QUANTHEON Lab — systematic investment research

Engineer ideas, backtest strategies, build portfolios

QUANTHEON Lab is an AI research desk: describe an investment idea in plain English and it turns it into editable visual logic, tests it and proposes a fix where it breaks. You score and investigate it across assets, then see how it belongs in a shared-capital portfolio. No code, free to start. Open the app.

One connected research record

• The Mandate — your inviolable limits in plain English: hard walls nothing can be promoted past.
• The Research Memo — every result as a one-page verdict you approve or reject.
• Track — run any saved study forward from the end of its backtest to today, on demand, in the Vault.

What you can do

• AI research desk — describe an idea in plain English and it builds the block graph, then critiques the logic and restructures the strategy itself.
• No-code visual builder — 54 indicators plus price and formula blocks, every rule reading back as a plain-English sentence.
• AI import — paste Pine Script, thinkScript, MQL or EasyLanguage, or just describe it, and it is rebuilt as an editable graph, authoring the custom indicators it needs.
• Adaptive exposure — tie position size to volatility, trend or breadth and the engine scales it bar by bar.
• Multi-asset backtest — one strategy across the entire S&P 500, plus ETFs and indices, in a single pass; results come back aggregated and per-asset, so you see what generalizes instead of the one ticker where it worked.
• Overfitting Polygraph — Walk-Forward, Monte-Carlo (1,000 paths) and parameter stability, fused with the Deflated-Sharpe haircut into one verdict: trustworthy or curve-fit.
• Validated optimization — grid or Bayesian search that only sees in-sample data, with the held-out tail kept as a gate and a haircut for the combinations you tried.
• Genetic experiments — mutate and breed strategy variants, each judged once on a sealed out-of-sample vault.
• Shared-capital portfolio — one cash pool where positions compete for capital, and a check of which allocation actually survives out-of-sample.

Free guides

• How to backtest a trading strategy (no code)
• What is overfitting in trading strategies?
• Walk-Forward analysis explained
• Monte-Carlo simulation for trading
• The Deflated Sharpe ratio (and PSR)
• The Strategy Score, explained
• Adaptive position sizing: scale exposure with a signal
• Custom indicators: build once, reuse everywhere
• Independent blend vs a shared-capital book
• Build and validate a strategy: full walkthrough
• Build and backtest a portfolio: full walkthrough

Free calculators

• Sharpe ratio · Sortino ratio · Max drawdown
• Position size · CAGR · Kelly criterion
• Deflated Sharpe ratio — price the search behind a backtest: how many configurations you tried, and whether the winner survives it.

Under the hood

• The technical whitepaper — typed-AST strategies, a deterministic no-look-ahead kernel, the anti-overfitting stack and the allocation intelligence.
• Or browse the full library: Learn — guides & calculators

Loading the Lab… if nothing appears, enable JavaScript.