SIZING UP THE TAXPAYER'S NEEDS

By Raymond G. Russolillo

In Brief

Client Needs Come First

Many advisors devise elaborate financial plans based on the best theoretical tax strategy. These plans are often impractical because they to not consider the actual needs of the person receiving the advice.

A pragmatic approach to retirement planning should start with information gathering that considers the taxpayer's needs. This would include determination of--

* cash flow needs

* potential beneficiaries

* unique circumstances

* whether a "grandfather" election was made.

Once the information is gathered, various assumptions have to be made. They include the effective tax rate now and in the future, the effective estate tax rate, and the assumed growth rate of both qualified plan and personal assets. In analyzing qualified plan assets, a "net to beneficiary" paradigm should be used--a net of tax approach that assumes retirement plan distributions are either reinvested or consumed. In using software to calculate the effect of various strategies, care should be exercised in inputting the various assumptions.

The author presents an example of such planning with results shown in an exhibit.

Over the years, much has been written and many retirement plans devised in the so-called "MBA" mode. CPAs and other financial planners devise all sorts of elaborate methods to keep tax deferral going well over 50 or 60 years without ever considering the actual needs and goals of the client. This theoretical approach often "puts the cart before the horse," offering solutions to problems that have no practical relevance. This outcome is perfectly understandable because distribution planning ranks right up there with estate planning in terms of providing sophisticated solutions for seemingly simple, yet unexpectedly complex problems. In fairness to practitioners, the complexity of the tax rules coupled with the phenomenal potential for long-term growth through extended tax deferral virtually requires sophisticated analysis to inform the client of what he or she is missing if the "wrong" choice is made! All too often, practitioners quantify solutions that look good on paper but have no practical appeal. There is a need for a pragmatic approach to retirement planning. Presented here is a framework for the analysis of retirement plan assets and distribution options while considering the overriding objectives of the plan participant.

Information Gathering

Cash Flow Needs. Retirement plan assets have ballooned in recent years. For the average employee or self-employed individual, retirement assets consisting of pension plans, 401(k) and 403(b) plans, individual retirement accounts, tax deferred annuities, and the like, make up an increasingly large, if not dominant, portion of his or her net worth. Therefore, it has become increasingly important for planners to try to permanently optimize those characteristics that make investment in such plans so attractive. When reviewing retirement assets, however, advisors need to consider an overriding question: "What type of cash flow is needed for his or her living expenses?" Before we can even approach the analysis of the retirement assets and distribution options, we need to evaluate cash flow needs. As mundane as it sounds, budgeting and goal setting is much more important than sophisticated but impractical plans. In fact, we find the vast majority of Americans need to dip into their retirement accounts sooner rather than later to maintain any semblance of their preretirement lifestyles in retirement. So, the question is, do fancy tax strategies really help these folks? Maybe, but then again, maybe not.

Beneficiaries. Next, the analysis turns not to lifetime distributions, but to testamentary dispositions. The advisor needs to understand the cast of potential characters who may be deemed worthy of being named a beneficiary of a qualified plan. It is dangerous to assume who the beneficiaries should or should not be. There are pros and cons for selecting spouses, children, grandchildren (or trusts for their benefit), other relatives or friends, or even charities or charitable trusts (see Exhibit 1). Each of these possibilities offers distinct advantages and disadvantages when it comes to minimum distribution calculations and ultimate disposition at death. But all of these planning possibilities are rendered meaningless if the realities of family situations are ignored. For example, the taxpayer, who happens to be a grandfather, is not on speaking terms with his grandchild. Does it make sense to suggest naming that grandchild as a beneficiary if there is no chance the advice will be followed? The advisor must figure out the lay of the land first, then gather the ages of those potential beneficiaries who may, indeed, land on the taxpayer's financial planning radar screen.

Client-Specific Data. The next step is to gather together all of the data needed to analyze the taxpayer's situation. This includes copies of the latest statements from IRA plans; 401(k) and 403(b) plans; money purchase, profit sharing, and other types of defined contribution plans; defined benefit plans; Keoghs; and SEP-IRAs. These plans each have their own minimum distribution requirements, but can be linked when paying out the overall minimum. In addition, beneficiary designations and payout methods can vary from plan to plan, and any meaningful course of action must include data on all plans existing for the benefit of the taxpayer.

