Universal/Variable Life Insurance versus Similar Unbundled Investment Strategies

Author(s): Stephen P. D'Arcy and Keun Chang Lee

Source: The Journal of Risk and Insurance, Vol. 54, No. 3 (Sep., 1987), pp. 452-477
Published by: American Risk and Insurance Association
Stable URL: http://www.jstor.org/stable/253362 .
Accessed: 08/05/2014 22:53

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

American Risk and Insurance Association is collaborating with JSTOR to digitize, preserve and extend access
to The Journal of Risk and Insurance.

http://www.jstor.org

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
 UniversallVariable Life Insurance Versus
Similar Unbundled Investment Strategies
Stephen 1'. D'Arcy
KeuntChang Lee

A IBSTRACT
Universal/variable life insurance combines the tax advantages of cash value life
insurance wvithinvestment in money market, hond, or equity funds. lespite expense
loatinigs and( stirsunder charges o(n univers.tl/variable life policics. tIhistax treatment
often generates a gi-mear aIter tax retuirnthan alternativc investm;ientstrategies. This
paper provi(les a method for calculating relative after tax proceeds for
universal/varinble life and comparable investment strategies based on the provisions of
the Tax Reform Act of 1986. In general, uaniversal/variabiclife insurance policies inust
he kept ini force for at least eight years before providing a greater return than
comparable investmsenltstrategies.

Inltrotullction
Universal life itisuratice, introduced in 1979, and universal/variable (also
known as flexible pretnium variable) lire insurance, approved by thleSecurities
and Exchange Commtiiissioniti November, 1984, provide thle tax sheltered
treatment of investment earnings inherent itt cash value life insurance policies
with the insured retaining somneor all of thieinvestment risk. In both policies
tlheinvestment mediun is similar to that offered to non-insuratncepurchasers.
In universal lire policies, the cash value is invested in a floating rate fund
similar to a nmoneymarket fund; for universal/variable life insurance policies,
thie c.ash value can be invested in any or a variety of alternatives generally
including stock market, long termibond, and money mnarketrtunds.

'Slephet IY'Arcyis ani Assistant Professor in the University of Illinois at Urbania-Chanmpaign

Department of Finance. I le also earned a lh degree at tlheUniversity of Illinois and is a Fellow
of the Casualty Actuarial Society. lr. D'Arcy has previously published numerous articles in this
Journal, the Proceedings of tMleCasualty Actuarial Society, the Journal of Insuratnce Regulation,
and the Journal of Economics antidBusiness. tie is also a co-author with Neil Doherty of an S. S.
Huebner Foundation monograph entitled The Financial Theory of Pricing Property-Liability
Itisurance Contracts.
Keun Chang Lee is a doctoral student in finance at tlie University of Illinois at
Urbana-Champaign.

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
 I a ab/l Li/c In..11
UnIvriw/ jIll ,nlr( 453

The typical universal lile policy includes an expcilsC loading, either flat rate
or as a percentage of pzremiums, and an insurance charge based onl the
insured's mortality risk, with tile remainder invested in a cash value account
that earns a floating rate of interest. Some universal life policies include
surrender charges, either in lieu of or in addition to, front end loads. Thle
surrender charges reduce over timie to encourage policy retention. Premiums
for all universal life policies are not predetermined; within fairly wide limits
the insured has flexibility in the amount of premiums paid. Changes in short
ternminterest rates, in theory, directly afflect the return on tile policy's cash
value. In practice, some universal life insurers invest the proceeds in longer
term assets and credit the policy with the coupon rates of return achieved,
rather than the total rate of return which would include gains or losses on
investments. This practice is anl attractive competitive strategy while interest
rates are falling, but would generate an uncompetitive, from tile interest rate
standpoint, product when interest rates rise. Insurers that do not match assets
to liabilities risk incurring substantial investment losses on universal life
insurance. However, for the policyholder, interest rates will fluctuate based on
market conditions similar to money market fund investments. Death benefits
on universal life policies equal either thle initial face value of tile policy or the
face value plus any cash value. Policyholders can borrow flunds from universal
life insurance policies as fromttraditional cash value life insurance. Thlieloans
are niotconsidered taxable income. However, interest paid on policy loans for
universal life insurance is not an allowable tax deduction. Due to the
indeterminate nature of' universal life insurance premiums, the requirement
that four of the first seven premiums be paid in full cannot be met [121.
In addition to transferring some investment risk to policyholders, universal
life insurance also requires the policyholder to assume the risk of shifts in
mortality rates. The mortality charges are priced in a similar fashion to
indeterminate premium annual renewable termi insurance, with a current,
nonguaranfteed , rate struc( irc subject to guaranteed maximums. I1fmortality
improves, the mortality charges can be, but are not required to be, lowered.
The current mortality charges can be raised if mortality experience worsens or
for other reasons. For example, rather than lowering tile highly visible interest
rate on a universal life insurance policy, anl insurer could raise the mortality
charges, which are less noticeable and mioredifficult to compare.
A number of other differences amiong, insurers exist on universal life
insurance. Some insurers credit tile current interest rate on the entire
unborrowed cash value, whereas others pay that rate only on balances in
excess of stipulated levels. Thle rate of return credited on policies is based on
an external index for some insurers, based onl portfolio performance for
others and set by tile Board of Directors in still others. For some universal life
insurancepolicies, the interest rate charged on loans is predetermined and tile
cash value that is collateral for such loans earns a predetermined, but lower,
rate. Onl other policies the loan rate fluctuates above, but in line with, tile

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
454 The Journal of Risk and Insurance

credited intcrest rate. All universal life insurance policies include guaranteed
minimium interest rates that vary from 3 to 6 percent.
Universal/variable life insurance policies are similar in structure to
universal life with a wider array of investment options and no minimum
guaranteed rate of return. They differ fromtvariable life policies considerably,
notably in the discretionary prenmiumilevels, the distinct expense loadings, and
the term insurance rate structure for the mortality risk. All investment
choices, equity funds, bond funds, and specialized investment pools, are
similar to investments generally available to the public outside of a life
insurance policy, although competing investments do not have tile same tax
treatment. Unlike standard insurance accounting that uses amortized, rather
than market, values and coupon, rather than total, rates of return, tile
performance of the investment funds associated with universal/variable life
insurance is consistent with other investment funds. As universal/variable life
insurance encompasses the basic features of universal life, with additional
investment options, the term universal/variable will be used to apply to both
policy types.
The tax advantage of life insurance policies becomes increasingly important
the longer the policy is kept in force. Taxes on investment earnings are
deferred until tile cash value is withdrawn. If the policy is surrendered for tile
cash value, only the excess of cash value over all premiums paid is taxable;
investment earnings that are offset by expense loadings and insurance costs
are never taxed. If the cash value is paid as part of the death benefit, no
income tax is payable onl any investment earnings accrued prior to death.
Since the tax advantage of life insurance policies increases with tile holding
period of tile policy, there is often a specific holding period after which
investment in tile universal/variable life insurance policy dominates an
alternative unbundled investment strategy without the life insurance tax
advantage. Policies held for shorter periods underperform alternative
investments, primarily due to the expense loading inherent in the life
insurance policies.
In this paper universal/variable life insurance is compared with a number of
alternative strategies. The alternative investment strategies involve purchasing
term insurance and investing the difference in money market flunds, bond
funnds, equity flnids, deep discount bonds, deferred annuities, municipal
bonds, or through an individual retirement account in a money market, bond,
or equity fund. The specific tax advantages of each investment option are
explained and included in the analysis to determine the optimal investment
strategy based on tile values of the parameters and the holding period.

