Analysis of Actuarial Valuation Report
Appraising the community interest in a pension plan upon dissolution of marriage is not a well-defined process. It is necessary to know what series of benefits is being valued.
- Is it the series of monthly benefits expected to be paid starting at the earliest retirement age on the assumption the employee terminated service now?
- Is it the monthly benefit expected if the employee continues in service and does not receive any salary or benefit level increases?
- Or is it the monthly benefit expected if the employee continues in service and receives salary and benefit increases?
- What age is assumed for retirement?
- If the plan has a provision for a post-retirement cost-of-living adjustment, has this benefit been valued?
- Has the valuation been made as of the date of separation without apportionment between separate and community property, or as of the date of trial with the community interest apportioned between separate and community property?
- What technique was used to make the apportionment?
- What assumptions were used?
- Are the assumptions consistent?
If the life expectancy method has been used, be careful! Use of the life expectancy method suggests that a quasi-actuary has made the calculations. The method has been widely condemned by both actuaries and courts. Properly used, the life expectancy method can sometimes produce valid approximations; however,
a review of reports made by quasi-actuaries using this technique more often than not discloses mistakes. For example, in several cases the
actuarial present value developed has been the value at retirement, not the value now (the actuarial present value). In other cases, the
technique has been applied in a manner that assumes all the monthly pension payments are held back and paid at the time of death.
It is popularly believed that the expectation of life is widely used in actuarial calculations. In reality, it is of interest to actuaries as an index for comparing different mortality tables. The life expectancy method is unnecessary, improper, and indefensible, but unfortunately
too firmly entrenched in legal and judicial publications for removal and obliteration. Inadequate data and benefit documentation can seriously impair the validity of an actuarial valuation. The available booklets describing pension plan provisions are occasionally incomplete, ambiguous,
or out of date. The actuary is dependent on his attorney-client to obtain current plan information and accurate statistical data on which to base a valuation.
A. Actuarial Value of What Benefits?
1. Introduction
a. Community Property Defined
Community property is property acquired by either spouse during marriage as a result of community property labor or purchase with community property funds. Benefits resulting from service under a pension plan during marriage qualify as community property. Code #5118 (now Family Law Code #771).
b. “Community” times “Separate” — Character ?
What is the character of property whose value is dependent upon the product of community property and separate property? Part are a function of service, which is community property, and salary above the salary at the time of marriage separation which is separate property. Typically, if employment continues after marriage separation, benefits for all years of service will be based upon a salary
earned postseparation that is greater than the salary at the time
of marriage dissolution. The excess salary is, by definition,
separate property. The service for the community period is, by
definition, community property. Considered alone, neither the
excess salary nor the community service has any value for pension
purposes. However, taken together, a value is created. What is
the character of a hybrid asset that requires both separate property
(service) and separate property (postseparation salary increase)
to have value? The purpose of this section is to show that many
of the determinations necessary for a conclusive answer to this
question have not been answered and to argue that this hybrid
benefit is community property and should be valued for community
property purposes.
c. Equal Division Requirement
The court is required to make an equal disposition of community
property between the spouses. For pension benefits, the court
can select between the approaches of retained jurisdiction and
actuarial cash-out. If, in practice, these different approaches
are allowed to produce results of differing values, can both remain
valid given the requirement of an equal disposition?
(1) Under Retained Jurisdiction
If the court decides to distribute the community interest in
the benefits as each pension payment is made (retained jurisdiction)
and if employment continues, the law is clear that the community
shares in benefit increases resulting from post separation salary/benefit
increases.
(2) Under Cash-Out
Should the dissolution actuary, in making a valuation for cash-out
purposes, value the benefit stream expected to be received if
the court selects the alternative retained jurisdiction method
of disposition? In other words, should the dissolution actuary
develop an equivalence in value between the approaches? Or is
the assumption (that the actuarial cash-out approach should
value the same stream of benefits expected if the court elects
to retain jurisdiction) simply arbitrary and without foundation?
Is the valuation for cash-out purposes to be limited to a consideration
of what is available if the employee spouse terminates employment
now? Is the cash-out an alternative that the non-employee spouse
can choose in lieu of receiving an equal share of the community
property asset?
