A method of prediction is presented to aid compression of sequences of complex-valued samples. The focus is on using prediction to reduce the average magnitude of residual values after prediction (not on the subsequent compression of the residual sequence). The prediction method has low computational complexity, so as to keep power consumption in implementations of the method low. The new method presented applies specifically to sequences that occupy a significant percentage of the sampling bandwidth; something that existing, simple prediction methods fail to adequately address. The new method, labeled "residual-as-prediction" here, produces residual sequences with reduced mean magnitude compared to the original sequence, even for sequences whose bandwidth is up to 85% of the sampling bandwidth.