Literature Review
Prior to the development of life insurance policies that left the investment
risk with tile insured, analysis of life insurance purchase decisions and
competing investment alternatives (buy term and invest the difference)

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
 Universal! Variable Life Insurance 455

compared anl interest rate guarantee against a hypothetical investment return

[4, 8, 14, 15 pp. 135-45, 17, and 191. Variations onl investment rates of return
affected one side of the equation only. More recently, Myers and Pritchett [181
examined the rate of return over 20 years onl differential premiums between
those paid onl participating and nonpzarticip)atingp)olicies issued in 1959. The
achieved rate of return depended heavily on the length of timie tile policy was
kept in force. For policies kept in force for the full 20 year period, returns
exceeded those available on competing investments.
Another study comj)aring investment options bzetweena tax advantaged
insurance product, in this case an annuity, and alternative investments was
performed by Adelman and Dorfman [1l. Although this study ignored tile
then current capital gains treatment of equity investment alternatives, tile
effect of (lifferent tax levels was measured. Again the holding leriod proved to
be an important factor in evaluating thie niore advantageous investment.
Broverman also concluded that the holding p)eriod was a key determinant of
the internal rate of return in an analysis of single premium life and annuity
contracts [6].
Warshawsky analyzed the impact of' the 1959 Life Insurance Company
Income Tax Act on tlheafter tax rate of return onl investments for life insurers
and foundthlat higher interest rates increased the differential be vweenlife
insurance savings and alternative savings vehicles 1261. For thie period
1977-1981, life insurance policies yielded a rate of return 2.0 percentage
points below comparable long term bond investments. In 1982 the Tax Equity
and Fiscal Responsibility Act temporarily revised the life insurance company
taxation formula to reduce this differential. In 1984 tlhe TFaxReform Act
included a major overhaul of the taxation of lile insurers to return tile tax
calculation to a total income base, as opposed to separating underwriting and
investment income [7].
This analysis of life insurance purchase (ecisions for universal/variable life
includes the same rate of return forecast onl both the life insurance policy and
the competing investment alternative. As B3elth[31 notes, rates of return onl the
savings component of universal life insurance di ffer depending onl whether
expense loadings are treated as a protection lement or a savings element. If
the expense loading is regarded as a savings element, the rate of return may be
negative, whereas if tie expense loading is .allocated to tile protection element
of the policy, the rate of return could be quile high relative to alternative
investments. The proper procedure in this calculation is to attribute a portion
of the expense loading, in line wvithcompetitive values, to tlhe protection
elenient with the remainder allocated to tlie savings element.

Ilves(tmnent Value l)etermnination

The after tax surrender value of anl investment in a universal/variable life

insurancepolicy that has a death benefit equal to the initial face value plus tile
cash value can be determined as follows:

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
456 7/h, Journtal of Ri.sk ((i Insurunce

(l-S,)L((l
-ei)l'0-- )(l +r+d)dP i if UVL c pi
(1) UVL =

17 Cx, I it if
(I
-tx'|)( ~S,,E( -edp'-g )1s+ i |
n II

-EPj +?EPi
i r. I j _ I
ot herwise

where P; = preminum paid in year i

n = niumberof years (lhepolicy is kept in force (holding pcriod)
S. = surrender charge at thieend of the holding period
e; = front end expense loading (as percentage of premium) in year i
g = index of conipetitiveness of terni insurance through universal life policy
F = face value of the policy in thousands
Cx = cost of term insurance for policyholder age x per $1000 of coverage
r = annual net rate of return for comparable investment fund
d = differential between policy interest rate and comparable investment fund
rate
tX= marginaltax rate for policyholder at age x'

The amount invested in the cash value each year is the premium less an
expense loading, and less (lte cost of insurance. Thlecash value earns a rate of
return, r + d, that tracks below, at, or above comlparable investment rates of
return. This rate reflects the net rate of return credited onl the cash value,
which is not the same as the gross rate of return, ignoring expense loadings,
cited by some insurers. The investment eartiings are not taxed until the policy
is surrendered. The investmentivalue may be reduced by a surrender charge,
which is a portion of (the total cash value at surrender. If, at that time, thle
withdrawal value does not exceed the total premiums paid, no income tax
liability exists. If thie withdrawal value does exceed teie premiums paid, thle
excess is taxed at tile inisuired'scurrcnl marginal tax ratc.
One alternative investment strategy that is comparable to investing the cash
value at money market futid rates, is a strategy of buying term insurance and
investing the remaining sutm iti a motney market fund. The value of this
investment would be:

'Soie of the investient alternatives generate taxable income continuously whereas others
create taxable income onily on withdrawal of funds. Investment in money market, bond (par value
or deep discoint), or stock funds generate taxable income throughout the year. Thus, tile
appropriate tax rate is 1, , for each year's determination. For example, in thie first year thie
moneymarketfund producesinterest ihatis taxatileat teieinvestor'sinitial(t I, =t tax rate.
Other investment alternatives, such as the universal/variable life insurance policy, an tRA
investment or a decferredannuitv would generate taxable income only when capital is withdrawn,
at which time thletax rate would be t, ,F. For an IRA surrendered at the end of thiefirst year, tihe
appropriate tax rate is the investor's tax rate a year after the itiitial year, or t, , . Thus, different
subscripts to the tax rate apply depending upon whether the investment income is taxed currently
or only on \vithtdrawal.

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
 U/uicr.%aI/ I ah/el Life Insluram
tii 457

(2) 11,N,,=
13FIDN x (I -- Ft- ? (1 - t , i ))]

The investment proceeds are taxed each year under this alternative, reducing
the current yield. No expense loading is deducted from tile amount to be
invested as mioney market funds are no load funds. The cost of insurance is
simply the lowest priced coverage available in a renewable term policy. Note
that this can be higher than, equal to, or lower than the rate charged in the
universal life policy depending on whether, g, tlie index of competitiveness of
the insurance costs through tlie Universal life policy, is less than, equal to, or
greater than one. The rate ol retiurnis simply the standard nioney market fund
rate. This same calC~lllliOll would apply if the alternative investment were in
par value long termi bonds under which all of the return is interest income.
The rate of return, r, would likely be higher than for the short termirate, but
the tax consequences would be (tle same.
Another alternative investment strategy would be to buy termiinsurance and
invest (lie remaining sum11i
in tax free municipal bonds either short or long
term. The value of this investment would be:

,D I
(3) 13TI)I = (Pi- F Cx, )(I +nir)n I

where il ratio of municipal bond yields to taxable bond yields for similar
=
maturities
No taxes are involved in this determination and no expense loadings apply if
a no load fund is selected. However, municipal bonds yield less thati similar
taxable investments in light of this tax advantage.
A third alternative investment strategy would be to invest the difference,
after purchasing terni insurance, in a deferred annuity. The value of this
investment would be:
It
= (I -SA,,) (I -ea)(Pl - F- C, ,)(t + r), I
n

if BTIDA< K (Pi-F C'X-)

(4) BTIDI)A

= (I-l, * ,,- I 4)[(I-SA,,)E(I-ca;)(Iz*- I t, )3(t+r)0 --i I

( .-FC
- ,
(1 li - F- C, ) otherwise

where H = . I it'
f xn < 6()
0 otherwise
SA,, = surrender charge at withdrawal on the deferred annuity
ea, = front end expense loading at year i on the deferred annuity
The tax treatment of deferred annuities is somewhat similar to theat of a
uiniversal/variablelife insurance policy. If the withdrawal value does not
exceed the total premiums paid (total investment less termninsurance costs), no
income taxes are owed. Otherwise the excess is taxe(l at the policyholder's

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
458 7Te Johurnaul (if R!i.?k ,id Insirawwe

currcnt tax rate. Howvever,a penalty of 10 percent of tile taxable amount

withdrawn is applied unless the policylholder is 59'/2 or older, disabled or
deceased, or the proceeds are withdrawvnin the fornmof an annuity.
A fourth alternative would be to invest the remaining sum in deep discount
bonds. These bonds pay a coupon rate below current interest rates for similar
investments and therefore sell at a (liscount fromliface value. The interest
income is fully taxable, but the amortization of the discount is not taxed until
the bond is sold or miatures.The value of this investment would be:

11 (P1-i
(5) 13l)T = FCyP )[VJ(lA Ir(lI-t , -)+r2)

