Mortality Tables for Determining Present Value Under Defined Benefit Pension Plans, 46388-46411 [2017-21485]

Agencies

• Internal Revenue Service
• DEPARTMENT OF THE TREASURY

[Federal Register Volume 82, Number 192 (Thursday, October 5, 2017)]
[Rules and Regulations]
[Pages 46388-46411]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 2017-21485]


=======================================================================
-----------------------------------------------------------------------

DEPARTMENT OF THE TREASURY

Internal Revenue Service

26 CFR Part 1

[TD9826]
RIN 1545-BM71


Mortality Tables for Determining Present Value Under Defined 
Benefit Pension Plans

AGENCY: Internal Revenue Service (IRS), Treasury.

ACTION: Final regulations.

-----------------------------------------------------------------------

SUMMARY: This document contains final regulations prescribing mortality 
tables to be used by most defined benefit pension plans. The tables 
specify the probability of survival year-by-year for an individual 
based on age, gender, and other factors. This information is used 
(together with other actuarial assumptions) to calculate the present 
value of a stream of expected future benefit payments for purposes of 
determining the minimum funding requirements for a defined benefit 
plan. These mortality tables are also relevant in determining the 
minimum required amount of a lump-sum distribution from such a plan. In 
addition, this document contains final regulations updating the 
requirements that a plan sponsor must meet to obtain IRS approval to 
use mortality tables specific to the plan for minimum funding purposes 
(instead of using the generally applicable mortality tables). These 
regulations affect participants in, beneficiaries of, employers 
maintaining, and administrators of certain retirement plans.

DATES: 
    Effective date: These regulations are effective on October 5, 2017.
    Applicability date: These regulations apply to plan years beginning 
on or after January 1, 2018.

FOR FURTHER INFORMATION CONTACT: Concerning the regulations, Arslan 
Malik at (202) 317-6700; concerning the construction of the base 
mortality tables and the static mortality tables for 2018, Michael 
Spaid at (206) 946-3480.

SUPPLEMENTARY INFORMATION: 

Background

A. Generally Applicable Mortality Tables

    Section 412 of the Internal Revenue Code (Code) prescribes minimum 
funding requirements for defined benefit pension plans. Section 430 
specifies the minimum funding requirements that apply generally to 
defined benefit plans that are not multiemployer plans.\1\ Section 
430(a) defines the minimum required contribution for a plan year by 
reference to the plan's funding target for the plan year. Under section 
430(d)(1), a plan's funding target for a plan year generally is the 
present value of all benefits accrued or earned under the plan as of 
the first day of that plan year.
---------------------------------------------------------------------------

    \1\ Section 302 of the Employee Retirement Income Security Act 
of 1974, Public Law 93-406, as amended (ERISA), sets forth funding 
rules that are parallel to those in section 412 of the Code, and 
section 303 of ERISA sets forth additional funding rules for defined 
benefit plans (other than multiemployer plans) that are parallel to 
those in section 430 of the Code. Under section 101 of 
Reorganization Plan No. 4 of 1978 (43 FR 47713) and section 302 of 
ERISA, the Secretary of the Treasury has interpretive jurisdiction 
over the subject matter addressed in these regulations for purposes 
of ERISA, as well as the Code. Thus, these Treasury regulations 
issued under section 430 of the Code also apply for purposes of 
section 303 of ERISA.
---------------------------------------------------------------------------

    Section 430(h)(3) contains rules regarding the mortality tables to 
be used under section 430. Under section 430(h)(3)(A), except as 
provided in section 430(h)(3)(C) or (D), the Secretary is to prescribe 
by regulation mortality tables to be used in determining any present 
value or making any computation under section 430. Those mortality 
tables must be based on the actual mortality experience of pension plan 
participants and projected trends in that experience. In prescribing 
those mortality tables, the Secretary is required to take into account 
results of available independent studies of mortality of individuals 
covered by pension plans.\2\ Under section

[[Page 46389]]

430(h)(3)(B), the Secretary is required to make revisions to any table 
in effect under section 430(h)(3)(A) at least every 10 years to reflect 
actual mortality experience of pension plan participants and projected 
trends in that experience.
---------------------------------------------------------------------------

    \2\ The standards prescribed for developing these mortality 
tables are the same as the standards that are prescribed for 
developing mortality tables for multiemployer plans under section 
431(c)(6)(D)(iv)(II) (which are used to determine current liability 
as part of the calculation of the minimum full funding limitation 
under section 431(c)(6)(B)). These standards also apply for purposes 
of determining current liability as part of the calculation of the 
minimum full funding limitation under section 433(c)(7)(C) for a 
CSEC plan (as defined in section 414(y)).
---------------------------------------------------------------------------

    Section 430(h)(3)(C) prescribes rules for a plan sponsor's use of 
substitute mortality tables reflecting the specific mortality 
experience of a plan's population instead of using the generally 
applicable mortality tables. Under section 430(h)(3)(C), the plan 
sponsor may request the Secretary's approval to use plan-specific 
substitute mortality tables that meet requirements specified in the 
statute.
    Section 430(h)(3)(D) provides for the use of separate mortality 
tables with respect to certain individuals who are entitled to benefits 
on account of disability. These separate mortality tables are permitted 
to be used with respect to disabled individuals in lieu of the 
generally applicable mortality tables provided pursuant to section 
430(h)(3)(A) or the substitute mortality tables under section 
430(h)(3)(C). The Secretary is to establish separate tables for 
individuals with disabilities occurring in plan years beginning before 
January 1, 1995, and in later plan years, with the mortality tables for 
individuals with disabilities occurring in those later plan years 
applying only to individuals who are disabled within the meaning of 
Title II of the Social Security Act.
    Final regulations (TD 9419) under section 430(h)(3) were published 
in the Federal Register for July 31, 2008 (73 FR 44632). The final 
regulations issued in 2008 include rules regarding generally applicable 
mortality tables, which are set forth in Sec.  1.430(h)(3)-1 (the 2008 
general mortality table regulations), as well as rules regarding 
substitute mortality tables, which are set forth in Sec.  1.430(h)(3)-2 
(the 2008 substitute mortality table regulations).
    The 2008 general mortality table regulations prescribe a base 
mortality table and a set of mortality improvement rates that are used 
to project mortality rates for years after the year 2000. The mortality 
tables included in those regulations are based on the mortality tables 
included in the RP-2000 Mortality Tables Report (based on an experience 
study for the period 1990-1994 and referred to in this preamble as the 
RP-2000 mortality tables) released by the Society of Actuaries in July 
2000 (updated in May 2001). The mortality improvement rates included in 
those regulations are the Scale AA Projection Factors (which are based 
on mortality improvement for the period 1977 through 1993), which were 
included in the RP-2000 Mortality Tables Report for use in conjunction 
with the RP-2000 mortality tables.\3\
---------------------------------------------------------------------------

