AI Workspace

Using the ML trading algorithm and backtests

Work with the monthly ML rebalance snapshot, compare it to the S&P 500, and validate it with walk-forward backtests.

The monthly rebalance snapshot

The ML model publishes a fixed monthly portfolio snapshot. It uses a fixed reference amount, starts from the beginning of the year, and is benchmarked against the S&P 500. On the first of each month the platform publishes the new portfolio; you review the fixed allocation and compare it to the benchmark.

Running a historical backtest

To validate the approach, use the historical validation section: set a start and end date, optionally label the run, and start the backtest. The history accordion tracks each run with its status and a progress bar, and you can cancel a run in progress.

Reading walk-forward metrics

Expand a completed run to read out-of-sample, walk-forward results — the numbers that matter for judging robustness rather than a pretty in-sample curve.