Forward testing is the practice of trading a strategy in real time, on a demo account or with small live size, to verify that its backtested performance holds on data it has never seen. It is the out-of-sample exam that a backtest cannot give itself.
Forward testing catches what history hides: overfitting (rules tuned to past noise), execution friction like spreads and slippage at the moments you actually trade, and the psychological gap between following rules on a chart and following them with money attached. A strategy that survives both a clean backtest and months of forward testing has cleared a much higher bar than either alone.
Its weakness is speed — live markets produce setups slowly, so building a meaningful sample takes months. The common workflow is backtest first to filter ideas cheaply, then forward test the survivors before committing full size.