Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.joinmobius.com/llms.txt

Use this file to discover all available pages before exploring further.

A backtest is a simulation of your trading strategy run against historical market data. Instead of risking real capital to test an idea, you replay the past and see how your strategy would have performed. Mobius backtests are powered by the same AI decision engine used in live trading, so what you test is what you deploy.

How the backtest engine works

1

Data fetch

Mobius fetches OHLCV (Open, High, Low, Close, Volume) bars for your selected asset and date range from Alpaca Markets. For crypto, daily history is sourced from Yahoo Finance; intraday crypto bars use the Alpaca crypto feed.
2

Indicator calculation

All technical indicators referenced in your strategy (RSI, EMA, MACD, Bollinger Bands, etc.) are computed from the raw price data using TA-Lib. Alternative data (Reddit sentiment, congressional trades, fundamentals) is fetched from the appropriate provider and merged into the same timeline.
3

AI decision loop

For each bar in the date range, the AI receives: the current date, price, indicator values, and current position. It returns one of three actions — buy, sell, or hold — along with a reason and quantity. This decision loop runs bar by bar across the full backtest window.
4

Trade simulation

Buy and sell decisions are executed against the simulated portfolio. Commissions and slippage are applied as configured. The portfolio value is tracked at each bar.
5

Results

At the end of the simulation, Mobius calculates performance metrics and renders the equity curve.

Performance metrics

MetricWhat it measures
Total returnPercentage gain or loss over the entire backtest period
Annualized returnReturn normalized to a one-year basis
Sharpe ratioRisk-adjusted return — higher is better (above 1.0 is generally good)
Max drawdownLargest peak-to-trough decline in portfolio value
Win ratePercentage of trades that closed with a profit
Total tradesNumber of completed buy/sell pairs
Average tradeAverage profit or loss per trade

Supported timeframes

Bar frequencyBest forMax backtest period
1 minScalping, intraday30 days
5 minShort-term intraday180 days
15 minSwing intraday2 years
30 minSwing3 years
1 hourSwing / trend5 years
DailyLong-term trend5 years

Important limitations

Past performance does not predict future results. A strategy that backtests well may perform differently in live markets due to regime changes, liquidity differences, and market impact.
  • Backtest data reflects historical prices — it does not account for corporate actions (splits, mergers) in all cases.
  • Alternative data (Reddit sentiment, congressional trades) may have limited history before 2020.
  • Fundamental data updates quarterly, so intraday fundamental strategies have lower granularity.

Next steps

Configure a backtest

Set capital, date range, slippage, and commission before running.

Interpret results

Understand every metric in the backtest results panel.