<t> int_u=0^u=t1 (t2-u) du equals to t1*t2/2 , so that's where this part of the term comes from.I'm not sure about this one, though, but in your general convexity adjustment rule, is it possible you miss the 1/(t2-t1) factor and the -1 as in1/(t2-t1)*( exp( int_u=0^u=t1 ( b(u,t2) - b(u,t1) )*b(u,t2)...