Walk-forward direction testing on your currently loaded candles.
This dataset is synthetic. Results are a software exercise, not real market performance. Tests compare candle close direction; no profits, broker payouts, spreads, fees or execution delays are modeled.
Test configuration
320 loaded candles60-candle warm-upSimulated data
Put your strategy through history
Each prediction sees only candles available at that moment. The next selected horizon determines direction accuracy, not profit.
How this test works
After a 60-candle warm-up, the engine evaluates each completed candle using up to 240 preceding candles, including that completed candle. This rolling window keeps large imports practical and never uses future data for a prediction. An UP prediction is correct when the close at the chosen horizon is higher; DOWN requires a lower close. Equal closes are ties. WAIT signals are skipped and excluded from accuracy. The result is directional accuracy, not win rate at Quotex, expected return, or proof of a profitable strategy.