AI backtesting
Backtest a US stock or ETF strategy without code
Describe the rules in plain language. Stratos runs them on historical bars, shows the path, and lets you paper test. Live brokerage execution is coming. It is not available today.
Who this is for
A trader who already has a US stock or ETF setup, and wants to see what those rules did on history before risking a brokerage account. You can say the rules to Orbit or enter them in Builder. You do not write Python for either path.
The workflow
- Name the symbol, bar size, date range, and the entry, exit, and risk rules. Vague language such as “buy strength” is not a test until the signal and the exit are defined. The product page shows one concrete SPY example. That example is a rule, not a result.
- Run the historical simulation. ATLAS uses open, high, low, close, and volume. A standard entry is judged after a bar closes, and the simulated fill is the next bar’s open. Stops and targets can exit inside a bar, including the entry bar. The backtesting guide walks the same steps in order.
- Set slippage and commission if you want costs in the path. Both start at zero. The default comparison is buy and hold. You can choose one benchmark ticker.
- Read the equity curve, drawdown, and trade count. Save the strategy and paper test it with virtual capital when you want a forward check on live prices. Paper results are still simulated.
What the AI does
Orbit turns what you say into Builder fields, and it explains a finished backtest from that run’s numbers. You can change every field before you run it. The simulation is ATLAS on historical bars. It does not predict the next move, and it does not place a brokerage order.
What this is not
Natural-language, no-code backtesting is not unique to Stratos. TrendSpider’s Strategy Tester already offers it, including a path they describe as a live broker bot. The Stratos and TrendSpider page cites their current product pages and says where each one fits.
Stratos does not support options, futures, or forex. Major crypto pairs such as BTC/USD and ETH/USD are available, and the pages in this set are about US stocks and ETFs. A finished backtest can be opened from a read-only share link. Those share pages stay out of the search index. The public explanation of a test is a research article, published only when a real run answers a specific question.
Common questions
Can I backtest a US stock or ETF strategy without writing code?
Yes. Describe the entry, exit, and risk rules in plain language and Orbit can fill the Builder, or set the fields yourself. ATLAS then simulates those rules on historical bars. Builder and SmartLab modes do not require Python.
Does Stratos place live brokerage orders?
No. Live brokerage execution is coming. It is not available today. Paper trading uses virtual capital. An alert emails you once and then turns off. It does not submit an order.
Is natural-language backtesting unique to Stratos?
No. TrendSpider’s Strategy Tester already lets traders describe a strategy in natural language or build it with point-and-click menus and backtest it without code. Source: https://trendspider.com/product/strategy-development-and-backtesting-tools/. Stratos is the research path for writing US stock and ETF rules, inspecting a historical test, and paper testing.
What can a Stratos backtest prove?
It shows what the stated rules did on the selected history, with the cost and comparison settings you chose. It is not a record of brokerage trades and not a forecast. Past performance does not guarantee future results.