You are nearly right Jim, Tensile Strength does play a part in this but only in deciding which grade of bolt to use against the desired clamping force required. The tensile strength of a bolt is actually the point at which it breaks in two (to keep it simple), the proof load (yield) is the point at which the bolt starts to go beyond the elastic limit, when bolts are tested for use, the key pass criteria is to have the minimum proof load applied to the bolt and for the length of the bolt to be unchanged to the un-stressed length, if it is longer it is said to have gone into permanent set and failed.
When tightening bolts on normal joints you usually apply only 60 - 80% of the proof load, that way you are guaranteed not to send the bolt into yield and plastic deformation.
That said sometimes it is important to achieve maximum clamping force, which is where tightening to yield comes into play, in order to achieve this these bolts have to be made from really good alloy steel, closely controlled on heat treatment for both core and surface properties, and then the shanks are sometimes ground to ensure the body diameter is almost identical (but not much benefit, as failure occurs in the smallest diameter - threads) on each bolt and then coated with manganese phosphate to ensure a consistent coefficient of friction, and then these bolts are then torqued up to just below the yield limit, and then a degree of turn added, called angle, this is normally done on a special machine with transducers to accurately monitor the torque.
I did say i was going to do a write up on bolting and materials, must get round to it.