(3) Some Viewpoints
Some commentators, in discussing the considerations supporting
the actuarial approach from the viewpoint of the employed spouse,
state that using the actuarial approach precludes the non-employee
spouse from sharing higher pensions resulting from increases
in pay. For example, in the Treatment of Retirement Benefits
from the syllabus of the Annual Los Angeles Bar Association
Family Law Section Symposium, Judge Richard E. Denner states
that there is no authority statute or case for the use of projected
salary in the calculation of the actuarial present value of
retirement benefits. This interpretation means that the determination
of the actuarial present value of retirement benefits (cash-out
value) is independent of what is expected to occur if jurisdiction
is retained. In other words, the actuarial or cash-out approach
calls for the valuation of a different stream of benefits from
those involved under the alternative approach to disposition
of retained jurisdiction. Under retained jurisdiction, the law
is well established that the community shares in benefit enhancements
due to future salary/benefit increases. If the court can select
among different approaches to valuation, how can it be possible
for these different approaches to produce results of different
value? On the other hand, if the actuarial or cash-out approach
is supposed to develop the best estimate of the present value
of the future benefit stream expected if the court retains jurisdiction,
then all factors, including future salary or benefit increases,
should be considered in the calculations. The effect of future
salary or benefit increases on an anticipated basis should be
part of the actuarial present value calculation, and the community
should receive consideration for benefits resulting from these
anticipated increases.
2. Footnote 9
Gillmore [1981] The purpose of this note is to provide some actuarial
background that may result in a better understanding of both footnote
9 of Gillmore [1981] and the actuarial valuation process. Footnote
9 reads as follows: “The non-employee spouse, of course, cannot
have it both ways. The decision to ask for distribution of the
retirement benefits before the employee spouse actually retires
constitutes an irrevocable election to give up increased payments
in the future which might accrue due to increased age, longer
service, and a higher salary”. (In re Marriage of Luciano supra
104 Cal App.3d at p.961, citation omitted.) Thus, if Vera chooses
to receive her share of the retirement benefits immediately, she
will forfeit her right to share in the increased value of those
benefits in the future. If the Court retains jurisdiction and
supervises the disposition of the future payments, then it is
operating after-the-fact with actual, known facts. The person
is either alive to receive the benefit payment or is not alive.
Both the amount and occurrence of the benefit payment is known.
On the other hand, in the cash-out, or actuarial present value
process, the actuary does not know in advance the amount of the
payment or whether the person will be alive to receive the payment.
The actuary must substitute after-the-fact known facts with before-the-fact
assumptions. For a retirement plan to pay benefits, it is necessary
that certain conditions be met. Typically, for payments of basic
retirement benefits, it is necessary for the person to be alive.
After-the-fact, this event is either a “yes” or “no” situation;
before-the-fact, this is a contingent event. Dealing with the
consequences of contingent events is a hallmark of the actuarial
profession. An actuary treats a contingent event as a probability
and weights the consequence of the event by the probability of
its occurrence. For example, if the payment of $100 is contingent
on an event that has one chance in four of happening, the probability
is 25% and the expected, or actuarial value, of the payment is
25% of $100, or $25. This expected value is then discounted from
the time of payment back to the date of appraisal, and the result
is known as the actuarial present value for that particular benefit
payment. The actuarial present value process replaces after-the-fact
knowns with before-the-fact estimates. Each factor involved after-the-fact
has its before-the-fact counterpart. If, after-the-fact, the magnitude
of a pension payment is based upon a compensation period that
occurs after the Gillmore election, then the actuarial present
value process replaces this after-the-fact compensation period
with its before-the-fact estimate. If the implied “actual” is
added, Footnote 9 would read: “The non-employee spouse, of course,
cannot have it both ways. The decision to ask for distribution
of the retirement benefits before the employee spouse actually
retires constitutes and irrevocable election to give up increased
payments in the future which might accrue due to actual increased
age, actual longer service, and an actual higher salary”. (In
re Marriage of Luciano supra 104 Cal App.3d at p.961, citation
omitted.) Thus if Vera chooses to receive her share of the retirement
benefits immediately, she will forfeit her right to share in the
increased value of those benefits in the future. With the implied
“actual” added, the intent of the footnote becomes clear. If the
non-employee spouse elects to receive a stream of benefits now
(and this stream of benefits is reduced to an actuarial present
value), the non-employee spouse in the actuarial present value
process already has received benefit due to assumed increased
age, assumed longer service, and assumed higher salary. The non-employee
spouse cannot have “it” both ways and receive additional increases
based on actual events; however, the non-employee spouse should
have “it” one way or the other–either after the fact (retained
jurisdiction( or before the fact (actuarial cash-out). It is clear
that if the retained jurisdiction approach is used, the community
receives consideration for benefits resulting from actual future
salary or benefit increases. Obviously, the community should not
receive benefit from both assumed and actual future salary increases.
