Quickstart

Go from a blank account to a reviewed paper strategy without connecting a brokerage. This path introduces the core workflow while keeping real capital out of scope.

Build and test your first strategy

1. Create an account and open Terminal

   Sign in, accept any required terms, and open Terminal. Terminal is the workspace for chat, charts, generated strategy code, and backtest results.

2. Describe one testable idea

   Ask for a symbol, timeframe, entry, exit, and position-sizing rule. Use the starter prompt above as a base, then change one assumption at a time.

3. Inspect the generated rules

   Read the strategy code and confirm the assets, bar size, conditions, sizing, and risk controls match your request. Ask Astral to explain or revise anything unclear.

4. Run and review the backtest

   Inspect total return, drawdown, risk-adjusted metrics, trade count, the equity curve, and individual orders. Use Analysis and Replay to understand when the result was earned and where it failed.

5. Save the strategy

   Save a version you want to revisit. Saving creates a private reusable asset; it does not publish or deploy it.

6. Deploy to Astral Paper

   Choose Paper in the deployment flow, select a notification profile, and enter starting cash. Paper deployment simulates ongoing activity and does not place orders at a brokerage.

What each action means

Continue learning

How to review the generated strategy

Ask Astral to restate the strategy in plain language, then match each sentence to the inspectable rules. Confirm:

• SPY is the only intended asset;
• the bar size is daily;
• the fast and slow averages use the intended price series;
• entry requires the upward cross rather than merely being above;
• exit uses the reverse cross;
• position sizing means 10% of equity at the relevant time;
• no additional stop, target, schedule, or condition was added implicitly.

If any item is unclear, revise before interpreting the backtest.

What to look for in the first backtest

Do not begin by asking whether the return is “good.” First check that trades occur where the rule says they should. Use the order list and Replay to sample several entries and exits. Then inspect trade count, drawdown, equity path, and whether one period produced most of the result.

Try one controlled revision, such as a different slow-average length. Keep the dates and capital constant. This demonstrates how to compare strategy versions without turning the exercise into a search for the best-looking parameter.

Observe the paper deployment

After activation, find the strategy in its monitoring surface and confirm the mode says Astral Paper. Record the starting cash, notification profile, activation time, and strategy version. A daily strategy may remain quiet until a completed daily bar produces a new condition.

Paper deployment creates an ongoing simulated record. It does not place a broker order and does not inherit the historical backtest balance.

When you are ready to explore further

Add complexity only when you can still explain the behavior. Useful next experiments include comparing a benchmark, adding one risk rule, testing another market regime, or researching whether the moving-average behavior differs across related assets.

Avoid connecting a broker merely to continue the tutorial. Broker connectivity adds account authorization, live order, and position-management responsibilities covered in the broker and manual-trading guides.