June 5th, 2015, 10:22 am
First of all, it is Ho, not Hoo. Second, given that you get to choose the functional form for the discount factors, you may cleverly choose one that is analytically differentiable. In fact, come to think of it, it would be hard not to. Third, yes, you are right. Instantaneous forward rate dynamics turned out to be a bit of a side track for interest rate modeling, and both the math and finance get easier if you focus your modeling on discount factors (aka zero coupon bond prices). That being said, HJM made pretty serious strides in this framework, and their work is important on several levels. The '86 Ho Lee paper is only interesting from the perspective of the history of interest rate modeling, and as squarely set in a discrete time binomial setting did not in fact address instantaneous forward rates.