    \3\ The RP-2000 Mortality Tables Report is available at https://www.soa.org/experience-studies/2000-2004/research-rp-2000-mortality-tables/.
---------------------------------------------------------------------------

    The 2008 general mortality table regulations prescribe the use of 
generational mortality tables,\4\ but include an option for plans to 
use static mortality tables. The static mortality tables (which are 
updated annually) use a single mortality table for all years of birth 
to approximate the present value that would be determined using the 
generational mortality tables. The 2008 general mortality table 
regulations set forth static mortality tables for valuation dates 
occurring in 2008 and provide that static mortality tables for 
valuation dates occurring in later years will be published in the 
Internal Revenue Bulletin. Static mortality tables for valuation dates 
occurring during 2009-2013 were published in Notice 2008-85 (2008-1 
C.B. 747 (March 28, 2008)). Updated static mortality tables for 
valuation dates occurring during 2014 and 2015 were published in Notice 
2013-49 (2013-32 I.R.B. 127 (July 10, 2013)). Updated static mortality 
tables for valuation dates occurring in 2016 and 2017 were published in 
Notice 2015-53 (2015-33 I.R.B. 190 (July 31, 2015)) and Notice 2016-50 
(2016-38 I.R.B. 371 (September 2, 2016)), respectively.
---------------------------------------------------------------------------

    \4\ Generational mortality tables are a series of mortality 
tables, one for each year of birth, each of which fully reflects 
projected trends in mortality for individuals who are born in a 
particular year.
---------------------------------------------------------------------------

    Notice 2013-49 also requested comments on whether a separate 
disability mortality table is still warranted with respect to 
participants who became disabled before 1995. In addition, Notice 2013-
49 noted that the Treasury Department (Treasury) and the IRS were aware 
that the Society of Actuaries was conducting a mortality study of 
pension plan participants and specifically requested comments on 
whether other studies of actual mortality experience of pension plan 
participants and projected trends of that experience were available 
that should be considered for use in developing mortality tables for 
future use under section 430(h)(3).
    In October 2014, the Retirement Plans Experience Committee (RPEC) 
of the Society of Actuaries issued a new mortality study of 
participants in private pension plans. The study, which is based on 
mortality experience for the years 2004 to 2008, is referred to as the 
RP-2014 Mortality Tables Report (and sets forth mortality tables that 
are referred to as the RP-2014 mortality tables). The RP-2014 Mortality 
Tables Report, as revised November 2014, is available at www.soa.org/Research/Experience-Study/pension/research-2014-rp.aspx. At the same 
time, RPEC issued a companion study of mortality improvement, referred 
to as the Mortality Improvement Scale MP-2014 Report (which sets forth 
mortality improvement rates that are referred to as the Scale MP-2014 
rates). As described in the Mortality Improvement Scale MP-2014 Report 
(available at www.soa.org/Research/Experience-Study/pension/research-2014-mp.aspx), the Scale MP-2014 rates were based on mortality 
improvement experience for the general population through 2009.
    In October 2015, RPEC released an update to the Scale MP-2014 
rates. The updated rates, referred to as the Scale MP-2015 rates, were 
released as part of the Mortality Improvement Scale MP-2015 Report 
(available at https://www.soa.org/Research/Experience-Study/Pension/research-2015-mp.aspx). The Scale MP-2015 rates were created using 
historical data for mortality improvement for the general population 
through 2011 and the same model and parameters that were used to 
produce the Scale MP-2014 rates. In conjunction with the release of the 
updated rates, RPEC indicated the intent to reflect the latest data 
available by providing future annual updates to the model as soon as 
practicable following the public release of updated data upon which the 
model is constructed.
    In October 2016, RPEC released a further update to the Scale MP-
2014 rates, which are referred to as the Scale MP-2016 rates. The Scale 
MP-2016 rates take into account data for mortality improvement for the 
general population for 2012 and 2013, along with an estimate of 
mortality rates for 2014. As described in the Mortality Improvement 
Scale MP-2016 Report (available at www.soa.org/Research/Experience-Study/Pension/research-2016-mp.aspx), in developing the Scale MP-2016 
rates, RPEC changed some of the parameters from those that were used in 
developing the Scale MP-2014 rates.

B. Plan-Specific Substitute Mortality Tables

    Section 430(h)(3)(C) permits a plan sponsor to request approval by 
the Secretary to use plan-specific substitute mortality tables in lieu 
of the generally applicable mortality tables. If the Secretary 
determines that the proposed

[[Page 46390]]

tables meet the statutory standards and approves the request, the 
substitute mortality tables are used to determine present values and 
make computations under section 430 during the period of consecutive 
plan years (not to exceed 10) specified in the request. Under the 
statute, a substitute mortality table may be used for a plan, only if: 
(1) The plan has a sufficient number of plan participants and has been 
maintained for a sufficient period of time to have credible mortality 
information necessary to create a substitute mortality table; and (2) 
the table reflects the actual mortality experience of the plan's 
participants and projected trends in general mortality experience. 
Except as provided by the Secretary, a plan sponsor may not use 
substitute mortality tables for any plan unless substitute mortality 
tables are established and used for each plan maintained by the plan 
sponsor or a member of its controlled group.
    The 2008 substitute mortality table regulations provide for review 
by the Commissioner of a plan sponsor's request for approval to use 
substitute mortality tables. Under those regulations, to use substitute 
mortality tables with respect to a plan, a plan sponsor must submit a 
written request to the Commissioner that demonstrates that those 
substitute mortality tables comply with applicable requirements. A 
request for approval to use substitute mortality tables must specify 
the first plan year and the term of years for which the tables are 
requested to be used.
    Substitute mortality tables may not be used for a plan year unless 
the plan sponsor submits the request at least 7 months before the first 
day of the first plan year for which the substitute mortality tables 
are to apply. The Commissioner has 180 days to review a request for 
approval to use substitute mortality tables. If the Commissioner does 
not deny the request within this 180-day period, the request is deemed 
to have been approved unless the Commissioner and the plan sponsor have 
agreed to extend that period.
    Under the 2008 substitute mortality table regulations, substitute 
mortality tables for a plan must be established separately for each 
gender, and a substitute mortality table may be established for a 
gender only if there is credible mortality experience with respect to 
that gender. If the mortality experience for one gender is credible but 
the mortality experience for the other gender is not credible, the 
substitute mortality tables are used for the gender that has credible 
mortality experience, and the generally applicable mortality tables are 
used for the gender that does not have credible mortality experience.
    Under the 2008 substitute mortality table regulations, there is 
credible mortality experience with respect to a gender if and only if, 
over the period covered by the experience study, there are at least 
1,000 deaths of individuals of that gender.\5\ For this purpose, the 
minimum length of the experience study period is 2 years and the 
maximum length of the experience study period generally is 5 years. 
Furthermore, the last day of the final year reflected in the experience 
data must be less than three years before the first day of the first 
plan year for which the substitute mortality tables are to apply.
---------------------------------------------------------------------------

