walk forward method

Is better test the strategy in WF with original strategy or the optimized? And in the second option, is better simple optimization or the sequencial? Thanks

Walk-forward optimization does not validate a single parameter set (original or optimized), but rather the robustness of the strategy across re-optimized parameters over time.

Simple optimization evaluates all parameters at once by searching the full parameter space simultaneously, aiming to find the global best combination in a single pass. It’s straightforward and can capture parameter interactions but it quickly becomes computationally expensive as the number of parameters grows.

Sequential optimization optimizes parameters step by step (or in small groups), fixing previously optimized values before moving on. This makes it much faster and more practical for large or noisy problems but it risks missing the true global optimum because early choices constrain later ones.