
MALI Platform
AI Stock Research
Backtesting Lab
The Backtesting Lab runs a strategy over history on up to ten tickers and shows where its returns and its risk came from. Walk-forward windows and five dated crisis periods test it outside the easy years.
Plan: Pro and EdgeIn the app: Trading Hub → Backtesting Lab
What you can do
- Choose Momentum, Mean Reversion, Breakout, Trend Following, Volatility Compression, Regime Adaptive or a Multi-Strategy Blend.
- Size positions as a fixed share of capital, by the Kelly criterion, or to a 15% annual volatility target.
- Turn on walk-forward optimisation: a year of training, then a quarter of testing, rolled forward.
- Rerun it on the 2008 crisis, the dot-com bust, the February 2018 volatility shock, the 2020 COVID crash or the 2022 rate shock.
- Read total return, CAGR, Sharpe, Sortino, Calmar, max drawdown, hit rate, profit factor and exposure, with equity curves and attribution by ticker.
What it doesn't do
- Its summary figures average the tickers equally, and its max drawdown is the worst single ticker's; it is not a combined portfolio.
- It tests MALI's strategies; there is no free-form rule language.
- It places no orders.
Data: Daily prices, adjusted for splits and dividends
Questions
- What does a backtest assume about costs?
- 0.1% commission and 0.05% slippage per trade by default, starting from 10,000 per ticker.
- How much history does a test need?
- At least 30 daily bars. The panel starts at January 2020 unless you change it.
Related features
Scenario LabYour portfolio, at today's weights, through seven historical market episodes.Market ScannerRule-based signals on live prices across stocks, ETFs, crypto and FX, on every plan.Options Chains, IV Surface and Strategy GeneratorChains with Greeks, an IV surface for any ticker, payoffs and an AI strategy generator.