-tX, t8(t 4 1(l -- t j ) f r.)-I(l ?+r,(- t, j )]

wiherer, = current yiekl

r= = amiiortization of discount

The tax (leferral ot tl1eamnortizationo' Itle (liscount has served to maintain the
markct value of lokv yield bonds such that r14 r2 < r for similar investments.
A final alternative investment strategy comparable to a fixed income
investment allocation in a universal/variable life insurance policy is to invest
the difference, after purchasing term insurance, in anl Individual Retirement
Account (IRA) or other similar salary reduction plan.2 Investments made ill
such plans uipto an allowable inaximumiare deducted from gross income and
investment income is tax deferred. When money representing tax deductible
contributions and interest earnings is withdrawn from the IRA, it is taxed in
its entirely at the individual's then current income tax rate. Withdrawals prior
to age 59'/2 are also subject to a 10 perccenttax penalty unless occurring as the
result of the death, disability or early retirement after age 54. The maximum
annual contribution to an IRA for individuals not covered by a pension plan
or with all a-djusted gross income below $25,000 f'or a single taxpayer or
$40,000 for a married taxpayer is $2,000. For individuals with incomes over
these levels, thle deductible IRA contribution is reduced by 20 percent of the
excess adjusted gross income. Thus, single taxpayers with adjusted gross
incomes over $35,000 and married over $50,000 cannot inake any deductible
contributlions to an IRA if they arc covered by ati employcr's pension plan.
Thcy can, hiowevcr, miake nondeductible contributions, subject to tile IRA

2'Plans that function similarly to an1IRA include 403(b) plans for employees of tax-exempt or
educamionaloigaui7fltioiis and 401-k plans for enpltoyces of private firms that offer this benefit.
Each of tliese tax stieltered plans has spccial rules dcfining (lie maximnumn
allowable contribution
in ternmsof salary with an absolute uipperlinit. The example developed in the paper is based on
1he IRA riltes for maximium contribution since more individuals are eligible for an IRA. If an
individual is also eligible to contribute to one or hc ottier (ax sheltered plans, the aggregate
maximum may be increased. The taxation of tluip sumi withdrawals from the plans differ, also,
with the entire proceeds froin an IRA or 403(b) plan subject to taxation but five year forward
averaging applicable to 401-k distributions. The IRA withdrawal rules are used in this program.

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
 (tlmverS(1l'a (i t jUlI't' Lif/i InsullraIn((e 459

maximums, which earn tax deferred interest. The after tax withdrawal value
of an IRA is:

= I) Vtiu[ E N] i EN
(6) IrTII),H ( -

+ EMax - irN i I( ~
where -I- .1 if x + n) < 60
t) otherwise

Di = maximuimti IRA coniril)utioni

dedutlctible
IRAcontribmnion
N, = nlondedIttibile = NIui4(2,00X-DI),
P,- F C I , - D - tt

Any deductible amount invested in teie IRA is larger than simply thle
difference betwveenwhat would halvebeen invested in a tiniversal/variable life
insurance policy aniedthe cost of tierm insurance as a result of the tax
deductibility of IRA invest ments. A deductible contribution to al IRA
reduces taxes by t iI times the investment. This tax savings would then
bie available to increase the investment in the IRA. Thus, the amount
(i - F, CX i ) is dividedl by (I - t, the total amount
,) to (ICtermllinle
.availableto invest in ain IRA that wotild be equivalent to an investment of
(P - F CX ,) in a nondeductible investment. The difference, if any,
between the $2,00() maximum annual IRA investment and the individual
investor's deductible IRA contribution can be put into anl IRA with tax
deferral of interest. Triis interest is taxable when witidrawn and subject to the
l() percent early withdrawal penalty, if applicable. Nondeductible IRA
contributions are not taxed on withdrawval, but all withdrawals are treated as
proportional distributions of deductible and nondeductible contributions and
interest. Contributions in excess of $2,00() are not tax sheltered in any way,
bit are treate(l similarly to non-IRA money market filled investments.
The same analysis can be performed assuming the policyholder elects
investment in equity funds, whichleave different (ax treatment fromt fixed
income funds. In a stock market fund realized capital gains and dividends are
taxed currently at ordinary income tax rates. Unrealized gains are not taxed
until shares of the I'fund re sold. For equity gains in a universal/variable life
insurancepolicy or anrIRA, taxes are (lelerred on dividends and capital gains
until the investment is withdrawn.
The alternative stock fund investment strategy includes tax deferral
;advantagesnot found in a money market fund or fixed income investment.
The valuie of this alternative is:
It ,. n to

(7) Ds -
1}11) (P. - F C>, ,}(Q1 - Q2)w-1x,(IQ 2 nIQ)]
"here Q, = I + sr(l -t, ,t

O. = ( - s)r
s = proportion of r producedIby realized capital gains and dividends

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
460 71, .Iotirnal of Risk nId Insuirance

If the stock mutual Fund did not generate ally realized capital gains or
dividends, no taxes would be payable until the shares were sold. If realized
gains or dividends were generate(, the investor has the option of reinvesting
those amounts or receiving them in a cash distribution. Since tile taxes owed
would always be less than the cash distribution, the investor can pay the taxes
out of the distribution an(d reinvest the remainder back in the stock fund.
Thus, the basis in the fidn would reduce only by any taxes paid and the
investor would not retain any excess cash. Under this procedure, no shares
would have to be sold to pay taxes. This situation is preferred because any sale
of shares would involve taxes on any unrealized (by the fund) gains, which
would result in additional taxes payable.
Universal/variable life insurance contains an additional tax advantage not
included in this holding period analysis. If the investor dies, all of the
proceeds received as a death benefit, if paid to the beneficiary in a lumipsurn,
are free froni income taxation. Proceeds paid to a beneficiary from an IRA
account are taxable income. The tax treatment of proceeds from a deep
liscoulit bonidor stock FUn(lare more complex. The beneficiary's basis for tax
purposes is increased from the owner's basis to hliemarket value at death.
Thus, any unrealized appreciation of these funds prior to the death of the
investor would not be taxed.
Life insurance comparisons are normally event specific. The situations
compared in this research involve keeping a policy in force for a specified
holding period and then withdrawing the proceeds either in a lump sum (prior
to retirement) or as periodic payments (after retirement). Comparisons based
on the policyholder's assumed demise at specific tinmeswould yield different
results, although are more speculative than the holding period comparisons
illustrated herein. Since the additional tax advantage based omithe death of tile
policyholder is in favor of universal/variable life insurance, this analysis
serves as a conservative comparison of (lhebenefits of this investment strategy.

Ca(egories of Invcs(meni(AIlerinafives
Any comparison of projected investment results must limit the investment
alternatives to those with similar risk characteristics. It would not be realistic
to compare returns on a 6 percent short term government security with a
projected 20 percent illiquid and speculative real estate investment. The higher
return on the real estate investment results from greater risk, lower liquidity,
and longer required holding period. In this study three types of investment
alternatives are examined: short term money market accounts, intermediate to
long term bonds, and equity funds. Short ternmmoney market accounts
generally maintain an average maturity of 40 to 90 days and provide a rate of
return that fluctuates frequently in line with changes in short term interest
rates. Although the principal is not guaranteed in many such accounts, tile
short maturities tend to minimize this risk. An individual can invest in short
termimoney market accounts through many universal/variable life insurance
policies, bank accounts, money market funds, anidshort ternmmunicipal bond

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
 ni, sa/I Variable Life. I/nSuraIance 461

ftinds. Thllesetypes ol investments arc generally conisidered to be the least risky

investments and consequently provi(Ie the lowest rates of return.
Intermediate to long term bond funds invest in longer term fixed income
investments and generally have anl average maturity of five to fifteen years.
Interest rates onl these investments are usually highier than short term
investments, although at times the yield curve is inverted so long term rates are
lower than short term rates. One reason that long term rates are usually higher
than short term rates is that if interest rates were to rise then the value of
outstanding bonds would fall, generating a loss of principal. TFhehigher rate
of return onl longer term issues compensates the investor for assuming this
additional risk. An individual can invest in longer term fixed income issues
through soeC universal/variable life insurance policies, bank certificates of
deposit, bond mutual Iuinds, deferred annuities, deep discount bonds, and
municipal bond funds. For the purposes of this research, it is assumed that no
load deep discount b)ond funds exist; no load municipal bond funds and
taxable bond funds do exist.
Investment in common stock is considered to be even riskier thall long term
bonds. Stock prices fluctuate constantly with no guaranteed rate of return
over any particular holding lperiod. Dividends, unlike interest payments, can
le altered by the company's Board of D)irectors.Over long periods of time the
average return produced by equity investments exceeds long term bond
returns. However, over shorter periods stock averages, anrdeven over longer
periods individual stocks, have underperformed bond investments. Investors
can participate in equity investments through universal/variable life insurance
policies, purchasing individual common stocks or investing in stock mutual
funds, many of which have no expense loads.