    \5\ The 1,000-death threshold for credible mortality experience 
under the regulations was intended to provide a high degree of 
confidence that the plan's past mortality experience will be 
predictive of its future mortality, and is consistent with relevant 
actuarial literature (see, for example, Thomas N. Herzog, 
Introduction to Credibility Theory (1999); Stuart A. Klugman, et. 
al., Loss Models: From Data to Decisions (2004)).
---------------------------------------------------------------------------

    Under the 2008 substitute mortality table regulations, development 
of substitute mortality tables for a plan requires creation of a base 
table and identification of a base year, which are then used to 
determine the substitute mortality tables. The base table must be 
developed from a study of the mortality experience of the plan using 
amounts-weighted data. Under those regulations, a plan's substitute 
mortality tables must be generational mortality tables that are 
determined using the base mortality tables developed from the 
experience study and the Scale AA Projection Factors (that is, using 
the same basis for mortality improvement that is used under 2008 
general mortality table regulations).
    Under the 2008 substitute mortality table regulations, the use of 
substitute mortality tables is terminated early in certain 
circumstances, including pursuant to a replacement of the generally 
applicable mortality tables. The early termination pursuant to such a 
replacement must be effective as of a date specified in guidance 
published in the Internal Revenue Bulletin.
    Rev. Proc. 2008-62 (2008-2 C.B. I.R.B. 935) sets forth the 
procedure by which a plan sponsor of a defined benefit plan may request 
and obtain approval to use plan-specific substitute mortality tables in 
accordance with section 430(h)(3)(C). The revenue procedure specifies 
the information that must be provided in a request for approval to use 
substitute mortality tables and specifies two alternative acceptable 
methods of construction for base substitute mortality tables. Under 
section 11 of Rev. Proc. 2008-62, a base table for a population may be 
created from the unadjusted base table for the population through the 
application of a graduation method generally used by the actuarial 
profession in the United States.\6\ Section 12 of Rev. Proc. 2008-62 
provides for an alternative method of constructing a base table through 
the application of a fixed percentage to the mortality rates of a 
standard mortality table, projected to the base year.
---------------------------------------------------------------------------

    \6\ The revenue procedure identifies the Whittaker-Henderson 
Type B graduation method or the Karup-King graduation method as 
acceptable methods.
---------------------------------------------------------------------------

    Section 503 of the Bipartisan Budget Act of 2015, Public Law 114-74 
(129 Stat. 584 (2015)), which was enacted November 2, 2015, provides 
for changes to the rules on the use of substitute mortality tables. 
Under that section, ``the determination of whether plans have credible 
information shall be made in accordance with established actuarial 
credibility theory, which (1) is materially different from the rules 
under [section 430(h)(3)(C)], including Revenue Procedure 2007-37,\7\ 
that are in effect on [November 2, 2015]; and (2) permits the use of 
tables that reflect adjustments to the tables described in [section 
430(h)(3)(A) and (B)]'' if those adjustments are based on the actual 
experience of the pension plan maintained by the plan sponsor.
---------------------------------------------------------------------------

    \7\ Rev. Proc. 2007-37, 2007-1 CB 1433, was not in effect on 
November 2, 2015. It was issued in 2007 in conjunction with proposed 
regulations regarding substitute mortality tables (REG-143601-06, 72 
FR 29456) and was replaced by Rev. Proc. 2008-62 when those 
regulations were finalized in 2008.
---------------------------------------------------------------------------

    Proposed regulations regarding revisions to mortality tables under 
section 430(h)(3) (REG-112324-15) were published in the Federal 
Register on December 29, 2016 (81 FR 95911). The proposed regulations 
contain revisions to the generally applicable mortality tables (based 
on the RP-2014 Mortality Tables Report), as well as new rules regarding 
substitute mortality tables that reflect section 503 of the Bipartisan 
Budget Act of 2015. Comments were received on the proposed regulations, 
and a public hearing was held on April 13, 2017. After consideration of 
the comments, the proposed regulations are adopted by this Treasury 
decision subject to certain changes, the most significant of which are 
described in this preamble under the heading Explanation of Provisions.

Explanation of Provisions

I. Overview

    These final regulations revise the methodology for developing the

[[Page 46391]]

generally applicable mortality tables that are used to determine 
present value or make any computation under section 430. Pursuant to 
section 417(e)(3)(B), a modified version of these tables will be used 
for purposes of determining the amount of a single-sum distribution (or 
another accelerated form of distribution). This revised methodology for 
developing tables under section 430(h)(3)(A) is being issued pursuant 
to the requirement under section 430(h)(3)(B) to revise the mortality 
tables used under section 430 at least every 10 years to reflect the 
actual mortality experience of pension plan participants and projected 
trends in that experience. As under the 2008 general mortality table 
regulations, the methodology involves the separate determination of 
base tables and the projection of mortality improvement.
    These regulations also revise the rules regarding substitute 
mortality tables. This revision is being made pursuant to section 503 
of the Bipartisan Budget Act of 2015, which requires that the 
determination of whether a plan has credible information be made in 
accordance with established actuarial credibility theory. Following 
enactment of that requirement, Treasury and the IRS undertook a review 
of actuarial literature regarding credibility theory and consulted with 
experts on that topic from the Society of Actuaries. Based on that 
review and analysis, these regulations set forth a method for 
developing substitute mortality tables that is materially different 
from the method that is required under the 2008 substitute mortality 
table regulations and the associated revenue procedure. The method for 
developing substitute mortality tables that is set forth in the final 
regulations is simpler than the graduation method that applies under 
the 2008 substitute mortality table regulations and also accommodates 
the use of substitute mortality tables for plans with smaller 
populations that have only partially credible mortality experience.