To do so would give the community two bites of the same apple.
However, if the apple is there to bite, and it is clear that it
may be there, the community should get a bite regardless of the
approach used for disposition. Under retained jurisdiction, the
bite comes from the result of actual salary or benefit increases.
Under the actuarial or cash-out approach, the bite comes from
the results of assumed future salary or benefit increases.
3. Benefit Stream to Reduce to Present Value
What monthly benefit stream should the dissolution actuary reduce
to present value? Brown [1976] held that all pension benefits,
whether vested or nonvested, earned during the community period
are community property. Vesting, as used in a marriage dissolution
proceeding, occurs when the employee is certain to receive benefits,
whether employment is continued, although survivorship to a particular
age may be a further requirement before actual payment. Brown
directed the trial court to consider the possibility that . .
. termination of employment may destroy (pension) rights before
they mature. In our hypothetical case, the Gillmore [1981] situation
would occur at employee’s age 50 (benefits matured), when the
employee could either retire with a monthly benefit stream of
$400 or continue employment until age 55, when a monthly benefit
stream of $825 would be available to the community. In this situation,
should the analysis be limited to the monthly benefit stream of
the $400 available and conclude that (since the non-employee spouse
has an immediate right to the actuarial present value of this
benefit stream) no consideration should be given to the monthly
benefit stream of $825 available to the community at age 55 if
employment continues? The court does not have the discretion to
defer the start of distribution of the community interest until
the employee spouse reaches age 55, if the non-employee spouse
exercises Gillmore rights at the time of the dissolution. However,
if the non-employee spouse does not exercise this right, and the
court is left with the discretion to dispose of the pension benefits,
should the court limit its consideration to the actuarial present
value of the matured benefit available? At employee’s age 50,
even though the Case II benefit is mature, the Case III situation
remains as a nonvested benefit. Should this Case III benefit be
considered by the court per Brown [1976]?
a. Hypothetical 1
(1) Disposition 1 If employment ends at age 40, the community
divides a monthly benefit stream of $200.00 starting at age
50, the earliest possible retirement age; each payment is contingent
upon the survival of the employee spouse.
(2) Disposition 2 If employment continues to age 50, the community
divides a monthly benefit stream of $400.00 starting at age
50, with each payment contingent upon the survival of the employee
spouse.
(3) Disposition 3 If employment continues to age 55, the community
divides a monthly benefit stream of $825.00 starting at age
55, with each payment contingent upon survival of the employee
spouse.
b. Hypothetical 2
Consider an employee, now age 54, who has 29 years of service,
all during the community period, covered by a plan with a monthly
benefit formula of 2% per year of service times high-3 salary
(the highest average monthly salary over a three year period)
with deferred retirement at age 62, or longevity retirement
at age 55 with 30 years of service. The monthly benefit available
now is a deferred benefit, starting at employee age 62, of 58.00%
of high-3 salary. The benefit is vested but not matured.
If the non-employee spouse wants his/her share now, should
the actuarial analysis be limited to the benefit available now
here, a monthly benefit stream of 58.00% of salary starting
at employee’s age 62? Or should the nonvested longevity benefit
that starts at age 55 be valued with discounting to reflect
the possibility that 30 years of service will not be completed?
Under retained jurisdiction, if employment continues the one
additional year, the community will have a 29/30ths interest
in a monthly benefit of 60.00% of the member’s high-3 salary
starting at employee’s age 55. If the non-employee spouse wants
his/her share now, is the community value limited to what is
available now, in this case, the monthly benefit stream starting
at employee’s age 62?
Or should the court follow the direction of Brown and consider
the actuarial present value of the community benefit stream
starting at age 55, with some reduction for the possibility
that the additional year of service may not occur? Since benefits
in this hypothetical case are based on high-3 salary, if employment
continues an additional year, how should the high-3 salary on
which the benefits are based be estimated? Should the high-3
salary be calculated on the assumption that the next year’s
salary will be the same as the current salary? Or should an
estimate be made of the salary that is expected to be paid,
if the additional year of service is worked? Or should the average
salary at the time of trial, an assumption that generally implies
a decrease in future salary, be used?
c. Hypothetical 3
Ask your actuary or pension plan appraiser for an opinion on
the following hypothetical case:
“I have a client, age 40, who has been covered under the 2.0%
at 50 California Highway Patrol formula for the past 20 years.
