INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… AUTHOR:
Patrick J. Collins, Principal
DATE:
455 Market Street, Suite 1450
San Francisco, California 94105
PHONE:
415.291.3000
TOLL-FREE:
FAX:
877.291.2205
415.291.3015
http://www.schultzcollins.com
July 2013
INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 Introduction Background
Successfully managing a modest-sized family trust portfolio during a bear market can be difficult when the trust is making
distributions. Both personal and corporate trustees face the task of providing adequate and sustainable cash flow to the current
beneficiary from a diminishing pool of capital and meeting obligations to remaindermen. Although bear market environments
generally increase the desirability of capital preservation strategies, winding down equity risk in a low-interest rate environment
may place the trustee in a further bind. As global stock values sink, a flight to quality often means acceptance of extraordinarily
low yields on principal-guaranteed investments. Even at low rates of inflation, historically low yields on default-free fixed income
investments cannot protect against erosion of purchasing power.
Unfortunately, this type of economic climate creates a favorable habitat for a variety of reach-for-yield strategies which,
historically, have been as safe as storing nitroglycerine in a mix-master. The pressure to pursue such strategies may increase by
virtue of the presence of current beneficiaries anxious to continue generous distribution streams and remainder beneficiaries
expecting to receive generous terminal wealth distributions. Even total-return distribution formulae (based on a percentage-ofcorpus policy or on the power to adjust principal and income) may not be able to solve the dilemma of too little resources and too
much demand. In a nutshell, the trustee faces a situation where interested parties demand the removal of fifteen pounds from a
ten pound sack.
Recessionary environments also spawn litigation risk and regret over marketing materials. The Settlor’s files may even be papered
with brochures asserting that a trust department offers well-constructed portfolios capable of weathering adverse economic
conditions during down-market cycles. Marketing efforts may suggest that institutional resources make the trustee a superb choice
for managing portfolios on a risk-controlled basis. Advertisements may even claim that perceptive personnel monitor markets
continuously to spot opportunities for investment growth or to avoid threats to financial security. Trust department executives may
nevertheless be asked to explain how a portfolio’s value dropped by forty percent after fees, taxes and distributions, and they may
be invited to reconcile performance in a bear market with marketing promises.
The Sample Trust
A modest-sized family trust entails investing a limited amount of capital to fund trust objectives. Trusts vary in their objective.
Some will permit current distributions only in emergencies and focus on growth for remainder interests. Some will be silent on any
tilt toward either current beneficiaries or remaindermen. Still others will provide for the support of current beneficiaries in
reasonable comfort or only for their basic needs, even to the exhaustion of the interest of remainderman. While trusts that are
silent on favoring either current beneficiaries or remainderman and trigger a pure duty of impartiality present key issues addressed
at Section IX, it will be easiest to illustrate the central metrics recommended here by focusing on trusts that are clearly intended
to provide primarily for the basic needs of the current beneficiary.
The Questions
With that focus, the following questions arise: If the trust’s value falls as the result of a bear market investment climate, how does
the trustee monitor the portfolio to determine if threshold financial objectives continue to remain feasible? At what level of wealth
does the trustee determine that it is unlikely that the trust has a reasonable probability of successfully meeting critical objectives
over the remaining planning horizon? How does the trustee communicate a dynamically unfolding problem to the beneficiaries?
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 2 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 What solution steps are prudent for a trustee seeking to maintain a sustainable wealth management and portfolio distribution
strategy in the face of limited resources? What are reasonable steps to avoid litigation by irate beneficiaries?
Effective monitoring and communication are key to avoiding a worst-case outcome in which the trustee must inform the current
beneficiary that funds are no longer available to pay for minimum expenses, not to mention those historically covered by the trust,
and advise the remaindermen that there may be no remainder. Such a monitoring system consists of something more than merely
checking boxes on a trust department procedures list so that the file documents the stock and bond holdings, the weighting of
stock sectors, minutes of the yearly review committee meeting, compliance with tax reporting and return preparation requirements,
etc. Paradoxically, it is the smaller-sized trusts that may have the greatest need for sophisticated surveillance and economic
analysis because of the potentially grave economic consequences of significant declines in trust principal. This requirement, of
course, presents a further dimension to the theme of “bad choices” because it constitutes a commercial challenge to trustees who
must also keep a close eye on the profitability of their service packages.
I. A Sample Fact Pattern Conceptually, it is useful to consider trust management as an activity within “free boundary” problems. The following fact pattern
may be helpful as a point of reference. Assume a current beneficiary who is a widow age 75 and in good health. She requires a
minimum, inflation-adjusted income of $6,000 per month, and—fortunately for her—the governing trust agreement provides for
meeting her minimum income needs. Two years ago, the market value of her trust assets was $2 million, and the annual pre-tax
distribution was $72,000. Trustee fees and investment costs were $20,000, for a total yearly outflow of $92,000. The distribution
equals 4.6% of portfolio value. Today, the market value of trust assets is $1.3 million—a 35% decrease—and the pre-tax distribution
is $74,000, consisting of the constant dollar equivalent of $72,000. There are also $13,000 in fees and investment costs. The total
annual $87,000 outflow equals 6.7% of portfolio value. Initial simulation output from a risk model used by the trustee indicated a
probability of expected success in sustaining value greater than 90%. Rerunning the model today indicates that there is an
approximately 20% chance that the current beneficiary will outlive resources and the remaindermen will receive no financial
benefits. How close is the trust to reaching a wealth level that makes financial success unlikely? Is it prudent for the trustee to
“stay the course”? What planning options are available and how should the trustee communicate them to the beneficiaries? Section
VII of this paper [“Locating the Free Boundary”] revisits this fact pattern.
II. Trust Management as a Free Boundary Problem Considering management of modest-sized family trusts as a free boundary problem provides a host of conceptual, practical, and
commercial advantages to the trustee. One class of free boundary problems, known as “Stefan” problems, involves estimating the
1
demarcation between two regions where the line of demarcation is not fixed. A classic example is estimating the location of the
boundary between solid ice and liquid water when the temperature drops below freezing. In winter, the depth of a Minnesota
lake’s boundary between ice and water fluctuates according to the random variable of water temperature. In cold weather, the ice
pack is thick; in warm weather, it is thin. Analogizing to trust management, the trustee faces the problem of determining the line
of demarcation between two regions—a region of wealth surplus, in which the current value of the trust portfolio is able to support
financial objectives; and a region of wealth deficit, in which the current value of the trust portfolio is not able to support financial
objectives. We call the first region the “feasible region” and the second region the “unfeasible region.” A bull market tends to
move a trust farther into the feasible region—the region of wealth surplus. A bear market, however, tends to move a trust toward
FOOTNOTE 1: Friedman, Avner, “Free Boundary Problems in Science and Technology,” Notices of the American Mathematical Society (September, 2000), p. 854.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 3 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 the line of demarcation—the free boundary—that separates the regions. As wealth depletion pushes farther and farther toward the
region of unfeasibility, the consequences of trustee actions or inactions are magnified in the sense that an imprudent asset
management strategy may not only generate losses, but losses from which the portfolio can never recover.
The trustee has to know in effect the thickness of the ice pack, lest undue optimism regarding the temperature variable induces
the fatal mistake of conducting trust affairs on thin ice. Letting a trust portfolio continue on a path which may lead to a break in
the ice is a dangerous strategy. Once the ice cracks and one falls through and is unable to climb out, no matter what the
temperature is on the next day, the disaster is irreversible. When a portfolio runs out of money, even spectacular subsequent
market performance is of no use.
The following graph presents the geometry of the free boundary concept as a trust moves from the region of wealth surplus toward
the region of wealth deficit:
The boundary is “free” in the sense that there is not a fixed dollar value that can act as the minimum required lower bound for all
family trusts. It is the ratio of trust wealth to costs and distributional demands that determines the boundary’s location. This ratio
differs for each trust and, in some cases, might require the trustee to look beyond the four corners of the trust document to
2
consider other beneficiary resources. The boundary location changes constantly as a function of beneficiary aging, health changes,
inflation rates, liability discounting rates, and other factors. In a bear market, the interaction of investment results with these
factors will tend to push the portfolio value towards the boundary at either a slow velocity or, in the event of a perfect storm, at a
rapid velocity. The free boundary continues to be the point which determines the “feasibility condition”—the dollar value that
separates the prospect for a successful financial outcome from the prospect for trust failure and possible beneficiary ruin.
Although penetrating the boundary is not an event that generates an explicit signal—there may be many thousands of dollars
remaining in the portfolio at the time the boundary is breached—it nevertheless is an event that the trustee should not take lightly.
FOOTNOTE 2: The assumption here is that the trustee has determined the required minimum or lower-bound periodic distribution, irrespective of whether it is expressed in
nominal or constant dollar terms, that the trust must provide lest the current beneficiary experience catastrophic economic failure. Thus, the demands on the trust are net
of any other sources of income.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 4 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 In terms of trust stewardship, it is perhaps the single most critical piece of information that the trustee should know. Keeping a
close eye on the feasibility condition puts a premium on intelligent trust monitoring. While the usual information a trustee
generates concerning a trust is useful, it is crucial to know how likely it is that even a one standard deviation move to the downside
of the forecasted mean for trust performance could prove to be an economically non-survivable event. Hence the importance of
monitoring the trust in terms of the probability of its future viability—instead of the probability of hitting purely extraneous
performance metrics such as a market index. Beneficiaries spend dollars—not market indexes. They need to know whether they are
in trouble, not whether equities outperformed the S&P.
III. Trust Monitoring and Surveillance Policy In terms of monitoring, we suggest that trustees focus on the legal and financial issues in the following questions:
• Where is the free boundary line or, equivalently, what is the current dollar value of the free boundary line?
• How far is the portfolio’s current value from the line?
• Has the trust’s risk tolerance changed with the change in its dollar value?
• What is the probability that the trust will move into the region of economic unfeasibility?
• If principal value continues to move toward the free boundary, is it prudent for the trustee to continue to pursue its
“disciplined investment strategy”—i.e., stay-the-course?
• Is it a breach of fiduciary duty to allow the portfolio to drift into a region of unfeasibility?
• If a trust has drifted into the region of unfeasibility, what is the probability that trust wealth will rebound and return, at a
future date, to the region of feasibility?
• Is an asset management strategy based on hope for such a rebound—“trust the market”—prudent?
This set of issues may differ considerably from some of the questions that trust departments commonly pose regarding their
operations: Is the portfolio sufficiently diversified? Is the trust following its asset allocation guideline? Has the distribution formula
been correctly calculated? Do quarterly reports provide detailed information regarding investment transactions, costs,
disbursements and current trust value? Are specific Settlor directions communicated and executed? The items of concern to risk
managers in trust departments often flow directly from the recommendations of regulators such as the Office of the Comptroller of
the Currency. These “best practice” recommendations often focus on mitigating litigation risk and, hence, on preserving bank or
trust company capital rather than on assuring the most favorable outcomes for trust beneficiaries. It is, however, critical to know—
no matter how much runway lies ahead—that it is not enough.
Risk that is vague and ill-defined is not conducive to effective decision making. A monitoring system that clearly illustrates ranges
of probable outcomes along a decision path can greatly facilitate decisions. Reference to the free boundary may be the easiest and
clearest communication tool.
Beyond the obvious observation that a trustee with discretionary distribution powers should pay close attention to the economic
consequences of diminishing the asset base that supports future wealth accumulations and distributions, it is worth noting that
both investment strategies and asset allocations may differ considerably as a function of distributional stresses. The stress of
distributing a fixed amount in all economic environments differs from the stress of distributing a fluctuating amount dependent on
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 5 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 investment performance. Although the key to effective coordination of trust management strategies is the monitoring and
surveillance program, one finds surprisingly little discussion on this topic. Most commentary focuses on tools and techniques for
enhancing the value of a portfolio in wealth accumulation—as opposed to decumulation—mode. By contrast, the case we consider
here is a modest-sized family trust which must make periodic distributions despite experiencing poor investment performance. This
case is especially helpful in that the value of a credible monitoring program is best illustrated when things are going poorly.
IV. Some Important Preliminary Insights The “modern portfolio theory” work that originated with Tobin, Markowitz, Sharpe, Treynor and others generally assumes either a
portfolio operating within a single period with no cash flows, or a portfolio operating over a multi-period horizon with the goal of
optimizing accumulation of terminal wealth at a level of risk acceptable to the investor. As academic theory began to shape
investment practice, academic insights became a wellspring of investment advice for a “representative” long-term investor
attempting to accumulate wealth over market cycles of varying length. Commentators noted the difficulty in predicting the
evolution of the real economy and, because the functional relationship between the real economy and financial assets constantly
changes, the difficulty of predicting the stock and bond markets. As a result, market timing is generally considered a risky proposition. Rather than accumulating wealth through a stock picking or treasure hunting strategy, many commentators recommended
concentrating on asset allocation. They conclude that setting long-term return expectations through strategic asset allocation is
the primary determinant of investment performance, diversification is a key element in portfolio risk control, and the best
investment strategy is to avoid the high-cost game of trying to beat the market and simply stay-the-course.
It is interesting to observe how the financial planning profession translated academic models into generalized investment maxims
as stay-the-course became the mantra of many in the profession. Variations on commonly used justifications for stay-the-course
trust investment policy include: “we’re in it for the long-term,” “the portfolio is designed to provide attractive returns throughout
the complete market cycle,” “stocks have always outperformed bonds over 20-year periods,” “the average recession lasts only 36
months,” and “if you are out of the market, you may miss the sudden upside moves that provide much of the investment rewards.”
The common theme is that each statement deals only with the asset side of trust administration. An effective monitoring system,
by contrast, demands consideration of both assets and distribution needs—or liabilities. The central issue is whether the assets are
sufficient to continue payments to beneficiaries throughout the trust’s planning horizon.
Charles Ellis, for example, provides a classic statement of investment policy in terms of a stay-the-course philosophy: “The
principal reason for articulating long-term investment policy explicitly and in writing is to enable the client and portfolio manager
to protect the portfolio from ad hoc revisions of sound long-term policy, and to help them hold to long-term policy when short-term
3
exigencies are most distressing and the policy is most in doubt.” The initial strategic asset allocation sets the expected long-term
return of the portfolio at a tolerable volatility level. There is a presumption that, absent significant changes in investor circumstances,
investment policy should remain largely static and that spending policy should also be merely a reflection of the long-term return
estimate. The asset-only point of view, however, is changing as the investment profession moves from static to dynamic investment
policy. The CFA Institute, for example, acknowledges that “asset allocation policies are likely to change over time as characteristics of the investor change and as market circumstances vary” and that “…volatility as a descriptive measure of risk may be
4
irrelevant beyond an absolute level of loss that would completely derail an investment portfolio.” Reference to “…an absolute
FOOTNOTE 3: Ellis, Charles D., Investment Policy (Irwin Publishing, 1993), p. 55.
FOOTNOTE 4: CFA Institute 2010
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 6 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 level of loss that would completely derail an investment portfolio…” is the free boundary and raises a key question: what should be
a trust’s investment policy as it approaches this absolute level?
The rationale for staying the course—adhering closely to a diversified target asset allocation—may represent a generally successful
investment strategy for meeting long-term wealth accumulation objectives. It does not, however, adequately address wealth
decumulation objectives—where trust resources are used to fund periodic distributions to the current beneficiary—together with
wealth preservation objectives—where nominal or inflation-adjusted principal is supposed to pass to the remaindermen. In the
presence of cash distributions, reliance on long-term average return may be inadequate. Although the trust’s actual rate of return
may converge to expected return given a sufficiently long planning horizon, this provides no assurances that the trust’s dollar
wealth will converge to an amount sufficient to avoid portfolio depletion.
5
The order of returns—as opposed to the long-term average of returns—matters whenever there are cash flows. The term “sequence
risk” describes this asset management problem. If early returns are high when there are many dollars in the portfolio, future
negative returns are not as much of a threat to long-term wealth. However, if early returns are negative, there is not only a
decline in the market value of assets, there are fewer resources with which to fund periodic distributions. An average rate of
return target thus no longer provides a reasonable monitoring benchmark. People spend dollars, not rates of return. Over time,
when faced with the presence of distributions, a trust can be exactly on track with respect to a portfolio’s expected average return,
but wildly off track with respect to the portfolio’s anticipated dollar value. It is the dollar value, however, that must support
financial objectives.
