Model usage up to 76% cheaper on annual plansSee pricing

Quant Backtester

helio / quant-backtester
Quant Backtester

Describe the strategy logic, the instrument, and the parameters. Your AI teammate implements and backtests it, then scans for signals during market hours so you're not watching a screen all day.

What is Quant Backtester?

  • ·Turns a strategy description into tested code: Implements the described logic and runs it against a specified historical window — never a guess dressed up as a result.
  • ·Labels its own assumptions: Data range and known limitations stated plainly, so a strong backtest reads as a hypothesis, not a guarantee.
  • ·Scans continuously during market hours: Watches signal conditions as they change, not on a fixed check-in schedule.
  • ·Alerts, never executes: Every signal produces a prompt for a human decision — no trade is ever placed automatically.

What does Quant Backtester produce?

🧮Backtest implementationStrategy logic turned into runnable code against historical data.
📊Performance metricsReported with the data range and assumptions labeled.
🔔Signal alertFires the moment an entry/exit condition is met, not on a fixed schedule.
🛡️Risk-rule checkPosition size and stop-loss rules evaluated alongside every signal.

The SOP, step by step.

  1. 01

    Take the strategy description

    Logic, instrument, and parameters, in plain language.

  2. 02

    Implement the backtest

    Turn the description into runnable code against historical data.

  3. 03

    Run and measure

    Performance metrics over the specified historical window.

  4. 04

    Label the assumptions

    Data range and known limitations stated plainly — never presented as a guarantee.

  5. 05

    Load the signal conditions

    The specific rules that define an entry or exit.

  6. 06

    Scan continuously during market hours

    Not on a fixed interval — as conditions actually change.

  7. 07

    Check risk rules

    Position size and stop-loss conditions evaluated alongside the signal.

  8. 08

    Alert, don't execute

    A signal produces an alert for you to act on — it never places a trade on its own.

What it produces, in practice.

Illustrative example generated for this page — actual output depends on your real data.

📉 Backtest Report · Moving-Average Strategy

Sample · illustrative

Window: 2020-01 – 2025-12 (assumes 0.03% commission, 0.1% slippage).

MetricValue
Annualized return14.2%
Max drawdown-22.6%
Sharpe ratio1.03
Win rate47%

Parameter sensitivity: stable for the short-window parameter between 3–7 days; performance degrades sharply past 10 days — flagged as a possible overfit.

Risk rules: position sizing and stop-loss are built into the backtest logic.

Live scan: checks entry conditions every 30 minutes during market hours; a signal produces an alert, never a trade.

See it happen

Watch Quant set up a new automation.

Q
QuantAI
Ask anything…
Automations+ New
NameWhenStatus

What Quant Backtester connects to.

Market data API

Coming soon

Pulls historical and live pricing data for the instruments you're tracking.

Local code directory

Connected

Implements and version-controls the backtest alongside your existing research code.

Questions

Does this give investment advice?
No. It implements and backtests strategies you describe and reports what happened — it never recommends a trade or claims a result will repeat. Every real decision is yours.
Can it execute trades automatically?
No. Signal detection produces an alert only. Trade execution is a deliberate, separate step you control.
How do I know the backtest isn't overfit?
Every result is labeled with its data range and assumptions. Treat a strong backtest as a hypothesis to validate further, not a guarantee — the report says so explicitly.
What data sources does it support?
It connects to whatever market data API or historical dataset you provide access to — it doesn't come bundled with a specific paid data source.
Try for free

Stop watching the market by hand.

Describe your strategy — Helio backtests it and scans for signals during market hours. You still make the call.

Ask about this automation