September 01, 2013


The American Taxpayer Relief Act of 2012 (ATRA) changed the landscape for estate planning—not just by setting the estate tax exclusion at $5.25 million per person in 2013, but, more importantly, by indexing this amount for inflation in future years. A rising exclusion amount will shelter a growing portion of a client’s estate from transfer taxes. In some cases, this will render ineffective the estate-planning practices of prior years. Indeed, under the new tax law, there is a heightened risk of overplanning: giving away assets required for other purposes that might ultimately be sheltered from estate tax.

In this article, we explain why the interplay between inflation, the applicable exclusion, and client spending has to be taken into account when considering the merits of any wealth-transfer strategy. In our view, estate-freeze strategies that can transfer wealth with little or no use of the applicable exclusion amount may prove particularly valuable.1 The example we consider here is the zeroed-out grantor retained annuity trust (GRAT).

Core Capital – A Critical Number

Of course, nobody wants to give away assets that may be needed to support lifestyle spending in future years. That is why, before considering a wealth transfer strategy, a client needs to know how much capital to retain. This assessment necessarily requires an understanding of the client’s spending and spending horizon, and of the risk/return characteristics of the assets that the client will rely on for spending. At Bernstein, we call this core capital—the amount needed to sustain spending goals, grown with inflation, for the rest of an individual’s life.

Clients have to sustain core spending even in dismal markets, so when we use our proprietary Wealth Forecasting SystemSM to stress-test markets, inflation, and longevity, we seek a very high level of confidence that their money will last. Wealth over and above what is determined to be core capital is considered surplus capital—the amount that can be transferred off a client’s balance sheet without jeopardizing long-term spending security. An important question that needs to be addressed post-ATRA is how much (if any) surplus capital can stay in the estate without creating estate-tax exposure.

Case Study

To bring to life the interplay between inflation and client spending, let us consider the case of the Forsytes, both age 75. They have $20 million of liquid investments in a moderate portfolio comprising 60% diversified global stocks and 40% intermediate-term bonds. They envision passing the wealth they have accumulated down to their children and would like a tax-efficient plan that will not sacrifice their own long-term security. The couple currently spends $875,000 a year and would like to continue spending at this level for the rest of their lives. Can they reduce their estate-tax exposure over time without jeopardizing their lifestyle?

Let us first consider the impact of outright gifts on the likelihood of meeting their spending plan.


We estimate that the odds are extremely high that the Forsytes can fund their spending into their mid-90s—if they hold on to their assets (Display 1). Making a current gift much beyond $1 million erodes those odds considerably. If the couple were to make a gift of $5 million, the odds of not meeting their spending plan would jump to better than one-in-three. Translation: There is not much surplus capital to work with at the present. But if they are unwilling to make significant gifts, will their current estate-tax exposure persist?

Not necessarily. Future inflation is likely to increase the Forsytes’ expenses, which will put downward pressure on the value of their portfolio and whittle away the estate-tax exposure over time. At the same time, under ATRA, inflation will cause the applicable exclusion to tick up.

The lines in Display 2 depict our median forecast for the value of the couple’s portfolio, after spending and inflation, relative to their combined applicable exclusion amount of $10.5 million. Since we are showing inflation-adjusted dollars in the display, the value of the applicable exclusion appears as a constant horizontal line. After supporting their spending, their portfolio is projected to decline in inflation-adjusted dollars and to converge with the applicable exclusion amount between years 15 and 20. In other words, now that ATRA has permanently indexed the applicable exclusion to inflation, the Forsytes’ significant current estate-tax problem may resolve itself—given enough time.

Of course, the markets are not nearly as linear as the display would seem to indicate. There is a range of possible outcomes that should be evaluated to provide a more complete picture. As Display 3 illustrates, we estimate that there is a 42% chance that the estate will exceed the exclusion amount in year 20 and a 10% chance that the assets will be worth $21.2 million – $1.2 million above their current value in today’s dollars. In other words, market appreciation could push the Forsytes’ estate into territory where the estate tax will take a serious toll.


Strong markets are certainly a good problem to have. The Forsytes need a solution that can transfer future upside without giving away core assets that they will need if markets disappoint. As an alternative to an outright gift, what if our couple considered the merits of a zeroed-out GRAT? If equity markets turn out to be particularly strong, contributing $6 million of their equity portfolio (half of their $12 million in total equities) to a properly structured 10-year term GRAT would shift a significant portion of future growth (Display 4). And they will not lose access to the contributed core capital, which will be returned to them in the form of annuity payments.


