Living The Same, Longer

by

David J. Drucker

December 2007

The first time it happened was in the 1990s when Lynn Hopewell said, “Why not Monte Carlo simulations?”

Hopewell, an ex-CIA communications engineer, entered the financial planning business in the 1980s, and after experimenting with popular planning methodologies for a decade and uncovering their inherent faults, brought his Harvard MBA education to bear on the problem. In an October 1997 article in the Journal of Financial Planning entitled, “Decision Making Under Conditions of Uncertainty: A Wakeup Call for the Financial Planning Profession,” Hopewell (who passed away in March 2006) questioned the use of straight-line investment return assumptions by every existing financial planning software package (at the time).

He said, in so many words, “We’ve already got a better system that statisticians have known about and applied for decades—Monte Carlo simulations. Why are we using straight-line assumptions when Monte Carlo simulations already exist?” Today, you would be hard-pressed to find financial planning software that doesn’t include a Monte Carlo option.

There’s a phenomenon at work here, which is that our industry has matured in something of a vacuum and is in danger of developing “best practices” that don’t always serve the best interests of advisors or their clients. Real innovation often comes from eclectic influences outside the industry, which is why we should pay attention to what Dr. Laurence Kotlikoff has to say when—as Hopewell did—he asserts that we’re doing it all wrong.

What Dr. Kotlikoff doesn’t like about modern-day planning software is evident in the way we use it. The drill usually goes something like this: We input demographic information for the clients, along with their income, expenses, assets and assumptions including inflation rate and return on investment. Our software then shows us the levels of assets that will be saved or consumed between now and death. If we lower our spending assumptions or fiddle with other variables, assets will endure for longer or shorter periods. By varying our many assumptions, we attempt to develop strategies to help our clients achieve their goals without prematurely consuming all their assets.

“There are several problems with this,” says Kotlikoff. “With [traditional planning software], risk comes from investments and translates into running out of money; with my methodology, the risk isn’t running out of money, but having less to live on. Traditional planning methodologies stick with the client’s assumed living expense target even if the market crashes and the client’s assets drop. Or, they direct the client to spend 4% of his assets every year no matter what happens [externally]. In real life, the client would adjust his living expenses downward in response to an external event like a market crash.” In other words, traditional planning software doesn’t take into account human reactions; it goes only as far as “static thinking” can take it.

To correct for this inadequacy, Dr. Kotlikoff’s own financial planning software—E$Planner (www.ESPlanner.com)— relies on “dynamic programming,” a mathematical process for solving problems involving sequential decisions. “Dynamic programming considers the future decisions we are likely to make in order to determine a recommended course of action in the present,” says Kotlikoff, “thereby improving upon traditional planning methods by more nearly approximating the way we make decisions in real life.”

When Hopewell introduced Monte Carlo simulations to the financial planning world, he was drawing on a statistical sampling method that had been around since at least 1930 (although it didn’t begin to be widely used until 1945 with the advent of the first electronic computers). Similarly, dynamic programming has been around since the 1940s when it was first developed by mathematician Richard Bellman. He used it a decade later in his work with the RAND Corporation, and it became widely used by economists such as Paul Samuelson and Eugene Fama in the late 1960s and early 1970s.

Kotlikoff, like Hopewell, has simply introduced to the planning profession from the outside world a better way of doing what we do — with a twist. “We have a patent for the particular type of dynamic programming used in E$Planner,” says Kotlikoff. “We have a few dynamic programs iterating with each other.” Without that refinement, it would take hours to run a plan, much as Monte Carlo simulations sometimes took minutes instead of seconds in the software packages in which they were first adopted.

So if traditional planning software focuses on the sufficiency of a client’s assets, E$Planner focuses on the sustainability of the client’s living standard and on his optimal life insurance coverage. In fact, by all appearances, it is the implementation of Nobel Prize-winning economist Franco Modigliani’s “life cycle consumption smoothing model.”

So how could anyone refute E$Planner’s approach? The software has an authoritative foundation and its approach is exceedingly logical. Yet, E$Planner is surrounded by controversy. Why? Because, in many cases, the program tells clients to save less and buy less life insurance than would most traditional planning analysis. Advisors like the concept behind E$Planner, but choke on the results because they’re sometimes so out of tune with the answers advisors are used to seeing.

In his February 2006 paper entitled, “Is Conventional Financial Planning Good for Your Financial Health?” Dr. Kotlikoff compares the results of E$Planner with four Web-based retirement calculators from Fidelity, American Funds, Vanguard (using Financial Engines) and TIAA-CREF, and found these calculators’ recommended savings levels to be between 36% and 78% higher than those of E$Planner.

Why does this occur? Because all financial planning is geared to consumption smoothing, that is, determining what living standard can be sustained both pre- and post-retirement. Says Kotlikoff, “However, [using traditional methods] the client is forced to set his own retirement spending targets consistent with consumption smoothing, and this is incredibly difficult to do. Even small targeting mistakes of as little as 10% can lead to enormous mistakes in recommended savings and insurance levels.”

What exactly then is the output from E$Planner that forms the crux of the advisor’s recommendations to his clients? Consumption. Kotlikoff defines consumption as all expenditures by the household other than special expenditures (e.g., buying a new car every X number of years), housing, taxes and life insurance premiums. By taking into account the client’s current and future economic resources, current and projected demographics, current and projected housing spending, special expenditures, taxes, life insurance premiums and any expectations the client has about increasing his future standard of living, E$Planner spits out a recommended annual consumption amount the client should be able to maintain indefinitely.
Recommended annual consumption also forms the benchmark for “what-if” comparisons. The user can test the effectiveness of competing investment strategies, Social Security benefit timing decisions and the like by comparing the recommended annual consumption levels for one strategy versus another.

