Reality Bites

Investment counselor cries “Foul!” over lawyers' projections for total return trusts

I am responding to the article “Using Total Return Trusts,” by Laurence J. Kline and Karl R. Anderson, in the February issue. As an investment professional, I have watched with dismay the drift toward this type of trust. Evidently — and as the article demonstrates — many fiduciary professionals are unaware of some of the basic mathematical tenets that determine the health of investment portfolios over time. My primary concern centers around the table on page 32, which purports to compare the results from an all-income trust invested at 50/50 stocks and bonds and a total return trust (TRT) paying 4 percent on principal but invested in an 80/20 mix.

I have run projections under the identical scenario described in the article for the total return trust, but using actual market returns from 1965-1984. (See “The Way It Really Was,” this page and “Forcasting Trust Resources,” p. 7.) In the actual market return situation — which incorporates negative years as well as positive ones — there is considerable variability in the income provided to the beneficiary, and the final result after 20 years is a market value that is barely higher than that for the all income trust shown in the article. Note that income paid to the beneficiary under the 4 percent rule fluctuates by tens of thousands of dollars — in the downward direction — several times during this period. In fact, under the all income trust scenario in the table, the total payouts over 20 years are $3,126,701 versus $3,078,972 under the market projections — a considerable shortfall for the beneficiary.

No doubt that if we were to apply actual market return numbers to the all income trust scenario, it too would look considerably worse than presented in the article. The income assumption in the article is that bond returns average 5 percent and all of that is income and thus paid out. Today, that assumption might be untenable with the 10-year U.S. Treasury hovering at only 4 percent. Five percent is awfully hard to find today.

In my practice as an investment advisor, we frequently use projections similar to the ones I have presented here to describe to our clients the effects of market volatility on their wealth, especially when withdrawals from the investment program commence. Describing results under withdrawal scenarios with straight-line return assumptions is mathematically fallacious. I wish the practice were abolished. It can lead to a bankrupt client, or at the least, a portfolio that will no longer support the draws “required.” This is because negative market returns (not accounted for in a straight line/always positive market return scenario) take value from the portfolio right alongside the client and result in what I call, “negative compounding,” essentially, running backward very fast.

Lest you think that the 1965 to 1984 period is unusual, consider this: the period includes bull markets, bear markets (especially the bear market of 1973-1974, which was nearly identical in magnitude to the 2000-2002 bear market), Republican rule, Democratic rule, inflation, disinflation, war, peace, high interest rates and rates comparable to today's rates. This period is also fairly modern, capturing the Fed's ability to manage the economy efficiently — as opposed to the Depression era, when mistakes were made. If there is anything unusual about this period it is that equity and bond returns at the tail end were very, very good. We may not even have that wind at our backs now. More likely, these markets will move oppositely, and surely there is not enough yield in bonds to provide much capital appreciation from that asset class over the next 20 years.

In addition to my concerns around the calculations presented in Kline and Anderson's article, I am also concerned that the 4 percent and 5 percent payout provisions for these trusts are becoming incorporated in state statutes. Institutionalizing return numbers, just like making iron clad investment edicts in the language of a trust document, will backfire eventually. If we are headed the same way as Japan, toward zero interest rates (and there is considerable demographic evidence to support that theory) then, in time, even a 4 percent payout will look onerous.

Practitioners had best carefully consider the possibility of unintended consequences when pursuing total return trusts.
— Michelle Dicus,
president,
Cascade Investment Advisors, Inc.,
Portland, Ore.

Authors Respond

Thank you for these comments regarding our article “Using Total Return Trusts.” We are not investment specialists and cannot predict how the market will perform in the future. We understand that investment returns do not occur on a “straight-line” basis. The table in our article was meant solely as an illustration of how a total return trust could outperform an all income trust — based on certain assumptions of investment returns.

Our assumed investment return for the stock portion was based on the after-tax return of the Vanguard 500 Index Fund (VFINX) for the last 10 years. Our assumed bond return was based on the after-tax return of the Vanguard Total Bond Market Index Fund (VBMFX) for the same period. All of Vanguard's historical information is available on their web site, and their after-tax returns were based on the highest marginal rate in each year.

We understand that the market is volatile, and that consultation with investment experts like you is a crucial part of the estate-planning effort. Also, a trustee that elects to use a total return trust must keep in mind the relative needs of the beneficiaries and the investment risks in determining the percentage payout.

The points in your letter are well taken. We believe this type of dialogue among planning professionals is very important.
— Laurence J. Kline, partner and
Karl R. Anderson, associate,
Hoogendoorn & Talbot LLP, Chicago

THE WAY IT REALLY WAS

Using Kline and Anderson's scenario, these would be the results, using the actual market returns from 1965 to 1984

Year	Bond Return	Stock Return
1965	1.02%	12.45%
1966	4.69	-10.06
1967	1.01	23.98
1968	4.54	11.06
1969	4.54	-8.5
1970	16.86	4.01
1971	8.72	14.31
1972	5.16	18.98
1973	4.61	-14.66
1974	5.69	-26.47
1975	7.83	37.2
1976	12.87	23.84
1977	1.41	-7.18
1978	3.49	6.56
1979	4.09	18.44
1980	3.91	32.42
1981	9.45	-4.91
1982	29.1	21.41
1983	7.41	22.51
1984	14.02	6.27
Source: Ibbotson & Associates

FORECASTING TRUST RESOURCES

Looking at resources in a total return trust under an actual market scenario

Year	Beginning Balance	Additions/Withdrawals	Return for Year	Earnings	Ending Balance
2005	3,150,000	-120,000	10.16	320,166	3,350,166
2006	3,350,166	-134,007	-7.11	-238,197	2,977,963
2007	2,977,963	-119,119	19.39	577,308	3,436,152
2008	3,436,152	-137,447	9.76	335,231	3,633,937
2009	3,633,937	-145,357	-6.95	-252,486	3,236,093
2010	3,236,093	-129,444	6.58	212,935	3,319,585
2011	3,319,585	-132,783	13.19	437,920	3,624,721
2012	3,624,721	-144,989	16.22	587,785	4,067,517
2013	4,067,517	-162,701	10.81	-439,536	3,465,280
2014	3,465,280	-138,611	-20.04	-694,373	2,632,296
2015	2,632,296	-105,292	31.33	824,593	3,351,597
2016	3,351,597	-134,064	21.65	725,487	3,943,020
2017	3,943,020	-157,721	-5.46	-215,368	3,569,932
2018	3,569,932	-142,797	5.95	212,268	3,639,403
2019	3,639,403	-145,576	15.57	566,655	4,060,481
2020	4,060,481	-162,419	26.72	1,084,879	4,982,942
2021	4,982,941	-199,318	-2.04	-101,552	4,682,072
2022	4,682,072	-187,283	22.95	1,074,442	5,569,230
2023	5,569,230	-222,769	19.49	1,085,443	6,431,904
2024	6,431,904	-257,276	7.82	502,975	6,677,603
Note: These projections assume that the account is invested 20 percent in bonds and 80 percent in stocks, with 4 percent of the prior year's ending value paid out every year, except the first. Source: Michelle Dicus