Federal taxes on income, gifts and estates reduce the fruits of investment success, but otherwise seem unrelated to investment decision-making. Our tax law and culture do not permit U.S. taxpayers to treat their tax returns as an opening bid. In theory, every tax question is expected to have a right answer, however obscure that answer may be. Contrast this mindset with the marketplace of investing, where differences of opinion are more plentiful than opportunities, and investors constantly strive to gain the upper hand by applying better analytical tools.

Thus, tax risk analysis usually bears little resemblance to investment risk analysis. Many wealthy taxpayers measure their risk tolerance using a scale that registers only broad points of view, such as a personal sense of how much money their family needs, how fair the tax is, how well the government spends tax dollars, and how they feel about government and philanthropy in general and their tax advisor in particular. Perhaps most importantly, wealthy taxpayers consider whether a dispute with the Internal Revenue Service would be a distraction, embarrassment, expense or mark against their personal integrity. Conversely, some see a tax dispute as an opportunity to best the IRS. There are taxpayers who work to save taxes as if a dollar saved in taxes was more than a dollar earned. Others leave tax dollars on the table that they would never pass up if it was a matter of avoiding investment risk.

In other words, tax risk analysis typically has been based on emotional factors. It's time taxpayers took a much more objective approach — one more akin to investment risk analysis. Here's why and how.


Tax planning is a necessary evil. Reducing taxes is critical to financial success over an extended period of time, and tax risk can be minimized — but not avoided. Risk in tax planning is never zero.

Annual income taxes, coupled with a wealth-transfer tax on capital for every generation, will substantially erode net financial returns, unless reduced by proper planning. Without tax planning, the real world tends to produce investment returns that trend to near zero when measured across more than one generation after taxes and inflation. For example, even if an equity portfolio were to provide a healthy 10 percent annual return on average over the next several decades, only about 6.5 percent is likely to be left after income taxes and expenses. If about half of that annual increase in net worth goes to pay wealth transfer taxes, then the remaining fully taxed return (3.25 percent) barely covers a projected inflation rate of 3 percent each year. Tax planning can reduce this constant flow of money to the government — but not without risk.

In view of the complexity of today's tax laws and their inconsistent application by the IRS and the courts, it would be unrealistic to plan to substantially reduce taxes without facing the risk of a bad result. Just think how volatile tax law is. For example, grantor remainder annuity trusts can play a central role in wealth tax planning due to their low gift tax cost. Despite the GRAT's critical role — or perhaps because of it — the IRS has vacillated on how it separates the good, the bad and the ugly when it comes to GRATs. The Service originally allowed donors to zero out their taxable gifts by retaining a fixed-term annuity, then disqualified that form of annuity — but approved an alternative, the so-called “two-life” or “spousal annuity.” The IRS then reversed course again, and attacked that alternative as well. When the Service lost in court on the fixed-term annuity, it continued to attack the two-life annuity. Worse yet, the IRS actually claimed the two-life annuity was “abusive” because it was a “contingent” interest and thus quite unlike the fixed-term annuity. This argument was advanced even though the IRS earlier had claimed that the fixed-term annuity also was “abusive” on that same ground. It claimed that allowing two-life annuities in GRATs would mean the beginning of the end of the transfer tax system. Accepting the Service's arguments, the U.S. Tax Court and U.S. Court of Appeals for the Seventh Circuit reached the same ironic conclusion: A two-life annuity that leaves some gift tax owed is the kind of tax planning Congress considered “abusive,” but a fixed-term annuity that eliminates the gift tax entirely is not abusive.1

Given such developments, it is easy to conclude that tax risk is never zero whenever reducing taxes is the goal.2 So the better question is how to analyze and manage the tax risk.


Decision-making in tax planning is not entirely quantitative, but neither is it so random that it defies a risk/return analysis.

The best approach to wealth tax planning seems to depend upon a bewildering array of intangible or subjective factors: Is the plan too complicated? Is it flexible enough? Does it enhance or undermine the taxpayer's other wealth-management plans and practices, as well as investment goals for cash flow, liquidity and diversification? Does it require personal involvement to manage the plan? How long will it be before the results are known?