"Grandfather" Election. Finally, the advisor must be cognizant of the so-called "grandfather" election that may have been made with the 1987 or 1988 income tax return. The 1986 Tax Act instituted the 15% "success" tax on excess (lifetime) distributions from and excess (estate) accumulations in qualified plans. However, an exception was carved out for individuals with $562,500 or more in total qualified plan assets as of August 1, 1986, who made an election on their 1987 or 1988 income tax return to exempt those existing assets from ever being subject to the excise tax. This grandfather amount may have an effect on the calculation of the success tax. Typically, because of the legislated size of the threshold amount, the grandfather amount must be very large for it to have any effect on the success tax, either during lifetime or at death. However, it must be considered to make the analysis proper.

Making Assumptions

Once all the hard data is gathered, the fun part begins. Certain computational assumptions must be made and documented, including the effective income tax rate for the taxpayer now and in the future, the effective estate tax rate, and the assumed growth rate of both qualified plan and personal assets. The importance of these assumptions cannot be overstated. They must make sense, given the taxpayer's wealth and prospects for the future. They must be realistic with respect to the rates of return available in the marketplace for taxable and tax deferred alternatives. Both the advisor and the taxpayer must be flexible about the assumptions so that "sensitivity analysis" may be performed (determining the results under varying alternative scenarios and assumptions). The advisor must be diligent about documenting the assumptions, because changes in the assumptions can sometimes lead to unexpected results.

Growth Rate Inside and Outside the Plan. Some advisors have suggested that turnover within a portfolio, coupled with assumptions about transactions costs and capital gain or ordinary income tax treatment, will have a significant impact on the outcome of any strategy. This is true as strategies unfold, but almost impossible to plan for. Growth rate is a simplifying assumption and should be net of transactions costs. In actuality, this is how investors view their returns. They do not concern themselves with turnover issues except to the extent a brokerage account may be churned, or a money management fee may be excessive, negatively impacting the net return. If that is the case, movement to a new broker or manager is in order. The issue is administrative in nature and not a planning concern, because, as economists say, the effects can be "assumed away" through a net growth assumption.

Growth rates do not necessarily have to vary between qualified plans and personal portfolios. In fact, I like to use the same growth rate assumptions for both sides of the equation under the theory that investment decisions should be made based on a review of the taxpayer's growth and income requirements and risk profile. Therefore, I find the asset allocation decision should produce similar goals in both a taxable account and under a tax deferred umbrella. Some advisors prefer to use a triple-tax-free rate outside the qualified plan. This is fine, if it mirrors reality, as long as a taxable growth rate (presumably higher) is used inside the plan.

Tax Rate Assumption. Caution requires that apples and oranges are not compared. If state taxes are included in the analysis (and they should be), any calculations concerning lump sum distributions and special averaging taxes must be adjusted accordingly. In addition, it is important that the taxpayer's marginal effective tax rate is used rather than the average tax rate. All planning should be done at the margin, allowing each dollar of additional income to bear tax at the highest marginal rate, and each additional dollar of deduction to reduce tax at the highest marginal rate. Using an average rate instead of a marginal rate effectively "spreads" the tax hit over income that rightfully should be taxed at lower rates and is in direct contradiction of the progressive tax rate structure.

Estate Ramifications. Because of the complex tax deferred structure of qualified plans, distributions cannot and should not be analyzed in a vacuum. It is tempting to use a straight present value computation to value the stream of minimum distributions from a plan and compare it to a lump sum liquidating value today. The problem with this simplistic approach is that it does not factor estate taxes into the equation. It works only if there are assurances that the client will live for his/her life expectancy, a rather aggressive assumption. A better approach is one in which the growth and distribution options available during an account owner's life are analyzed in conjunction with the owner's estate planning objectives. Because of the structure of the current tax laws, retirement plan assets are often the worst assets to have in an estate. They are subject to estate, income, and possible excise taxes, which can confiscate up to 85% or 90% of the decedent's tax deferred assets in one fell swoop. Without explicitly addressing the estate ramifications in the analysis, there is the risk of producing an inadequate and ultimately useless report.

