QL Simulator — rehearse the portfolio before you own it
The verdict, up front: QuantLogix now has a simulation lab. QL Simulator (Institutional tier) lets you sculpt a what-if book — or seed it from your paper wallet or linked brokerage — and answers the only question that matters before you commit capital: what does owning this actually feel like? 5,000 Monte Carlo futures bootstrapped from years of real daily history, five named crisis replays (2008, COVID, the 2022 bear, two flash selloffs), and an A/B verdict that prices every edit against your baseline in plain language: what the change buys, and what it gives up.
Why this isn't another backtest
Backtests tell you what a strategy did. The Simulator tells you the distribution of what your book could do next — and it refuses to hide its own uncertainty:
Regime-switching Monte Carlo. Every historical day of your book gets labeled with a market state (volatility cluster × trend). Each simulated day first evolves that regime through the empirical transition matrix, then samples from history that lived in the simulated state — so regime shifts happen mid-path at exactly their historical hazard. Run it conditioned on today's regime, or force the stressed twin and watch the fan widen.
The Regime Lens. The live quant stack — the regime model, the standing risk-on/risk-off tape call, the index trend report, the six-pillar mood composite (tape, volatility, breadth, credit, news narrative, positioning) — votes on which regime the simulation should condition on. Continuous market monitoring, wired straight into the math.
The model challenge. Three models argue over the same book: the historical bootstrap (real fat tails), the textbook lognormal (the normal-world shortcut), and a zero-drift challenger (momentum removed). The worst actual crisis replay stands under all three as the deterministic floor. Where they disagree tells you which number to distrust.
Validation first. Every run ships a confidence tag built from the engine's own leakage-free calibration backtest, history depth, beta-proxy share, and model disagreement. When the read is shaky, the product says so.
Cost of entry. The execution engine prices building the book — spread plus square-root market impact on a 10% participation schedule — and flags names too illiquid to enter cleanly.
For developers and agents
The same engine is on the platform surface: POST /api/v1/simulate for Institutional API keys, and the simulate_portfolio MCP tool for connected agents. Dollar-weighted book in, forward distribution + regime context + model challenge + confidence tag out. Stateless — nothing you simulate is stored.
Track record, as always
The Simulator grades itself the way everything at QuantLogix does: the calibration backtest that powers its confidence tags is a rolling-origin walk over your book's own history, and the same simulation machinery now runs as a shadow risk gate inside the autonomous paper bot — its verdicts are being counterfactually graded weekly before any enforcement is switched on. Simulation you can interrogate, not a black box.
QL Simulator is hypothetical, educational research — not investment advice, a recommendation, or a projection of actual future performance. Simulated results resample real history and do not include taxes, fees, or dividends withheld. Nothing you enter is stored.