If, on the other hand, market returns are poor, the Forsytes will not transfer wealth to heirs through the GRAT plan. The GRAT only works if the contributed assets appreciate in value. Since a properly structured GRAT uses very little or none of the applicable exclusion amount, the heirs would be no worse off with a GRAT in place should markets prove challenging. That is one of the great features of this strategy.


A zeroed-out GRAT is a potential all-weather hedge that can save estate taxes without jeopardizing long-term spending nearly as much as a large outright gift would. In the best 10% of future market outcomes, the GRAT plan is estimated to cut the taxable portion of the couple’s estate from $10.7 million to $1.3 million over the next 20 years—that is nearly $4 million in federal estate-tax savings at today’s 40% transfer tax rate (Display 5). The savings could be even more substantial if the Forsytes reside in a state that levies a separate death tax. The upshot is that the potential estate-tax savings of the strategy still provide the couple a 90% chance of meeting their spending plan through age 95.


No plan is without flaws. The benefits afforded by utilizing a GRAT strategy as illustrated in our case study are only realized if the grantor outlives the term of the trust. While this mortality risk can be hedged by acquiring term life insurance and/or shortening the term of the GRAT, neither is without cost. In addition, many estate-freeze strategies, including GRATs, are driven by the availability of very low interest rates. As rates rise from their historic lows, it will not be possible to shift as much appreciation away from the estate, which will render these strategies less attractive.

It is clear that in the post-ATRA world, spending and inflation are extremely important variables to consider. Careful modeling of their effects can help clients avoid a serious pitfall of overplanning: giving away assets they might eventually need for themselves. Under the new tax law, there is a high premium on Goldilocks planning—not too little and not too much.

Notes on Wealth Forecasting System

The Bernstein Wealth Forecasting SystemSM (WFS) is designed to assist investors in making a range of key decisions, including setting their long-term allocation of financial assets. The WFS consists of a four-step process: (1) Client Profile Input: the client’s asset allocation, income, expenses, cash withdrawals, tax rate, risk-tolerance goals, and other factors; (2) Client Scenarios: in effect, questions the client would like our guidance on, which may touch on issues such as which vehicles are best for intergenerational and philanthropic giving, what his/her cash-flow stream is likely to be, whether his/her portfolio can beat inflation long term, when to retire, and how different asset allocations might impact his/her long-term security; (3) The Capital Markets Engine: our proprietary model that uses our research and historical data to create a vast range of market returns, taking into account the linkages within and among the capital markets (not Bernstein portfolios), as well as their unpredictability; and (4) A Probability Distribution of Outcomes: based on the assets invested pursuant to the stated asset allocation, 90% of the estimated returns and asset values the client could expect to experience, represented within a range established by the 5th and 95th percentiles of probability. However, outcomes outside this range are expected to occur 10% of the time; thus, the range does not establish the boundaries for all outcomes. Further, we often focus on the 10th, 50th, and 90th percentiles to represent the upside, median, and downside cases. Asset-class projections used in this paper are derived from the following: US diversified stocks are represented by the S&P 500 index, with an assumed 20-year mean annual return of 8.9%, based on simulations with initial market conditions as of March 31, 2013; US value stocks by the S&P/Barra Value Index (mean annual return of 9.1%); US growth stocks by the S&P/Barra Growth Index (mean annual return of 8.9%); developed international stocks by the Morgan Stanley Capital International (MSCI) EAFE Index of major markets in Europe, Australasia, and the Far East, with countries weighted by market capitalization and currency positions unhedged (mean return of 10.1%); emerging markets stocks by the MSCI Emerging Markets Index (mean return of 10.1%); municipal bonds by diversified AA-rated securities with seven-year maturities (compounding rate of 2.6%); and inflation by the Consumer Price Index (compounding rate of 2.7%). Expected market returns on bonds are derived taking into account yield and other criteria. An important assumption is that stocks will, over time, outperform long-term bonds by a reasonable amount, although this is by no means a certainty. Moreover, actual future results may not be consonant with Bernstein’s estimates of the range of market returns, as these returns are subject to a variety of economic, market, and other variables. Accordingly, this analysis should not be construed as a promise of actual future results, the actual range of future results, or the actual probability that these results will be realized.


Contact the authors:

Stephen S. Schilling, CFA – Bernstein Global Wealth Management –

Matthew J. Teich, CFA, CFP® – Bernstein Global Wealth Management –


1 Bernstein does not provide tax, legal, or accounting advice. In considering this material, you should discuss your individual circumstances with professionals in those areas before making any decisions.