The statistic regarding the amount of shrinkage in qualified plans in an estate, however, can be misleading. Income taxes will always be due on retirement assets. Death does not change that. Timing, reinvestment, and tax deferral are the critical points here. While it is tempting on the part of practitioners to draw a parallel between qualified plan assets and personally held assets, the comparison is unfair. Presumably, any tax due on personally held assets has already been paid. Additional capital appreciation escapes income taxation at death, but not estate taxation. The whole package must be analyzed simultaneously to adequately quantify the effect on the client's financial situation.

"Net to Beneficiary" Model. For this reason alone, it is imperative that qualified plan assets be analyzed using a "net to beneficiary" paradigm. Multiple calculations are required to determine distributions net of tax, reinvested into a personal account with its own tax characteristics, and valued, again net of tax, at the death of the account owner and primary beneficiary. Alternatively, distributions may be assumed to be consumed currently and not reinvested. The "net to beneficiary" paradigm captures the effect of both current taxation on non-plan assets and the effects of long-term tax deferral in the face of successful tax implications for qualified plan assets.

The "net to beneficiary" paradigm makes a very significant assumption about the use of retirement plan distributions--either reinvested or consumed. This simplifying assumption must be addressed to be able to compare apples to apples. If distributions from qualified plans are consumed each year as distributed rather than reinvested, it must also be assumed the personal fund created from the after-tax proceeds of a lump sum distribution will be consumed in a similar systematic way. If the assumption is, instead, that distributions from retirement plans will be reinvested, net of tax, into a personal fund and allowed to grow in value over time, any comparative lump sum distributions must also be assumed to be reinvested, net of tax, into a similar fund to grow in value over time. In each year of the planning horizon, we calculate the net to beneficiary is calculated assuming that both the owner and the primary beneficiary die during the year. Without parallel assumptions regarding reinvestment, comparative analysis would be meaningless.

This critical assumption is, unfortunately, conceptually difficult for many advisors to grasp and, once grasped, equally difficult to explain. However, without consistent application of the paradigm, the resulting analysis is not only meaningless, it is dangerous!

I have also found, with the software I use for these computations, that it is easier to assume reinvestment of distributions into a personal fund. This approach will show an ever-increasing net to beneficiary amount over the planning horizon. But, again, is this approach truly a realistic assessment? For high-net-worth individuals it may be. Most people, however, do not have enough of an income or asset base to ignore their retirement plans as a source of funds for living expenses. This goes back to my basic premise--the taxpayer's goals, objectives, and financial situation are paramount. Once understood, the planning becomes much more pragmatic.

Software

Various software exists in the market for calculating the effects of various strategies. For this, I used Leimberg's Pension and Excise Tax Planner, version 3.04. This program is professional software and not suitable for public consumption. Its calculation engine is adequate for the job, but a solid knowledge of retirement planning is crucial to ensure proper results. Earlier on, I mentioned the need to determine and document certain computational assumptions. At this point in the analysis, it is crucial that each of these assumptions is entered in the proper place. For example, at the input screen for the plan data, Leimberg asks for growth rate assumptions in and out of the qualified plan. This is fairly obvious and easy to see. The tax rate screens, however, are in another section of the program. A special effort must be made to go to these screens to adjust the tax rate assumptions for current and future years. It is also at these screens that the decision to use five or ten year forward averaging in a particular scenario is made. The special averaging conventions are based on Federal law and must be adjusted for state implications. The message is clear. Inputs and assumptions must be triple checked to ensure proper results with Leimberg or any other program.

An Example

A simple case may be illustrative. A qualified plan owner, age 701Ž2 and resident in New York state, has just one plan with $4,000,000 in assets. Assume she has no other information about her goals or financial situation (cash flow needs, beneficiary designations, etc.). This removes the need to consider various beneficiaries with differing ages and the ancillary decision of using the recalculation or term certain methods of measuring life expectancies. The sole goal, at this point, is to maximize her "net to beneficiary" amount. Again, for simplicity, assume reinvestment of all distributions. The plan can say--

* take required minimum distributions (RMD), pay the taxes, and reinvest the net proceeds;

* distribute the so-called "threshold amount," if it is greater than the required minimum distribution. The threshold amount is the annual amount that can be distributed from qualified plans in total without incurring the 15% success tax. For 1997, 1998, and 1999, the success tax has been suspended for lifetime distributions but not for excess accumulations at death. The purpose of annually withdrawing the threshold amount as opposed to the minimum distribution is to lower the total amount remaining in the client's qualified plans at death to minimize any success tax due in the estate. For purposes of the analysis, a "bogey" can be constructed using the amount that would constitute the threshold amount had the law not be suspended. In our example, the threshold amount turns out to be less than the required minimum distribution under all scenarios, therefore, it does not apply here.