II. Generally Applicable Mortality Tables

A. Base Mortality Tables

    Under these regulations, the generally applicable base mortality 
tables are derived from the tables contained in the RP-2014 Mortality 
Tables Report. In response to Notice 2013-49, commenters generally 
recommended that the RP-2014 mortality tables form the basis for the 
mortality tables used under section 430. After reviewing the RP-2014 
mortality tables, the accompanying report published by the Society of 
Actuaries, and related public comments, Treasury and the IRS determined 
that the experience study used to develop the RP-2014 mortality tables 
is the best available study of the actual mortality experience of 
pension plan participants (other than disabled individuals). As a 
result, Treasury and the IRS issued proposed regulations that use the 
RP-2014 mortality tables as the foundation for the base mortality 
tables used to project the mortality rates of pension plan participants 
(other than disabled individuals).\8\
---------------------------------------------------------------------------

    \8\ Mortality tables that may be used as an alternative to the 
tables provided in these regulations with respect to certain 
disabled individuals are provided in Rev. Rul. 96-7 (1996-1 CB 59).
---------------------------------------------------------------------------

    Most commenters supported the selection of the base mortality 
tables in the proposed regulations. One commenter opposed this 
selection but did not suggest any alternative. Accordingly, the base 
mortality tables under these final regulations are the same as in the 
proposed regulations. Like the mortality tables provided in the 2008 
general mortality table regulations, the mortality tables under these 
final regulations are gender-distinct because of significant 
differences between expected male mortality and expected female 
mortality. In addition, as under the 2008 general mortality table 
regulations, these regulations set forth separate mortality rates for 
annuitants and nonannuitants.
    The base tables that are set forth in these final regulations are 
used to develop the mortality tables for future years. These base 
tables have a base year of 2006 (the central year of the experience 
study used to develop the mortality tables in the RP-2014 Mortality 
Tables Report). These base tables generally have the same mortality 
rates as the RP-2014 mortality tables after factoring out the mortality 
improvements from 2007 to 2014 (calculated using the Scale MP-2014 
rates). However, these base tables also include nonannuitant mortality 
rates for ages below age 18 and above age 80 and annuitant mortality 
rates for ages below age 50. This generally is the same approach that 
was used to develop the base tables included in the 2008 general 
mortality table regulations.

B. Mortality Improvement and Static Mortality Tables

    The proposed regulations, like the 2008 general mortality table 
regulations, provided that expected trends in mortality experience must 
be taken into account through the use of either generational or 
annually updated static mortality tables. In accordance with section 
430(h)(3)(B), the proposed regulations updated the mortality 
improvement rates used under the 2008 general mortality table 
regulations. To select up-to-date mortality improvement rates, Treasury 
and the IRS reviewed the Mortality Improvement Scale MP-2014 Report, 
related public comments, the data sources cited in those comments, the 
Mortality Improvement Scale MP-2015 Report, the Mortality Improvement 
Scale MP-2016 Report, and other published data sources.\9\ After that 
review, Treasury and the IRS issued proposed regulations that applied 
the MP-2016 rates to develop mortality tables for use in 2018.
---------------------------------------------------------------------------

    \9\ See the August 2013 Literature Review and Assessment of 
Mortality Improvement Rates in the U.S. Population: Past Experience 
and Future Long-Term Trends, available at www.soa.org/Files/Research/Exp-Study/research-2013-lit-review.pdf.
---------------------------------------------------------------------------

    Some commenters supported the selection of the mortality 
improvement rates in the proposed regulations, while other commenters 
expressed concerns about the selection of those rates. Those commenters 
who expressed concern about the mortality improvement rates noted that 
the selection of a long-term mortality improvement rate assumption is 
inherently speculative and cautioned against using the assumptions 
regarding the rate of long-term mortality improvement that were used by 
RPEC (which are a long-term rate of 1.0 percent per year for ages 85 
and younger, grading down to 0.85 percent at age 95, and further 
grading down to 0 at age 115). Instead of the RPEC assumptions, these 
commenters suggested that Treasury and the IRS use assumptions 
regarding the rate of long-term mortality improvement that are closer 
to the rates that are used by the Office of the Actuary within the 
Social Security Administration. Those rates also vary by age group, and 
the documentation accompanying the 2017 report of the Board of Trustees 
of the Federal Old-Age, Survivors Insurance and Disability Insurance 
Trust Funds indicates that, under the intermediate assumptions (which 
reflect the Trustees' best estimates of future experience), the 
weighted average over all ages of those assumed long-term mortality 
improvement rates is 0.72 percent per year.\10\
---------------------------------------------------------------------------

    \10\ See ``The Long-Range Demographic Assumptions for the 2017 
Trustees Report,'' Office of the Chief Actuary, Social Security 
Administration (July 13, 2017), at Mortality, page 17 (available at 
https://www.ssa.gov/OACT/TR/2017/2017_Long-Range_Demographic_Assumptions.pdf).
---------------------------------------------------------------------------

    Treasury and the IRS carefully considered the assumptions used by 
Office of the Actuary within the Social Security Administration and 
compared it with the long-term assumptions

[[Page 46392]]

currently recommended by RPEC. In evaluating the merits of each, 
Treasury and the IRS took into consideration the views of the Technical 
Panel of the Social Security Advisory Board. The Social Security 
Advisory Board is an independent federal government agency, and the 
Technical Panel, which is comprised of actuaries, economists and 
demographers, is charged by the Advisory Board with reviewing the 
assumptions and methods used in the annual report of the Board of 
Trustees of the Federal Old-Age, Survivors Insurance and Disability 
Insurance Trust Funds. The Technical Panel, which issues a report every 
4 years, has consistently recommended that the mortality improvement 
assumption used by the Office of the Actuary be increased.\11\ In 
addition, the Congressional Budget Office uses a faster rate of 
mortality improvement in evaluating Social Security solvency than the 
Office of the Actuary.\12\
---------------------------------------------------------------------------