Yet, despite these non-quantitative factors, there must be a way to improve the decision-making process for tax planning. After all, investment planning uses objective analysis and also must consider factors that cannot be readily measured. These tools do not lead to cleanly controlled results in investment planning (it's still investing), but it would be difficult, if not impossible, to make thoughtful decisions on investments without using them, explicitly or implicitly.


Fundamental concepts similar to those used in investment decision-making can also be used in performing a risk/return analysis in tax planning. The “return” can be calculated by estimating the various costs of the plan and its potential outcomes. This return can be computed as an “internal rate of return” (IRR) for each outcome that is within a reasonable range of likelihood. A tax plan that carries significant risk with great variation in the potential results must show a high “tax benefit” return on the “tax investment” — a high IRR — to justify the risk of going forward, just as a volatile or speculative investment must offer a high potential return to compete with more predictable investment opportunities.

A topical example is readily available in the scheduled repeal of the estate tax, for one year only, as of the year 2010. This repeal might become permanent, or at least lead to permanent estate tax cuts, but expert predictions differ as to important details. How can this uncertainty be taken into account when deciding whether to take proactive steps to save estate taxes for a wealthy taxpayer? Why, for example, make gifts or implement a freeze to save estate taxes that might disappear in any event? The best answer to that question is to apply a risk/reward analysis that does not require a “yes” or “no” answer but can deal with “maybes.”

Such an analysis can be used to compare the costs of the plan against the future estate tax savings to determine whether the cash-on-cash “investment return” in saving taxes at some future date is high enough to justify the risk that the “investment cost” could be wasted due to estate tax repeal. The investment cost includes any gift taxes, and any income taxes paid to raise cash to pay gift taxes. But it also includes the legal and appraisal fees as well as other transaction and maintenance costs for the plan, plus the quantitative impact, if any, on the taxpayer's other plans and practices apart from taxes. In measuring the comparative tax savings at the back end of the plan, it is usually appropriate to reduce all assets to cash under each alternative plan, to account for the imbedded capital gains tax liability for those plans that skip the estate tax and thus skip the step-up in basis. Reducing assets to cash allows an apples-to-apples comparison, even if the assets may in fact not be sold.

This analysis provides a meaningful scale to weigh costs and outcomes that cannot be estimated with precision. A high expected rate of return can reinforce the decision to proceed with a plan — in the face of varying projected rates of return over a range of different estimated costs and outcomes, and future events that may render the plan unnecessary or completely ineffective (due to changes in the tax law, family circumstances or the taxpayer's investment and business activities).3 At the same time, a modest projected rate of return would ordinarily justify rejecting a plan — unless the data used in the analysis is unusually reliable, the range of outcomes is highly predictable, and the impact on the taxpayer's other choices and activities is limited. This relationship between acceptable rates of return and the risk of unpredictable outcomes is the same tradeoff that leads the venture capital investor to demand much higher projected returns than the bond investor.

Such a rate-of-return analysis is useful in deciding whether to adopt a plan that requires a large multi-year upfront commitment, in comparing the tax plan with competing investment opportunities, and in considering a plan that requires important compromises in the taxpayer's personal goals or carries other subjective disadvantages.4 A major commitment cannot be justified if a higher rate of return with less risk and dislocation can be found elsewhere, whether in tax planning or actual investing. Conversely, a plan with several disadvantages becomes more appealing if the analysis shows a projected IRR of 40 percent or more, as returns of that size are hard to find anywhere else.

On a more general level, the risk/reward analysis in tax planning parallels the evaluation of an investment opportunity in that:

  • Delay and timidity are threats to successful tax planning. Opportunities are rarely timeless, and a good plan takes best advantage of existing circumstances.

  • A tax plan should be designed to coincide with the taxpayer's personal objectives and risk profile. The plan should fit the facts, and the facts should be managed in advance to minimize risk.

