In class we used the Bulow-Klemperer Theorem to prove an approximation result. For n agents with i.i.d. and regularly distributed values, the revenue of the second-price auction is a n/(n-1) approximation to the revenue optimal mechanism. Specifically, writing this approximation bound as (1+) the is decreasing in n as 1/n. For monotone hazard rate distributions, i.e., where f(x)/(1-F(x)) is monotonically non-decreasing, this bound can be improved to be exponentially decreasing in n, i.e., where is O(c^) for some constant c. State and prove this result for the best constant c that you can identify.