My client is earning $2,500 per month. If employment of my client
continues until age 50, my client expects to receive salary
increases of 5.0% per year. My client has an offer to change
employment. The retirement formula of the potential employer
is the same as the retirement formula for the CHP. Should my
client give any consideration to the economic consequences of
changing employment with respect to the 20 years of service
already invested in the retirement program?”
Without exception, any actuary will agree that if employment
continues and salary increases are granted, the 20 years of
service already worked have additional value. Why then, if the
hypothetical case is modified to reflect a family law situation,
should the conclusion of the actuary change? If the basic idea
is that all benefits that flow from community service are community
property and subject to valuation, why then should the benefit
based upon community service and future salary increases that
result from continued employment be ignored?
B. Valuing the Pension Benefits
After determining the contingent (actuarial), interest (economic)
and the monthly benefit amounts, the dissolution actuary combines
them to determine the value of the benefits.
1. Valuation Date
Generally, the valuation date should be as close to the date of
trial as possible unless an alternate valuation date has been approved.
See Civil Code #4800 (now Family Law Code #63), Marsden [1982],
and Hayden [1981]. Shattuck [1982] suggests that valuation be made
as of the date of marriage separation. However, to be consistent
with Gillmore [1981], Civil Code #4800 (now Family Law Code #63)
and sound actuarial principals, Shattuck [1982] needs to be interpreted
as having the community interest stop at the date of marriage separation.
That is, the community interest has to be apportioned out from the
total interest.
2. Two Scenarios
As a general practice, the dissolution actuary should consider
two scenarios if the participant spouse is currently employed. The
first scenario considers that employment may terminate now, and
the second that employment may continue until the participant is
eligible for retirement. If the participant has terminated employment
with a vested right to a deferred pension benefit, is eligible to
retire, or has already retired, consideration is limited to one
scenario.
a. Employment Termination Now
The first scenario assumes employment will terminate now. The
benefit valued is the deferred pension benefit payable at the
earliest possible retirement age permitted by the pension plan.
As the service on which the benefit is based ceases, no projection
of salary or benefit level is made.
b. Employment Continues Until Retirement
The second scenario assumes that employment will continue until
the earliest date possible for retirement and values a pension
benefit (longevity pension benefit) projected to this date. The
pension benefit is based on the assumptions that the participant
continues in service and receives increases in either salary or
benefit level. The concept supporting the second scenario has
been called the momentum concept and is described in Adams [1976],
Judd [1977], Anderson [1976], and Andreen [1978].
3. Momentum Concept
Momentum generates an extra community interest which is the difference
between the value developed in the second scenario and the value
developed in the first scenario after adjustment for nonvested benefits.
The difference in value represents the potential increase in the
value of the community interest due to postseparation efforts built
on the foundation of community effort. It may be reasonable to assume
that the community has an interest in this extra value proportional
to service already rendered to total career service. Example Actuarial
present value of benefits if employment continues to retirement,
the longevity or projected benefit $100,000 Actuarial present value
of benefits if employment terminates now, the deferred retirement
benefit. $70,000 Additional value resulting from continued employment
“momentum” $30,000 Years Already Worked 20 Future Years to Retirement
10 Total Years 30 “Time-rule” apportionment of $30,000 additional
value to community interest $20,000
4. Justification of Momentum Concept
The momentum of a pension benefit that arises from effort during
the community period can be likened to the building of goodwill
in a business, particularly in the case where the pension benefit
is based on a final-pay formula. The benefits based on the period
of employment during the community period are enhanced by the future
final salary. Even for non final pay plans, the community interest
is enhanced during periods of inflation, because if the employee
remains in service to retirement the employer will often increase
the benefit level, either directly or indirectly, to offset inflation.
C.
Assumptions
1. Economic Assumptions
a. Interest
An interest rate for discounting to present value is required.