  • Successful wealth-tax planning takes advantage of opportunities, but also avoids relying entirely on one strategy that purports to deliver permanent solutions in a short time span. Diversification among different strategies reduces risk. Patient planners have more choices available.

  • Most investment news tends to be mere noise obscuring the long-term view that leads to successful investment decisions, caution renowned investors as well as scholars of investor psychology. This admonition to screen out topical events that are unimportant or unreliable in the long run has a parallel in wealth transfer tax planning.5

  • Tax planning ideas can become popular simply because they promise readily available off-the-shelf benefits, without any change in personal business or investment strategy. Such planning is more likely to attract a challenge by the IRS and such promises often prove to be just as unrealistic as hot investment tips that provide easy money.


The crossover utility of these analytical tools does not mean, however, that the factors that affect the risk/return tradeoffs in investing follow the same pattern in tax planning. Of greatest importance, the duration of the “investment horizon” in wealth-tax planning — the time before the plan generates a return — does not lend itself as readily to quantitative risk management, and thus requires even more attention than in investment decision-making.

Duration is a key factor in investment analysis, but its related risks can be managed or at least understood in making an investment. For example, the returns on many asset classes become more predictable when averaged over long periods of time. Due to this factor and the benefits of compounding, an investor who starts early is more assured of achieving his goals, all else being equal. Moreover, the relationship between duration and investment return is reasonably well understood because of the large quantity of historical data for various asset classes, which allows for thoughtful analysis, though obviously not precise predictions.

There are no parallels in wealth-tax planning for understanding and managing duration risk. Unlike the investor who selects asset classes and rebalances periodically, a taxpayer cannot as readily convert from one tax plan to another based on a regular review. Nor can the taxpayer choose the duration in the first place. A taxpayer usually cannot control when the estate tax would otherwise be due and the savings realized. Thus the planning must address life spans of varying durations, because life expectancy tables provide predictions based on averages, not glimpses into the future.

In wealth-tax planning, furthermore, the longer the duration of the plan, the more likely the results will vary. These less-predictable results are usually caused by changes in the tax law.6 The situation is akin to an investor asked to commit to an illiquid investment without a clear exit strategy. This is not just a theoretical risk. Look at the last 30 years, which seems like a long time but is indeed the estate-tax-planning horizon for a successful 50-year-old who expects to live to age 80-plus.

During the last 30 years, the tax law has seen the coming, going and return of the generation-skipping tax (passed in 1976, but postponed until its replacement in 1986 with various effective dates) as well as fits and starts in developing “anti-freeze” rules (first in 1987, then in 1988, then as Chapter 14 in 1990). We saw the repeal of the step-up in basis in 1976, then the repeal of the repeal, and now the potential for changes to the step-up in basis if the estate tax is repealed. On a lesser scale, we have seen both special exemptions for the first $100,000 of a qualified plan benefit, now repealed, and special additional income taxes of a sort on “excess benefits” in qualified plans, now repealed. Sometimes fundamental changes have taken hold from the outset, as in the case of the integration of the estate and gift tax in 1976, the unlimited marital deduction in 1981 and the below-market loan rules in 1984. But each of these bold strokes actually took shape through a thousand smaller brushstrokes as the change was implemented. Similarly, the privileged tax status of life insurance has survived — but the care and feeding of policies has been disrupted from time to time, whether by the attacks on Crummey powers and split dollar or by the complications that beset life insurance when used in a generation-skipping trust.

Duration will be shorter in some wealth-transfer tax plans, even for younger taxpayers. For example, if the taxpayer seeks to pass assets when they can be most useful to the next generation, then the true goal may be reducing taxes on gifts — shortening the time over which risk/return is measured. Also, the most critical step in some plans is taken early in the process, such as making gifts at low values to minimize later estate taxes, as compared to saving estate taxes by relying on a favorable transfer at death. Yet this duration is still longer than in most personal income tax planning, because the horizon for a wealth tax plan almost always extends, in some way, well beyond the next April 15.