Parameter Values
The objective of this research project is to determine if universal/variable
life insurance policies dominate similar investmerit strategies in money market,
fixed income amidequity funds outside of life insurance policies for thie range
of parameters available. The values of tlhe parameters Used to evaluate a
universal/variable life insurance policy vary significantly depending tupon the
potential policyholder and teie specific policy. Standard values are determined
that represent the typical universal/variable life insurance policy contract.
These standard values are used to compare universal/variable life with other
investment opportunities. However, meanlyof these parameters change over
time, across insurers, or among policyholders. The rapidity and importance of
such changes is noted by Heath and Witteruore 191.The parameters can be
varied through use of the personal computer program developed for this
research. Determination of (lte standard values are discussed in this section.
The rate of return, r, used ill this analysis is (lte money market or fixed
inconie interest rate or thie equity fund total rate of return. This valuie
indicates the rate payable on a competing investment alternative; it could be
considered either the average rate paid by money market, fixed income, or

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
462 7I1(e.I,,r,, ani1dInsurance
ol IRisA-

cquity Inmids,or (licrate paid by a pairticularfuind. The relevant rale of' return
is that experienced alter (lie investment choice, uniliversal/variablelife or buy
term, is made. Thus, it is a forecasted value, not a historical value, that
indicates the preferred investment. As such, a range of values of r should be
examined by a potential policyholder. Prior to (lte mid-1970s, interest rates
were both lower and more stable (hlan since that 'time 1101. Neither money
market funds nor universal/variable life insurance would have been viable
financial instruments for individual investors prior to the mid-1970s. The
actual rates of return achieved on, short terim government bills, long term
corporatc bonds, andeq(uitics from 1976 tirough 1985 are summarized in (he
Appendix. Based on these data, tlhe standard rate of return is 9 percent for
money market investments, 11.5 percent for bond investments, and 14 percent
for equity investments. Howvever, money market fund returns and other
interest rates have fallen in recent years. The mid-1986 average return on
mioney market funds was 6.5 percent [201. To illustrate the effect of low
interest rates onl universal/variable life insurance, a separate run of tile money
market investment selection using a 6.5 percent rate of return is also included.
The tax rate, t, is the individual's marginal tax rate each year under the buy
termistrategy or when the cash value is withdrawnl under the life insurance
strategies. The tax rate can vary during the preretirement period but after
retirement it is assumed to be a constant value. The Tax Reform Act of 1986
establishes three effective marginal tax rates, 15, 28 and 33 percent, for years
1988 and after. Inl this analysis, the individual is assumed to be ill tile 28
percent tax bracket over the entire period. Any other pattern of tax rates could
be input.
Expense loadings onl universal/variable life insurance policies take a variety
of forms, including a flat fee per policy, a charge based onl the amount of
coverage, a percentage of the investment, or a combination of these charges.
In somie cases expense loadings are constant over the life of the contract
vwhereas other policies reduce expenses after the first year [251. Ilnthis analysis
for the front loaded policy the expense loading, ej, is determined as a
percentage of annual investments (see Appendixi]. For the standard value the
initial expense rate is set at 15 percent wviththe renewvalexpense rate 7.5
percent.
Some universal/variable life insurance policies include surrender charges
that reduce the policylolder's cash valuieif it is withdrawnlprior to a stipulated
holding period. Based onl anlanalysis of informat ion onl 131 insurer's policies,
31 percent had no surrender charges, 3 percent had surrender charges for the
first year only, 18 percent included charges for the first five years, 28 percent
for the first tell years, 15 percent for fifteen years, and 5 percent for twenty
years [251. The average surender charges were 53.1 for the first year, 9.6
percent in the fifth year and 1.8 percent in the tenth year [see Appendix]. For
the back loaded policy, the standard surrender value is a declining function of
the holding period starting at 50.0 percent in year one and reducing by 10.0
percentage points per year until year 5, and then reducing by 1.6 percent per

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
 L'iii i'i lI/ rti' b/eL 1/c hI..11,InI(i( 463

year. Thus, thie surrender charge is 10.() percent for thle fifth year and 2.0
percent for the tenth year. No surrender charge applies after thie tenth year.
The interest rate differential, d, indicates how the interest rate credited on
the universal/variable life policy compares wvithrates of return available onl
comparable investments. Some universal life insurance policies have an
interest rate that is tied to anl external interest rate level, such as 90 day
Treasury bills. Other insurers base the interest rate onl hlow their own
investment portfolio performs, providing either portfolio rates, liexv money
rates, or investment year rates. A final group determine the interest rate based
onl an1 insurer decision t hat is inot tied directly to any performance reisults.
Based onl 126 universal life insurers for which teie inerest rate deterilliliation
was disclosed, 12 percent relied onl anl external index, 79 percent used all
insurer yield value, and 9 percent determined thie rate independent of ally
performance guide [251. For insurer decision cases tlhe lolicyholder has nlo
guarantee iat tile insurer will not alter past patterns of interest levels, but ally
change would affect all policyholders. For universal/variable life insurance
policies tile rate of return eared onl cash values is not controllable by the
insurer, but depends onl short term, bond, or equity investment performance.
Adminimstrat ive expenses, taxation of life insu rers, and invest ment policy mallly
generate a differential betweeni th1ereturn Carniedby thle insurance fuLlnd and
other public funds with similar risk characteristics. After these policies heave
established a track record, investmentlperformance could be analyzed to
project a differential value. Given tlie current lack of an investment record for
uiniiversal/variablelife insurance funds, expense loadings could be compared
to project a difference. Trhe differential is in percentage point terilmsand
represents tile difference between thleuniversal/variable life rate of returIl and
the comparable fund rate of returnl.Thle standard value for tlhedifferential is
zero. At tlheend of 1984, universal life insurers were crediting anl interest rate
approximately 1.5 percentage points above short term interest rates [251,
reflecting the lag effect ill changing credited interest rates and a reluctance to
lower this key value. Over ally lengthy lperiod, this differential should be zero
or slightly negative.
Portfolio turnover also aflfects (lhe relative attractiveness of investment in a
universal/variable life insurance policy. Ally gains realized by tlhe investment
fund in this policy are tax defclrred until tlhe policy is surrendered and then
taxed at ordinary income rates to tle extent cash value exceeds premiums
paid. In tlhecompeting equity investnuent, realized capital gains and dividends
are taxed currently all oreinary income ,ax rates. Stock funds have a wide
range of portfolio turnover rates. A samipble of finds examined indicated
values of 20 percent to inl excess of' 200 percent. Higher turnover and dividend
yield increases (lte current taxation onl the competing investment strategy and
improves the position of universal/variable life.
For this analysis, (lte standard rate of retirn onl stock investments is 14
percent and (lte standard valuie of s, (the proportion of r produced by realized
capital gains and dividends, is 80 percent. Trheremaining proportion of r is
deferred until the intiual Rfnd is sold.

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
464 7TheJournal of Risk and Insurance

The available capital per year, P, is the amount the policylholder wants to
invest in either anl insurance policy or the buy term and invest the difference
strategy. One advantage of tie newvlire insurance policies is the flexibility tile
policyholder has with regard to premium payments. Within fairly large limits
tile policyholder can select any investment level and alter the amount at will.
Generally the minimum allowed investment is the amount necessary to cover
mortality costs, although some policies allow no payment if the cash value is
large enough so that mortality costs can be paid by a reduction in cash value.
The maximum contribution level is determined by provisions ill tile Tax
Reform Act of 1984 (DEFRA) that establish a new definition of life insurance
for tax purposes. In order to qualify for favorable tax treatment, a policy
must meet one of two tests. The first test, termed the cash value accumulation
test, prohibits the cash value from exceeding the net single premium for future
death benefits, based on a four percent interest rate or the policy guaranteed
rate if higher, and policy mortality rates or the mortality rates used ill
determining statutory reserves. The second test has twcoparts. The guideline
premium regulation prevents aggregate premiums paid front exceeding
guideline premiums based on interest and mortality rates as determined in tile
cash value accumulation test. Also, the death benefit must be at least an age
based percentage of the cash value. For a policyholder age 40 or less, the
death benefit must equal or exceed 250 percent of the cash value in a universal
life policy; for insureds over 40, the percentage reduces each year until age 75.
Insureds age 75 to 90 must have a death benefit at least 105 percent of the cash
value. The required percentage reduces by one point easch year over age 90
until leveling at 100 by age 95. For this analysis annual investment levels are
assumed constant throughout the policy term. The standard premium level is
$1 ,000. Il 1984, the average universal life insurance premium was $978 1131.
The maximum deductible IRA contribution, ), is calculated based on the
provisions of the 1986 Tax Reform Act and the investor's filing status and
adjusted gross income. It call range from 0 to $2,000. The standard value used
for the illustrative runs is 0; thus any IRA investments are nondeductible
although interest is tax deferred.
The face valuiC of' tie life insurance policy, F, is tile amount of coverage
nitially purchased. This IanalysisdeternIines thle(c ali bemefit by sitinilmingthe
policy face atndthe cash valuc. Thus, tlie mortality cost is based oil a constant
amount of coverage. The standard face value is $100,000. Tile average value
for uliiversal life policies in 1984 was $82,000 1131.
The initial age of the policyholder, x, is used to determitie the mortality cost
in the life policy and the cost of term insurance in the buy term strategy. Term
insurance rates are calculated based oil the average current rates quoted by the
seven largest universal/variable life insurance writers [251, which equaled
almost exactly the average rate quoted by 62 insurers writing indeterminate
premium aninual renewable ternminsurance [11I. The standard value for the
policyholder's initial age is 35.
Thle relative rate of return on tax free municipal bonds versus similar
taxable issues varies over tiue andl across maturities. From the end of 1985

