The forward fan 5,000 bootstrap paths
Run a simulation to roll the futures. The fan shows where the book's cumulative return lands — the line is the median path, the shaded band holds the middle 80% of outcomes.
Simulation Lab
QL Simulator
Rehearse the portfolio before you own it — fail safely in simulation, not in your brokerage account. Type a what-if book, or seed it from your real one, and QL rolls 5,000 Monte Carlo futures from years of actual daily returns, replays the book through 2008, COVID and the 2022 bear, and prices your edit against the baseline. Same market history, two books — only the decision changes, and every result ships with a validation-first confidence read.
🔒 Stateless — what-if books are simulated in memory and never stored🏛️ Institutional tier
The book under test dollars per position · up to 30 names
| Ticker | Value (USD) | Weight |
|---|
Book: $0
Private sleeve roster companies · dollars invested · valuation, growth, sentiment, market & liquidity factored · up to 20 names
Type a company name from the tracked roster (OpenAI, Anthropic, Stripe…). No mark? Cost basis is used and labeled.
Waiting for a book
Each row is a dollar allocation — the simulator prices the book, pulls each name's real daily history, bootstraps 5,000 correlated forward paths, and replays five named crises with per-name betas for anything too young to have traded through them. The first run becomes your baseline; edit the book and run again to see the A/B verdict. Minimum one priced position.
Regime lens continuous market monitoring — the quant stack's live reads pick the conditioning
Reading the live regime model, the standing risk-on/off tape call, the index trend report, and the six-pillar mood composite (tape, volatility, breadth, credit, news narrative, positioning)…
Crisis replay real windows, the book's own history
Run a simulation to replay this book through 2008, COVID, the 2022 bear, and two flash selloffs — plus hypothetical −10% / −20% / rate-shock scenarios mapped through each name's beta.
Model challenge competing models argue over the same book — disagreement is the signal
Run a simulation and three models fight it out at your horizon: 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.
Private sleeve stage-calibrated outcome model · blended with the public book at the horizon
History replay the book as it would have traded — monthly rebalance
Run a simulation and the book is also replayed through its own history: the equity curve you would have lived, with CAGR, volatility, Sharpe and the deepest drawdown — beside SPY over the same window.
Walk-forward validation forecast vs what actually happened — leakage-free folds
The forward fan is only worth trusting if bands like it kept their promises in the past. Each fold forecasts the horizon from data available at the time, then grades reality against the band: dot inside the band = promise kept. Run a simulation to see the folds.
QL Intelligence desk read the numbers, argued by the desk — then hand the run to your team
Run a simulation first
Honesty panel validation comes first — how much to trust this simulation
The Monte Carlo engine back-tests its own forecast cone on your book's history (leakage-free rolling origin) and reports whether the cone has been honest — well-calibrated, over-confident, or under-confident. Every run is then tagged with a confidence read built from those checks, so you know where to trust the numbers. Run a simulation to see it.
QL Simulator is hypothetical, educational research — not investment advice, a recommendation, or a projection of actual future performance. Simulated results are built by resampling each book's real daily return history (historical bootstrap) and by replaying named historical windows; they do not include taxes, fees, slippage, or dividends withheld, and books that never existed were never actually tradable. Crisis replays use each holding's own traded history where it exists and a market-beta proxy where it does not (proxied names are flagged). Nothing you enter is stored.