This can be illustrated by the following example: if you start with $1.00 and lose 20% in period one, you need to earn 25% in period
two simply to get back to even [0.80*1.25 = $1.00]. The average return over the two periods is a positive 2.5%; the average gain in
dollar wealth, however, is zero. Even worse, whenever a portfolio experiences periods of gains and losses, distributional demands
act as a further decrement to dollar wealth. This is because, during periods of losses, the distributions act as downside multipliers—
i.e., there are fewer dollars left in the portfolio to “rebound” in a period of subsequent gain. Conversely, during periods of gains,
distributions act as a cap—the full dollar value of the portfolio cannot participate in the gain because assets must be liquidated to
raise cash for distributions. Multiplying losses and capping gains is a process known as “Feeding the Bear.” It highlights the need for
the coordination of asset allocation with distribution policy. Such coordination is difficult without an effective monitoring and
surveillance program that focuses on the free boundary point where the trust fails.
V. Shortfall Risk and the Free Boundary Assume a modest-sized family trust is losing value in a bear market. What is the likelihood that the trust will be unable to discharge
its financial objectives? This is a question that demands attention to the long-term evolution of the portfolio in the presence of
possibly inflation-adjusted distributions. The answer to this question quantifies “shortfall risk”—in our example, the probability of
portfolio depletion while the current beneficiary remains alive, and the magnitude of shortfall risk—the duration of time the
current beneficiary spends without sufficient financial resources. Shortfall risk focuses on the likelihood of depletion in the future.
As shortfall risk increases, a monitoring system can anticipate the problem and can help the trustee determine what corrective
actions can be taken today lest the trust suffer catastrophic failure in a future year. A related question asks what current market
value is equal to or greater than the present value of the trust’s future liabilities? This is a solvency question; or, in terms of the
FOOTNOTE 5: To be clear, we are not recommending that portfolios in decumulation mode eschew diversification. Rather, the presence of cash outflows increases the risk
of all portfolios—diversified or otherwise. Monitoring assures that the risk of the portfolio remains well synchronized with investment decisions.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 7 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 free boundary problem, it is the amount of trust wealth that separates the region of feasibility from the region of unfeasibility. It is
not a future-oriented prediction; it is a question of current solvency. Determining shortfall risk and quantifying solvency by locating
the free boundary are both important components of trustee risk assessment, and both provide important feedback information
regarding asset management dangers and opportunities.
The boundary location constantly shifts on account of multiple variables, typically receding as investment returns become positive
and drawing closer as investment returns become negative. Deterioration in current beneficiary health pushes the boundary away;
a cure for her mortality-impairing ailment pushes the boundary closer. A lower or higher than expected inflation rate may also
influence the location of the free boundary. A trust, the value of which falls below the free boundary, may reenter the region of
feasibility as a result of a change in beneficiary circumstances or as a result of a strong market recovery. However, falling below
the boundary implies that, at the present time, the trust is technically insolvent—it lacks the money necessary to fund the
beneficiary’s threshold consumption target for the planning horizon. Falling below the boundary does not mean the trust is
currently broke—its current dollar value may be substantial. However, relative to its minimum or lower-bound distribution target,
it has violated the “feasibility condition”: assets are less than the present value of liabilities.
The feasibility condition [assets ≥ liabilities] is not easy to calculate for a specific trust. Each trust has its asset allocation, fee
structure, distribution target, beneficiary profile, tax liabilities, and so forth. It is easy to see that the current value of trust assets
constantly changes as investment results are realized. In a single week, month or quarter, the fair market value may rocket up or
down depending on the volatility of trust-owned investments. Changes in inflation and interest rates will change the discount rate
for future cash payments; life expectancy changes can increase or decrease the planning horizon. Indeed, the planning horizon—
and thus the present value of the aggregate liability—changes merely with the passage of time.
Trust wealth thus remains in the region of feasibility when:
The Present Value of Assets ≥ Stochastic Present Value of Distributions + Stochastic Present Value of Remainder Interest +
Stochastic Present Value of Fees and Investment Expenses.
The free boundary for the purposes here is located at the exact point where the present value of assets equals the stochastic
6
present value of distributions required to fund the current beneficiary. Of course, the amount by which assets exceed liabilities
equals a cushion or investment surplus.
Trustee efforts to develop appropriate responses and to communicate them effectively may benefit from seeing the components of
trust wealth. At any point in time:
Total Trust Assets = The Share to be Distributed to the Current Beneficiary + The Share to be Distributed to the Remaindermen +
Taxes, Trustee Fees and Other Costs.
The remainder interest is the economic equivalent of an investment surplus. The term “distribution surplus” is also helpful because
it highlights the concept that as long as the remainder interest is positive, there are sufficient assets to pay the current beneficiary
the projected threshold distribution target. As the surplus shrinks, the risk to the trust’s ability to fund the current beneficiary’s
threshold needs increases; as the surplus shrinks the risk to the remaindermen’s expected terminal distribution also increases. As
the surplus shrinks, risk increases at an increasing rate.
FOOTNOTE 6: Our free boundary definition is arbitrary in the sense that it could also encompass investment expenses and remaindermen expectations.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 8 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 Assume a $6 million trust with a Present Value of distribution projection equal to $3 million. The value of the Trust
7
Asset/Remainder Interest Ratio is 6 ÷ 3 = 2. At this wealth level, a 1% change in the value of trust assets translates in a 2% change
in the value of the remainder interest. Assume, further, a 25% decrease in the value of trust assets to $4.5 million. The ratio is now
4.5 ÷ 1.5 = 3. This translates into a change in the value of the remainder interest of 3% for each 1% change in the value of trust
assets. Finally, assume a further decrease in portfolio value to $3.5 million. The ratio value is now 7. This is a “leverage factor” of
seven to one! As the risk to the remaindermen increases, does the trustee stay-the-course with respect to the trust’s asset allocation?
The risk tolerance of the remainder beneficiaries is almost certainly changing as trust wealth decreases, and, given that this is a
modest-sized family trust, the current beneficiary is also likely to be experiencing discomfort. The wealth to distribution ratio is
another important data point in determining a prudent course.
8
Consider a trust asset allocation that has an expected return of 8% and a standard deviation of 16% [95% probability of a singleperiod return between -24% and +40%]. If the Wealth/Surplus Ratio is 1.5 (relatively modest distribution demands relative to
current trust wealth), the risk to the investment cushion is also modest [multiply all values by 1.5 so that there is a 95% probability
of a single-period change in surplus (remainder interest) between -36% and +60%]. If, however, the ratio value is 4 (relatively high
distribution demands relative to current wealth), the risk to the investment cushion is great [95% probability of a single-period
change in the value of surplus between -96% and +160%]. The expected economic value of the remainder interest could be
eliminated in a single period. The return-to-risk tradeoff embodied in the asset allocation is not per se inappropriate. The trustee,
however, should not take a static initial asset allocation, perhaps approved by all beneficiaries at the time of trust inception, as a
set-in-stone set of marching orders. Rather, the trustee must closely monitor the dynamic interactions among asset price changes,
the strategic asset allocation, and the interests of current and remainder beneficiaries. Paradoxically, to a remainderman
beneficiary, an asset management plan that calls for periodic rebalancing to the strategic asset allocation target may not seem like
a stay-the-course approach. In a decreasing market, the risk to the remainder interest emphatically does not “stay the same.”
As current trust wealth declines, some trustees put a premium on strategies to increase future expected return. The sooner a trust
recovers investment losses, the happier all interested parties. However, strategies calling for higher growth portfolios are more
volatile. While the higher expected growth may reverse investment losses, the higher volatility also increases the probability that
wealth will continue to decline. There is a risk to increasing portfolio risk. At the limit, if the only response to losses is to seek
investment gains by increasing risk, asset management becomes mere gambling. Return and variance move together, and any
attempt to increase return will also increase the probability of failure. Losses do not mandate increasing risk in order to recoup the
losses.
VI. Locating the Free Boundary How do you determine the free boundary location—i.e., the “level of loss that would completely derail an investment portfolio”?
The present value of threshold lifetime income for the current beneficiary is an actuarial calculation. An appropriate benchmark
measure of required wealth is the current price of a single premium immediate annuity generating an income stream sufficient to
FOOTNOTE 7: The ratio is actually a ratio of two distributions rather than two numbers. The numerator reflects the probable range of short-term asset values, and the
denominator represents the probable range of short-term changes in the Present Value of trust liabilities. The current value of trust assets is a reasonable proxy for the
numerator, and the mean of the liability distribution [50th percentile of aggregate distributions to the current beneficiary discounted for inflation] is a reasonable proxy for
the denominator. The ratio does not consider fees and investment charges which would represent an increase in the denominator’s value.
FOOTNOTE 8: The risk tolerance of a beneficiary class is not the same as the risk tolerance of the trust. Trustee actions must reflect the trust’s risk tolerance which, in large
measure, is a function of the Settlor’s preferences and constraints as set forth in the governing documents.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 9 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 9
meet the minimum distribution target. If there is not sufficient wealth to purchase an annuity, the trust is in trouble because
there is a danger that it cannot sustain minimum lifetime income to the current beneficiary let alone provide terminal wealth to
the remaindermen.
Returning to the sample fact pattern of Section II, the family trust portfolio has declined in value from $2 million to $1.3 million.
10
Its shortfall risk, as determined by simulation of a credible risk model, has increased from approximately 10% to 20%. Given the
trust’s threshold distribution demands [an inflation-adjusted $6,000 per month], the trustee is interested in knowing how close the
trust is to reaching a wealth level that would unambiguously place it in the region of unfeasibility. Given the term structure of
11
interest rates in, for example, March 2011, the following single-premium immediate annuity monthly lifetime benefits payable to
the current beneficiary could be purchased for $1,000,000:
• Fixed Nominal Benefit: $7,362 to $7,600 per month
• 3% per year Graded Benefit: $5,901 to $6,123 per month
• CPI-linked Benefit: $5,827 to $5,858 per month.
It appears that the present value of the current beneficiary’s interest is approximately $1,000,000 while the present value of the
remainder interest is approximately $300,000. That is to say, the trust is currently in the region of feasibility at a $300,000
distance from the free boundary.
In practice, monitoring programs seem to report results relative to a benchmark—did the equity position beat the S&P 500 stock
index or did the fixed-income position beat the Aggregate U.S. Bond Market index? A more helpful monitoring system would
indicate if changes are facilitating the attainment of trust goals. Settlors seldom direct trustees to “beat the market.”
VII. Asset Management Choices At this point in our examples, the trustee can communicate helpful information to interested parties. The bear market combined
with investment costs, trustee fees and previous distributions have pushed the portfolio’s value from a high-water mark of $2 million
to its current $1.3 million value. Despite the fact that the force of mortality now operates for a current beneficiary aged 75 rather
12
than 73, the trust’s long-term shortfall risk has doubled. If bear market returns persist, the ultimate feasibility of critical financial
objectives may no longer remain viable. However, if the market recovers recent losses, the free boundary will recede and the
FOOTNOTE 9: Annuity is defined as a single-premium, immediate life annuity with no period certain or refund features. See Dus, Ivica, Maurer, Raimond and Mitchell, Olivia
S., “Betting on Death and Capital Markets in Retirement: a Shortfall Risk Analysis of Life Annuities versus Phased Withdrawal Plans,” University of Michigan Retirement
Research Center (RB 2005 – 053) for a discussion of five types of annuities that fit the general definition of a single premium immediate life annuity.
FOOTNOTE 10: The issue of determining the credibility of a risk model is beyond the scope of this discussion. Suffice it to say that the common practice of using a Monte
Carlo simulation to generate multiple portfolio evolutions under the assumption that investment returns are normally distributed produces spurious results. For a more
complete discussion of risk models and shortfall risk assessment, see Collins, Patrick J., “Trustee Asset Management Elections: Portfolio Performance Evaluation and
Preferencing Criteria,” The Banking Law Journal (February, 2011), pp. 136 – 171.
FOOTNOTE 11: Benefit amounts are the high and low monthly payments offered by insurance carriers participating in Vanguard’s no-load, “institutionally-priced” annuity
center. It is relatively straightforward to incorporate an annuity pricing algorithm into a trust portfolio monitoring application in order to approximate monthly benefits as a
function of expenses, benefit structures, the force of mortality, and the current term structure of interest rates. “Impaired life” annuities are written by several insurers.
These instruments offer increased monthly benefits to annuitants suffering from impairments that are expected to shorten their lifespan. See, for example, Brown, Robert
L., and Scahill, Patricia L., “Issues in the Issuance of Enhanced Annuities,” available on the Society Of Actuaries website at soa.org [2010]:
www.soa.org/library/journals/.../apf-2007-10-brown-scahill.pdf.
FOOTNOTE 12: Shortfall risk from a legal perspective is discussed in Fast, Steven M., Gianopulos, Christiana N., & Macauley, Leiha, “Prudence—From Fuzzy to Precise” ALIABA Course of Study: Representing Estate and Trust Beneficiaries and Fiduciaries, 171 (July 19-20, 2007); and DiCarlo, Donald P. & Fast, Steven M., “Prudence: What Are
the Odds: 3 to 1?” ALI-ABA Course of Study: Representing Estate and Trust Beneficiaries and Fiduciaries, 477 (July 17-18, 2008).
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 10 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 trust’s economic crisis may subside. There is no way to know which path will unfold and, of course, the trustee is not a guarantor
of investment results.
The present-value cost of funding the trusts critical objectives is the monetary value of the free boundary. Does the trustee
purchase an annuity and invest the $300,000 surplus in a growth-oriented portfolio dedicated to growing the remainder interest?
This solution is a variation on the classic CAPM two-fund solution first articulated as the “Separation Theorem” by James Tobin.
Should the trustee segregate assets and manage a two-fund portfolio—an annuity dedicated to the economic interests of the
current beneficiary, and a portfolio of higher risk assets dedicated to the economic interests of the remaindermen? The remaindermen’s interest, under this solution, is no longer put at risk by the demands to fund periodic distributions; and the current
13
beneficiary may welcome the mitigation of longevity risk.
The actuarial argument is that a single premium immediate annuity places the current beneficiary on the optimal consumption path
14
given that the trust owns a limited amount of investment capital. The current beneficiary captures the mortality risk premium
(the extra periodic payment available to each annuitant as the deceased annuitants default their income streams to the remaining
living annuitants) and thus leaves more funds available to invest in a growth-oriented portfolio of risky assets. Annuities can maximize
the cash flows available from a limited amount of trust principal. However, from another perspective, an annuity may be seen as
the ultimate “autopilot” distribution formula which permanently locks in the budget constraint—precluding any possibility of future
enjoyment of bull market returns, and impairing the trustee’s ability to fulfill the duty to control trust property and exercise
appropriate discretion over it.
As noted, management of modest-sized trust’s generally involves an exchange of one type of risk for another. With an annuity
strategy, the trustee decides to consider capital sacrifice in exchange for income support to the current beneficiary. The actuarial
solution may view an annuity purchase in terms of hedging. Just as a trustee might hedge downside market risk through acquisition
of a derivative contract (short futures, long puts, etc.), so also acquisition of an annuity contract is a method by which the trustee
can hedge the systematic risk of longevity. In exchange for this asset management alternative, however, the trustee abandons the
potential for future investment gains and enters into an illiquid and irrevocable financial arrangement with an insurer. A trustee’s
decision to pursue an actuarial solution—even within the structure of a two-fund solution of a risk-free annuity portfolio for the
FOOTNOTE 13: An annuity, in this context, approximates a risk-free asset in that it is an instrument specifically designed to generate a level of lifetime income (assuming
reasonable fees and charges) greater than available through replicating a risk-free bond portfolio—e.g., a series of zero-coupon treasuries. The annuity, as a contractual
obligation of an insurance carrier, is not risk free because there is a probability of default. State insurance guarantee funds may mitigate losses up to a specified amount,
but such funds are not direct obligations of state governments. The annuity can generate higher income because of its “mortality premium”—short-lived annuitants forfeit
future income to long-lived annuitants. In terms of the replicating portfolio, the series of zero-coupon bonds have a 100% probability of default at an uncertain future
date—the annuitant’s date of death. The default—cessation of future payments—creates a wealth transfer from the deceased annuitant to the remaining participants in the
annuity pool. It does not represent a forfeiture of principal to the insurance carrier, although this is a commonly held belief.
