Right on the ball diamondjim. Although any bolt designer worth his salt should design the thread engagement length long enough to ensure that tensile rupture occurs at a lower load than thread strip (i.e. shear of the thread teeth)
Swall - when you say the shank stress area is smaller than the thread stress area, how are you calculating thread stress area, there are several methods for this depending upon the predominant failure mode. 'Stress areas' are not always the same when condsidering hi-temp and low-temp load situations, partly because the residual benefit of cold work in a rolled thread (and all con-rod bolts will have rolled threads) is different in the two cases. What in fact is done for con-rod bolts is actual tensile tests on the bolt, that way the user is sure of designing either the shank or the thread slightly stronger, as he wishes.