    \11\ See ``2015 Technical Panel on Assumptions and Methods 
Report to the Social Security Advisory Board,'' available at 
www.ssab.gov/Details-Page/ArticleID/656/2015-Technical-Panel-on-Assumptions-and-Methods-A-Report-to-the-Board-September-2015.
    \12\ See Comparing CBO's Long-Term Projections with Those of the 
Social Security Trustees: Hearing before the Subcommittee on Social 
Security, Committee on Ways and Means, U.S. House of 
Representatives, 114th Cong. September 21, 2016 (Testimony of Keith 
Hall), available at https://www.cbo.gov/sites/default/files/114th-congress-2015-2016/reports/51988-socialsecuritytestimony.pdf.
---------------------------------------------------------------------------

    After review and consideration of the comments, the documentation 
accompanying the Trustees' Report, and the views of the Technical Panel 
of the Social Security Advisory Board, Treasury and the IRS have 
concluded that the procedures that RPEC used to develop the Scale MP-
2016 rates generate mortality improvement rates that currently are the 
appropriate rates for use in developing mortality tables to be used for 
purposes of pension funding. Accordingly, these regulations provide 
that the mortality improvement rates for valuation dates in 2018 are 
the Scale MP-2016 rates.
    Treasury and the IRS understand that RPEC expects to issue updated 
mortality improvement rates that reflect new data for mortality 
improvement trends for the general population on an annual basis. As 
noted by the commenters, while the rate of mortality improvement has 
fluctuated significantly on a year-to-year basis, there has been a 
significant reduction in the rate of improvement over the past few 
years compared to the rate of improvement for the past 25 years. RPEC 
has indicated the intent to continually review the methodology used in 
its mortality improvement model in an effort to improve the overall 
effectiveness of the model, especially with respect to year-over-year 
stability and forecast accuracy, and it has identified the assumed 
long-term rate of mortality improvement and graduation techniques as 
two of the items included in this review. In establishing the mortality 
improvement rates to be used under section 430(h)(3) for valuation 
dates in years after 2018, Treasury and the IRS will continue to take 
into account RPEC's updates (including any modifications to RPEC's 
methodology), as well as other sources of data or analyses regarding 
mortality improvement. These regulations provide that the mortality 
improvement rates applicable for those future valuation dates will be 
specified in guidance to be published in the Internal Revenue Bulletin. 
See Sec.  601.601(d)(2)(ii)(b) of this chapter. If Treasury and the IRS 
determine that significant revisions to the mortality improvement rates 
are appropriate, the revisions may first be proposed in a new 
rulemaking in order to allow for public comment before the rule is 
finalized.
    Some commenters asked that Treasury and the IRS commit to providing 
the mortality improvement rates for a calendar year at least 12 months 
before the start of that year. Treasury and the IRS understand that a 
significant motivation for this request is to avoid the issuance of new 
mortality improvement rates in the early part of a calendar year 
(because issuance of new mortality improvement rates at that time could 
result in the need to revise calculations that have already been made 
in the course of preparing a plan sponsor's financial statement as of 
the previous December 31). While Treasury and the IRS intend that the 
mortality improvement rates for a calendar year generally will be 
issued more than 12 months in advance of that year, the final 
regulations do not include a provision requiring that the mortality 
improvement rates for a calendar year be issued within this timeframe. 
Retaining the flexibility to issue mortality improvement rates closer 
to the date they would become effective will allow additional time for 
the possibility that certain revisions to the mortality improvement 
rates will first be published in proposed form.
    Other commenters requested that Treasury and the IRS consider 
updating the mortality tables on a less frequent basis than annually. 
Although the RPEC indicated its intent to issue updated mortality 
improvement rates on an annual basis, the final regulations do not 
require the mortality improvement rates under section 430(h)(3) to be 
updated annually. However, to minimize the discontinuities in mortality 
rates that could arise from infrequent updates, Treasury and the IRS 
contemplate that generally the rates will be updated annually. If the 
changes from one year to the next are minimal, Treasury and the IRS may 
choose not to update the rates for that year.
    As under the 2008 general mortality table regulations, the proposed 
regulations take into account the limitations of some current actuarial 
software that is not designed to use generational mortality tables and 
continue to permit the use of static mortality tables. These static 
mortality tables, when used to determine the present value of an 
annuity, approximate the present value that would be determined using 
the generational mortality tables. All but one commenter supported the 
option to use static mortality tables, and these final regulations 
provide for this option. These static tables consist of separate 
gender-specific tables, which are updated annually. The static 
mortality tables that will be used for 2018 are included in these 
regulations. For later years, updated static mortality tables will be 
set forth in guidance published in the Internal Revenue Bulletin. See 
Sec.  601.601(d)(2)(ii)(b) of this chapter.

C. Use of Section 430 Mortality Tables for Other Provisions

    Section 417(e)(3) generally provides that the present value of 
certain benefits under a qualified pension plan (including single-sum 
distributions) must not be less than the present value of the accrued 
benefit using applicable interest rates and the applicable mortality 
table. Section 417(e)(3)(B) defines the term ``applicable mortality 
table'' as the mortality table specified for the plan year for minimum 
funding purposes under section 430(h)(3)(A) (without regard to the 
rules for substitute mortality tables under section 430(h)(3)(C) or 
mortality tables for disabled individuals under section 430(h)(3)(D)), 
modified as appropriate by the Secretary. The modifications to the 
section 430(h)(3)(A) mortality table used to determine the section 
417(e)(3)(B) applicable mortality table are not addressed in these 
regulations and are currently provided in Revenue Ruling 2007-67 (2007-
2 C.B. 1047).
    As under the proposed regulations, the final regulations provide 
that the same mortality assumptions that apply for purposes of section 
430(h)(3)(A) and Sec.  1.430(h)(3)-1(a)(2) are used to determine a 
plan's current liability for purposes of applying the full-funding

[[Page 46393]]

rules of section 431(c)(6) (in the case of a multiemployer plan) and 
section 433(c)(7)(C) (in the case of a CSEC plan). For this purpose, a 
multiemployer plan or CSEC plan is permitted to apply either the 
annually-adjusted static mortality tables or the generational mortality 
tables.

