy presented to a trader must first survive contact with historical data. The platform's models are trained, tested and refined against GB market history before they ever touch live capital.
The interface is deliberately unflashy. Charts, probabilities and risk bounds are shown as they are calculated, so that decisions stay grounded in what the data actually supports rather than in sentiment or speculation.
No part of the platform is described as "magic" or "guaranteed." Each stage below does one specific, checkable job.
Kestrel Vermira synthesises tick-level price, volume and volatility data for GB-listed instruments into a single probability model. The model updates continuously as new data arrives, producing a probability-weighted view of likely price movement rather than a fixed forecast.
Before any signal reaches an account, it is calibrated against stated risk parameters, including maximum drawdown and daily loss limits. Position size is adjusted to fit the risk budget, not the reverse.
Approved signals are executed through standard API connections to supported brokers. The execution layer accounts for spread, slippage tolerance and order-book depth before a trade is confirmed, keeping latency to a minimum.
Each strategy passes through the same pipeline before deployment. Nothing skips a stage, and nothing goes live untested.
Historical and live price, volume and order-flow data for GB equities, index futures and major currency pairs is collected and normalised for consistency.
Candidate strategies are run against the historical dataset across multiple market regimes, including periods of elevated volatility, before any further refinement.
Parameters are adjusted to reduce drawdown and overfitting risk, then re-tested on data the model has not previously seen.
Validated strategies are connected to live execution with pre-set risk limits, and continue to be monitored against ongoing market behaviour.
Backtesting draws on an extensive historical dataset covering FTSE-listed equities, GB index futures and major currency pairs, refreshed daily so that strategy validation reflects current market structure, not only past conditions.
The platform does not publish live account returns. The figures below illustrate the type of output a trader reviews, not a record of actual performance.
Used to weigh return against volatility, so a strategy is not favoured purely for producing larger, riskier swings.
Tracks the depth and duration of losing periods in the backtest, which informs the position-sizing limits applied later.
Identifies periods where volatility tends to persist, helping the model adjust exposure ahead of turbulent stretches rather than after them.
The underlying models stay the same; the parameters, time horizon and risk tolerance change with each profile.
Trades on short-horizon signals with tight risk bounds. Prioritises low-latency execution and rapid position turnover across liquid GB instruments.
Holds positions over days to weeks, relying on probability-weighted trend signals and drawdown controls rather than constant monitoring.
Applies conservative position sizing and stricter loss limits, using backtest transparency to satisfy internal governance and reporting requirements.
Kestrel Vermira connects to supported UK brokers through their standard trading APIs. You authorise access with your own credentials; the platform does not hold client funds and routes orders directly through your existing account.
Latency depends on your broker's API and network conditions, not only on Kestrel Vermira. The execution layer is built to minimise internal processing delay, and order confirmation details are logged for every trade.
Each strategy is tested against historical GB market data covering multiple volatility regimes, then validated on a separate, unseen data segment before deployment. Parameters are not tuned directly on the validation set.
No. Kestrel Vermira provides analysis, signals and execution infrastructure. Trading decisions and capital remain with you, and no past backtest result is presented as a guarantee of future performance.
Review the methodology, connect a broker account when you're ready, and keep full visibility of how each signal was generated.