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
 1/nlil ^'('rsu// J jna/le'
I 1.1/c
i ln i,,s ic 465

through mid-1986, tax tree mioney market runds were yielding onl average 63
to 67 percent of taxable Fund returns and long termnmunicipals were yielding
appzroximately83 l)ercent of comparable maturity industrial bonds [16, 20,
241. The lower maximnumtax rates uenderthe Tax Reform Act of 1986 will
affect this relationship. The standard value of the relationship miused in this
study is .75.
Each deep discount bond lhas its own maturity date, coupon and discount
from face value. To thle authors' knowledge, no mutual fund specializes in
deep discount bond investments although no practical reason prevents this
specialization. Without no load inutul funds providing this investment
medium, an individual investor would have to make individual purchases of
bonds incurring significant transaction costs. For the purposes of this study it
is assumed that a no load fund investing in deep discount bonds exists. It is
also assumed that the current yield rate, r,, is one half thle comparable
investment rate of return, r, and appreciation, r, is 40 percent of r. The total
return onl the deep discount bond fund is less than comparable investments as
a result of the tax deferral of the amortization.
The deferred annuity contains its own expense loading and surrender charge
scale. Deferred annuities differ as widely as universal/variable life insurance
policies in regard to interest rates, expense loadings, and surrender charges.
To simplify the comparison, typical values were selected for the deferred
annuity option. The interest rate is the samieas available onl taxable bonds; tile
expense loading is 10 percent of all premiums; thiesurrender charge starts at
100 percent and declines linearly to zero after ten years.
Prior to making the decision of %whether to buy a universal/variable life
insurance policy or to buy teri insurance and invest the difference in a money
market, bond, or stock fund, the prospective policyholder would know( the
face value of the policy desired (F), the maximium deductible IRA
contribution (D), his or hierage (x) and current tax rate (t). These values do
not depend onl the insurer or the policy. Also, for each life insurance policy
considlered, thle indiviClUal can determine the expense loadings (el) and
surrendercharges (S.,), how the rates compare Vithlbasic termiinsurance rates
(g), and any differential between similar investments and the interest rates
credited for thie policy (d). The decisionmnakermust estimate fuaturetax rates
(t) and rates of return (r), the currently taxable earnings (s) in comparable
stock fuLnldsuenderthieequity investmuentoption, t(ie tax free municipal bond
interest rate relativity (in), the deep discount bonrdrate factors (r, and r2) and
thie expense coml)onents of any deferred annuities (SA,, and eaj) and decide
the amount to ilnvest(1').

Comparison of Investient Resulls

Two universal/variable life insurance policies are compared with alternative
investment choices. The first policy (U VLF) is a front loaded
universal/variable life policy with no surrender charges. The second policy
(UVLB) is a back loaded policy that has surrender charges but no premium

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
466 The Journal of Risk and Insurance

loadings. These two policy types represenit ilheextremes of expelnse loadings.

Many insurers combine front loadings with surrender charges. Any
combination of expense loadings call be input into tlie program, but these two
examples were chosen to demonstrate the effect of different expense factors.
The after tax lump sumisurrender values for holding periods of one to thirty
years are shown. Aln alternative to a liumpsum withdrawal from all insurance
policy or other investment is the selection of periodic payments. A stream of
payments often meets the financial needs of all older individual more closely
than one large payment. For universal/variable life insurance, IRAs, and
deferred annuities, the tax consequences of periodic payments are more
favorable than lunip sum payments as the income is taxed only when it is
received. For holding periods of 30 years or longer, after tax periodic
payments derived from the accumulated values of the alternative investments
are displayed. The terminal surrender value of the particular investment is
taxed, if applicable, and tewn used to purchase all individual life annuity with
no period certain.) The after tax portion of the annuity payments are
displayed.
Annuities are taxed in such a way that tlie expected value of tile return of
the investor's basis is tax free with tlie remainder of the periodic payment
taxed as ordinary income. The ittntaxcdportion of each payment, terled tile
exclusion ratio, is determined by (dividing thie basis by the product of thie
annual paymcnt times liheinsured's life expectancy based onl Internal Revenue
Service Mortality Tables. A cash value life insurance policy, including
universal/variable, call be converted to all annuity as a tax free exchange
bascd onl IRS Sectioni 1035. Tile basis of this annuity is tlhe total of all
premiums paid wvithoutany reduction for the mortality costs. For IRAs the
basis is tlie total of nondeductible contributions. The basis on the deferred
annuity is tlie stum of tlie annual investments less thiecost of life insurance
each year. The basis for annuitized mioney market, bond, municipal bond,
deep discount bond, or stock investments is tlie after tax lump sum values, as
these investiments have to be terminated, and taxed, prior to purchasing an
aiiiiuity.
If tlie individual investor (lecides to invest in short termilloney market
accounts, the investment alternatives are univcrsal/variable life insurance or
buying term insurance andIinvesting the difference in a money market fund,
short terimmu iiicipal bond lu11(1, or tIirougha notideductible IRA ini a mioney
market rund. The after tax surrender valuies for the various investment
alternatives are (lisplayecI in Table I for the standard parameter values. The
values listed in each row represent tile amount that tlhe individual would
receive after taxes if tlie account were completely withdrawn within tile first 30
years (by age 65) or annuitized thereafter. Each columin represents an
alternative investment strategy.

'The annuity rates were derived from representative current male rates, as listed in Best's
Fliteraft 151.

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
 UiiiICI.U/. ~1 ,l/I /(,.hfCIi.',u11(11ce 467

I; t l )lt' I

Com.~nparisono M
larket
NMoney In vestmiiciul Alternalives

Starting age 35
Retirement age 65
Anmount of capital/yr($) 1000
Maximum tax deductible IRA contribution($) 0
Face valuc of poliCy($) I00000
Holding period(yr) 35
Marginal tax rate l)efore rciremehnt age(0/Vo) 28
aifter retirement age(1/o) 28
Rate of Ret urn(%/o) 9
Interest differential onl UVL1L & LIUVI
-1( 0/0)
Index of competitiveness onl UVLF & UVL13 I
Initial expense loading on U VI F( /0) 15
Renewal expense loading onl UVI I (W/o) 7.5
Duration of surrender charge onl UVI.13(yr) 10
Surrender charge at year l(?,o) 50
Surrender charge at year 5(0/o) 10
Annual decrement aftcr year 5(') 1.6
Index of return on imuinicipal bonld .75

