I am referring to Loss Development Using Credibility by Brosius.
At Page 7 Option 2, it says that MSE = 6c^2 - 20c + 18,
and minimum is found at c = 5/3
At page 8, it says that when c = 5/3, MSE = 10/3, which is not the case if we pluck back c = 5/3 into the equation. In fact, it give us 4/3. The difference is 2, which coincidentally, is the expected value of outstanding claim count.
Am I missing something here?:-?
At Page 7 Option 2, it says that MSE = 6c^2 - 20c + 18,
and minimum is found at c = 5/3
At page 8, it says that when c = 5/3, MSE = 10/3, which is not the case if we pluck back c = 5/3 into the equation. In fact, it give us 4/3. The difference is 2, which coincidentally, is the expected value of outstanding claim count.
Am I missing something here?:-?
MSE of Link Ratio Estimate