I thought transform boundaries were the big earthquake risk (not that there is none at convergent boundaries).

At least that's what I remember from my 8th grade earth science class. I could be really wrong.