Alter Tax Lump Suin Wit hdrawal Amotunt

Y'eur II VLt lInf 1'V I.l11ninf ItttII)InIIf ltAiimAr ttitt)rnl

1 751 457 893 886 896

2 1,640 1,14() 1,834 1,812 1,841
3 2,599 2,0)75 2,825 2,783 2,840
4 3,632 3,289 3,868 3,801 3,894
5 4,747 4,8 16 4,969 4,873 5,0)09
6 5,951 6,0)94 6, 130) 6,004 6,188
7 7,175 7,417 7,346 7,1 89 7,426
8 8,454 8,849 8,620 8,435 8,728
9 9,809 10,399 9,958 9,748 10,099
10 11,246 12,079 11,363 11,134 11,542
11 12,772 13,945 12,838 12,601 13,062
12 14,388 15,725 14,379 14,146 14,654
13 16,I1)2 17,618 15,98' 15,777 16,323
14 17,922 19,634 17,674 17,503 18,075
15 19,859 21,784 19,437 19,333 19,914
16 21,922 24,079 21,284 21,279 21,848
17 24,114 26,524 23,206 23,336 23,867
18 26,444 29,130 2 5,208 25,517 25,978
19 28,926 31,913 27,297 27,835 28,187
20 31,574 34,888 29,475 30,302 30,501
21 34,402 38,073 31,750 32,934 32,926

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
468 7iI(' .liiio l of Risk candInsiIr/cl-

22 37,401 41,461 34,093 35,712 35,434

23 40,585 45,070 36,508 38,652 38,032
24 43,973 48,920 39,00() 41,771 40,726
25 47,581 53,032 41,575 50,047 43,522
26 51,430 57,431 44,236 54,157 46,427
27 55,502 62,102 46,936 58,501 49,394
28 59,817 67,069 49,677 63,100 52,427
29 64,397 72,360 52,463 67,979 55,531
30 69,264 78,004 55,295 73,163 58,711

After Tax Annuity Withdrawal Aniount

Year UT'IFmmr ltw litinmr lllTlDlnmf IRAmmr lwrllmb

30 8,499 9,639 6,225 9,293 6,610

31 9,352 10,627 6,704 10,220 7,141
32 10,282 11,708 7,204 11,225 7,701
33 11,307 12,904 7,735 12,330 8,299
34 12,417 14,203 8,285 13,518 8,923
35 13,624 15,622 8,849 14,810 9,570

For an individual who holds the investment for only one year, investing
$1,000 in a front loaded universal/variable life insurance policy at the
beginning of the year returns $751 at tile end of the year plus the value of the
protection during tile year. For a back loaded policy, the amount available to
be withdrawn after one year is $457. For a holding period of only one year the
optimal strategy is to buy term insurance and invest the difference in a short
term municipal bond fund, which would return $896, plus the value of the
protection (luring the year. Buying term insurance and investing the difference
in a money market fund produces almost as high a terminal value. Tile penalty
tax on the interest for premature IRA withdrawals reduces tile terminal value
of this option, although it is higher than the uiniversal/variable life alternatives
initially.
Buying terni insurance and investing in a municipal bond fund remains the
optimal investment strategy for tile first seven years. For holding periods
longer than scven years, buying a back loaded universal/variable life
insurance policy is the optimal investment strategy. After the tax penalty on
premature withdrawals is removed, in year 25 for this example, the IRA
alternative becomes thiesecond most attractive investment.
For years 30 through 35, thieafter tax values of annuitizing the investments
into a life income, no refund, are displayed. Although each investment
alternative is annuitized, tile investor's basis in tile annuity differs depending
on the investment medium. Tile basis is used to determine tile tax free portion
of each payment. For tile universal/variable life insurance policies the basis is
the total of all premiums paid, $30,000 in this example at year 30. For the
money market fund and the municipal bond fund, the basis is the entire lunmp
sum withdrawvalvalue. For tile IRA account the basis is the total of all

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
 Univermsll Voriat/e tIle linsunrnce 469

nonmdeductibleinvesiments not inciluding tlhe cost of insurance coverage

purchased. For each of (lie years displayed, the back loaded universal/variable
life insurance policy provides the largest after tax annuity payment. The
reasons this policy produces ilhe highest values are the tax deferral on
investment income and the (ax free status of earnings equal to the mortality
costs.
In this run tlheinvestor is unable to put any deductible cont ributions into anl
IRA. In other runs, n1ot included here, where IRA contributions were
deductible, lihe IRA alternative dominated universal/variable life after
l
preimaturcwithdrawal peitallties enided. Thus, if an individual is not already
putting the maximumi deductible anmountsinto an IRA, or comparable tax
sheltered investment suich as 401-k or 403(b), this investment option would be
best for long termnsavings. Otherwise, tile back loaded universal/variable life
insurance policy is tlie optimal investment strategy for holding periods of
eight years or more.
Although short teriminterest rates have averaged about 9 percent over thle
last decade, by mid-1986 they had fallen to approximately 6.5 percent. To
illustrate thie effect of lower initerestrates onl the investment alternatives, the
progra.1 was reruinilsing all of' the samneparameter values except lowverintg tlhe
interest rate to 6.5 percent. The results of this run are listed in Table 2. Thle
relationships among tlhe investment alternatives remain similar with thle
municipal bonld flund representing the optinial investment for tlhe first seven
years, after which the back loaded universal/variable life insurance policy
dIoniiiates.

Tabl)e2
Comparison of M-loneyMarket Inuivesitmewi
Alterntmives
Alt parametervalues are the same as TFableI except the rate of returtr = 6.5%/o
After Tax Lump Sum Withdrawal Aiounit
Year Uv tl f LIVt .
UNT1 r l I IIIf
ItFI1 MtAnisif it rii)nll)

1 734 447 878 873 880

2 1,584 1,101 1,787 1,771 1,792
3 2,480 1,978 2,728 2,697 2,739
4 3,422 3,096 3,702 3,652 3,719
5 4,414 4,474 4,71(0 4,638 4,738
6 5,460 5,619 5,756 5,658 5,795
7 6,554 6,852 6,830) 6,70)5 6,884
8 7,699 8,129 7,93 5 7,782 8,006
9 8,899 9,439 9,072 8,891 9,164
10 10,113 10,827 10,243 10,034 10,359
11 11,343 12,335 11,449 11,215 11,591
12 12,614 13,728 12,682 12,426 12,855
13 13,926 15,170 13,942 13,669 14,150
14 15,284 16,666 15,232 14,947 15,478
IS 16,690 18,220 16,552 16,264 16,841

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
470 7Te Journal of Risk aind Insurace

16 18,148 19,834 17,905 17,623 18,241

17 19,650 21,503 19,277 19,013 19,665
18 21,199 23,230 20,669 20,437 21,115
19 22,799 25,020 22,083 21,900 22,592
20 24,453 26,876 23,519 23,403 24,097
21 26,164 28,802 24,979 24,950 25,631
22 27,910 30,777 26,429 26,512 27,162
23 29,694 32,805 27,868 28,091 28,688
24 31,520 34,890 29,296 29,690 30,212
25 33,388 37,035 30,714 34,050 31,731
26 35,302 39,244 32,119 35,970 33,245
27 37,226 41,482 33,458 37,882 34,702
28 39,161 43,751 34,729 39,787 36,098
29 41,107 46,053 35,928 41,684 37,431
30 43,065 48,390 37,052 43,574 38,697

After Tax Annuity Withdrawal Amount

Year wi[vi.nminr vtwtImnmur Illlimirn 1RAmmf It FID),nh

30 5,082 5,776 4,171 5,437 4,356

31 5,440 6,202 4,390 5,805 4,597
32 5,809 6,645 4,602 6,179 4,832
33 6,196 7,114 4,809 6,564 5,067
34 6,590 7,596 5,003 6,947 5,289
35 6,990 8,094 5,177 7,331 5,494

If thle individual investor is willing to accept the risk of investing in longer

term bonds, additional investment alternatives exist and a higher rate of return
is likely. The investment alternatives include a front loaded or back loaded
universal/variable life insurance policy with the cash value invested in bonds
or buying term insurance and investing the difference in a bond fund, deep
discount bonds, a deferred annuity, long term municipal bonds, or through an
IRA in a bond fund. The after tax surrender values for the various investment
alternatives are listed in Table 3 for the standard parameter values, including
a rate of retrin of 11.5 percent.