This presentation does not advocate for or against an annuity purchase by the trustee, but notes that an annuity’s current price provides an objective and independent
benchmark that serves as a proxy for the free boundary location. Trust purchases of annuities remains controversial. Consider the commentary in the 2011 edition of Loring
and Rounds: A Trustee’s Handbook [§6.2.2]: “A trustee who purchases with trust principal a commercial annuity contract of the life of the current beneficiary is investing in
what amounts to a speculative wasting asset. On its face, at least, it would appear that the remaindermen are being disadvantaged, particularly if the annuity payments
are to be paid to the current beneficiary. Also, ‘at one time, such a purchase might have been deemed improper, as an unsecured loan to the life insurance company.’
While today, purchase with principal of an annuity contract is not per se imprudent, in most actual situations it probably is. Of course, if the trustee is authorized to invade
principal, such an investment may be less problematic. Still, a trustee who contemplates making such an investment should document in advance why he considers it a
prudent thing to do under the circumstances, as well as why his duty of loyalty to the remaindermen and his duty to give personal attention to the affairs of the trust are
not implicated. Certainly the trustee would be ill-advised to rely on the advice and counsel of the insurance broker as to what his fiduciary duties are with respect to the
purchase and administration of the annuity contract.”
FOOTNOTE 14: An optimal consumption path is not necessarily financed by a constant income stream compared with a smoothed consumption—is the ability to keep the
marginal utility of consumption steady, not the ability to keep the dollar value of expenditures steady. This essay does not explore this topic because it defines
expenditures as those required to maintain a minimum threshold standard of living.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 11 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 current beneficiary and a growth-oriented portfolio for the remaindermen—may distort the Settlor’s intentions. Likewise, trading
current wealth for an equivalent lifetime income stream may make the trust more vulnerable to contingencies should the current
beneficiary encounter medical emergencies or other needs not met by regular payments over a lifetime. Are trust beneficiaries
better off with a decision to keep exposure to a dynamically changing equity risk premium or a decision to lock in a mortality risk
premium sufficient to guarantee adequate future income? The exchange of risks must be prudently evaluated by the trustee.
A bear market forces interested parties to consider a variety of planning options, none of which are ideal:
1.
The trustee can stay-the-course and continue to invest according to the risk/reward guidelines established by the
initial investment policy. The danger in such a course is that, in the words of the CFA Institute, the trust may suffer
“…an absolute level of loss that would completely derail an investment portfolio….” If the dollar value of trust assets
penetrates the free boundary, the trustee can hope for a market recovery of sufficient magnitude to restore the
portfolio’s long-term viability. However, there is no guarantee of such a result. Indeed, litigation risks may increase
for trustees who redefine “prudence” as “hope.”
2.
A trustee may conclude that investment portfolios with low Wealth-to-Consumption Ratios require higher expected
returns, even at the risk of speeding up the time to portfolio depletion. The trustee can elect to increase the trust’s
exposure to growth-oriented assets in order to make up recent losses. This is especially helpful if the bear market
terminates and there is a switch to a more favorable investment climate. However, increasing expected future
returns comes at the cost of increasing portfolio volatility. If the favorable conditions fail to materialize, shortfall
probabilities may increase dramatically.
3.
The trustee can pursue a two-fund solution. This might take the form of moving the bulk of portfolio assets to cash in
order to mitigate further investment losses. However, the opportunity cost of remaining in cash during low-yield
environments may be so high that such an attempt to preserve principal merely exacerbates longevity risk. This
choice is also not free from litigation risks, especially if the portfolio remains in cash during a period of strong market
15
recovery.
4.
If an effective monitoring and surveillance system is in place, the trustee might elect to implement a dynamic asset
allocation program where equity risk exposure is a function either of a pre-set floor or of the distance from the free
boundary location. As the bear market unfolds, a dynamic system would typically move the portfolio toward cash.
Such a system—also commonly referred to as Constant Proportion Portfolio Insurance—is not commonly found in trust
departments; which generally elect to implement buy and hold or constant asset mix portfolio management
approaches. A reason for such a decision is generally the costs and complexities of dynamic allocation systems.
Modest-sized trusts may incur far greater trustee administrative expenses under a dynamic allocation approach, with
the extra costs perhaps vitiating the risk control advantages.
5.
16
Another form of a two-fund solution is a division of the trust corpus into an annuity to provide secure lifetime income
to the current beneficiary and a performance-oriented portfolio to provide growth opportunities for the remainder-
FOOTNOTE 15: Menoncin, Francesco and Scaillet, Olivier, “Mortality risk and real optimal asset allocation for pension funds,” FAME Research Paper Series from International
Center for Financial Asset Management and Engineering (2003), reviews multiple fund solutions ranging from Tobin’s two-fund solution and Merton’s three-fund solution to
their own model, which is a five-fund solution.
FOOTNOTE 16: A more detailed discussion of dynamic asset allocation approaches for trust management is found at Collins, Patrick J. & Stampfli, Josh, “Managing Private
Wealth: Matching Investment Policy to Investor Risk Preferences,” The Banking Law Journal (November/December 2009), pp. 923 – 958.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 12 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 men share. An early death of the current beneficiary, however, might prompt inquiries regarding the prudence of a
17
trustee’s voluntarily electing current capital sacrifice to secure a future contingent income stream. This is a buy-anannuity / invest-the-difference approach, which is parallel to a buy-term-insurance / invest-the-difference approach
to asset accumulation.
6.
The worst of all possible worlds is for a trustee to discover that the free boundary has been breached—i.e., the
current market value of assets is less than the stochastic present value of liabilities. Use of the annuity cost metric for
the free boundary makes sense when there are surplus funds. However, in the case of insufficient funds, asset
management becomes more of a gambling question, i.e., do the interested parties want to accept a certain but
unhappy outcome or risk a worse disaster in an effort to outperform an annuity with a more risky asset portfolio? The
untenable nature of this situation highlights the importance of the trustee’s portfolio monitoring and surveillance
policies. When the trust portfolio’s value hits the free boundary, the remainder beneficiary is effectively ‘stopped
out’ of any further interest in the trust assuming a decision to secure the current beneficiary’s interests with an
annuity.
18
Developing prudent responses in the face of unattractive alternatives requires trustees to communicate clearly with interested
parties. A disclosure of the nature and magnitude of the hurdles currently facing the trust, and an intelligent discussion of planning
alternatives, can help interested parties achieve consensus on how best to proceed. Although there are issues concerning the force
of a “sign-off” on an investment plan by an unsophisticated and underinformed beneficiary, having the agreement of all parties is
not only a good way to reduce the possibility of litigation, but it is a good way to foster a helpful working relationship among the
parties.
19
A New Reporting Paradigm Trustees have a duty to report. This duty is typically phrased as an obligation to keep beneficiaries reasonably informed. The
purpose is to enable beneficiaries to protect their interests. (A collateral benefit is that reporting may surface a problem more
quickly and allow it to be addressed before it grows to be more serious.) The key question, however, is: what information is
adequate to allow beneficiaries to be reasonably informed and protect their interests? As noted earlier, trustees typically provide
information about trust holdings, yields, sector concentrations, asset allocation, and sometimes even performance (maybe not
after expense) relative benchmarks like the S&P 500. All of this information is appropriate, but this information is nevertheless very
hard for a non-expert to evaluate in terms of the real issues affecting a beneficiary’s actual interests.
What counts to a beneficiary is none of this information. Moreover, to the extent a trustee seeks protection based on disclosure, it
is critical that the trustee be able to demonstrate that the beneficiary in fact had what we normally would consider to be adequate
information to provide informed consent. None of this information may be adequate even for that.
FOOTNOTE 17: No investor has access to an actuarially fair annuity market. Thus, purchase of an annuity contract involves an analysis of loads, expenses and fees. The
Restatement (Third) of Trusts cautions trustees to incur only “justified” expenses.
FOOTNOTE 18: The change in the remainderman’s risk tolerance as the portfolio value approaches the free boundary is of great interest. The remainderman faces increased
leverage as the trust approaches this threshold. Paradoxically the remainderman might favor increased leverage at this point, as he is effectively stopped-out at zero. If the
portfolio remains below the threshold, he gets no money—so by flipping a really big coin, he increases his expected value by increasing his terminal share if the coin flips
his way while suffering no additional downside if the flip goes against him. Of course, knowing this, the remainderman’s preferences might shift to lower risk as he will lose
a beneficial interest in a positive share of trust value as total assets approach the free boundary. It may be better to be a bit more “prudent” and conservatively maintain a
cushion rather than be aggressive and risk a downside move that may be fatal to his interests.
FOOTNOTE 19: Acquisition of an annuity to guarantee lifetime income to the current beneficiary may be a situation requiring advanced notice to trust beneficiaries.
Restatement (Third) of Trusts [§79, comment d] suggests advanced notice for “important adjustments being considered in investment or other management strategies….”
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 13 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 As indicated here, what counts is whether the trust is likely to succeed or fail and its distance from the free boundary that signals
probable future failure. There are a number of ways for the distance to disaster to be described. The proposal here is that reporting to
beneficiaries plainly incorporate a measure of the trust’s feasibility—one way or another.
A parallel issue is the affirmative duty to declare to beneficiaries that the free boundary is approaching or—worse—been reached or
breached. It is easily arguable that the trustee has a duty to understand the probability that the trust will fail before it does fail
and to disclose that probability to beneficiaries. Doing so would meet the requirement articulated in Section 82 of the Restatement
(Third) of Trusts to keep beneficiaries informed as to “significant developments concerning the trust and its administration,
particularly material information needed by beneficiaries for the protection of their interests.”
Courts have often found that there is an affirmative duty to warn or alert beneficiaries to non-routine transactions in advance. See
Allard v. Pacific National Bank, 99 Wash.2d 394, 404-05 (1983) (trustee breached its fiduciary duty to inform where the trustee sold
the sole asset of the trust and failed to give advance notice to beneficiaries, even though trustee did not need beneficiaries’
consent to sell property). See also Barber v. Barber, 837 P.2d 714, 717 (Alaska 1992) (lower court’s approval of settlement
involving sale of trust property reversed on the grounds that contingent beneficiary did not receive notice); In re Green Charitable
Trust, 172 Mich.App 298, 317 (1988) (trustee had a duty to inform beneficiaries before selling the single most valuable trust asset).
Thus there may be cases where a trustee’s exercise of discretion to distribute to current beneficiaries, with the knowledge that
remaining principal will not recover to the level required by the trust agreement, falls into this “nonroutine” category.
VIII. Finally, Impartiality The trust case examined here in detail was selected because it facilitated illustration of the application of the free boundary
problem to trusts. Trusts that do not provide a trustee with a clear direction to prefer the interests of current beneficiaries over
remaindermen, or visa versa, present a far more difficult challenge to trustees. These trusts cannot, as in the sample case here,
provide for the needs of the current beneficiary to the exclusion of the interest of remaindermen, because they are subject to the
duty of impartiality. That may mean, of course, disappointing impartially.
The duty of impartiality requires that, whenever there are two or more beneficiaries of a trust (whether concurrent or successive),
the trustee deal with respective interests in a way that advantages none over the others. Restatement (Third) of Trusts § 183
(1992). How does a trustee do that? First, the trustee carefully evaluates the circumstances of the trust to be certain there is no
“tilt” in favor of one or another interest. Second, the trustee follows the trust terms. If income only is permitted to current
beneficiaries, then the trustee holds to that limit as defined by any applicable power to adjust or unitrust policy. Third, the trustee
exercises any discretionary power in accordance with applicable state law fiduciary standards, even though it may inadequately
provide for the clear needs of the current beneficiary or shrink the principal that will remain at termination below its original
value—let alone that value adjusted for inflation. Finally, if all things are really equal in the competing claims of current and
remainder beneficiaries, and the duty of impartiality is to be applied literally, the prudent trustee will set the free boundary where
it will allow both (i) the current beneficiary to receive distributions that conform to the dictates of the applicable state law power
to adjust or unitrust policy and looks at total return over that amount as belonging one-half to current beneficiaries and one-half to
remainder beneficiaries. To accomplish that, the trustee sets that standard as the free boundary, monitors it, manages to it—and
clearly discloses to beneficiaries the condition of the trust in relation to it.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 14 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 Appendix I: Chronological Survey of Academic Literature The literature on the topics of asset allocation, longevity risk and portfolio sustainability is extensive. Although hundreds of authors
have made contributions to our understanding of how portfolios behave in the presence of distributions, we outline academically
rigorous studies that have a direct bearing on issues faced by trustees seeking a sound, defensible and administratively reasonable
approach to asset management. We restrict the survey to academic literature not because the numerous articles appearing in
trade-oriented publications lack value; rather because the academic literature is, for the most part, free of sales and marketing
agendas. As such, it represents an independent, objective source of credible information. Two supplemental literature surveys
might prove useful: (1) “Post Retirement Financial Strategies from the Perspective an Individual Who is Approaching Retirement
Age” [2010] by Arnold F. Shapiro and, (2) “Longevity Risk Quantification and Management: A Review of Relevant Literature,” [2008]
by Thomas Crawford, Richard de Haan, and Chad Runchey.
Given an intended audience that may lack the background to slog through pages of mathematical / actuarial / statistical calculations,
any summary of mathematically complex articles must inevitably result in oversimplifications which, to the authors of such articles,
may seem to distort their work. Although we offer no defense against this charge other than “good faith,” nevertheless we suspect
that the benefit of presenting, in non-technical language, a survey of research articles that may otherwise escape notice outweighs
any embedded faults. Although none of the articles directly address trust-owned portfolios, from time-to-time we take the liberty
of extending the discussion in this direction for the sake of increased relevance. Finally, as you peruse this survey, please note that
conclusions reflect insights derived by examining the mathematical models developed by the authors. However, all models have
limitations, and an article’s conclusions do not necessary translate into investment planning prescriptions for actual investors.
Many academic commentaries trace their origin to the 1965 paper by Menahem E. Yaari [“Uncertain Lifetime, Life Insurance, and
the Theory of the Consumer”] which demonstrates that investors lacking a bequest objective and having access to actuarially fair
annuities in a complete market setting—i.e., insurance or financial instruments span all economic risks faced by the investor—will
hold all wealth in “actuarial notes.” Under the Yaari complete market model, annuities put the investor on the “optimal feasible
consumption path.” Yaari’s paper is also an important source for studies using a portfolio shortfall probability risk approach: “The
chance-constrained programming approach requires that the constraint (in this case the wealth constraint) be met with probability
λ or more, where λ is some number fixed in advance, say .95….” However, Yaari, like the authors of most early studies, chooses to
explore investment and insurance alternatives in terms of their ability to maximize investor utility (welfare) over the applicable
planning horizon at an appropriate discount rate.
