The methodology, stated plainly

Forecasts are produced at 1-day, 5-day, 21-day and 63-day horizons, matched to the swing structure of the platform.

The predictive layer uses deliberately diverse model families rather than an ensemble of near-identical gradient-boosting variants. Diversity in inductive bias is the point of an ensemble; correlated learners provide none.

Model outputs are consolidated via Bernstein Online Aggregation — an online-learning aggregator with formal second-order regret guarantees — which re-weights each model as its realised performance accumulates, rather than fixing weights at training time. This is the mechanism behind the homepage's claim: the system re-weights what it believes, every trading day.

Prediction intervals are produced through Conformal Quantile Regression — quantile bundles combined with split conformal prediction under Adaptive Conformal Inference — yielding distribution-free intervals with finite-sample coverage guarantees that hold without assuming the return distribution takes any particular form. Every model is continuously monitored on conformal interval coverage, calibration slope and CRPS (the continuous ranked probability score), so a forecast's reliability is itself a measured quantity.

Construction is organised as template → policy → portfolio: a single strategy template, fully parameterised by policy, can power thousands of independently configured portfolios. The platform publishes three model Books, one per risk profile; every client Book is built from its own risk profile, mandate and preferences, and every parameter lives in the policy, versioned and reproducible.

Portfolios are built on the belief distributions published by the predictive layer — not point forecasts — and resolved against portfolio state: sizing by conviction, equal weight or the Kelly criterion as each Book's policy sets it, with each position's uncertainty drawn from its own forecast interval; horizon conflict resolution; binding bounds on gross and net exposure; tiered drawdown circuit breakers; a regime overlay that can only reduce risk, never add it; and risk-profile presets, so the same research output produces different portfolios for different mandates.

Trading costs are modelled per stock: a bid-ask spread estimated from daily prices (never below one cent, capped at 60 basis points), a 0.5-basis-point commission, and an Almgren-Chriss market-impact charge for orders above 0.5% of a stock's daily volume, at standard default parameters. There are no live fills yet, so none of these is fitted to real executions. Where estimated cost exceeds the predicted statistical edge, the instruction is suppressed. The edge is evaluated net, not gross. The $10M median daily dollar-volume floor exists specifically so that impact costs remain modelable at realistic sizes.

VaR is not subadditive — a VaR budget can be exhausted by a diversifying trade. CVaR is subadditive and positively homogeneous, which yields an exact Euler decomposition of the portfolio budget into per-position contributions: “this order spent 12 basis points of risk budget” is an arithmetic fact, not an attribution heuristic. Portfolio tail risk is estimated by filtered historical simulation — chosen for identifiability at the platform's actual sample sizes, because a measured statistic with a limit can be backtested and an unidentifiable dynamic model cannot.

Covariance comes from a factor model estimated across the whole universe, not from the few names in one book, because a book of fifteen names on a year of data cannot identify a full correlation matrix. Tail risk is measured by filtered historical simulation, which keeps the real shape of past losses and rescales them to today's volatility, and fat tails are fitted only where there is enough data to fit them. The risk budget is set in CVaR rather than VaR, because CVaR adds up exactly across positions. Concentration is measured as correlation-adjusted effective N, and liquidity horizons are re-estimated under stress. Reverse stress testing is closed-form: the smallest shock that breaches a limit is computed, not guessed.

Limits live in a versioned, dated, approved limit book with four-eyes change control; regime overlays are tighten-only by construction — no override can ever widen a limit; kill-switch governance, restricted-list pre-trade compliance and an append-only audit trail are enforced in software. Risk evaluation is fully deterministic — no randomness, no wall-clock caches — so every verdict is exactly replayable.

Model skill is evaluated on a Model × Context lattice — roughly 4,000 evidence cells per day — across sectors, liquidity buckets and horizons, with empirical-Bayes shrinkage so that a thin cell is pooled toward the prior and cannot over-claim: an apparently spectacular result on eight observations shrinks to nearly nothing, and says so.

Rolling statistics flatter themselves: a rolling IC series can carry lag-1 autocorrelation near 0.97, so months of daily values contain only a handful of independent observations. The evidence layer therefore derives non-rolling, per-day metrics and reports effective sample sizes computed from them — a deliberately harder standard.

Every monitored question is a standing, pre-registered hypothesis: metric, direction and minimum effect are declared before any evidence is read, and pre-registration is enforced in software, not advisory. Verdicts are corrected for multiplicity (Benjamini-Hochberg across each day's hypothesis family), inference uses block-bootstrap methods that respect serial dependence, and robustness is tested with the Deflated Sharpe Ratio in both directions — because a low probability of skill is evidence of skill not established, not of harm demonstrated, and the two must never be conflated.

Page-Hinkley change detectors monitor every model-horizon pair for degradation, and distribution shift is vetoed with two-sample Kolmogorov-Smirnov tests — chosen over population-stability heuristics whose thresholds were calibrated on samples three orders of magnitude larger than a trading day provides.

The platform publishes three model Books, one per risk profile; every client Book is built from its own profile and inherits the same evidence chain: deterministic walk-forward validation of its policy, factor attribution against its benchmark, and exact CVaR budget attribution per order.

A months-long live window is sufficient to demonstrate that a system operates as designed under real frictions. It is not sufficient to establish a statistically significant edge, and we do not present it as if it were.

Any published performance figure ships with: the exact observation window (start and end dates); gross and net of all costs, fees, commissions and slippage; annualised volatility and maximum drawdown; Sharpe or Sortino ratio; average and maximum gross and net exposure; beta and factor attribution versus benchmark; turnover; and the benchmark stated as total or price return.