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
 Tabl)e3
lond hsiment
Comparison of' 1B Alternatilses
All paramteirvalu.es are the same as Table 1 exceptllotlows:
as
Rafe of returti (0/6) 11.5
Rate of ret urni on l)D13 (0/0)
Imerest 5.75
Appreciation 4.6
Expense loading on Deferred Annuity (%1o) I(

First ycar stirrender charge ott 1)A (1'/o) 100

Nutnmberol years 1(
Surrender charge on DA (O) 0 I00
After Tax Lump Sutm \Wihdrawal Atmot

Year UVI l l z1l1l llX U\ 1.1alhond l ll1D) .. ,,zl 1l\t)bolld 11FD1 11)1 11|111)(1a1

1 768 468 908 899 902 0 911

2 1,697 1,18() 1,88 1 1,854 1,860) 177 1,890)
3 2,722 2,1 75 2,924 2,871 2,88() 559 2,943
4 3,852 3,492 4,041 3,958 3,965 1,179 4,075
5 5,073 5,13() 5,240) 5,1 25 5,1 22 2,072 5,293
6 6,348 6,493 6,527 6,379 6,356 3,284 6,605
7 7,722 7,995 7,900 7,723 7,665 4,857 8,010
8 9,20)6 9,656 9,366 9,168 9,055 6,676 9,516
9 10,815 11,495 10),934 10,727 10,533 8,482 11,132
10 12,561 13,534 12,611 12,413 12,1(07 10,621 12,867
I1 14,460 15,850 14,407 14,241 13,782 13,141 14,730
12 16,524 18,133 16,321 16,219 15,560 14,947 16,724
13 18,769 20,623 18,362 18,364 17,447 16,902 18,859
14 21,217 23,345 20,542 20,698 19,454 19,026 21,147
15 23,892 26,324 22,871 23,243 21,590) 21,338 23,601
16 26,819 29,592 25,362 26,026 23,867 23,863 26,236
17 30,018 33,169 28,0)15 29,059 26,282 26,612 29,052
18 33,518 37,092 30,841 32,373 28,847 29,613 32,067
19 37,355 41,401 33,857 36,005 3 1,576 32,898 35,296
20 41,568 46,139 37,0)77 39,990 34,482 36,499 38,759
21 46,200 51,356 40,5 19 44,372 37,581 40,456 42,474
22 51,271 57,081 44,164 49,163 40,854 44,778 46,429
23 56,834 63,372 48,030 54,4 12 44,3 16 49,5 10 50,643
24 62,944 70,295 52, 135 60,177 47,986 54,703 55,140
25 69,665 77,921 56,500 74,936 5 1,881 67,844 59,944
26 77,067 86,332 61,144 83,058 56,014 75,150 65,080
27 85,187 95,578 66,038 91,975 60,335 83,157 70,524
28 94,107 105,754 7 1,201 101,779 64,857 91,944 76,300
29 103,921 116,967 76,656 112,572 69,599 101,602 82,439
30 114,730 129,337 82,426 124,468 74,578 112,230 88,970

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
472 The Journal of Risk an:(l!nsuranice

After Tax Anntity Withdrawal Amoulllt

Yrer til1Fhond UV'llBhond It l1 Dhond fl1Ahond I1TID(Ilh ItIII)dl ttTlII)mh

30 14,429 16,334 9,279 15,983 8,396 14,383 10,016

31 16,297 18,472 10,202 18,055 9,197 16,233 11,054
32 18,404 20,886 11,206 20,388 10,061 18,313 12,189
33 20,806 23,643 12,312 23,045 11,004 20,678 13,447
34 23,500 26,739 13,509 26,021 12,015 23,322 14,817
35 26,545 30,243 14,798 29,383 13,092 26,305 16,302

For a holding period of seven years or less, the optimal investment is to buy
term insurance and invest the difference in a municipal bond fund. For
holding periods of eight years or longer, buying a back loaded
universal/variable life insurance policy would be optimal.
For the investor who is willing to accept the risks associated with investment
in the stock market, four comparable investments exist: back and front loaded
universal/variable life insurance policies with thie cash value invested in
equities, buying term in1suranceand investing thiedifference in a stock Inutual
fulid, or buying teriminsurance anl investing the difference through anl IRA
in a stock mutual fund. The expected value of returns from eqttity investments
are higher than bond investiments, altlough greater variation occurs. For this
example, stock returns are projected to be 14 perceintannually. The terminal
investment values of these alternatives are listed in Table 4.

Frable 4

Comparison of Equityhivesncnm
I Alternatives
All parameter Values are the same as Table I except as follows:
Rate of return (%o) 14
Realized taxable portion on 13TIDstock (%o) 80

After Tax Lump Stin Withdrawal Amountt

Ycar tIV
IN.IFsck I.I IM.fsick 11 il)slock IllAslock

1 785 478 924 912

2 1,755 1,221 1,929 1,896
3 2,849 2,278 3,026 2,963
4 4,060) 3,705 4,221 4,123
5 5,345 5,410 5,527 5,392
6 6,763 6,925 6,954 6,786
7 8,324 8,633 8,505 8,312
8 10),0)49 10),561 10,193 9,990
9 11,961 12,745 12,0)34 11,844
10 14,087 15,224 14,041 13,898
11 16,455 18,102 16,233 16,180
12 19,0)92 21,031 18,618 18,715
13 22,036 24,308 21,216 21,538
14 25,330) 27,982 24,049 24,691

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
 Universal/Variable Life Insurance 473

15 29,022 32,106 27,142 28,222

16 33,168 36,746 30,523 32,186
17 37,822 41,962 34,206 36,630
18 43,053 47,835 38,222 41,624
19 48,944 54,456 42,609 47,247
20 55,586 61,931 47,403 53,589
21 63,084 70,379 52,648 60,753
22 71,531 79,909 58,354 68,820
23 81,059 90,672 64,570 77,921
24 91,822 102,843 71,352 88,205
25 103,991 116,616 78,758 113,631
26 117,763 132,217 86,849 128,942
27 133,321 149,860 95,630 146,255
28 150,914 169,830 105,171 165,894
29 170,829 192,455 115,553 188,045
30 193,389 218,104 126,863 213,206
After Tax Annuity WithdrawalAmount
Year UVLFstock UVLBtsock BTIDstock IRAstock

30 24,689 27,913 14,282 27,556

31 28,579 32,335 16,040 31,909
32 33,088 37,466 18,009 36,954
33 38,364 43,473 20,237 42,854
34 44,452 50,409 22,724 49,657
35 51,525 58,473 25,490 57,563
The optimal investment for holding periods of six years or less is to buy
term insurance and invest the difference in an equity mutual fund. For holding
periods of seven or more years, investing in a back loaded universal/variable
life insurance policy would be optimal.
Universal life insurance policies have been written only since *1979;
universal/variable life insurance policies have been offered only since 1984.
Thus, the industry has not compiled accurate retention rates for these policies.
The likelihood of retaining a policy for seven years or more simply cannot be
calculated based on experience. The policyholder has to make a judgment
about his or her likely holding period. However, lapse rates for all existing life
insurance policies indicate that in 1984, 23 percent of policies in force for less
than two years lapsed and 10 percent of policies in force for two years or more
lapsed [21. Based on these data, most policyholders do not view life insurance
as a long term investment. However, this study demonstrates that unless the
policies are kept in force for at least seven or eight years, alternative
investment strategies would be preferable. If a policy is to be kept in force this
long, back loaded policies will tend to dominate front loaded policies as the
surrender charges are usually either minimal or entirely eliminated after seven
to ten years. However, if the policy had to be surrendered in the early years,

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
474 A1
t Juwnal 0/ kiA. and
Inm/ wanc

tle investmenlt value of' thle back loaded policy would be below thle fronit
loaded policy. For holding periods of 25 years or longer, nondeductible IRA
investments are preferable to the front loaded universal/variable life policy,
but still dominated by the back loaded policy.

I)escriptioni of Computer Program

The program used to derive thleexamples included in this paper is written in

BASIC and runs onl 113NIFPCsand compatible computers. The program call
be used to calculate the optimal investment selection for the specific
parameters for a particular investor, rate of return forecast, or specific
universal/variable life insurance policy. The results indicate the lump sullm
withdrawal values at each holding lperiod from one year to retirement and
after tax periodic payments resulting from annuitizing tile withdrawal values
after retirement.
The program includes values for typical termi insurance rates fromnage 20
through 99 based on current indeterminate rate term policies for the largest
writers of this coverage. Tile restrictions onl cash value accumulations and
death benefits for uliversal/variable life insurance policies as codified ill thle
1984 Tax Reforin Act are included in the program. If the universal/variable
life insurance premiums are not sufficient to cover the policy expenses and
mortality costs, the cash value is reduced to cover the deficit. For tile
alternative investments, if the annual investment is not sufficient to pay for
the term insurance, thie accumulated investment value, subject to taxation if
applicable, is used to offset the difference.
In running thie program, the investor is first asked for the desired
investment medium, money market instruments, bonds, or equities. This
choice determines the comparable alternatives to a universal/variable life
insurance policy. The program then requests tile necessary parameters for tile
universal/variable life insurance policy and tile alternative investments. Thle
results are hot d
displayed on tile screen and hiplpt to a tile for later analysis.