Because of the typically long time span involved in wealth tax planning, the risk/reward analysis would not be as useful if it simply showed the estimated net present value of the tax savings. That would disguise the lack of precision and predictability. It is far more useful to review a qualitative picture of all the things that can go wrong (and right) and to calculate a set of IRRs showing the range of expected returns for various alternatives, so the risk of each can be compared with the potential returns. In this way, the subtle nature of the possibilities for disappointment — and success — become more apparent. The decision-maker is no longer distracted by the clever details of the plan and the seemingly black-and-white nature of the outcome. Instead, clients can make more objective choices from a menu of maybes.


  1. This trail of confusion finally stopped, or at least paused, when the U.S. Court of Appeals for the Ninth Circuit reversed the U.S. Tax Court, and found the two-life annuity just as qualified as a fixed-term annuity. For a brief recap of this tortured tale of how not to construe a gift tax statute, see Donald D. Kozusko and Steve Vetter, “Commentary on Schott v. Commissioner,” 28 Tax Mgmt. Est., Gifts & Tr. J. 165 (May 8, 2003).
  2. Inconsistency, of course, can be caused by the sheer complexity of the issue. For example, among all the tax advisors in the United States, those who would claim to thoroughly understand how the IRS applies the “subtraction method” under IRC Section 2701 would probably all fit inside a Mini Cooper, and that handful would probably disagree among themselves.
  3. For example, many “estate freezes” produced little benefit during the 1970s and early 1980s due to flat valuations, but then dramatically improved in the late 1980s, except for real estate.
  4. A family limited partnership (FLP) must pass scrutiny under a long list of rules to provide a transfer-tax benefit, including the nine topics of concern addressed in the Greenberg article cited in footnote 5, plus a few more depending on the facts (for example, IRC Sections 2035(a)(2), 2036(b), 2701 and 6166). Avoiding all these rules may push the design and operation of the entity in directions that cause the family managers to compromise on their other objectives: for example, as to what investments should be held in the partnership, who should invest in it and on what terms, and how and when the returns should be distributed.
  5. It would have been distracting to a fault to have followed each twist and turn over the last decade as the legal system tried to develop a coherent view of valuation discounts and FLPs. Any plan that depended heavily on the latest court decision would have missed the underlying trend that has favored taxpayers who are careful but not timid. See Louis A. Mezzullo, “Is Strangi a Strange Result or a Blueprint for Future IRS Successes Against FLPs?,” 99 Journal of Taxation (July 2003); Richard H. Greenberg, “How to Handle Issues Raised by the IRS in the Family Limited Partnership Arena,” Business Entities (January/February 2003); J. Joseph Korpics, “For whom Does Kimbell Toll — Does Section 2036(a)(2) Pose a New Danger to FLPs?” 98 Journal of Taxation 162 (March 2003). The law is much closer to real-world valuations than it was before the IRS purported to abandon its insistence on family attribution in Rev. Rul. 93-12, 1993-1 CB 202.
  6. A certain tax treatment may be grandfathered if there are adverse changes in the tax law, but grandfathering does not provide protection unless it will apply to all critical steps in the tax plan. That protection may also be illusory, since it may turn out to be easy to destroy the grandfather privilege in the ordinary course of maintaining the plan (e.g., the “substantial modification” exception to the grandfather rule under IRC Section 2703, and the exceptions to the effective date rules under the generation-skipping tax). Also note that “technical corrections” are typically not passed with grandfather protection, and yet these can follow the original change in the law by several months or even years, as in the 1988 change that added IRC Section 2642(f) to the generation-skipping tax of 1986. In addition, the lapse of time inherent in a tax on death-time transfers can make it difficult to interpret effective date provisions because of intervening events or unanticipated consequences. See Gallenstein v. U.S., 975 F.2d 286 (6th Cir. 1992) construing the 1977 change to Section 2040 as applicable to property bought in 1955 and passing at death in 1987, after further change to the Code in 1981.