The trial court need not accept the interest rate used by the
expert, but may have the value computed using an interest rate
the court finds appropriate — Bergman [1985]. Sheldon [1981]
held the trial court determines the interest rate to be used,
based on the credibility of experts. (1) PBGC and annuity purchase
rates Pension Benefit Guaranty Corporation (PBGC) interest and
mortality tables are often used by actuaries outside California
and an argument can be made for using current annuity purchase
rates, typical California practice is to assume a level interest
rate for discounting purposes.
b. Post-Retirement Increases for Cost of Living
Any automatic post-retirement cost-of-living increases provided
by the plan will be paid. However, a maximum to the post-retirement
cost-of-living needs to be made by the actuary.
c. Pre-Retirement Benefit or Salary Increases
Before retirement, projected benefits are based on an assumption
that future salary increases at 3.00% per year. If projected benefits
are not based directly on salary, an assumption that the benefit
level per year of service will increase at 3.00% per year is made.
2. Actuarial Assumptions
a. Mortality Rate
“The mortality basis selected shall be one considered representative
of the experience of the pension plan members in general, modified
if appropriate to reflect a medically determinable impaired state
of health of the particular plan member. Tables commonly referred
to as “population” mortality tables normally are not appropriate,
whether with respect to a plan member, former plan member or a
spouse. The table should be a sex-distinct table. Unisex mortality
is not permitted.” — from Standard of Practice for the Computation
of the Capitalized Value of Pension Entitlements on Marriage Breakdown
for Purposes of Lump-Sum Equalization Payments of the Canadian
Institute of Actuaries. Pension mortality tables are more appropriate
than general population mortality tables for determining the community
interest in pension plans on dissolution of marriage because they
reflect the expected mortality rates of employed persons rather
than those of the general population. It should be noted that
many pension plan appraisers from other, non actuarial professions,
usually use the less appropriate population (United States Life)
tables. Other pension plan appraisers, also from non actuarial
professions, use annuity purchase rates. Such rates include insurance
agent commisions, allowance for anti-selection against the company,
premium taxes, and insurance company expenses. Pension (annuitant)
mortality tables often used for calculations in dissolution of
marriage cases are:UP-94 Table
(1) UP-94 Table
The 1994 Unisured Pensioner Mortality Table updated the UP-84
Mortality to reflect the decreased rates of mortality (that
is, improved mortality), indicating that the current mortality
for uninsured pensioners based on data from 1985 through 1989
was about 14 to 18% less than that expected under the UP-84
Mortality Table. The UP-94 Table is one of the first mortality
tables to factor in generational mortality, which recognized
the trend of mortality improvement and dynamically projects
and incorporates those improvements.
(2) 1983 GAM (Group Annuity Mortality) Table
Developed specifically for use in the valuation of pension
plans, the 1983 GAM Table replaces the 1971 GAM table;
(3) UP-1984 Table
The Unisex-Pension 1984 Table was based on the mortality experience
of twelve large uninsured pension plans of which female experience
representated about 20% of the pensioner death experience. A
one-year set forward in age represent the mortality experience
of an all male group. Likewise, a five year age setback may
be considered appropriate for an all female group. Use of an
age setback is an actuarial technique to modify an existing
mortality table by using mortality rates for a younger (setback)
age than the actual age involved.
(4) CalPERS Table
This mortality table, based upon the experience of the California
Public Employees Retirement System, was developed specifically
for use in the valuation of the CalPERS plan.
b. Sex
Verlinde [1987] accepted actuarial values based upon sex-distinct
mortality rates and held that a wife is not denied equal protection
of the law when the court considers her statistically greater
longevity.
c. Poor Health of Either Spouse
(1) Modify the Mortality Table
Modification of the mortality table was approved in Bergman
[1985] and Shattuck [1982]. If either spouse is in poor health,
a special mortality table, a modification of an existing mortality
table or a cash-flow analysis is indicated. Shattuck [1982]
held that the court will take into account the “inherent uncertainties
such as the parties’ probable life expectancies and their health
… embrace everything that has reasonable relevancy…”.
(2) Life Underwriter v. Medical Doctor
An insurance company risk underwriter (not an insurance salesperson)
is the professional person best qualified to review medical
records and advise the actuary on appropriate modifications
to an existing mortality table. The typical medical doctor does
not have the statistical experience necessary to make a valid
assessment of the effect of medical impairments on longevity.
Most dissolution actuaries have established a working relationship
with a life underwriter for reviewing medical records. Generally,
there is no need for the underwriter to testify as it is customary
for an actuary’s opinion to based on the advice provided by
the underwriter.
(3) Other Considerations
Shattuck [1982] ordered the taking into account the “inherrant
uncertanies such as the parties’ probable life expectancies
and their health, and the fact that an early death might preclude
any pension payments at all. It will embrace everything that
has reasonable relevacy to the expectany’s present value.”