WEBVTT

00:00:00.090 --> 00:00:02.023
An annuity with zero growth should

00:00:02.036 --> 00:00:03.814
match the same level payments.

00:00:04.253 --> 00:00:06.405
But this actuarial library returns about

00:00:06.418 --> 00:00:08.441
thirteen point six two instead of

00:00:08.454 --> 00:00:09.820
fourteen point nine four.

00:00:10.570 --> 00:00:12.851
The example uses seventy-five percent

00:00:12.864 --> 00:00:14.474
of a standard mortality table.

00:00:14.913 --> 00:00:17.309
The growth calculation silently rebuilds

00:00:17.322 --> 00:00:19.294
that table at one hundred percent.

00:00:19.732 --> 00:00:21.536
So the survival assumptions change.

00:00:22.316 --> 00:00:24.313
The proposed repair keeps copies of

00:00:24.326 --> 00:00:26.878
the supplied mortality data and changes

00:00:26.891 --> 00:00:28.089
only the interest rate.

00:00:28.514 --> 00:00:30.795
It also handles custom survivor tables.

00:00:31.583 --> 00:00:33.928
Six hundred thirty direct cash-flow checks

00:00:33.941 --> 00:00:35.307
pass after the repair.

00:00:35.745 --> 00:00:37.575
The report and patch are public,

00:00:38.000 --> 00:00:39.637
and the author has been contacted.

00:00:40.062 --> 00:00:42.717
These synthetic tests establish a calculation

00:00:42.730 --> 00:00:45.268
defect, without claiming customer losses.