Indeed, one of the main criticisms of using multiple ANOVAs to test multiple independent variables is that we increase the risk of committing type I errors. That said, I would also point out that MANOVAs require more power than ANOVAs do making it harder to find significant effects. In other words, we may increase the risk of type II errors. This can be especially problematic if the two dependent variables are un-related to one another as one effect may wash out the effect on another dependent variable. In your example MANOVA you discuss examining how different genders and prices affect buying behavior. From this description it seems like you only have one dependent variable; buying behavior. What is you second dependent variable?