By 1997, Ho, Milevsky & Robinson [“Asset Allocation Via the Conditional First Time Exit”] phrase the optimization problem in terms
of projecting the time at which a portfolio maintaining a fixed vector of asset allocation weights and earning a stochastic return
suffers depletion under the stress of its distribution requirements. The optimal asset allocation postpones the time of depletion to
the furthest feasible future date. The paper decomposes wealth into asset and liability components: portfolio value at any future
date consists of the compound growth of initial wealth less the accumulated value of an annuity due. The date of ruin—the
“conditional first time exit”—is the limit of time for which wealth is greater than consumption demands. Beyond this time, the
investor will run out of money. Although the authors are primarily interested in solving for the date of financial ruin, their
approach is generally conformable to the free boundary approach. Additionally, the paper was one of the first to argue the need
for a dynamic asset allocation policy in the face of a static spending policy—especially where there is a lower bound designating an
acceptable spending level.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 15 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 Milevsky published a follow on study in 1998 [“Optimal Asset Allocation Towards the End of the Life Cycle: To Annuitize or Not to
Annuitize”] which considers the importance of the Wealth to Consumption ratio in planning for optimal asset allocation. This paper
is an important source for a series of studies by various authors on the general topic of the optimal time to exercise an option to
annuitize financial wealth. The thesis of the paper is that a threshold condition for annuitization occurs when annuity mortality
credits overtake annuity costs. The optimal time to annuitize is expressed in option valuation terms. The recommended course is to
review the portfolio after each period. If there is a wealth surplus—i.e., the portfolio’s current market value is greater than the
present value of projected future consumption, then it is beneficial to delay exercise of the annuitization option. A primary
justification for delay is that the option’s time value could be destroyed by early exercise. Delay, however, is not risk free because
the future costs of an annuity may rise as a function of interest rate changes. The delay strategy, however, implies that wealth
must not be allowed to drop below the present value of consumption. An individual is better off deferring purchase of an annuity if
his “after consumption wealth” at the beginning of period x=1 is greater than the annuity’s cost at the beginning of period x+1.
Milevsky’s argument parallels the free boundary approach in the case where initial wealth is above the lower bound. Whereas the
purchase of an annuity is an irrevocable decision to pay a non-refundable sum to an insurance company, and whereas the decision
eliminates liquidity and the ability to make bequests, the decision to annuitize should be deferred for as long as possible provided
that the risk of failing to acquire an adequate lifetime income stream remains within tolerable levels. Later studies by Milevsky and
others [e.g., the 2002 essay “Optimal Asset Allocation and The Real Option to Delay Annuitization: It’s Not Now-or-Never,” coauthored with Virginia R. Young], de-emphasize the importance of the Wealth/Consumption Ratio in favor of a pure option-toannuitize valuation approach. Holding investments in a risky asset portfolio beyond the optimal stopping time could result in a
further deterioration in portfolio values to the detriment of the investor.
Philip Dybvig’s 1999 paper [“Using Asset Allocation to Protect Spending”] makes a valuable contribution to the topic of portfolio
sustainability in that it is one of the first essays to explore the linkage between asset allocation policy and portfolio distribution
policy. Dybvig argues that asset allocation and spending decisions must be made jointly. Common practice is to link spending to the
long-term expected return of the strategic asset allocation. This is a static linkage and does not provide any dynamic feedback to
the investor. Dybvig advocates a sequential (year-by-year) decision making process in which asset allocation changes dynamically to
reflect changes in the value of wealth, interest rates, spending objectives, etc.
Several papers explore the question of whether annuitization provides an optimal income stream. For example, the 1999 study by
Mitchell, Poterba, Warshawsky and Brown [“New Evidence on the Money’s Worth of Individual Annuities”] notes that annuities are
backed by insurance company owned bond portfolios that are acquired in the capital markets. Whereas the duration of an annuity
is, in general, greater than the duration of an underlying bond portfolio, a trustee who buys a nominal dollar annuity exchanges
systematic longevity risk for systematic interest rate risk. The 2001 essay by Peter Albrecht and Raimond Maurer [“Self-Annuitization,
Ruin Risk in Retirement and Asset Allocation: The Annuity Benchmark”] concludes that annuities purchased in low interest rate
environments produce modest payouts that, in most cases, can be achieved through self-annuitization of a risky asset portfolio.
However, at older ages and in higher interest rate environments, an attempt to match an annuity payout with distributions from a
risky asset portfolio runs a much higher risk of ruin. These observations have implications for investors in down market economies
where interest rates tend to be low. Jeffrey R. Brown, Olivia Mitchell and James Poterba’s 2001 study [“The Role of Real Annuities
and Indexed Bonds in Individual Accounts Retirement Program”] points out that maintaining the constant dollar value of
consumption may not be the optimal consumption path. Smoothed consumption is the ability to keep the marginal utility of
expenditures steady—not the dollar value of such expenditures. Constant (real) annuity income may not reflect the time
preferences of retired investors. Interestingly, J.R. Brown’s paper also published in 2001 [“Private Pensions, Mortality Risk, and the
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 16 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 Decision to Annuitize”] suggests a possible solution. If there is sufficient surplus wealth, the investor may wish to annuitize most
wealth so that there is a choice to spend most or all of the periodic payment if the time preference discount rate is high or to save
and reinvest a portion of each payment if the time preference discount rate is low. Brown’s observations suggest that annuitization
might be considered even when current wealth is well above the free boundary location.
James M. Poterba presented a paper entitled “Annuity Markets and Retirement Security” at the Third Annual Conference of the
Retirement Research Consortium on May 17 – 18, 2001 in Washington DC. Poterba links the topic of the optimal retirement income
stream with the concept of option-based “optimal stopping time” for annuitization. Optimal stopping time approaches to asset
management decision making—i.e., the optimal time to switch from a risky asset portfolio to annuity-based income—constitutes an
important alternative to the free boundary approach. Poterba notes that annuitizing all wealth at a single moment (“an optimal
stopping time”) may not be optimal. It is, in his opinion, a type of annuity market timing. Poterba develops the theme of ‘time
diversification’ of annuity purchases. Variation in bond returns generates a substantial variation in annuity payout rates over time;
and, therefore, annuitizing all wealth at a single moment is a risky strategy. It is interesting to contrast Poterba’s view with that of
John Americks and Paul Yakoboski [“Reducing Retirement Income Risks: The Role of Annuitization”]: “…there is an inherent
tradeoff between maintaining a stock of assets and supporting a flow of income….” The annuity contract, according to Americks
and Yakoboski, is valuable in so far as it allows retirees to achieve the greatest efficiency in spending money throughout retirement.
Ten years later [2011], Moshe Milevsky and H, Huang [“Spending Retirement On Planet Vulcan: The Impact of Longevity Risk
Aversion on Optimal Withdrawal Rates”] echo Poterba’s arguments: “…the utility-maximizing retiree is not willing to reduce their
initial standard of living simply because of a small probability they will reach age 105….They deal with longevity risk by setting
aside a financial reserve AND by planning to reduce consumption if that risk materializes in proportion to the survival probability
and linked to their risk aversion.” The authors quote Irving Fisher (1930): “… [the investor has] a high degree of impatience for
income because he expects to die and he thinks: instead of piling up for the remote future, why shouldn’t I enjoy myself during the
few years that remain?”
J. Michael Orszag, acting either as sole author or as a co-author, contributed several studies in 2002. In the paper “Discrete-time
Drawdown Analytics: Annuities and Drawdown in a Retirement Income Model,” the author takes an option valuation approach in
that the optimal time to annuitize is the point at which the annuity produces an income stream higher than a portfolio drawdown
program. Thus it is the ratio of consumption from the portfolio to consumption provided by the annuity that is determinative of the
optimal time for annuitization—not the wealth-to-consumption ratio. Whereas annuity mortality credits become greater at older
ages, the author generally recommends a top-down approach in which annuitization is delayed until an optimal stopping time.
However, annuities lock in the budget constraint. This means that there is a risk of losing potentially higher income at later ages if
annuitization occurs early in retirement. In a study co-authored with Sandeep Kapur, Orszag develops the concept of the ‘annuity
premium’—the spread between annuity yields and long-term government bond yields. Capturing this spread requires “capital
sacrifice.” Thus, a trustee decision to acquire an immediate payout annuity to protect the current beneficiary’s standard of living
involves defending capital sacrifice. Annuities “crowd out” bonds from the portfolio because of the benefits of the annuity
premium. If, however, equity is used to purchase an annuity, the decision is to exchange the expected equity risk premium for the
annuity premium. Orszag and Kapur contend that whenever the annuity premium exceeds the expected equity risk premium, the
option to annuitize should be exercised. This is a preference-free calculation—a question of stochastic dominance rather than
investor utility or target income feasibility. However, a third essay by Orszag [“Ruin in Retirement: Running Out of Money in
Drawdown Programs”] reintroduces a preferencing criteria by establishing a lower-bound consumption target. Orszag solves a
differential equation to determine the income withdrawal rate target that is the reciprocal of the consumption target. The
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 17 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 withdrawal rate calculation, however, ignores the influence of stochastic mortality. It is defined as the rate that will drive the
portfolio value to zero after x number of years. Orszag further notes that annuities are more attractive than bonds in low interest
rate environments because the annuity mortality premium becomes a greater percentage of total current income. Orszag’s view
contrasts, for example, with that expressed by Peng Chen and Moshe Milevsky [“Merging Asset Allocation and Longevity Insurance:
An Optimal Perspective on Payout Annuities”]. Chen and Milevsky argue that payouts on annuities are a function of the prevailing
interest rate at the time the contract is executed. As a consequence, “locking-in a fixed annuity is implicitly a market timing play.”
Unfortunately, a study of profound importance to trustees wishing to place their asset management activities on firm ground is
seldom quoted in the literature. In 2003 four Dutch actuaries, S. Vanduffel, J. Dhaene, M. Goovaerts, and R. Kaas explore the topic
of how much money (“reserves”) an insurance company needs in order assure funding adequacy for a stream of future liability
payments at a given confidence level [“The hurdle-race problem”]. The authors develop an approach to portfolio management that
forms a solid basis for monitoring portfolio sufficiency and for implementing procedures to protect the fund’s ability to provide the
required future payments. If one wishes to meet the payment obligations with certainty, the replicating portfolio consists of n
zero-coupon bonds assuming that the liability is deterministic. The paper, however, also calculates the optimal reserve when the
investment portfolio generates stochastic rather than certain returns. In this case, the objective is to determine the reserve or,
provision “…such that the probability that we will be able to meet our future obligations will be sufficiently large. Conversely, if
the level of the provision is given, our methodology will enable us to compute the probability that we will be able to meet our
future obligations under the given investment strategy.” When the level of the provision or reserve is given—such as the case for an
irrevocable trust—“…the optimal investment strategy could be determined as the one leading to the maximal probability that we
will be able to meet our future obligations.” An interesting variation on the “hurdle race” approach to asset management is the
2008 essay by Jason S. Scott “The Longevity Annuity: An Annuity for Everyone?”] in which Scott suggests that the initial “provision”
might be relatively small if the investor is willing to purchase a contingent payout life annuity that provides a periodic income only
to the pool of contract holders who survive to an advanced age.
The Hurdle-Race Problem paper details how to calculate the provision when the future payments are known and when the provision
is invested in a stochastic return process. Given a stream of liability payments, the provision must have a value equal to or greater
than zero at the end of the applicable planning horizon. The reserve is adequate if its level is greater than the stochastic present
value [PV] of the payments to be made. The liability is deterministic but the present value of the liability payment stream is
subject to changes in the discount rate. Therefore the PV of the liability is also stochastic. A limitation of defining the optimal
reserve in terms of “reaching the finish” is that there may be situations where the interim value of the reserve falls below a
threshold level which may violate regulatory requirements or which may, in general, represent an undesirable situation. The
problem becomes one of calculating the optimal reserve in terms of not only the ultimate goal but also in terms of the period-toperiod reserve value: “…the conditions that year-to-year the provision Rj is larger than a given deterministic value Vj with a
sufficiently large probability. These additional requirements are the ‘hurdles’ that have to be taken.” The authors’ approach shares
many similarities with the free boundary framework for prudent asset management. In particular, its emphasis on meeting interim
“hurdles” or solvency checkpoints as part of a safety first portfolio management approach, meshes quite well with the geometric
expression of the feasible and infeasible regions outlined above.
We can draw a direct line from Yaari’s 1965 article to the 2003 study by David Blake, Andrew Cairns and Kevin Dowd
[“Pensionmetrics 2: stochastic pension plan design during the distribution phase”]. Yaari, as noted, opts to use a utility-of-wealth
approach to determining the optimal consumption path for investors with limited resources. However, in order to achieve
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 18 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 mathematical tractability, Yaari, as well as most commentators following in his footsteps, assumes that the utility of wealth
function remains constant despite changes in the level of wealth. Although it had long been recognized that constant relative risk
aversion applies to investors only under a limited number of conditions [for a helpful discussion see “Utility with Decreasing Risk
Aversion,” Gary Venter, 1983], nevertheless, through the end of the 20th century, many studies assess mathematical models based
on the twin assumptions of log-normal risky asset return distributions (geometric Brownian evolutions) and constant relative risk
aversion on the part of investors. These assumptions have profound consequences. If return distributions are similar to the bell
curve, then a sequence of below-the-mean returns is merely a series of unlucky draws from a stable distribution. Realized results
carry little information regarding the desirability of holding the risky asset portfolio in future periods, and changes in investment
strategy may be unwarranted because the nature of the underlying distribution remains unchanged—i.e., the probability of future
sequences of positive or negative returns remains unchanged. Likewise, the assumption of constant relative risk aversion assumes
that an investor maintains full willingness to continue to undertake a risky venture despite the pressure on current surplus—trust
resources above the free boundary line. Operating together, these twin assumptions provide a powerful rationale for staying the
course, trusting in the restorative power of capital markets, and maintaining equanimity in the face of decreasing wealth. Blake,
Cairns and Dowd, by contrast, restore the focus to the wealth/consumption ratio. The authors argue that optimal annuitization is a
function of investment performance and the size of the wealth fund. Smaller sized accounts produce less income and the marginal
utility of the excess income from annuitization has greater value: “…the marginal utility of consumption gets large as the fund size
gets small.” The annuitization trigger changes from an optimal stopping time problem—annuitize when the annuity mortality
premium exceeds the equity risk premium—to a decision that is directly related to the size of the portfolio relative to the
distributional demands placed against it: “the size of the fund is directly related to the propensity to delay annuitization: the
larger the fund, the longer the delay.” We also refer readers to an excellent 2003 discussion [“Annuities and Individual Welfare”
Thomas Davidoff, Jeffrey Brown and Peter Diamond] regarding the economic assumption of complete markets—an assumption that
starts with Yaari and continues throughout the history of academic commentary. Davidoff, Brown and Diamond relax Yaari’s
complete market assumption and demonstrate that, absent a bequest motive, investors will annuitize all wealth as long as
annuities pay a higher return than assets of comparable risk.
Gabriele Stabile [“Optimal Timing of the Annuity Purchases: A Combined Stochastic Control and Optimal Stopping Problem” 2003]
defines a region for which it is not optimal to annuitize as well as a region in which the investor immediately annuitizes all wealth—
the Yaari solution. The regions are separated by a boundary where the boundary is defined as a utility-of-wealth value function
rather than a dollar-denominated free boundary. The optimal time for annuitization occurs when the dynamic programming
solution to the investor value function indicates that the investor preferences are best served through an actuarial as opposed to
an investment-oriented solution. If the investor’s utility-adjusted wealth remains above the lower threshold, then the investment
program continues—i.e., annuitization is deferred. If the investor approaches the boundary from below, however, the decision rule
is to purchase an annuity at the boundary. If the investor does not immediately annuitize at the boundary, the situation might
deteriorate further. The paper coauthored by Russell Gerrard, Steven Haberman and Elena Vigna in 2004 [“Optimal Investment
Choices Post Retirement in a Defined Contribution Pension Scheme”] provides insight into an interesting variation of the boundary
problem. The paper points out that the region of annuitization might occur either at a minimum boundary to protect against
further deterioration in portfolio distributions or at a maximum boundary which is defined as the size of the fund that would allow
for purchase of a large annuity benefit. If the benefit through annuitization was sufficient large, there would be little need for the
investor to continue investing in a risky asset portfolio—the utility of the extra dollar generated by investing grows vanishingly small.
Finally, we call attention to the 2007 essay [“Integrating Optimal Annuity Planning with Consumption-Investment Selections in
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 19 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 Retirement Planning”] by Aparna Gupta and Zhisheng Li. Their article is an important contribution to the annuity-as-safety-net
approach to optimal asset allocation. High levels of wealth diminish the demand for annuitization of financial assets. The model
suggests that there is an upper bound for the age of annuitization because of ‘brevity risk.’ This means that the length of the
planning horizon is a factor in the decision to annuitize. A sudden change in health may make an annuitization decision have poor
results which would decrease utility for both current and remaindermen beneficiaries.
