Actuarial Outpost > CAS Calculating the hetero adjustment
 Register Blogs Wiki FAQ Calendar Search Today's Posts Mark Forums Read
 FlashChat Actuarial Discussion Preliminary Exams CAS/SOA Exams Cyberchat Around the World Suggestions

 Upload your resume securely at https://www.dwsimpson.com to be contacted when our jobs meet your skills and objectives.

#1
01-10-2019, 01:41 AM
 AbedNadir Member CAS SOA Join Date: Mar 2014 Studying for FCAS Posts: 2,607

I am trying to replicate the Milliman example for ODP bootstrapping given by Shapland from the original paid triangle in R

Let's say I want to find the hetero adjustment factor for periods 1-3 and the other factors are 4-7 and 8-10. I have confirmed that my standardized residuals are correct using the H matrix. I have found the standard deviation of all the standardized residuals in periods 1-3, and I have found the standard deviation of all of the standardized residuals across all periods, excluding the null points on each end of the triangle that do not have a residual.

I then divide h for all values by h for periods 1-3, but I get 1.546.

Here are the values I found standard deviations for:

Just the periods 1-3

[1] 183.60702 133.82127 -131.24915 -43.39100 -64.92590 -57.36316 -148.73570 92.42188 -55.11517 -102.84746 244.50883 -223.31533
[13] 204.07214 -189.73109 148.75836 79.40858 72.84334 -97.06667 88.67449 -164.26735 138.44964 -190.01584 -140.95863 284.34219
[25] -27.97642 27.97642

and all periods

[1] 183.60702 133.82127 -131.24915 -371.06810 193.89249 591.28897 -272.45711 -117.94646 -118.19503 -43.39100 -64.92590 -57.36316
[13] 160.13799 -208.95238 -159.80436 306.65820 32.42232 118.19503 -148.73570 92.42188 -55.11517 -26.64048 272.21870 -476.41522
[25] 252.07613 75.73922 -102.84746 244.50883 -223.31533 658.09028 -460.58857 -84.67212 -318.35358 204.07214 -189.73109 148.75836
[37] -215.74333 -24.18163 206.31529 79.40858 72.84334 -97.06667 -231.49620 256.12390 88.67449 -164.26735 138.44964 -37.66763
[49] -190.01584 -140.95863 284.34219 -27.97642 27.97642

How do I get the right value?

Thanks
__________________

Last edited by AbedNadir; 01-10-2019 at 02:51 AM..
#2
01-10-2019, 01:39 PM
 DamSon Member CAS Join Date: Sep 2018 Location: NYC Studying for 8 Favorite beer: Alcoholic Posts: 218

Quote:
 Originally Posted by AbedNadir I am trying to replicate the Milliman example for ODP bootstrapping given by Shapland from the original paid triangle in R Let's say I want to find the hetero adjustment factor for periods 1-3 and the other factors are 4-7 and 8-10. I have confirmed that my standardized residuals are correct using the H matrix. I have found the standard deviation of all the standardized residuals in periods 1-3, and I have found the standard deviation of all of the standardized residuals across all periods, excluding the null points on each end of the triangle that do not have a residual. I then divide h for all values by h for periods 1-3, but I get 1.546. Here are the values I found standard deviations for: Just the periods 1-3 [1] 183.60702 133.82127 -131.24915 -43.39100 -64.92590 -57.36316 -148.73570 92.42188 -55.11517 -102.84746 244.50883 -223.31533 [13] 204.07214 -189.73109 148.75836 79.40858 72.84334 -97.06667 88.67449 -164.26735 138.44964 -190.01584 -140.95863 284.34219 [25] -27.97642 27.97642 and all periods [1] 183.60702 133.82127 -131.24915 -371.06810 193.89249 591.28897 -272.45711 -117.94646 -118.19503 -43.39100 -64.92590 -57.36316 [13] 160.13799 -208.95238 -159.80436 306.65820 32.42232 118.19503 -148.73570 92.42188 -55.11517 -26.64048 272.21870 -476.41522 [25] 252.07613 75.73922 -102.84746 244.50883 -223.31533 658.09028 -460.58857 -84.67212 -318.35358 204.07214 -189.73109 148.75836 [37] -215.74333 -24.18163 206.31529 79.40858 72.84334 -97.06667 -231.49620 256.12390 88.67449 -164.26735 138.44964 -37.66763 [49] -190.01584 -140.95863 284.34219 -27.97642 27.97642 How do I get the right value? Thanks
By right value do you mean the 1.547 from the Shapland file?
#3
01-10-2019, 06:47 PM
 AbedNadir Member CAS SOA Join Date: Mar 2014 Studying for FCAS Posts: 2,607

Quote:
 Originally Posted by DamSon By right value do you mean the 1.547 from the Shapland file?
Hi, yes, I see 1.585 in cell C142 on the bootstrap sheet. That is the value I should be looking at, no?
__________________
#4
01-11-2019, 10:21 AM
 DamSon Member CAS Join Date: Sep 2018 Location: NYC Studying for 8 Favorite beer: Alcoholic Posts: 218

Quote:
 Originally Posted by AbedNadir Hi, yes, I see 1.585 in cell C142 on the bootstrap sheet. That is the value I should be looking at, no?
I'm pretty sure you're using the scale parameter hetero adjustment then. I believe that's the default on the file. If you go to the inputs sheet, cell C39, change to StdDev and recalculate you should get 1.547.

I compared some of your residuals to the ones in the Residuals sheet they look very similar. Seems like you're doing the calculations correctly.
#5
01-11-2019, 08:53 PM
 AbedNadir Member CAS SOA Join Date: Mar 2014 Studying for FCAS Posts: 2,607

Quote:
 Originally Posted by DamSon I'm pretty sure you're using the scale parameter hetero adjustment then. I believe that's the default on the file. If you go to the inputs sheet, cell C39, change to StdDev and recalculate you should get 1.547. I compared some of your residuals to the ones in the Residuals sheet they look very similar. Seems like you're doing the calculations correctly.
damn son, you're right

thanks
__________________
#6
01-11-2019, 09:32 PM
 redearedslider Member CAS Join Date: Oct 2015 Posts: 13,970

I see what you did there
__________________
Quote:
 Originally Posted by Abraham Weishaus ASM does not have a discussion of stimulation, but considering how boring the manual is, maybe it would be a good idea.
#7
01-11-2019, 11:41 PM
 Colonel Smoothie Member CAS Join Date: Sep 2010 College: Jamba Juice University Favorite beer: AO Amber Ale Posts: 49,266

Quote:
 Originally Posted by redearedslider I see what you did there
plz
__________________
Recommended Readings for the EL Actuary || Recommended Readings for the EB Actuary

Quote:
 Originally Posted by Wigmeister General Don't you even think about sending me your resume. I'll turn it into an origami boulder and return it to you.