

FlashChat  Actuarial Discussion  Preliminary Exams  CAS/SOA Exams  Cyberchat  Around the World  Suggestions 

Thread Tools  Search this Thread  Display Modes 
#2




Lower bound of leverage
Is there any reason why the lower bound of a leveraged point h_ii is 1/n instead of 0?
If we look at the variance of the residual var(e_i) = sigma^2 x (1h_ii); isn't it possible for h_ii = 0 and thus var(e_i) = sigma^2? Thanks
__________________
ASA in progress 
#3




The proof of the lower bound is mainly matrix algebra that isn't required for the exam. To get some ideas, look at the formula of the leverage for simple linear regression, or equation (3.2.1) in the ACTEX manual. The second term, (x_i  bar(x))^2 / S_xx, is nonnegative, so the leverage is at least 1/n. Equality holds if and only if x_i is exactly equal to the sample mean bar(x).
__________________
Ambrose Lo, PhD, FSA, CERA Associate Professor of Actuarial Science (with tenure) Department of Statistics and Actuarial Science The University of Iowa ACTEX Manual for SOA Exam PA  ACTEX Manual for SOA Exam SRM  ACTEX Manual for CAS Exam MASI Textbook: Derivative Pricing: A ProblemBased Primer (useful for derivatives portion of Exam IFM) 
#6




Any reactions from the September sitting? I'm still spooked out that the SRM threads get no love...
I found that it was, more or less, what I expected. I nailed all fifteenish computational questions that came my way, would estimate that I also nailed 510 of the conceptual, and was able to narrow down the remaining 1015 conceptual questions down to two or three options. Wondering if anyone was surprised (pleasantly or unpleasantly) by the sitting or if, like me, it was kind of like you thought it might be. 
#7




Thanks for checking in. For me work picked up a lot before the exam and unfortunately I couldn't shift things around to get the necessary study time, so I pushed my sitting to January. Good news is that work has lighten up significantly and I plan to have sufficient time to prepare.
I'll be on this thread much more over the next few months and will definitely provide feedback on the exam.
__________________
ASA in progress 
#8




Going for the January sitting here, doesn't seem like much is happening on the SRM threads lol... A little over halfway through the ASM manual (I've seen negative stuff about their manual for this one but I've been using it with too much success on everything else to abandon it now). Really flying through the material, thinking about using Adapt and reading through ISL once I exhaust this manual.

#9




Actex Example 3.2.1  Can anyone provide guidance on this one? I understand the model solution, but I'm curious as to why I can't back into y1_hat using the property that the residuals sum to 0. The model solution solves for the parameters b_0 and b_1 using LSE and gets y_1hat using x_1, which makes sense, but I thought my approach would also be adequate however I'm coming with a different value for y_1hat.
__________________
ASA in progress 
Thread Tools  Search this Thread 
Display Modes  

