PDA

View Full Version : Extension of 75-80 Mortality table beyond age 70


Chuck
03-09-2005, 04:41 PM
The SOA Table manager only shows select rates thru age 70. I can't find my copy of the 1986 TSA they are supposedly in and I can't find it in the online SOA library.

The thing is, I have a spreadsheet with the 75-80 table that I got from somewhere, and it happens to have an extension of the select rates thru issue age 80 for Male ANB. But I don't remember whether I, or somebody else just as unofficial as I, made those up, or whether there exists an "official" extension of the 75-80 rates beyond age 70. Anybody know?

Thanks,
Chuck

Numbers Nerd
03-09-2005, 05:26 PM
I have a version of the 75-80 table that goes all the way to issue age 90. 1st year rates for males issue age 71-75 are:
9.53
11.40
13.52
15.94
18.71

Based on the name of the table, it appears to have been extended by a large reinsurance company. p.m. me if you want the full table.

Ranger
03-09-2005, 05:33 PM
I've got a different set of rates above age 70.

I don't recall where I got mine - but I thought there was an offical SOA ANB version, released some time after the younger part of the table, and then after that, someone did a paper and released the ALB version.

I might be completely wrong - it's been years since I looked into it.

I'll try digging into my files and see what I can find.

Chuck
03-09-2005, 05:42 PM
The numbers I have for 71-75 are...

8.67
9.50
10.40
11.38
12.47

I'm pretty sure these are made-up, as they appear to just apply the issue age 70 select ratios to the higher age ultimates and I doubt whether that makes sense.

Chuck

Ms. Statamatician
03-09-2005, 08:53 PM
Numbers Nerd: You have the values for the Manulife extension

Chuck: You have the values for the Tillinghast extension.

I am no expert, but I just did a bunch of work involving extensions of the table, and I would recognize those numbers in my sleep! If you need either, let me know, and I can send it.

friendlyFSA
03-10-2005, 10:58 AM
You can find more in the Advanced Age study published by the SOA Reinsurance Section. Here is the link, hope it works:

http://www.soa.org/ccm/cms-service/stream/asset/?asset_id=3780045&g11n.enc=UTF-8

ul guy
03-10-2005, 11:36 AM
Right on, Ms. Statamatician. The so-called Manulife extension also can be found in the SOA Transaction Reports 1997-1998.

Chuck
03-10-2005, 11:39 AM
Numbers Nerd: You have the values for the Manulife extension

Chuck: You have the values for the Tillinghast extension.

I am no expert, but I just did a bunch of work involving extensions of the table, and I would recognize those numbers in my sleep! If you need either, let me know, and I can send it.Thanks to all. FriendlyFSA's link worked fine and I found the 97-98 transactions, so I have all three sets now.

I guess the $64 question now is which extension makes the most sense for pricing purposes? I'm thinking the reinsurance studies do moreso than the Tillinghast extension because it seems unlikely that age 70 selection would continue as assumed by the Tillinghast numbers.

Any opinions?

Chuck

ul guy
03-10-2005, 11:50 AM
Check the slope of each with the SOA 90-95 or the 2001 CSO. Check with your reinsurer, assuming they are taking some portion of the mortality risk. There have been some articles (I think there were three in the "Product Matters" section newsletter) written about the slope of the 75-80, especially at the older ages, that may be of help.

Numbers Nerd
03-10-2005, 12:18 PM
Ditto what UL Guy said. Except that you should use the 2001 vbt, not the CSO. You want to use the basic table, before loadings are added.

urysohn
03-10-2005, 01:14 PM
Based on recent (and consistent) feedback we're getting from our reinsurers, none are likely sufficient. Old age mortality is a hot item, with many of the reinsurers reaching the conclusion that there should be additional loadings at ages 70+, above and beyond those included in any of the existing extensions.

KD
03-29-2005, 04:32 PM
Is the "Manulife" extension table short for "Manufacturers Life Extension" table? Thanks!

USA AOK
03-29-2006, 02:30 PM
I have a SOA 75-80 table with an extension and I was wondering if someone knew if it is a well known extension within the industry. I can't seem to find it anywhere. I know it isn't one of the two already mentioned in this thread.

1st year rates for males issue age 71-75 are:
12.14
16.72
21.64
26.41
29.44

Thanks for the help.

teapeaexcueexplustea
04-06-2006, 03:54 PM
Thanks to all. FriendlyFSA's link worked fine and I found the 97-98 transactions, so I have all three sets now.

I guess the $64 question now is which extension makes the most sense for pricing purposes? I'm thinking the reinsurance studies do moreso than the Tillinghast extension because it seems unlikely that age 70 selection would continue as assumed by the Tillinghast numbers.

Any opinions?

Chuck


The Tillinghast extension is an extremely flat one, based pretty much on polynomial extrapolation. The Manulife was done somewhat like that, but it is much more conservative. I would say to be conservative, use Manu. Though, from my experience, some companies, and even the recent SOA study have used the Tillinghast extension. PM me if you want more details.

pepeTHEgrate
04-06-2006, 09:01 PM
Which one is the best for you depends on how you will use the extention.

teapeaexcueexplustea
04-07-2006, 10:11 AM
Which one is the best for you depends on how you will use the extention.

I agree

mayreeh
06-15-2006, 03:13 PM
I have a table that appears to be the Tillinghast table. Is the T table published online so that I can check to ensure I have it right?

teapeaexcueexplustea
06-16-2006, 10:49 AM
I have never been able to find it published online. You could send it to me though, and I can tell you.

Chuck
06-16-2006, 05:01 PM
I did not get it from an official source, but I found that the extension I have (that I was told was the Tillinghast extension in this thread) was extended by maintaining a constant ratio of the select rate to the ultimate rate at a given attained age and duration. In other words

q(70+y,t)=q(70+y+t-1)*(q(70,t)/q(70+t-1)) for x and t =1,2,3 etc

where q(x,t) is the select and ultimate q at age x, duration t and
q(x) is the ultimate only q at age x.

I think that expresses the formula correctly anyway.

Chuck

Numbers Nerd
08-23-2006, 03:23 PM
I have a SOA 75-80 table with an extension and I was wondering if someone knew if it is a well known extension within the industry. I can't seem to find it anywhere. I know it isn't one of the two already mentioned in this thread.

1st year rates for males issue age 71-75 are:
12.14
16.72
21.64
26.41
29.44

Thanks for the help.

I've seen that "extension" recently. Axis software includes it in their official list of tables. But you know what they've done - they've simply used the Age 70 rates to extend to the later issue ages. 12.14 is age 70 duration 2, for example.

teapeaexcueexplustea
08-25-2006, 03:05 PM
I've seen that "extension" recently. Axis software includes it in their official list of tables. But you know what they've done - they've simply used the Age 70 rates to extend to the later issue ages. 12.14 is age 70 duration 2, for example.

Those are just the attained age rates.