Walk-Forward Optimization

Walk-Forward optimization is process of comparing strategy performance between two sets of data, one that is seen by the optimizer and another that is unseen. The data that is used for the actual optimization is In-Sample (seen) and the data that remains unseen is Out-Of-Sample. During a Walk-Forward Analysis (WFA), multiple walk-forward runs are performed. For each Walk-Forward run, the performance on the Out-Of-Sample (unseen) data will be compared with the performance on the In-Sample (seen) data. The walk-forward optimizer will apply different criteria to evaluate the consistency of the walk-forward and, based on this analysis, it will provide an overall rating of the system's robustness.

 Applicable to users of the TradeStation Walk-Forward Optimizer.