D. Effective Date

    The proposed regulations provide that the regulations will be 
effective beginning in 2018. Some commenters expressed concern that 
this effective date would not allow adequate time for compliance. One 
commenter requested that the effective date of the regulations be no 
sooner than 18 months after the regulations are finalized in order to 
give plan sponsors adequate time to plan for the higher level of 
contributions that will be required under the new mortality 
assumptions. Treasury and the IRS understand that most employers have 
been planning for the issuance of updated mortality tables for the 
purposes of section 430 since the RP-2014 Mortality Tables Report was 
issued in 2014 and many of those employers are already using updated 
mortality tables for financial reporting purposes. Furthermore, any 
additional required contributions for a plan resulting from the 
adoption of the new tables will not be due before September 15, 2019. 
For a plan with a calendar plan year, this date is 8\1/2\ months after 
the end of the first plan year for which the regulations apply.\13\
---------------------------------------------------------------------------

    \13\ However, new mortality tables may affect plan operations 
relating to the requirements of section 436 during the 2018 plan 
year.
---------------------------------------------------------------------------

    Moreover, as described in section II.C of this Explanation of 
Provisions, the amount of a single-sum distribution computed as the 
present value of an annuity is determined using a mortality table that 
is based on the generally applicable mortality tables used for minimum 
funding purposes. Thus, retaining the mortality tables under existing 
regulations for the 2018 plan year, as requested by some commenters, 
would result in inappropriately depressing the amount of single-sum 
distributions payable to affected participants during the 2018 plan 
year (resulting in a permanent loss of retirement assets for those 
participants). A 2013 study indicates that approximately 56 percent of 
retiring participants in a traditional defined benefit plan with an 
unlimited single-sum option choose that option.\14\ In addition, the 
Government Accountability Office recently issued a study examining the 
increase in ``lump-sum window'' offers--typically, limited time offers 
to participants who are no longer employed by the sponsor but still 
waiting for their pension benefit to begin in the future, or retirees 
already receiving their pension annuity payments.\15\ Similarly, Willis 
Towers Watson reported significant risk transferring activity in recent 
years through lump sum windows and other means.\16\
---------------------------------------------------------------------------

    \14\ Sudipto Banerjee, Ph.D., Employee Benefits Research 
Institute, Issue Brief: Annuity and Lump-Sum Decisions in Defined 
Benefit Plans: The Role of Plan Rules, January 2013, available at 
https://www.ebri.org/pdf/briefspdf/EBRI_IB_01-13.No381.LSDs2.pdf.
    \15\ U.S. Government Accountability Office, GAO 15-74, 
Participants Need Better Information When Offered Lump Sums That 
Replace Their Lifetime Benefits (January 2015). This report also 
notes the substantial financial advantages that exist for plan 
sponsors implementing lump sum windows and attributes the recent 
increase in lump-sum window offers, in part, to the outdated 
mortality tables in the current regulations, which result in reduced 
payments to plan participants. Id. at pp. 16-17.
    \16\ Brendan McFarland, After a few ups and downs, corporate 
pension funding levels showed little change in 2016: Late-year rise 
in interest rates and stock market performance mitigate earlier 
downturn in funded status, Willis Towers Watson Insider, Vol. 27, 
No. 2 (February 2017) at p. 3, available at https://www.towerswatson.com/en-US/Insights/Newsletters/Americas/insider.
---------------------------------------------------------------------------

    Because these rules were proposed in December of 2016 to be 
applicable as final regulations for plan years beginning on and after 
January 1, 2018, the Treasury Department and the IRS believe that many 
plan sponsors have had adequate time to set aside funds needed for 
additional pension contributions for the 2018 plan year. Furthermore, 
because the steps that plan sponsors will need to take to update their 
administrative systems in response to these final regulations are not 
significantly different from the steps they would need to take in 
response to the annual update of mortality tables that has previously 
occurred at this time of the year, the Treasury Department and the IRS 
believe that plan sponsors generally will have sufficient time to make 
any needed changes to these administrative systems.\17\
---------------------------------------------------------------------------

    \17\ See section 4. of the Regulatory Impact Assessment provided 
under the heading Special Analyses, Regulatory Planning and Review 
(Executive Orders 12866 and 13563) in this preamble for a discussion 
of needed changes in administrative systems.
---------------------------------------------------------------------------

    The final regulations generally retain the effective date that was 
proposed, and apply to plan years beginning on or after January 1, 
2018. In response to comments indicating that this effective date may 
create certain administrative or financial difficulties, the final 
regulations provide an option that may be used in certain circumstances 
for the 2018 plan year to apply the regulations that were formerly in 
effect. Specifically, for a plan for which substitute mortality tables 
are not used for the 2018 plan year, mortality tables determined in 
accordance with regulations previously in effect may be used for 
purposes of applying the rules of section 430 for a valuation date 
occurring during 2018 if the plan sponsor (1) concludes that the use of 
mortality tables determined in accordance with the final regulations 
for the plan year would be administratively impracticable or would 
result in an adverse business impact that is greater than de minimis, 
and (2) informs the actuary of the intent to apply this option. While 
this option provides significant flexibility to plan sponsors, the use 
of the option will not affect the mortality table used to determine 
minimum present value for distributions with annuity starting dates in 
stability periods that begin during 2018 (which is based on the 
generally applicable mortality tables under section 430(h)(3)(A) that 
apply if this option is not used). Therefore, the lump-sum 
distributions received by participants retiring in 2018 will 
appropriately reflect their expected longevity.

III. Plan-Specific Substitute Mortality Tables

A. Application of Established Actuarial Credibility Theory

    These final regulations contain a comprehensive revision of the 
rules regarding plan-specific substitute mortality tables for plans 
that are subject to the rules of section 430.\18\ These regulations 
carry over many of the rules regarding substitute mortality tables from 
the 2008 substitute mortality table regulations. However, after 
analyzing the actuarial literature regarding credibility theory, 
Treasury and the IRS made a number of changes to the rules relating to 
the development of substitute mortality tables. Specifically, these 
final regulations, like the proposed regulations, generally require 
that a substitute mortality table be constructed by multiplying the 
mortality rates from a projected version of the generally applicable 
base mortality table by a mortality ratio (that

[[Page 46394]]

is, a ratio of the actual deaths for the population to the expected 
deaths determined using the standard mortality tables for that 
population).
---------------------------------------------------------------------------

    \18\ There is no provision for defined benefit plans that are 
not subject to the requirements of section 430 (such as 
multiemployer plans) to request approval to use of substitute 
mortality tables. However, the mortality tables under section 
430(h)(3) are required to be used for those plans only for very 
limited purposes (and the mortality tables used for those plans for 
most purposes, while subject to the requirements of section 
431(c)(3) or 433(c)(3), are not tables specified by statute or 
regulations).
---------------------------------------------------------------------------

    Use of mortality ratios (rather than providing for the graduation 
of raw mortality rates as under the 2008 substitute mortality table 
regulations) will make it easier for plan sponsors to develop 
substitute mortality tables because it eliminates the need to apply a 
complex graduation technique. It also facilitates efficient IRS review 
of applications for approval to use substitute mortality tables, which 
is particularly important in light of the other major change made in 
the regulations (permitting the use of substitute mortality tables for 
a plan that has mortality experience that is only partially credible). 
As a result of the changes made in these regulations, Treasury and the 
IRS expect that significantly more plan sponsors will request approval 
to use substitute mortality tables.