* or, take a lump sum distribution, pay the five or ten year averaging tax, and reinvest the proceeds.

Exhibit 2 summarizes the "net to beneficiary" amount using RMD's, five- and ten-year averaging under various rates of return (ROR), and at the end of 1, 10, and 20 years. The valuation "event" at the end of each of these years is the death of the retirement plan owner. All income, estate, and excise taxes are paid, and the net amount available for heirs is indicated. At the end of year one, the account owner would have maximized his wealth had he opted for five-year forward averaging and invested his money outside of the plan. This occurs for two reasons:

* The tax rate under five-year averaging, combined in an effective rate for both Federal and state taxes is less than the ordinary rate that would be paid on minimum distributions. Given the size of the distribution, however, this differential is very small.

* Because the lump-sum distribution was paid and taxed prior to her death, there is no success tax on the amount remaining in the after-tax personal fund.

The same cannot be said for the RMD scenario, where an excise tax in excess of $400,000 would be due and payable in the estate on the excess accumulation in the plan. This situation reverses itself somewhere around year seven. As illustrated by the year 10 numbers, the net to beneficiary amount for the RMD scenario is significantly greater. This occurs because the deferral is maintained, despite large amounts of excise taxes being paid, both on distributions during life and on accumulations at death. Compound tax deferred growth ends up being more powerful than confiscatory taxation!

This is the case in most circumstances. The positive effects of compound tax deferred growth generally win out over the evils of the excise tax. Rules of thumb can be deceiving, though. Under certain circumstances, earlier withdrawal may be warranted. I analyzed such a case recently. The client was in a unique situation for several reasons. As he had been a participant in his plan for many years and had "pre-1974" participation in the plan, approximately 53.3% of his plan assets qualified for the flat "capital gains" tax rate of 20% by Federal and 5.4% by state. The balance of his distribution qualified for five-year averaging. The combination of these two tax advantages created an effective income tax rate of 32.7% compared with an effective rate of 44% for other ordinary income. Coupling this income tax advantage with the three-year suspension of the success tax, the client's generally conservative approach to investments, and his desire to retain his wife as the designated beneficiary made a compelling case for taxing the lump sum distribution currently. Based on our analysis, we found that it would take approximately 21 years for the RMD scenario to outpace the lump sum distribution scenario as defined by the net to beneficiary paradigm.

Focus on Goals and Objectives

The most difficult part of such an analysis is boiling down all the variables into a summary that will make sense for the taxpayer and enable him or her to make an informed choice. The analysis is subject to the vagaries of the market, changing interest rates, an assumed level of risk tolerance, an assumption as to the current consumption of assets, and fluctuating tax rates. In making a choice, the taxpayer will be making certain implicit decisions about economic conditions in general and his or her life in particular. In the above case, the client's choice of a lump sum distribution was the most conservative approach. He chose the certainty of current low tax rates and no success tax over the vagaries of a more aggressive approach in investment style. Similarly, by choosing his wife, who happened to be the same age, as his beneficiary, he limited the upside potential of long-term tax deferral. If it had fit his goals and objectives (it did not) he could have named one of his children or grandchildren as his beneficiary. This would change the whole analysis.

Because of the large number of variables involved in retirement planning, it is easy to spend many hours analyzing plans in an infinite number of combinations. By focusing on the taxpayer's goals and objectives, the choices can be narrowed dramatically. Sometimes, if goals are unrealistic or if a certain financial structure will be particularly beneficial, there's a duty to educate them about their choices. This may take the form of quantifying the data, or merely advising them that another option may be significantly more profitable. In any event, implementation of a well-designed plan signals the first level of success with any engagement, and implementation will never occur if the plan is not in line with the client's goals and objectives. *

Raymond G. Russolillo, CPA/PFS, CFP, is director, Personal Financial Services, Price Waterhouse, LLP. This article was prepared with the assistance of Todd Kunego and Dan A. Yu, consultants with the Personal Financial Services Practice of Price Waterhouse, LLP.