Milevsky reappears in our survey of academic literature. Building on a paper authored by Virginia Young in 2003 [“Optimal
investment strategy to minimize the probability of lifetime ruin”], Moshe Milevsky, Kristen S. Moore and Virginia Young co-authored
“Optimal Asset Allocation and Ruin-Minimization Annuitization Strategies: The Fixed Consumption Case,” in 2004. These papers also
return to a barrier control problem approach. When current wealth is greater than the market price of a target annuity income
stream, the investor will buy an annuity sufficient to lock in the targeted lifetime consumption. However, when wealth is insufficient to purchase such an annuity, the question becomes the optimal time at which to buy the annuity. “…the optimal annuitypurchasing scheme is a type of barrier control.” To the left of the barrier—wealth is below the stochastic present value of
consumption—the investor makes no annuity purchase; to the right of the barrier, the investor will buy an annuity sufficient to
guarantee the targeted periodic income.
By 2005, several research paths find an elegant synthesis in the article “Normative Target-Based Decision Making” by Ali E. Abbas
and James E. Matheson. Classical utility theory states that investors seek to optimize utility over the distribution of potential
outcomes. However, the presence of a lower bound creates a type of ‘step-utility function’ which divides outcomes into two
regions—acceptable and unacceptable. With suitable mathematical transforms, the utility of the investment project is equal to the
probability that the result of the investment is above the “aspiration level” which divides the two regions. Although the aspiration
level is the boundary between satisfactory and unsatisfactory, its location constantly changes because of “changes in the lottery
that the individual is facing.” For investment issues, the aspiration level changes with factors such as wealth level and liability
values. Optimal decision making, according to the authors, is based on the probability distribution faced by the investor.
Probability distributions may, however, change through time and the targets may have to be revised to reflect updated
information: “…pursuing a fixed goal may be operationally motivational when things are going smoothly, but when major impacts,
such as setbacks or new opportunities, create a need to re-evaluate alternatives, the normative approach demands determining
new targets…. Simply maximizing the probability of reaching the old target is no longer optimal.” This paper provides solid
underpinning for asserting the importance of active monitoring and portfolio surveillance in terms of the goals set forth by the
Settlor. We also note the 2003 paper by Michael Stutzer [“Asset Allocation Advice: Reconciling Expected Utility and Shortfall Risk”]
which also discusses differences between the classical approach of portfolio optimization based on investor utility maximization vs.
optimization based on shortfall probability minimization.
Both the Abbas & Matheson and the Stutzer papers argue that the presence of a consumption floor—what this essay terms a lower
boundary based on a minimum income stream required by the current beneficiary—changes the nature of the decision making
process. A 2005 paper by Geoffrey Kingston and Susan Thorp advances this argument further and, in doing so, clarifies the
relationship between investor utility (welfare) preferences and portfolio management elections. The authors discuss an underlying
cause motivating individuals to delay annuitization of wealth: because risky assets carry the expectation of high return, a longer
period of holding such assets “…offers people a chance to improve their budget constraint that evaporates after annuitization. So
even risk averse individuals may decide to delay in the expectation of creating more wealth and enjoying a higher long-term
income.” Individuals who (subjectively) anticipate a long life span may delay annuitization given the potential benefits of (1) lower
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 20 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 future annuity costs—if interest rates increase, and (2) higher returns from exposure to risky assets. Given the fact that an annuity
purchase decision is irreversible and that the real option to annuitize has time value, the Kingston-Thorp argument parallels the
Milevsky-Moore-Young assertion that an investor will delay annuitization until such time that the expected return from the annuity
contract exceeds that of other financial instruments exhibiting comparable risk. In contrast to the Milevsky-Moore-Young model
which assumes CRRA, Kingston and Thorp assume Hyperbolic Absolute Risk Aversion—HARA which is a utility function suggesting
that the investor is sensitive to changes in wealth. Further, they assume that the investor has a fixed consumption floor. Such a
floor is a proxy for standard of living habit formation—or, in terms of this essay, for a minimum required distribution to the current
beneficiary. The authors point out that the commonly used CRRA utility function is consistent only with a constant mix portfolio
management approach. HARA utility functions however can accommodate a buy and hold approach as well as convex payoff
approaches such as “portfolio insurance” and other dynamic asset allocation strategies. Any model using a CRRA utility function
assumes that investors derive utility from consumption irrespective of its absolute level. However, it is plausible to assume that
only consumption above a threshold level generates positive utility. Consumption below a “subsistence” level does not generate
positive value. When such a “non-zero consumption floor” is introduced into the model, a HARA utility function is required to solve
for the optimal strategy. The presence of a “consumption floor” changes the decision making process in that the goal can now be
expressed in terms of surplus optimization. In the Kingston-Thorp model, the minimum standard of living target is fully funded
(“escrowed”) by the annuity. The investor secures this floor income as soon as possible and, therefore, tends to favor early
exercise of the option to annuitize. It is a variation of the two-fund solution approach with surplus wealth invested in the risky
asset portfolio. “It follows that introducing a positive consumption floor has a similar effect to raising relative risk aversion. In
addition, the agent recognizes that it is ‘cheaper’ to store escrow wealth in an annuity rather than a bond portfolio (at least where
there are small enough loadings), creating another incentive to switch into complete annuitization at an earlier date.” Monitoring
is critical in that a key ratio for a trustee is the level of available surplus relative to the changing costs of securing an “acceptance
level” income stream for the current beneficiary. The two-fund strategy also harkens back to a 2002 work by R. Korn and M. Krekel
[“Optimal Portfolios with Fixed Consumption or Income Streams”].
One effect of the boundary control approach to the problem of optimal asset management is a shift in the control variable from
asset allocation (long-term expected return) to the liability or spending variable. The feasibility condition for successful asset
management shifts from the asset side to the liability side. Philip Dybvig’s 1999 argument that efficient asset management requires
a simultaneous solution of the asset allocation policy and the portfolio distribution policy reenters the academic discussion by the
end of 2005. Consider, briefly, two papers:
(1) The 2005 essay “A Sustainable Spending Rate without Simulation,” by Moshe A. Milevsky & Chris Robinson asserts that
retirement is feasible when the Stochastic Present Value [SPV] of a spending plan is greater than current wealth. SPV is affected by
age, asset allocation, and the spending target. The authors test which of these three levers of ‘retirement sustainability’ is of
greatest importance in the prevention of “the probability of retirement ruin.” In general, spending rates higher than 5% of initial
wealth produce unacceptably high probabilities for ruin. The spending decision dominates the asset allocation decisions at a 5%+
rate: “No matter what reasonable portfolio is chosen, asset allocation will not turn a bad situation into a good one.” This is
because return and variance move together and any attempt to increase return will also increase the failure rate. The two most
effective levers for controlling retirement success are postponement of portfolio distributions to a later age or reductions in
consumption targets. This paper represents a change in emphasis from Milevsky’s earlier option valuation approach. Monitoring is
now focused on wealth in excess of a floor value.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 21 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 (2) The 2006 essay by Gary Smith and Donald Gould [“Measuring and Controlling Shortfall Risk in Retirement”] tests the affect of a
flexible distribution policy on shortfall probability. Their model assumes a 50% elasticity of spending—Δ10% wealth relative to initial
wealth generates Δ5% spending relative to initial spending. The flexible spending policy dominates the fixed distribution rule
(smaller shortfall risk and higher terminal wealth across all allocations). Flexibility in distributions (elasticity of consumption) is a
critical factor in portfolio sustainability.
Follow on studies on flexible spending include the 2006 study by R. Gene Stout & John B. Mitchell [“Dynamic Retirement
Withdrawal Planning”], the 2008 study by R. Gene Stout [“Stochastic Optimization of Retirement Portfolio Asset Allocations and
Withdrawals”], the 2008 study by John J. Spitzer [“Retirement Withdrawals: An Analysis of the Benefits of Periodic ‘Midcourse’
Adjustments”], the 2011 study by Moshe Milevsky and H. Huang [“Spending Retirement on Planet Vulcan: the Impact of Longevity
Risk Aversion on Optimal Withdrawal Rates”], and the 2011 study by John B. Mitchell [“Retirement Withdrawals: Preventive
Reductions and Risk Management”].
Many of the early academic papers, although mathematically elegant, employ risk models with simplifying assumptions—complete
markets, a representative investor exhibiting constant relative risk aversion, a log-normal return distribution. The models rarely
pursue a target-based approach to investment decision making. Such an approach, however, is particularly germane to portfolios
with low wealth to consumption ratios—i.e., modest sized trusts. The issue of efficient portfolio management [Prudence] rapidly
shifted away from an asset-only focus towards a consideration of asset-liability management. A good example of this shift in focus
is the 2006 essay by Russell Gerrard, Steven Haberman and Elena Vigna [“The Management of Decumulation Risks in a Defined
Contribution Pension Plan”]. This paper updates their 2004 research cited above. The essay assumes that current wealth is
insufficient to purchase an annuity at the desired level of consumption and that the retiree elects to invest in risky assets with the
hope of achieving a more favorable future income stream. There is a subtle, yet important, rephrasing of the investment issues.
The problem is now expressed as a risk-ruin-to-achieve-future-wealth-goal where the objective is to maximize the probability of
attaining the goal while minimizing the probability of bankruptcy. The question, of course, is whether this is a prudent strategy to
follow for trust-owned portfolios. The authors argue for constant monitoring of fund size relative to its ability to support performance targets. This mirrors the “hurdle race” asset management approach discussed above. In the authors’ opinion, risky asset
positions should be maintained during times of a shortfall in wealth. In their model, over time, the shortfall is “cured” by
continued exposure to risk. This recommendation contrasts sharply with that of Milevsky & Robinson [2005]. The prudent trustee
should also take note of Sid Browne’s conclusion [“The Risk and Rewards of Minimizing Shortfall Probability,” 1999] that the risk of
a risky asset position increases as the time available to correct a shortfall decreases. The reader is also referred to a 2008 paper by
Russell Gerrard, Bjarne Hojgaard and Elena Vigna [“Choosing the Optimal Annuitization Time post Retirement”] in which wealth
falls into a region of “continuation” in which the investor does not annuitize or into a “stopping region” where risky assets are
converted into annuity income.
Not only is there a marked trend to relax many of the strict assumptions underlying mathematical models, but the number of
variables incorporated into risk models also grows larger. For example, in the 2006 essay entitled “Life-Cycle Asset Allocation with
Annuity Markets: Is longevity Insurance a Good Deal?” Wolfram Horneff, Raimond Maurer and Michael Stamos consider the impact of
economic shocks on the demand to hold non-liquid annuity contracts, human capital as a non-tradable asset that is a close
substitute for bonds, and other economic factors. Furthermore, their model is a continuous time barrier control type problem
which considers the option to annuitize throughout the investor’s entire life cycle as opposed only to a one-time option to annuitize
at retirement or a continuous option throughout only the retirement phase of the life cycle. In the 2008 paper by Wolfram J.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 22 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 Horneff, Raimond H. Maurer, Olivia S. Mitchell, Ivica Dus [“Following the Rules: Integrating Asset Allocation and Annuitization in
Retirement Portfolios”], the authors provide a comprehensive list of the important assumptions and variables that characterize the
risk models in the academic literature. One important issue that has recently drawn the attention of several commentators is the
question of insurance company default. Annuities are not risk-free assets. Excellent treatments of this issue are found in the 2007
essay by David F. Babbel and Craig B. Merrill [“Rational Decumulation”], in the 2008 essay by Phyllis C. Borzi and Martha Priddy
Patterson [“Regulating Markets for Retirement Payouts: Solvency, Supervision, and Credibility” found in Recalibrating Retirement
Spending and Saving—Eds. John Ameriks and Olivia S. Mitchell. Oxford Univ. Press], in the 2008 essay by Alicia H. Munnell [“The
Role of Government in Life-Cycle Saving and Investing” found in The Future of Life-Cycle Saving and Investing (Second edition) CFA
Institute eds. Zvi Bodie, Dennis McLeavey and Laurence B. Siegel], and in the 2010 legal essay by S. Andrew Pharies [“Primer on
Commercial Annuities for Trust and Estates Attorneys”].
The consequences of incorporation a broader set of options into the risk models is apparent in the 2007 essay by Moshe Milevsky
and Virginia Young [“Annuitization and Asset Allocation”]. The authors point out that when an investor is faced with an all-ornothing option to annuitize, the optimal time for annuitization of total wealth is when the expected return from the risky asset
portfolio equals the risk free rate plus the annuity mortality credits. However, under a CRRA assumption, “…the optimal time to
annuitize one’s wealth is independent of wealth and is, therefore, deterministic.” When the value of the option to annuitize equals
the expected value of the payoff from the underling portfolio, then exercise the option. This is a boundary-value problem. Beyond
this boundary, the option to delay runs the risk of consuming less in the future than if one exercised the option to annuitize
immediately. However, for an individual with the capability of annuitizing a fraction of wealth at various time intervals, “…the
individual’s optimal annuity purchasing is given by a barrier policy in that she will annuitize just enough of her wealth to stay on
one side of the barrier in wealth-annuity space.” If wealth is spent to purchase an annuity, periodic income increases. In this risk
model, the barrier exists in wealth-annuity space. The barrier’s location is where the marginal utility of annuity income equals the
marginal cost of spending down wealth to secure it. When the benefit of the income exceeds the benefit of retaining wealth, the
investor will exercise the option to annuitize a fraction of wealth to restore equilibrium.
Additional papers provide insight and opinion regarding the merits of a two-fund investment solution. The 2007 study by David F.
Babble and Craig B. Merrill [“Rational Decumulation”] suggests that the utility maximizing investor will not pursue a strategy that
leaves a positive probability of failing to support the threshold level of lifetime consumption. Penetrating this minimum produces
infinite disutility and, given the assumption that utility is additive across all economic states, such a strategy is irrational. Thus, the
model assumes that the investor allocates risk-free assets sufficient to support the minimum standard of living goal. In a multiperiod context, the risk-free asset is an inflation-adjusted annuity like Social Security. If the minimum consumption target requires
periodic income greater than that available through government or corporate pension benefits, the model assumes that an amount
of current wealth will be annuitized in order to fund the deficit. Investors will annuitize up to the point where the marginal utility
of an extra dollar of consumption equals the marginal disutility of spending down wealth to fund an annuity income. This expression of
the risk/reward tradeoff of the two-fund solution has similarities with the analysis that the trustee must conduct when attempting
to balance the interests of the current and remainder beneficiaries.
Published research over the recent years continues to extend the work of previous articles. The 2008 essay by Huaxiong Huang and
Moshe Milevsky [“Portfolio Choice and Mortality-Contingent Claims: The General HARA Case”] asserts that the minimum value of
income needed by the family unit over the life cycle is the main driver of the demand for either life insurance coverage or for
income annuities. The level of income—not marginal utility—is the appropriate measure of financial risk. Feng Li in his 2008 study
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 23 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 [“Ruin Problem in Retirement under Stochastic Return Rate and Mortality Rate and its Applications”] models the distribution of the
present value of a life annuity under stochastic interest rates and mortality. Assuming that current wealth permits the purchase an
annuity that provides sufficient periodic consumption [i.e., current wealth ≥ PV annuity], then electing a self-annuitization strategy
carries a positive probability of ruin. Thus, the PV annuity is an appropriate standard for measuring the risks of self-annuitization.