B. Development of Substitute Mortality Tables for Plans With Full 
Credibility

    The substitute mortality table for a population with full 
credibility must be determined by applying projected mortality 
improvement to a base substitute mortality table developed using an 
experience study of the population. Like the proposed regulations, the 
final regulations use the same general requirements for an experience 
study as under the 2008 substitute mortality table regulations but 
reflect certain changes from the proposed regulations in response to 
comments. Specifically, the experience study generally must cover a 
period of at least 2 (and no more than 5) consecutive 12-month periods 
that ends less than 3 years before the first day of the first plan year 
for which the substitute mortality tables are to apply, and must cover 
the same period for all populations within a plan.\19\ However the 
final regulations include an exception that permits the use of an 
earlier study period if the submission is made more than 1 year (and 
less than 2 years) before the first day of the first plan year for 
which the substitute mortality tables are proposed to apply. Under this 
exception, the last day of the experience study period is permitted to 
be 3 years or more before the first day of the first plan year for 
which the substitute mortality tables are proposed to apply, provided 
that the last day of the experience study period is less than 2 years 
before the application is submitted.
---------------------------------------------------------------------------

    \19\ As under the 2008 regulations and the proposed regulations, 
the final regulations provide for permissive aggregation of plans of 
a plan sponsor for purposes of developing and using substitute 
mortality tables. The final regulations clarify that if two or more 
plans with different plan years are aggregated, the experience study 
may consist of data that is collected over different periods for 
plans with different plan years, subject to certain conditions.
---------------------------------------------------------------------------

    A base substitute mortality table generally is determined by 
multiplying the mortality rates from the corresponding standard 
mortality table (that is, the generally applicable base mortality table 
for the population, projected with mortality improvement to the base 
year for the base substitute mortality table) by the mortality ratio 
for the population. For this purpose, the mortality improvement rates 
that apply for the calendar year during which the plan sponsor submits 
the request for approval to use substitute mortality tables are used to 
project the generally applicable base mortality table to the base year 
for the base substitute mortality table.\20\ The mortality ratio is 
determined as a fraction, the numerator of which is the number of 
actual deaths during the experience study period (with each death 
weighted by the benefit amount) and the denominator of which is the 
number of expected deaths during that period (determined using the 
standard mortality table) weighted by the benefit amount. An 
individual's benefit amount (which is used to determine amounts-
weighted mortality rates and for other purposes in the construction of 
base substitute mortality tables) is the individual's accrued benefit 
expressed in the form of an annual benefit commencing at normal 
retirement age (or at the current age, if later) if the individual has 
not commenced benefits, and the individual's annual payment if the 
individual has commenced benefits. Consistent with section 503 of the 
Bipartisan Budget Act of 2015 (and unlike Sec.  1.430(h)(3)-
2(c)(2)(ii)(D) of the 2008 substitute mortality table regulations, 
which provides that the Commissioner may permit the use of other 
recognized mortality tables to construct the base substitute mortality 
table), the regulations provide that the standard mortality table that 
must be used for this purpose is the generally applicable base 
mortality table projected with mortality improvement to the base year 
for the base substitute mortality table.
---------------------------------------------------------------------------

    \20\ If the plan sponsor submits such a request during 2017, 
then mortality improvement is reflected using the mortality 
improvement rates that generally apply for use for 2018, which are 
the Scale MP-2016 rates.
---------------------------------------------------------------------------

    Some commenters pointed out that multiplying mortality ratios for a 
population by the mortality rates in the applicable standard mortality 
table could yield inappropriate results at extremely old ages. In 
response to those comments, the final regulations provide that 
mortality rates under the base substitute mortality tables must be the 
same as the mortality rates under the standard mortality table for ages 
above 109 and that a modified mortality ratio is used for ages from 96 
through 109 (to accomplish a gradual transition to the standard 
mortality table while avoiding inappropriate results). If the mortality 
ratio for the population is greater than 1.0, the modified mortality 
ratio for an age within this range is equal to the mortality ratio for 
the population reduced by 1/15th of the excess of the mortality ratio 
over 1.0 for each year by which the age exceeds 95. If the mortality 
ratio for the population is less than 1.0, the modified mortality ratio 
for an age within this range is equal to the mortality ratio for the 
population increased by 1/15th of the excess of 1.0 over the mortality 
ratio for each year by which the age exceeds 95.

C. Standards for Full Credibility

    The proposed regulations revised the standard for full credibility 
of a population under the 2008 substitute mortality table regulations 
(which is 1,000 actual deaths for the relevant population during the 
experience study period) to better reflect established actuarial 
credibility theory. Under established actuarial credibility theory, the 
1,000-death threshold (which is a rounding down of the 1,082 actual 
deaths that would be needed for a 90% confidence level that the 
measured rate is within 5% of the underlying mortality rate) should 
apply to the credibility for a single mortality rate and not an entire 
mortality table.\21\ Moreover, the 1,000 death threshold did not take 
into account the well-established actuarial principle that mortality 
experience within a population will vary predictably based on the 
amount of the annuity (or life insurance, as applicable). The base 
tables for the generally applicable mortality tables were constructed 
on an amounts-weighted basis (under which the individuals with higher 
benefit amounts have a greater weight in the computation of the 
mortality rate for a

[[Page 46395]]

particular age); accordingly, substitute mortality tables should be 
constructed using the same principle.
---------------------------------------------------------------------------

    \21\ Although the use of a graduation technique under Rev. Proc. 
2008-62 enables a plan with fewer than 1,000 deaths at each age to 
have credible mortality experience that may be used to establish a 
substitute mortality table, the statutory instruction providing that 
the determination of whether a plan has credible mortality 
information be made in accordance with established actuarial 
credibility theory which is materially different than the rules in 
effect on November 2, 2015, led to the elimination of that 
technique.
---------------------------------------------------------------------------

