Principal Agent Models Part 1: Moral Hazard With Observability