Average Causal Effect Estimation Via Instrumental Variables: the No Simultaneous Heterogeneity Assumption