Risking continuation of managing a portfolio of risky assets in the hope that an annuity generating higher future income can be
purchased depends on the distribution of annuity values. The 2008 essay by Cassio M. Turra and Olivia S. Mitchell [“The Impact of
Health Status and Out-of-Pocket Medical Expenditures on Annuity Valuation”] is an in-depth exploration of the utility of annuitization in the face of uncertain liquidity shocks. Even if the annuitant is in good health, “our stylized life cycle model with uncertain
out-of-pocket medical expenses shows that annuities become less attractive to people facing such medical expenses.” This
observation raises the issue of the attractiveness of annuities in an incomplete market setting: “…when both adverse selection and
uncertain medical expenses are accounted for and annuity markets are incomplete, we show that annuity equivalent wealth values
are fairly low for people in poor health and about 25 percent higher for people in good health.” The economic consequences of
liquidity demands generated by a deterioration in health and physical well being is also the focus of the 2008 essay by Jon Ameriks,
Andres Caplin, Steven Laufer, and Stijn Van Nieuwerburgh [“Annuity Valuation, Long-Term Care, and Bequest Motives”]. The
authors conclude that the demand to annuitize decreases as wealth diminishes. That is to say, there is a positive correlation
between wealth and annuity demand. As trust resources are depleted, the ability to fund lifestyle expenses may diminish. At the
same time, an annuity purchase solution may increase the current beneficiary’s vulnerability to health shocks. At low wealth levels,
a serious medical shock would simultaneously deplete liquid assets, raise current expenses, and decrease the mortality-adjusted
future value of annuity payments. If, as the authors suggest, “retirement security…can be summed up simply as ‘having the
resources you need, when you need them,’” then standard annuities may be only a partial solution to security in the face of severe
health shocks: “such products do little to deal with retirees’ need for resources when emergencies arise, and they can even
exacerbate financial distress in exigent situations.” A risk model that suggests a decrease in the demand to annuitize a retirement
portfolio as wealth decreases also finds support in a 2007 essay by Milevsky and Young [“Annuitization and Asset Allocation”].
Milevsky and Young conclude that the demand for annuities is increasing with wealth, risk aversion, health assessments, and
portfolio volatility. However, the implications of a retirement income risk model may not be directly transferable to management
of trust owned assets. A Trust’s demand to hold annuities may increase as wealth decreases and the current income beneficiary’s
economic benefit becomes more tenuous. This is why sound and defensible judgment is a prerequisite to prudent asset
management.
Appendix II: Mitigating Longevity Risk: A Non-­‐Technical Review of Actuarial Solutions Longevity Risk
Longevity risk is the likelihood that a trust beneficiary will outlive a portfolio of financial assets tasked with providing periodic
income. Although the Social Security Administration’s general population mortality table evidences an increase in life expectancy
over the course of the last century, within the high-income, white-collar population group it is not uncommon for one or more
spouses to live well beyond age 90. This means that an investment portfolio may have a planning horizon greater than 35 years
assuming an income beneficiary in their mid 60s. Additionally, this population group has developed high living standards which, in
many cases, require substantial cash flows if they are to continue unabated throughout their lifespan. Thus, traditional investmentoriented approaches to providing sustainable and adequate long-term income are under pressure. Recent capital market volatility
calls into question the viability of using risky-asset portfolios as vehicles to generate stable cash flows. Although long-term average
returns from stocks and bonds appear to remain attractive, trustees may find that a sequence of negative returns at the start of
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 24 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 trust administration may deplete the portfolio to the point where it lacks sufficient dollars to recover from early catastrophic
declines. To recap: (1) the trust’s assets must support long-lived individuals from a growing population of white collar investors;
(2) this population of high wage earners often manifests preferred standards of living based on high-consumption lifestyle habit
formation; and, (3) the traditional investment-oriented methods of financing long-term cash flow appear to offer tenuous future
results. It is little wonder, therefore, that trustees are considering actuarial approaches either as supplements to investment
strategies or, in some cases, as substitutes for risky investment portfolios.
Annuity Basics
Many trustees have a passing acquaintance with two insurance-oriented products: the life annuity which promises to pay a
20
guaranteed income for the remaining life of the benficiary; and, the reverse annuity mortgage which promises to provide
income—perhaps for life—secured by a mortgage against a personal residence. The life annuity has been around since the Middle
21
Ages while the reverse annuity mortgage program first appeared in the 1980s as a private insurance plan. The reverse annuity
mortgage involves strategies to monetize an illiquid asset and merits a separate discussion. Therefore this discussion focuses on the
following actuarial products:
• The single premium immediate retirement annuity;
• The single premium deferred retirement annuity;
• The variable annuity with lifetime income guarantee rider; and,
• The ruin contingent deferred annuity.
Before looking at each type of product, it may prove helpful to review some basic points regarding an annuity. An annuity is a
contract in which an insurance company promises to make a sequence of periodic payments—usually defined as a sequence lasting
for life—in exchange either for a large single premium collected at the beginning of the contract’s term or for a series of smaller
premiums collected prior to the start of the annuity’s initial payment date. An annuity in which the entire premium is collected at
the beginning of the contract’s term and which begins payments shortly thereafter is called an ‘Immediate Annuity.’ The purchase
of an immediate annuity involves an irrevocable sacrifice of capital in an amount equal to the premium paid to the insurance
company. The purchaser trades a sum of money for an actuarially equivalent income stream which, in turn, is reduced by the dollar
amount of fees, commissions and expenses charged by the insurance carrier. The lower the sales and administrative costs, the
greater is the periodic income, all else equal. The insurance company invests the premiums in the expectation of earning a return
sufficient to cover the obligations to which it has committed itself. Unless there is a special premium refund feature or payment
guarantee provision within the annuity contract, the payments cease upon the death of the annuitant(s).
The basic structure of an annuity sometimes engenders misconceptions among investors unfamiliar with this type of product. One
sometimes hears annuities described as “giving your money to an insurance company that invests in the same stocks and bonds in
which you could invest; and, because all payments cease upon death, allowing the company to reap a windfall because an early
FOOTNOTE 20: The term ‘beneficiary’ can mean a single individual or a beneficiary plus spouse. Where the term encompasses two individuals, the life span measured is the
joint life span. For purposes of expositional clarity, this essay assumes that the annuitant, payer, and contract owner are the same, and that the owner, presumably the
trust, qualifies as a “natural person” under the provisions of U.S. tax law.
FOOTNOTE 21: Merchants often traded personal wealth to local monasteries in exchange for the promise of lifetime support and protection for themselves and their families
within the confines of the monastic institution. Many monasteries later found themselves in grave financial crisis because they failed to collect sufficient wealth to fund
long-term obligations. The reverse annuity mortgage plans of the 1980s were initially privately insured. It was not until the early 1990s that the Federal Housing Authority
initiated government involvement.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 25 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 demise means that you forfeit your money to the insurance carrier.” This incorrect description of the actuarial principles underpinning an annuity misses entirely the concepts of ‘risk pooling’ and ‘mortality credits.’ Here is an example: Assume that eleven
investors, each age 80, contribute $1,000 at the beginning of the year so that they can collectively purchase an $11,000 Certificate
of Deposit maturing in one year and paying a simple rate of interest of 3%. Each investor has the expectation of receiving $1,000 in
principal + $30 in interest one year from today. This calculation assumes that each member of the pool receives a pro-rata return
from a CD with a total maturity value of $11,330. Now, suppose the eleven investors decide that if any one fails to survive until the
22
CD’s maturity date, the share due to that investor will, instead, be distributed to the remaining living members. If one member
fails to survive the requisite period, the remaining members divide the $11,330 CD proceeds into ten shares each of which is worth
$1,133. By pooling risk, the survivors have reaped a mortality credit of $103—an extra return of 10.3%. The deceased pool members
forfeit their shares to the surviving members—not to the bank which issued the CD. The issuing bank paid a 3% rate of interest
which was sufficient to attract capital that, in turn, was loaned out to other bank customers at a higher rate of interest. The bank’s
profit expectations are already built into the CD contract and the bank remains indifferent as to the fate of individual pool members.
Likewise, an insurer guarantees the annuity payout for each contract irrespective of which annuitants—or how many annuitants—
survive during the period. It can do this because of annuity pricing principles. It sets the payouts so that they appear sufficiently
lucrative to attract capital (premiums), and are sufficiently conservative to reserve a slice of the expected mortality credits for
23
insurance company profits.
If an investor is indifferent about making future gifts or bequests, the annuity contract will, assuming
reasonable fees and expenses, always pay an annuity yield higher than the interest rates on comparable fixed income investments
promising a full or partial return of capital to the investor.
24
One further item is noteworthy—an annuity is not a risk free investment. Annuity payout guarantees are only as good as the
insurance company that backs them. Historically, there have been several spectacular insurance company insolvencies which have
resulted in annuity payment delays or even in a loss of a portion of the investment in the contract when annuity payments ceased.
25
Annuities are guaranteed only up to a limited actuarially-defined value specified by state insurance guarantee funds. Investors
may also be unaware that the financial guarantees are not a direct obligation of the state but rather are the pro-rata obligations of
individual insurance companies who market within the state.
An annuity has unique tax provisions which trustees should carefully consider prior to making a purchase decision. Generally, an
annuity enables the investor to avoid recognizing income on any accumulations remaining in the contract. Thus it provides a “tax
shield” enabling funds to accumulate in a tax-favored environment. Additionally, each periodic payment is split into a tax-free
return of principal part and a reportable income part. The percentages of periodic income that are taxable or tax free follow a
complicated set of rules enumerated in the U.S. Revenue Code. However, the portion of income deemed to be taxable is subject to
FOOTNOTE 22: This is a form of contract known as a tontine which is illegal under current law because it may encourage some participants to take steps to assure the
premature demise of other pool members. An insurance company, however, pools many individuals into contracts where the death of any one member does not impact the
payout promised to the remaining annuitants—i.e., there are no “death dividends” to reward the survivors.
FOOTNOTE 23: Technically, profits or losses emerge over time as experience dictates whether the liability reserve for the annuity payments was conservative (profits) or
inadequate (losses).
FOOTNOTE 24: It is, however, misleading to compare the annuity yield to the current yield paid by a bond or a CD. The latter instruments usually involve a return of
principal. The replicating portfolio for an annuity payout, however, is a series of zero-coupon bonds with a 100% probability of eventual default. This is truly an oranges to
apples comparison.
FOOTNOTE 25: In California, the current amount of contract ‘non-forfeiture’ value that is covered under state guarantee fund’s provisions is $85,000 per insurance carrier.
Thus, it is wise to diversify the annuity portfolio over several insurance companies unless the annuity is of a type that is segregated from creditor claims against the insurer.
Independent rating firms upgraded the insurance industry outlook from ‘negative’ to ‘neutral’ following the recovery from the global recession. Several carriers with a large
share of the U.S. annuity market have recently exited the marketplace because the financial guarantees embedded in annuity contracts threatened the companies’
financial condition. In the 1990s, both Japan and Europe saw multiple carrier insolvencies as mispriced guarantees thrust insurance companies into dire financial straits.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 26 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 ordinary income rates. Once the aggregate return of principal exceeds the basis in the contract—i.e., the amount of premium
paid—all distributions are taxed as fully reportable ordinary income. Proponents of accumulation-oriented annuities point to the
advantages of the tax shield that allows the investment to grow on a tax-deferred basis. Critics of annuities—both accumulation
annuities and immediate payout annuities—point to the tax law provisions that convert investment gains usually subject to low
capital gains rates if held outside an annuity to investment gains taxed at higher ordinary income rates when received from an annuity.
The Single Premium Immediate Retirement Annuity
The single premium immediate annuity [SPIA] is, for the purposes of this discussion, a contract that the trustee purchases either
at the start of trust administration or at a later date within the period of current beneficiary’s lifespan. As stated, the term
‘immediate’ signifies that the annuity payment stream begins shortly after collection of the initial lump sum premium payment—
usually within 30 days. Assuming sufficient financial wealth, the attraction of a single premium immediate annuity is the ability
to lock in a lifetime payment stream, perhaps on an inflation-adjusted basis, by exchanging investment and longevity risks for
insurance carrier guarantees—i.e., counterparty risk. The trustee can lock in the lifetime income either upon the commencement
of the trust—assuming a beneficiary of an age sufficient to make an annuity attractive; or, if the trustee prefers to wait and see
how events unfold in the capital markets, at a later date assuming that sufficient wealth remains to cover the annuity’s future
purchase price.
Although there are many ways to use SPIA contracts to generate retirement income, most are variations on two basic approaches—
as a product to provide a base level of income; and, as a safety net:
1. The trustee determines the minimum income required to sustain threshold expenses during the beneficiary’s life.
This minimum threshold may be less than the “aspirational” amount of income that the beneficiary hopes to have
available to fund future consumption. Nevertheless, the idea of locking in a threshold income amount at the start
of trust administration may appeal to certain beneficiaries. Provided that the annuity contract is reasonably
priced, the annuity’s risk pooling and mortality credits make the attainment of lifetime income cheaper than the
income stream available from government-guaranteed bond portfolios. In terms of financial economics, the
annuity portfolio’s pricing advantage “crowds out” the bond component of the portfolio. The portion of trust
assets not used to purchase the annuity income stream may be invested in a diversified stock-oriented portfolio
with the expectation that the portfolio will provide for future capital growth to fund excess consumption for the
current beneficiary or to enhance the interests of the remainder beneficiaries.
2. An alternate SPIA strategy is to consider the annuity in terms of a safety net against a decline in wealth of such
magnitude that it jeopardizes the portfolio’s ability to provide a sustainable and adequate lifetime cash flow.
Under this approach, the trustee delays purchasing an annuity in hopes that traditional stock and bond investments will generate returns at or above those required to fund both threshold and aspirational consumption
objectives. However, should the trust encounter a bear market of sufficient severity to compromise the portfolio’s
objectives, he can exercise the option to annuitize. The option to annuitize at a later age may make it cheaper
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 27 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 to purchase the annuity income assuming that interest rates and insurance company reserving requirements are
favorable at the time the option is exercised.
26
Although the concept of guaranteed lifetime income is attractive, many investors never exercise the option to buy a single
premium immediate retirement annuity. Given the annuity’s pricing advantages relative to more traditional bond investments,
economists refer to the lack of widespread public ownership of annuity products as “The Annuity Puzzle.” Theory dictates that the
product should be popular, reality says otherwise. There has been roughly forty-five years of research and mathematical model
building throughout university economics and finance departments in an effort to explain the puzzle. Some of the reasons put forth
to explain why annuities are not more widely utilized are:
• Distrust of insurance companies and insurance sales representatives;
• Presence of a threshold inflation-adjusted income stream provided through social security;
• Irrevocable loss of capital upon purchasing of the lifetime income stream;
• Importance of gifting and intergenerational bequest planning for retired investors, or importance of a remainder interest in
a trust context;
• Strong time preferences for consumption—current wealth should not be spent to provide funds against a remote contingency
that income might be needed at age 99; rather, the optimal consumption path shows higher early expenditures and lower
expenditures later in life should it prove to be long lasting;
• Fear of unexpected “liquidity shocks” such as extraordinary medical expenses that would require large reserves of liquid
capital;
• Fear of locking in a permanent budget constraint if most wealth is exchanged for annuitized income; and,
• Fear of ceding control of trust wealth with the attendant loss of trustee discretion.
It is interesting to note the repetitive use of the word “fear” in the context of exercising the option to provide for lifetime income
security. Clearly, the SPIA is a powerful financial tool; however, any purchase decision must involve careful planning and informed
consideration.
At this point, the reader might benefit from brief comments on annuity costs and interest rates. During times of lackluster
economic performance interest rates tend to be low. However, these are precisely the times when the insurance industry ramps up
its marketing campaign to highlight how yield-starved investors can capture attractive cash flows by buying annuities. Relative to
the income currently thrown off by CDs and government guaranteed bonds, an annuitized income stream seems almost too good to
be true. However, the intelligent trustee also realizes that when money is paid to the insurance carrier, the carrier must invest the
funds in the same low-yield capital markets faced by all investors. All else equal, when interest rates in the general economy are
low, insurance carriers do not offer the amount of lifetime income per premium dollar that they offer during periods of higher
interest rates. This makes sense because if the insurance carrier can earn only 2% on its assets during poor economies, they cannot
pay out as much as they could if their assets earn 6%. Yet it is during bad economies that annuities look best to those concerned
FOOTNOTE 26: It is cheaper to buy a lifetime income of $x dollars at age 75 than at age 65, all else equal. Generally speaking, only investors who consider themselves to be
in good health find SPIA contracts to be of interest. If the investor does not expect to enjoy a long life span, longevity risk is not an important factor in lifetime income
planning. A few carriers write SPIA contracts with high lifetime payouts for individuals in poor health. Actuaries call these contracts “substandard annuities.” When these
annuities are customized to compensate individuals who have been awarded court judgments for life impairing injuries, they are known as “structured settlements.”