    Using established actuarial credibility theory to evaluate whether 
a population has fully credible mortality experience entails the use of 
a threshold that takes into account the dispersion of benefits within 
the population. Accordingly, under the proposed regulations, the number 
of deaths that are needed for a population within a plan to have fully 
credible mortality experience is determined as the product of 1,082 and 
the benefit dispersion factor for the population.\22\ The benefit 
dispersion factor for a population is equal to the number of expected 
deaths for the population during the experience study period, 
multiplied by the sum of the mortality-weighted squares of the 
benefits, divided by the square of the sum of the mortality-weighted 
benefits.\23\
---------------------------------------------------------------------------

    \22\ This formula for the number of deaths needed for full 
credibility is based on the assumption that the distribution of 
releases from liability due to deaths follows a compound Poisson 
model. See www.actuaries.ca/members/publications/2002/202037e.pdf.
    \23\ See Gavin Benjamin, Selecting Mortality Tables: A 
Credibility Approach, available at www.soa.org/Files/Research/Projects/research-2008-benjamin.pdf.
---------------------------------------------------------------------------

    Commenters supported the actuarial soundness of the standard for 
fully credible mortality information under the proposed regulations, 
and the final regulations adopt the provisions of the proposed 
regulations regarding full credibility. At the request of commenters, 
the regulations include expressions of various formulas in mathematical 
notation to assist actuaries in making computations under the 
regulations.\24\
---------------------------------------------------------------------------

    \24\ In the proposed regulations, these formulas were stated as 
amounts to be computed separately for each age and then summed for 
the population. The final regulations instead state these formulas 
more concisely as amounts to be computed for the entire population. 
These two approaches yield mathematically identical results.
---------------------------------------------------------------------------

    One commenter noted that the increase of the threshold for full 
credibility (together with the inability to reflect the pattern of the 
plan's mortality experience at different ages) may produce substitute 
mortality tables that are substantially different than those that are 
currently in use. To address this concern, in part, the final 
regulations include an option to increase the credibility of a plan's 
mortality experience by basing it on the combined mortality experience 
of both genders.\25\
---------------------------------------------------------------------------

    \25\ This option is described in section III.F.2 of this 
Explanation of Provisions (Option to use combined male and female 
mortality experience).
---------------------------------------------------------------------------

D. Partial Credibility

    As under the proposed regulations, the final regulations permit 
substitute mortality tables to be used for a plan that does not have 
sufficient deaths to have fully credible mortality information. In 
accordance with established actuarial credibility theory, the 
substitute mortality table used for such a plan is the weighted average 
of the standard mortality table and the substitute mortality table that 
would be developed for the plan if it were to have fully credible 
mortality information. The weight for the substitute mortality table 
that would be developed for the plan if the plan were to have fully 
credible mortality information is the square root of a fraction, the 
numerator of which is the actual number of deaths for the population 
within the experience study period and the denominator of which is the 
number of deaths needed for the plan to have fully credible mortality 
information.

E. Controlled Group Consistency Requirement

    Under section 430(h)(3)(iv) there is a general consistency 
requirement for the use of substitute mortality tables with respect to 
all plans within a controlled group. Thus, use of substitute mortality 
tables for a plan is generally permitted only if substitute mortality 
tables are used for all plans subject to section 430 that are 
maintained within the controlled group of the plan sponsor.
    The 2008 substitute mortality table regulations set forth an 
exception from this consistency requirement for plans that did not have 
credible mortality experience. As a result of the change permitting the 
use of substitute mortality tables for plans that have only partially 
credible mortality information, Treasury and the IRS concluded that the 
exception should be modified so that it only applies to plans with a 
relatively small population. Accordingly, the regulations provide that 
a population does not have credible mortality information (and so a 
substitute mortality table is neither permitted nor required to be used 
for that population) if the actual number of deaths for that population 
during the experience study period is less than 100. For this purpose, 
the length of the experience study period must be the same length as 
the longest experience study period for any plan in the controlled 
group \26\ and must end less than 3 years before the first day of the 
first plan year for which the substitute mortality tables are to apply. 
Treasury and the IRS proposed the use of a threshold of 100 deaths to 
balance the benefit of the use of substitute mortality tables for a 
plan with a relatively small population (which would be small, given 
the relatively low weight assigned to that plan's partially credible 
mortality experience) against the burden of developing substitute 
mortality tables for that plan (which would be required to comply with 
the controlled group consistency requirement). No comments were 
received objecting to this threshold or recommending a different 
threshold. As a result, the final regulations adopt the proposed 100-
death threshold.
---------------------------------------------------------------------------

    \26\ If a plan has credible mortality information for one gender 
but not for the other gender, the length of the period of this 
demonstration is the length of the experience study for that plan.
---------------------------------------------------------------------------

F. Other Rules Relating to the Use of Substitute Mortality Tables

1. Multiple-Employer Plans
    In response to comments, the final regulations provide rules 
regarding the use of substitute mortality tables in connection with 
multiple-employer plans. Under the final regulations, the application 
for use of substitute mortality tables in the case of a multiple-
employer plan must be made by the plan administrator, and the 
substitute mortality tables must apply on a plan-wide basis (even if 
the plan is subject to the rules of section 413(c)(4)(A)).
    In addition, the final regulations provide special rules for the 
application of the controlled group consistency rule in the case of a 
multiple-employer plan. Under this special rule, an employer 
participating in a multiple-employer plan is treated as maintaining 
that plan if and only if the proportion of the plan's funding target 
attributable to the employees and former employees of the employer and 
members of the employer's controlled group is greater than 50 percent. 
Thus, such an employer is subject to the controlled group consistency 
rule with respect to the multiple-employer plan and any other plans 
subject to section 430 maintained by that employer (or any member of 
that employer's controlled group). By contrast, if the proportion of 
the multiple-employer plan's funding target attributable to the 
employees and former employees of the employer and members of the 
employer's controlled group is less than or equal to 50 percent, then 
that employer is not subject to the controlled group consistency rule 
with respect to the multiple-employer plan and any other plans subject 
to section 430 maintained by that employer (and any member of that 
employer's controlled group).

[[Page 46396]]

2. Option To Use Combined Male and Female Mortality Experience
    Some commenters requested the ability to develop and use substitute 
mortality tables based on the combined experience of both males and 
females in the plan, to increase the credibility of mortality 
experience for a smaller population. Treasury and the IRS have 
determined that this approach is consistent with established actuarial 
credibility theory. Accordingly, the final regulations provide that a 
single mortality ratio may be developed for both genders and then used 
to construct separate gender-specific base substitute mortality tables 
for the plan. If this option is applied for a plan, then substitute 
mortality tables used for all pla