Conclusiois
Universal/variable life insurance policies allow all investor to participate in
tile returns of a selected investment medimn through a life insurance policy.
Tlaxadvailltages ilnhlrclntill li c inmsuirantce
create t(le situaltion that purchase of
thcse policics, despie p)aying cx)pense loadings or surrender charges above
those in comparable investments, is tile prelerred choice, once the maximum
deductible amounts have been invested in an IRA or similar tax sheltered
investment, if the policy is hield long enough. The necessary holding period
depends on a number of values, some known to the policyholder-age, cost of
insurance, tax rate, and expense loading, and some unknown-rate of return
to be earned through tile insurance policy and the alternative investment and

4The authors would be pleased to provide a disk copy of lie program to amiyoriewho forwarids
a floppy disk.

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
tihe tax status of stock investment earnings. This analysis provides both a
method for determining the preferred investment and illustrates the necessary
holding period for t(ie universal/variable life policy to dominate under a
selection of parameter values. For typical values, thie universal/variable life
insurance policy dominates thie alternative investment strategies in seven or
eight years. A policyholder can estimate thielikelihood of keeping the policy in
force for the necessary holding period and decide which investment is
preferable.

It E.FE'E
NtlC ElS
1. Adelman, Saul W., and Dorman, Mark S., "A Comparison of TDA and
Non-TDA Investment Returns," The Journal of Risk anid Insiurance, Vol.
49, No. I (March, 1982), pp). 73-90.
2. American Council of Life Insurance, Life insurance Eact Book, Ulpdate,
1985, (American Council on Life Insurance: Washington, 1985), p. 24.
3. Belth, Joseph Mi., "A Note on l)isclosure of Realized Rate of Return for
Retirement Accumulations, Savings Accounts, and the Savings Conipo-
nent of Universal Life Insurance Policies," The Jouirnal of Risk and
Insurance, Vol. 49, No. 4 (D)ecember, 1982), pp. 613-17.
4. BeltIi, Joseph Ni., "Thie Rate of Return on elie Savings Element in
Cash-Value Life Insurance," The Jourl(JIl of Risk and insulrance,Vol. 35,
No. 4 (December, 1968), pi). 569-81.
5. Best, A. Mi. Comipany, Best's Flitcraft Coilnpend (Oldwich, NJ: A. M.
Best Company, 1986), pp. 644-46.
6. Broverman, Samuel, "The Rate of Return on Life Insurance and
Annuities," The Journal of Risk and Insmrance, Vol. 53, No. 3
(September, 1986), pp. 419-434.
7. Erxplanation of Tax Reform Act of /984 (Chicago, IL: Commerce
Clearing House, 1984), Chapter 9.
8. Ferrari, Robert J., "Investment Life Insurance Versus Term Insurance and
Separate Investment: A Determination of Expected Return Equivalents,"
The Journal of Risk and Insurance, Vol. 35, No. 2 (June, 1968), pp.
181-98.
9. Heath, Roger and Wittemore, David G. "Comparing Universal Life
Policies," Eml)hasis (November, 1985), pp1.1-3.
10. lbbotson, Roger (G. and Sinquefield, Rex A., Sto(ks, Bondv, Bills, alnd
Inflation: The Past (1926-76) aid the Future (1977-2000) (Charlottesville,
Va: The Financial Analysts Research Foundation, 1977).
II. "Indeterininate-Premium Annual Renewable Term Policy Comparison,"
Best's Review Life and Health Edition (September, 1985), pp. 84-85.
12. Internal Revenue Code, Section 264.
13. "Life Sales Showed 21%a Increase, Says LIMRA," The National
Underwriter Propert v-Casualtyv Insurance Editiont, Vol. 89, No. 11
(March 15, 1985), p. 32.

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
4 76 JIII, .10JIJI1111/
of I'is ( ,(
(IIlfd /rnsilune

14. Linton, MI. Albert, "Life Insurance as anl Investment," Lije an(d IealithI
Insurance Handbook, ed. Davis W. Gregg (Second Edition; Homewood,
IL: Richard D. Irwin, 1964), pp. 238-45.
15. Melhr, Robert 1. and Gustavson, Sandra G., Life Insuraince Thleoiy and
Pr-aclice (Plano, Texas: Business Publications, inc., 1984).
16. "Motney Market Mutual Funds," IVall Slreet Joiutntal (November 18,
1985).
17. Murray, Michael L., "Analyzing the Investment Value of Cash Value Life
Insurance," Tie Jomnlrll of Risk (I(1 Insuamnce,Vol. 43, No. I (March,
1976), pp. 121-28.
18. Myers, PhlyllisS. and Pritchett, S. Travis, "Rate of' Return onl Differential
Premiums for Selected Participating Life Insurance Contracts," The
Journal of Risk aind Insurar-nce,Vol. 50, No. 4 (December, 1983), pp.
569-86.
19. Schwarzschild, Stuart, "Rates of Return onl the Investment Differentials
between life Insurance Policies," The Jouirnal of Risk and Insuraince,
Vol. 35, No. 4 (December, 1968), pp. 583-95.
20. Sebastian, Pamela, "Tax-F-ree Money Fund Yields Stay Sweet," Wall
Street Journal (July 7, 1986), p. 21.
21. Standard & Poor's Statistical Service, Basic Statistics (New York:
Standard & P'oor's, 1985).
22. Standard & Poor's Statistical Service, Cutrrent Statistics (New York:
Standard & Poor's, 1986).
23. Standard & Poor's Statistical Service, Secutrfr)'Price Itidex Record (New
York: Standard & Poor's, 1984).
24. Stein, Roe & Farillhan, 7izx Exempt Bonld Fuid 77lild Quarterl Repolt
(September 30, 1985), p. 3.
25. "Universal Life Interest Sensitive Whole Life Policy Comparison," Best's
Review Life and lHealth Edition (May, 1985), pp). 156-77.
26. Warshawsky, Mark, "Life Insurance Savings and Trhe After-Tax Life
Insurance Rate of Return," The Journal of Risk and Insurl-anice, Vol. 52,
No. 4 (December, 1985), pp. 585-606.

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions
 Univer-sal/l,'(riale L.ife Insutr(ance' 477

APIIwIENDI)X

Rates of Retiurnon Investime Altcrnatives'

1976- 1985
Arithinctic G;nceimric Slandnri
InvesImcnt Miediuirm Mlcnn M~cen Mlininimini Mladxiimm N'aile

Six-month Treasury Bills 9.1 9.1 5.3 13.8 9.0

Long-term Corporate I I. 11.4 8.5 14.6 11.5
Bonds
Equities 14.7 14.0 -7.2 31.5 14.0

Universal/Variable Life lExpense Loadings and SurreindcrCharges"'

Sland:nrd
Tipe (of I: gprnc Mitean Mlinimimi N1a1xiniim Valiac

Initial ExxpenseLoading 15.0 3.0 61.6 15.0

Renewal Expense Loading 7.7 2.5 20.() 7.5
Surrender Change:
Year 1 53.1 1.7 100.0 50.0
Year 5 9.6 0.2 44.1 1(0.0
Year 10 1.8 0.1 10.9 2.0
*zero valiecs excellt(ld

1 hlicsourcc for tlc ratcs of return was Standard & Poor's Statistical Service. Six-month
treasury bill rates were in (21. p. 161 and (22 (May), p. 41. The long-terni corporate bond rates
were the annual average yield to matiurity for composiie bonds rated A 123, p. 2701 and [22
(January), p. 281. The equity return is the total return, dividends plus change in price, of the S &
P 500 (23, p. 125-1261 and (22 (May), p. 30).
hThe expense loadings and surrender charges were calculated from the data provided in Best's
Review (251 for 131 univcrsal life insurance policies.

This content downloaded from 169.229.32.137 on Thu, 8 May 2014 22:53:10 PM

All use subject to JSTOR Terms and Conditions