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 28 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 with adequate and sustainable income. In terms of timing, buying an annuity contract in a low-interest / recessionary economy is
27
probably the worst time to do so. This is ironic—the product looks best when the time to buy is the worst!
The annuity product provides lifetime cash flows and, therefore, exhibits interest rate sensitivity [bond duration risk measure] far
higher than many fixed income investments. Thus, when considering the factors contributing to the “annuity puzzle,” the fear of
locking in a permanent budget constraint assumes great significance in light of dynamic future interest rates. The purchase of an
annuity has been called a “pure interest rate bet” by financial economists. This essay began with the proposition that buying an
annuity is equivalent to exchanging investment and longevity risk for counterparty risk. We can now add the further caveat that
buying an annuity is also equivalent to exchanging the risks traditionally associated with portfolios of financial assets for increased
interest rate sensitivity. For the savvy trustee, Beta risk is exchanged for Duration risk.
The Single Premium Deferred Retirement Annuity
The Single Premium Deferred Annuity [SPDA] is a relatively new annuity product offered by only a few companies. At the time of
this essay, we are unaware of any commission free SPDA contract. This said, the SPDA may appeal to certain trusts. Unlike the SPIA,
the SPDA provides a lifetime income that starts ten, fifteen, or twenty years after the payment of the premium—hence, the use of
the term ‘deferred.’ Under this type of contract, the annuitant receives income only if he is alive at the future income start date.
For example, suppose a trustee elects to purchase a $20,000 income stream with a 2% per year benefit step up for a 65-year-old
male beneficiary. The income start date is age 85. If the beneficiary is still alive at age 85, he will receive a yearly income of
approximately $30,000 with a continuing 2% annual step up in benefits for the remainder of his life. However, if he is not alive on
the designated starting date, he will receive nothing. The SPDA contract has been dubbed “pure longevity insurance.”
Not surprisingly, these features result in significant cost reductions relative to an SPIA. It has been estimated that a $40,000 per
year income with a 2% step up in benefits for an annuitant at age 65 may cost approximately $800,000. The same income received
under a SPDA starting at age 85 costs approximately $100,000—a $700,000 savings. The reason for the cost savings are threefold:
(1) only approximately half of the annuitants will be alive to collect the benefits; (2) the insurance company can invest the premium
for a 20 year period prior to paying out benefits; and, (3) the number of years of expected benefit payments is less at age 85 than
at age 65.
This product can be attractive to trusts with investment portfolios large enough to make it unlikely that they will face an income
shortfall within the next fifteen to twenty years.
28
However, if the beneficiary survives beyond this period, continued distributions
may result in portfolio failure rates beyond the trust’s tolerance for risk. Given the above example, an trust investing a $1 million
portfolio needs to transfer $800,000 to the insurance company to secure a $40,000 per year income with a 2% per year benefit
increase under an SPIA contract. This leaves only $200,000 for investments and emergency fund reserves. By contrast, the trust
can purchase the comparable SPDA contract and retain $900,000 in financial assets. In this case, the risk is that the $900,000 is
insufficient to provide the needed income during the twenty-year interval between the trust distribution start date and the annuity
start date. Of course, there is also the risk that the beneficiary will not live until age 85 and will not collect any benefits. However,
FOOTNOTE 27: Holding premium deposit amounts, age, sex, and annuity underwriting pricing formulae constant, the changes in interest rates over time generates what
economists refer to as “the term structure of annuity payments.” Thus, for example, a trustee spending $100,000 to buy an annuity for healthy 65 year old female in 1995
could buy a much higher lifetime income than a comparable annuity purchase for a 65 year old in 2012. The only thing that changed was the level of interest rates.
FOOTNOTE 28: It is the ratio of trust wealth to periodic distributions that is the key factor in determining the likelihood that the trust will face future portfolio depletion.
Trusts with high ratios—lots of wealth and only modest distributions from the portfolio—will be unlikely to gain utility from the purchase of annuity products. Classical
equilibrium theory suggests that the optimal amount to annuitize occurs when the marginal utility of receiving a dollar in periodic annuitized income equals the marginal
utility of retaining a dollar so that it can be reinvested in the portfolio for future consumption growth or bequests.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 29 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 the cost of off-loading longevity risk through the SPDA strategy requires a substantially reduced initial premium outlay from the
investment portfolio.
The Variable Annuity with Lifetime Income Guarantee
Although this annuity product is, by far, the most widely owned contract, cynics might attribute this to the fact that its layers of
opaque fees provide generous ongoing commissions to the sales reps who recommend it. Fortunately, there are a few carriers
offering no-commission products that charge significantly lower fees.
The variable annuity [VA] first appeared in the U.S. during the 1970s. VAs are “fund linked” annuity contracts which, when used for
retirement income purposes, usually require payment of a single, up-front premium. The insurance company offers the contract
owner the right to invest the single premium into a menu of fund offerings where the menu consists of proprietary funds or funds
29
managed by one or more unrelated mutual fund companies. The SEC considers VAs to be securities and requires insurance agents
to hold the requisite securities licenses and to provide investors with a prospectus. One advantage of these products is that they
are not deemed to be part of the general assets of the insurance carrier and are, therefore, not subject to creditor claims in the
event of the insurer’s insolvency. In the current marketplace the investor can find both commission-based and no-commission
30
products—the cost of which can vary significantly. A second significant advantage is that VA contracts are liquid. The investor,
subject to applicable contract surrender charges, does not irrevocably forfeit the funds to an insurance carrier. If sudden expenses
arise, contract can be surrendered, partially or entirely, to provide for the investor’s needs.
The structure of these contracts is complex and the actuarial formulae that determine payouts can seem byzantine. Basically, a VA
contract is a series of one or more pooled investments comparable to mutual funds. The trustee allocates the aggregate investment
account across a menu of money market, bond and stock funds. Occasionally, VA contracts offer more exotic choices such as hedge
funds, commodities, and so forth. The aggregated account is “wrapped” in an annuity bundle so that it qualifies as an insurance
product. This is important because, in its capacity as an actuarial instrument backed by an insurance company, the contract can
(1) offer the owner the right to annuitize the “fund linked” value of the investment account to provide for a lifetime income; and,
(2) offer riders which provide additional financial guarantees for both minimum payoffs upon the death of the annuitant and for
minimum dollar income benefits during the lifetime of the annuitant.
31
This discussion of a VA contract’s arcane structure is important because regulators (SEC and NAIC ) approved the marketing and
sale of VA products based on assurances given to them by carriers that the financial guarantees offered through annuity policy
riders to would be “incidental,” and that such guarantees would not be sufficiently widespread to exercise a material impact on
insurance company profitability. Insurance firms are quite good when it comes to pricing actuarial guarantees; pricing financial
FOOTNOTE 29: The funds under management of outside investment companies may differ from the similarly-named mutual funds offered to investors. For example, in some
cases, the expense structure of the mutual funds may differ from the funds offered through the VA menu.
FOOTNOTE 30: The cost structure of VA contracts is complex and, in the main, falls outside the boundaries of this discussion. This said, cost matters—a lot; and the prudent
trustee should thoroughly investigate the myriad of implicit and explicit costs associated with this product. The author recalls an experience during an arbitration hearing
to petition for trustee removal. The trust’s income beneficiary wanted higher yields and alleged that the trustee was not acting prudently because he refused to purchase a
VA contract with income guarantees. The national head of the private investor wealth planning division for a major insurance-company-owned broker dealer steadfastly
refused to address the question of contract costs for a lengthy period of time under the theory that costs did not matter because the contract owner is entitled to
wonderful benefits net of costs. Caveat Emptor.
FOOTNOTE 31: NAIC is the National Association of Insurance Commissioners. Insurance companies are regulated primarily on the state level with the chief regulatory officer
occupying the office of state insurance commissioner. The fifty state commissioners, in turn, belong to the NAIC which acts as an advisory body regarding the need to keep
or amend current regulatory standards.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 30 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 32
guarantees, however, is a much different ballgame. The NAIC is currently evaluating a new type of annuity contract—the ruin
contingent deferred annuity which is the subject of the next section—and, to the consternation of several major VA carriers, is
also taking the opportunity to reassess the risks and reserving requirements of a broad range of financial guarantees offered as VA
contract riders.
33
The basic VA contract provides the option to annuitize the investment account to provide for lifetime income. However, the
amount of income is not based solely on the dollar value of the account. Rather it depends on the complex interaction of two
actuarial formulae: (1) the number of annuity units within the aggregate account—the value of an annuity unit rises and falls with
market performance; and, (2) the payout value from each annuity unit which is based on the contract’s assumed interest rate (AIR)
per unit. If a trust pays a premium of $1 million which is allocated across the menu of mutual funds, the payout for the remainder
of the annuitant’s life depends on the present value of annuity unit liquidation over life (a unit value of $1 means that the investor
purchases 1 million annuity units) rather than the present value of a dollar-denominated lifetime income steam. In a nutshell, an
actuary swaps annuity units for dollars when calculating the payout benefit. Whereas, in nominal terms, a dollar is always worth a
dollar, an annuity unit is worth whatever the market says it is worth. This means that the income beneficiary receives a steady
lifetime income consisting of annuity units; or, to put it otherwise, an unsteady lifetime income of dollars.
A contract’s AIR sets an assumed rate of return on each annuity unit. If the contract has a low AIR—e.g., 3.5%—that means that an
investment performance by the aggregate portfolio greater than 3.5% will push the dollar value of retirement income higher. If,
however, the contract has a high AIR—e.g., 5%—that means that the rate of return on the aggregate investment portfolio must be
greater than 5% before the annuitant receives any increase in the dollar value of his annuity check. Not only can the annuitant fail
to receive an increase in benefits for the first 3.5% to 5% of positive market performance, the expenses, fees and commissions
within the VA are paid by liquidating annuity units. Assuming a 5% AIR with a 3% total annual fee, the income beneficiary forgoes
34
any benefit increase for the first 8% of aggregate portfolio gain!
It was not misleading for VA carriers to tell regulatory authorities
that annuity guarantees were “incidental” with respect to their earnings and profits because it is highly unlikely that the average
investor would opt for annuitizing a VA contract for retirement income purposes. Clearly, if VA sales were to penetrate the ‘senior
market,’ dollar-denominated guarantees would have to be offered to investors—hence, the appearance of lifetime income
guarantee riders.
VA income riders became widespread in 2002 with the marketing of a Guaranteed Minimum Withdrawal Benefit [GMWB]. Prior
to that time, most guarantee riders assured the contract holder that the beneficiary would receive a minimum amount of guaranteed death benefits within a designated time period—possibly measured by actual life span—irrespective of the actual value of the
35
account. Although the GMWB rider comes in many permutations, the basic structure is a guarantee that the contract owner
can withdraw over a specified period of time a pre-set annual amount of money until the aggregate withdrawal amount equals
FOOTNOTE 32: Recent events saw AAA-rated firms like AIG Life and Hartford Life requiring federal bailout funds to survive the liabilities that emerged when their financial
guarantees were suddenly in-the-money. Hartford Life, the company with the largest share of VA sales, recently announced that it is exiting the VA market.
FOOTNOTE 33: It is estimated that approximately 75% of all VA contracts sold in 2005 included riders with supplemental financial guarantees. Prior to the recent great
recession, insurance carriers were in a race to offer more and more competitive products. Currently, it appears that they are in a race to sell products with increased rider
costs and/or decreased rider benefits.
FOOTNOTE 34: On the plus side, a VA contract with a 5% AIR provides an initial benefit per annuity unit that is higher than a contract with a 3.5% AIR. Stated otherwise, a
trust with a 5% AIR contract assumption is less likely to see substantial future growth in the dollar value of lifetime income but receives a greater initial income. Many VA
contract holders opt for a contract with a 5% AIR under the theory that a bird in the hand is worth more than two in the bush. Such distinctions, however, are easily blurred
during a sales presentation.
FOOTNOTE 35: The death benefit is adjusted for aggregate withdrawals prior to death.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 31 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 the initial premium paid into the contract—perhaps reduced for fees, commissions and expenses, and possibly increased at
specific future dates to reflect a higher account value on those dates. However, once the aggregate withdrawal amount reached
its upper bound, the contract terms provided for no further benefits under the rider. The guarantee could not assure a lifetime
income stream.
Beginning in 2005, insurance carriers offered a new guarantee rider to VA policy owners—the Guaranteed Lifetime Withdrawal
Benefit rider [GLWB]. Since the time of its introduction, this rider has become the most popular because it does not place an upper
bound on the aggregate amount of income payable to the investor.
36
Basically, the GLWB rider guarantees that the contract owner
can withdraw a pre-specified fraction of the initial premium—adjusted for expenses, commissions and fees—for life. The allowable
fraction of the initial adjusted premium is a function of age. Typically, a 65 year old receives a lifetime withdrawal guarantee of 4%
of the premium. Often, there are step-ups in benefits that can be implemented at designated future dates provided that the
account value net of previous withdrawals, fees, commissions and expenses attains a dollar value higher than the amount of the
initial adjusted premium. By 2005 the insurance industry was able to offer investors a contract with the following advantages:
• Liquidity
• Lifetime income
• Control over investment choices and asset allocation
• Tax advantages through deferring reportable gains for funds remaining in the annuity contract
• Tax exclusion of a portion of income received as a periodic annuity payment
• Mitigation of the impact of downside market risk through lifetime income continuation
• Ability to participate in market gains if bull markets pushed account values higher.
Given the list of purported advantages, VA sales skyrocketed.
Whenever something seems too good to be true, it’s time to read the fine print. A moment’s reflection suggests that insurance
carriers do not have the power to suspend the ‘law of one price’—financial instruments with equal payoffs must sell for equal
prices. If they do not, traders will quickly arbitrage away the profit making opportunity by buying the cheap instrument and
simultaneously selling the expensive. If all risk is eliminated—payoffs are guaranteed under all economic conditions—then the
expected return to the investor, gross of fees, is the risk free rate. Net of fees, however, an investor in a true risk-free contract
earns less than the risk-free rate. In the case of a VA contract supplemented with annuity guarantee riders, there are so many
complicated moving parts it may be difficult to see the forest for the trees. Bottom line: if someone is selling you a guarantee
worth 1% for a cost of 2%, the guarantee might not be worth what you pay for it.
A full decomposition of a VA contract is beyond the scope of this essay. However, trustees should be aware of the following items:
• The GLWB guarantee is a contingent guarantee—it is “in the money” only if two events occur concurrently: (1) the income
beneficiary remains alive; and, (2) the VA contract value falls to zero. If the two events fail to occur, the guarantee
provides no benefits.
FOOTNOTE 36: Unfortunately, the names of the various riders have not become standardized across the industry. From time-to-time the term GMWB is used to describe a
GLWB. Often, companies have proprietary trademarked designations for these riders which further adds to the confusion.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 32 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 • An investor is unlikely to receive both the downside risk protection and the ability to participate fully in a bull market.
Consider the following example: a newly purchased VA contract has an account value of $1 million and provides a GLWB
of 4%. The contract states that one year from now, if the contract’s account value is higher than $1 million, the owner can
opt to ratchet up the withdrawal right to 4% of the higher value. The force driving the account value higher—the bull
market—is offset by the withdrawals, fees and expenses—forces which drive the account value lower. Bottom line—investors
receive a benefit from the downside protection withdrawal guarantee only in the event of a simultaneous occurrence of two
future contingencies, and receive the benefit of participating in a bull market only if the market’s performance is
37
extraordinarily positive.
• Sometimes, a marketing pitch might suggest that a VA with ‘step-up’ or ‘ratcheting’ provisions can provide inflation
protection because of the potential for participation in market advances. This claim is extremely dubious given the
significant and ongoing drag on account values from costs and from withdrawals.
• The tax consequences of receiving periodic payments under a GLWB rider may differ from those generally applied to
payments received under an annuity contract. Payments under the rider may be deemed to be fully taxable ordinary
income to the extent that there is any gain in the contract. Under certain conditions, the rider may convert all capital
38
gains to ordinary income and may eviscerate the ability to use capital losses to offset other tax liabilities.
• The more risk averse the trust, the greater the appeal of an income guarantee, all else equal. In a VA contract, however,
the income guarantee is available only as a rider on an annuity chassis replete with layers of embedded costs. That is to say,
the trust cannot purchase the rider on a stand-alone basis; rather, it must be bundled with an investment program generating costs far higher than those incurred by operating a portfolio outside of the annuity context. Thus, the guarantee has
both an explicit cost—the amount the insurance company charges for the rider; and an implicit cost—the extra costs of
investing through the auspices of the insurance company.
• Are the extra costs worth it? This question should be answered on a case-by-case basis. However, some general observations
are possible. A VA annuity contract has costs that are absent from non-annuity investment programs. The extra costs, however,
provide two benefits: (1) a tax shield, and (2) the right to annuitize according to the actuarial formulae described earlier.
For an additional cost, the trust can secure a rider providing a contingent guarantee of dollar-denominated withdrawal right.
It is clear that a rational investor would not purchase an annuity with a 2% non-deductible ongoing fee to obtain the benefit
of a tax shield on an investment program generating a 1% pre-tax rate of return. Therefore, in an investment environment
which exhibits low interest rates on fixed income instruments, the costs of the tax shield are justified if the investment
portfolio has an expected pre-tax return in the neighborhood of 5% or higher. This return may be achievable with high-yield
(“junk”) bonds—but the annuity guarantees appeal primarily to risk-averse investors who are not attracted to this asset
class. Therefore, to justify the costs and to exploit the purported advantages of the annuity program, the trustee in a low
interest economy must look to stocks. But the higher the allocation to stocks, the more volatile is the account value. At the
end of the day, it may be cold comfort to trust beneficiaries that the withdrawal right is “in the money” while the actual
account value is zero. When the account value hits zero, the trust is stopped out from any further participation in financial
FOOTNOTE 37: These observations are not criticisms of the VA contract. I hope to receive no benefits whatsoever from the premium payment made to my automobile
insurance company because I do not wish to become involved in an accident. However, if the company is charging a premium of $2,000 per year to insure a vehicle that is
worth $1,500, I would consider the financial consequences of continuing this arrangement.
FOOTNOTE 38: A working committee of the NAIC recommends that the GLWB rider be reclassified as a “hybrid annuity.” If the recommendation is adopted, it is uncertain
how or if this would affect the taxation of benefits received under the rider.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 33 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 gains, has lost all liquidity, has permanently locked in a budget constraint, has no inflation protection, and is solely reliant
on the continued solvency of the insurer.
• Trustees should understand that the GLWB rider limits a VA contract’s liquidity. If an annuitant makes withdrawals in excess
of those permitted under the terms of the rider, the amount of future income withdrawal rights is subject to modification.
For example, if a VA contract holder requires a large withdrawal to pay for unexpected expenses, there is usually a pro-rata
adjustment in the amount of guaranteed future income provided under the terms of the rider. Additionally, the amount of
excess withdrawal may also be subject to contract surrender charges. Suppose that an trust holding a VA contract with a
4.5% GLWB rider for the benefit of an 85-year-old beneficiary has scrupulously kept annual withdrawals to 4.5% of the
original premium amount—as required by the terms of the rider. However, because of poor investment performance, the
contract with an initial account value of $500,000 is now worth only $100,000. The trustee needs to withdraw an additional
$30,000 for unexpected medical expenses. Many GLWB riders make a pro-rata adjustment in the guarantee—in this case a
30,000/100,000 reduction in the guaranteed future payments.
• Perhaps the most dubious element of a VA contract is the implicit assumption regarding the nature of the guarantee itself.
Assume that a trust receiving annual withdrawal checks under the terms of the rider never makes an excess withdrawal
from the contract. Further, assume that the account value has been fully depleted because of a long and highly virulent
bear market. The only way for the trustee to believe in the credit worthiness of the guarantee is to believe that the forces
resulting in a horrible economy have no impact on the solvency prospects of the insurance company. It strains credulity,
however, to believe that the financial condition of insurance carriers move independently from general economic trends. If
you possess the guarantee of a free seat on any Pan American airlines flight leaving the continental United States, the
current market value of such a guarantee is extremely small.
The individual components of a VA contract seem attractive when evaluated in isolation. However, the informed trustee considers
how the components act in tandem. This is a difficult task because of the complexity of the product and the lack of clear disclosure
by some product purveyors. It is certain that a typical commission-paying VA contract’s costs are a substantial long-term drag on
investment performance when compared to no-load mutual funds and exchange traded funds. For investment horizons of 20 years
or more, the account values within a VA contract can easily be less than two-thirds of those within a mutual fund portfolio with the
same initial value and the same underlying asset allocation. This means that 30% to 40% of wealth may be sacrificed to purchase
contingent guarantees. If trustees and beneficiaries feel comfortable with VA rider guarantees, they should be aware that the cost
of comfort can be quite high when purchasing high-load products.
The Ruin Contingent Deferred Annuity
The GLWB rider requires an insurance company to act as both guarantor and investment intermediary. The Ruin Contingent
Deferred Annuity [RCDA], by contrast, does not require the layers of fees associated with investment programs conducted under
insurance company auspices. With a RCDA, the trust retains broader—but not unlimited—control over the portfolio of financial
assets while buying only a guarantee of lifetime income continuation for the beneficiary. The RCDA strategy allows a trustee to
approximate a ‘buy-a-guarantee-and-invest-the difference’ type of portfolio management program. The ability to pursue this type
of investment strategy conveys some important advantages. Unlike the irrevocable sacrifice of capital required by SPIAs and SPDAs,
the RCDA merely requires the payment of an annual fee. Should the trust fail to pay the fee, coverage terminates without surrender
charges. Presumably, an trustee encountering a bear market early into the trust administration can protect the portfolio by paying
the annual fee for the income guarantee. If the sequence of market returns is favorable, the trust surplus makes continuation of
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 34 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 the program unnecessary; if the sequence of returns is unfavorable, the trustee may elect to continue paying for the contingent
guarantee until such time as the guarantee is in the money, or until the portfolio recovers sufficiently. The astute reader may
recognize that the RCDA is akin to the GLBW rider; but, the RCDA is available without having to purchase a high cost VA contract.
39
These types of contracts are relatively new, and only approximately thirty states have approved the contract for sale. Several firms
have recently withdrawn their product from the marketplace but assert that the product specifications are being reworked and that
they intend to reintroduce an updated version in the future. Perhaps the largest current market share belongs to a commission-free
RCDA sold through a Transamerica Life subsidiary and based on the claims paying ability of Transamerica Advisors Life Insurance
40
Company. The product is marketed under the name of ARIA.
41
The ARIA product requires that the investor’s portfolio consist of a
broad choice of designated no-load mutual funds and exchange-traded funds offered by firms like Vanguard, Schwab, DFA, PIMCO and
so forth. The portfolio is not managed by the insurance carrier, but remains under the trust’s control at the custodian of the trustee’s
42
choice. The decoupling of the guarantee from the underlying portfolio means that the contract does not provide a tax shield.
The amount of lifetime income that the insurance carrier will guarantee is, in part, a function of the annuitant’s age at the time of
the income guarantee election date. The cost of the lifetime income guarantee is, in part, a function of age, current interest rates,
and portfolio asset allocation. The risk of the asset allocation is strictly limited by capping the allowable weighting of risky assets
to maximum percentages. This means that the cost of guaranteeing a lifetime income from a high volatility portfolio may be
substantially higher than the cost incurred from a lower volatility portfolio.
43
The relationship between portfolio volatility and RCDA cost is one area of concern for regulators. Felix Schirripa, chief actuary with
the New Jersey Department of Banking and Insurance currently heads an NAIC committee tasked with a broad-scope review of VA
and RCDA guarantees. The committee’s concern cuts in two directions:
1. What are prudent reserving requirements for such guarantees? This addresses the issue of solvency risk to the
insurance carrier.
2. What is the value of the guarantee to the consumer if the insurance company caps portfolio volatility by
limiting the proportional weighting of risky assets within the portfolio? This addresses a variety of consumer
protection issues.
Although the probability that the company making the guarantee will remain in business is of more than passing interest to the
trustee, Schirripa’s actuarial models suggest that the annuitant may not receive meaningful or cost-effective longevity protection
because of the volatility / cost-of-protection relationship. Schirripa’s work highlights the importance of considering all contract
provisions operating in tandem rather than separately. The RCDA annuity product seems to share the same liquidity difficulties as
FOOTNOTE 39: The American Academy of Actuaries report (“An Overview of Contingent Deferred Annuities”) states: “A CDA is essentially a stand-alone guaranteed living
withdrawal benefit….”
FOOTNOTE 40: The insurer, part of the Aegon Americas group of companies, is rated A+ by the A.M. Best Company and AA- by Standard & Poor’s as of June 1, 2012. Aegon
N.V. is an international life insurance, pension and investment group based in The Hague, The Netherlands. The parent company or affiliates, however, may not back the
guarantee according to disclosure information: “the guaranteed lifetime payments are backed by the claims-paying ability of Transamerica Advisors Life Insurance Company.
They are not backed by any other entity….”
FOOTNOTE 41: ARIA is an acronym for Access to Registered Investment Advisors. The actuarial acronym SALB or ‘Stand Alone Living Benefit” program is also used to describe
the product.
FOOTNOTE 42: Fees for the lifetime income guarantee are billed directly to the trust and, unlike a VA contract, are not paid by liquidating investment positions from within
the covered account.
FOOTNOTE 43: The ARIA product is offered by prospectus which trustees should study carefully in order to understand contract provisions and options. These provisions
include the ability to exercise options via “step ups” or “ratcheting” formulae, payout adjustments in the event of large interest rate moves, maximum future fee
adjustments, and discounts for large portfolios. Although perhaps not as complex as VA contracts, RCDAs give trustees a plethora of fine print to digest.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 35 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 VA contracts—namely that the guaranteed cash flows can decrease significantly for trusts finding that they need to access the
investment accounts for unexpected beneficiary expenses.
In addition to the cost v. volatility question, Schirripa’s committee has put forth a variety of issues. These include:
44
1. Are RCDA contracts financial guarantees or annuity products?
2. Are the contracts covered by state guarantee funds; and, if so, do they have nonforteiture provisions?
Tax rules governing CDRA contracts are not yet fully determined. We understand that the IRS has issued a series of private letter
rulings suggesting that they will treat income received from CDRAs as “annuity income” subject to the applicable annuity taxation
45
rules in the Revenue Code. However, some commentators have opined that certain withdrawals under CDRA contracts will be
46
taxed at lower capital gains rates. Likewise, the jurisdictional bounds of the SEC remain uncertain.
The American Academy of Actuaries indicates “from a consumer’s perspective, CDAs can be a beneficial product because the
product offers a solution to longevity risk exposure.” However, the NAIC Working Committee headed by Felix Schirripa [“Contingent
Deferred Annuity (CDA) Subgroup] urges caution because “we have regulatory concerns surrounding both solvency and consumer
protection.” The committee’s recommendation is for the NAIC to determine the best course of action with respect to company
financial solvency requirements, regulatory authority, and consumer protections.
It appears as if the RCDA product represents, for some investors, a significant advance over the GLWB/VA package. However, there
are a variety of issues remaining to be settled. The informed trustee should be fully aware of the NAICs work agenda. Fortunately,
it appears that many uncertainties may reach satisfactory resolution in the near future. At that point, the decision whether to
acquire the product to protect some or all of the portfolio’s cash flow generating ability will become a more straightforward
cost/benefit analysis.
47
Conclusion Longevity risk-reducing products are here to stay because they address a legitimate demand from a growing population of whitecollar investors. Mutual funds may be getting into the act not only by forming strategic marketing alliances with insurance
companies but also by jointly developing products involving investment, actuarial and financial engineering (derivative overlay)
strategies. Whether these bundled products of the future will provide meaningful protection at reasonable costs is, of course, the
paramount question. The insurance industry is still recovering from the black eye received from selling Long Term Care Policies
offering untenable guarantees; the mutual fund world has come under scrutiny for its target-date / life-cycle retirement products
which sometimes bundle poorly performing funds with more attractive funds; and the financial engineering world has been accused
of marketing highly leveraged products that contributed significantly to the recent global recession. Whatever the outcome of the
continued R&D in this area, however, it is certain that acquiring longevity protection is, for many trustees, not a do-it-yourself activity.
FOOTNOTE 44: The NAIC Working Committee recommends that GLWBs be classified as ‘hybrid income annuities’ and that CDRAs be classified as ‘synthetic hybrid income
annuities.’
FOOTNOTE 45: Private Letter Rulings apply only to the specific taxpayer to whom they are issued.
FOOTNOTE 46: ARIA marketing material states that “benefit payments are subject to ordinary income tax” if paid under the terms of the ruin contingent deferred annuity
guarantee. Additionally, the material states that “the annuity has no cash value, surrender value or death benefit.”
FOOTNOTE 47: For example, the trustee might compare the current cost of an SPDA where the payout is contingent only upon survival, with the present value of the yearly
cost of an RCDA promising to pay a comparable income where the payout is contingent upon both survival and complete portfolio depletion.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. P AGE 36 INSTITUTIONAL INVESTORS: INSIGHT When Every Choice Is Bad… July 2013 Meet Your Author Patrick Collins: P R I N C I P A L Ph.D., CFA Patrick J Collins is a founding principal of Schultz Collins Lawson Chambers, Inc. and
serves as an investment consultant to individual and institutional investors. As a
Chartered Financial Analyst, Mr. Collins creates custom investment policies, advises and
guides clients on asset allocation decisions, and provides ongoing education through
writing and teaching. As a licensed Insurance Advisor he provides fee-based life
insurance policy analysis for individuals, trusts, corporate directors, and other fiduciaries. Mr. Collins also provides litigation
support services for disputes involving both investment and insurance issues.
Mr. Collins is an adjunct professor at the School of Management at the University of San Francisco where he teaches the portfolio
management course in the Masters in Financial Analysis degree program. He also currently teaches three courses for the CFA
program sponsored by the Security Analysts of San Francisco.
Mr. Collins is the author of numerous articles that have been published in leading industry publications including Trusts & Estates
Magazine, Real Property Probate & Trust Journal, the Journal of Asset Protection, the Journal of Financial Planning, Insurance Law,
California Trusts And Estates Quarterly, ACTEC Journal, The Journal of Financial Services Professionals, the Maryland Bar Journal,
the Banking Law Journal, Wealth Strategies Journal, and The Journal of Investing.
Mr. Collins received his undergraduate degree from College of the Holy Cross. He also earned a Ph.D. in English and a Professional
Designation in Financial Planning from the University of California of Berkeley. He holds a Chartered Life Underwriter designation
from the American College. Mr. Collins is a member of CFA Institute, The Security Analysts of San Francisco, and the Society of
Financial Service Professionals.
Rev July 2013
© 2013 Schultz Collins Lawson Chambers, Inc. All rights reserved.
This document does not constitute a recommendation of or solicitation to provide services. Readers should consult with legal, tax, or accounting
professionals before acting upon any information or analysis contained herein. Schultz Collins Lawson Chambers, Inc. does not provide tax or legal
advice. The information in this document is drawn from sources believed to be reliable. However, no endorsements are made as to the accuracy of
third party information. Opinions expressed are as of the date appearing on this material. This material may not be sold or printed for distribution
without the express written permission of Schultz Collins Lawson Chambers, Inc.
The information contained herein is not intended to be used as a general guide to investing, or as a source of any specific investment
recommendations, and makes no implied or express recommendations concerning the manner in which any assets should be handled. Portfolio risk
management does not imply low risk. Past performance is not indicative of future results, which may vary. The value of investments and the income
derived therefrom can go down as well as up. Future returns are not guaranteed, and a loss of principal may occur.
SCHULTZ COLLINS LAWSON CHAMBERS, INC. 455 Market Street, Suite 1450
San Francisco, California 94105
415.291.3000
415.291.3015
P H O NE :
FAX :
877.291.2205
www.schultzcollins.com T O L L-­‐ FR E E:
P AGE 37