Charlie Ryan, a research analyst with The Family Firm, a fee-only advisory firm in Bethesda, Md., has done approximately 20 plans with E$Planner and says it has worked well for clients. “Most clients have one general desire: to find the highest sustainable standard of living throughout their lives, and that’s just what E$Planner does. You tell it the client’s income, assets and mandatory expenditures like housing, transportation and non-life insurance, and it computes a level of allowable, non-mandatory spending that is constant throughout the client’s lifetime. E$Planner refers to this non-mandatory spending as ‘consumption’ or ‘standard of living.’”

Ryan offers an example of a client with no savings, $50,000 in income, mandatory housing expenses of $12,000 a year and taxes of $8,000 a year. “With $30,000 left, E$Planner will compute a sustainable, unchanging level of discretionary spending through retirement all the way to the end of the plan. For the sake of this example, let’s say that level is $20,000 with the remaining $10,000 going to savings.”

Next, says Ryan, let’s enter a new assumption into E$Planner, namely that the client will get a $30,000 pay raise in five years; how would that change E$Planner’s recommendations? “E$Planner will recommend a higher level of discretionary spending now, perhaps saying all income should be spent now with no current contributions to savings. Why? Because more savings can occur later from the client’s higher pay, and the idea is to smooth the client’s standard of living over time.”

Rick Miller of Sensible Financial Planning in Cambridge, Mass., is another big fan of E$Planner. “When I started in the business five years ago, I looked at everything; E$Planner was the only software that rapidly delivered comparative solutions for a variety of scenarios. I don’t have to guess at how much people can spend in retirement.”

Miller points out a feature of E$Planner that would seem to fit well with the emerging emphasis on decumulation planning for baby boomers. Anyone who’s worked with retired clients has heard questions like, “Can I afford to buy a new car?” or “If I spend $10,000 on my vacation this year, is that too much?” It’s easier to answer these questions if your software is telling you what your client’s sustainable standard of living is as opposed to how long his assets will last.

How does Miller explain E$Planner’s results to clients? “I say we’ve added up all your resources, spending on things you want to do that is ‘lumpy,’ or not what you would ordinarily spend in a year, as well as housing and other expenses that are likely to change in a predictable way over your lifetime. We subtract the total of these items from your resources, and what’s left is what you can spend on day-to-day living.” On that basis, says Miller, he can then go on to answer client what-if questions, like the impact of retiring earlier or later, sending kids to one college rather than another, and so on.

And speaking of college, E$Planner doesn’t have a college planning function, per se. Nor does it address a client’s possible need for disability insurance or the pros and cons of different types of executive compensation (stock options, restricted stock, deferred comp, etc.). It only does these things if the parameters of these modeling exercises can be fit into the E$Planner framework.

For example, E$Planner can answer the question, “Should I enroll in college, take a student loan, and forgo earning income for the next four years of my life in order to get a college degree that will enable me to earn more in the future?” Says Kotlikoff, “The question isn’t whether college grads will earn more once they start working. It’s whether they can achieve a higher living standard over their lifetimes given the costs of attending college.”
If, says Kotlikoff, we assume 18-year-old Rebecca is thinking about borrowing to attend a $40,000/year college and expects, after graduating, to earn $22,600 more than a high school graduate (the average earnings spread as computed by The College Board, which administers the SAT exams), the answer isn’t so clear-cut. Why? Because the college loans will drag down net consumption for years. Obviously, much depends on the cost of the education, Rebecca’s major and her resulting earnings potential, but just knowing college grads earn more than high school grads isn’t enough, by itself, to recommend the college experience. Says Kotlikoff, “Borrowing to raise your future labor earnings is not much different from borrowing to invest in the stock market. They both entail risk.”

If a planning issue’s parameters can’t be fit into the E$Planner framework, then the question goes unanswered. Bedda D’Angelo of Fiduciary Solutions in Durham, N.C., says, “E$Planner’s good with income taxes and cash flow, but there is no way to model different types of trusts and estate planning tools. In my opinion, E$Planner is less comprehensive than other commercial programs.”
Upon reviewing E$Planner for use in his company, Steven Weydert of Bowyer, Weydert Wealth Planning Partners Inc. in Park Ridge, Ill., also reached the conclusion that there are simply too many planning areas not covered by the software. “Consider this,” says Weydert. “Today there are nearly two dozen discussion forums on E$Planner’s Web site. I randomly clicked on the one entitled ‘How Do I Use E$Planner’ where I found eight more sub-topics. I then clicked on ‘How Do I Model Contributions to Health Savings Accounts?’ and the response from Professor Kotlikoff was: ‘We have not yet formally included HSAs in E$Planner. We will likely do so in the near future. In the meantime, enter the contributions you plan to make as ‘special expenditures’ that are excludable from AGI.”

In other words, Weydert found “work-arounds” for E$Planner but not specific attention to a number of important planning areas. In addition to HSAs, he says E$Planner lacks a specific planning option for reverse mortgages, nondeductible IRAs and 529 plans, to name just a few.
Is this as big a problem as D’Angelo and Weydert make it out to be? It really depends on the nature of your clients and on your philosophy of financial planning. Says Miller, “Is it comprehensive enough? I’m not sure because there’s no perfect financial planning program out there. As for E$Planner, I believe its benefits outweigh its deficits. It covers the crux of what we do.”

A competing software program—MoneyGuidePro—has acquired thousands of users without being comprehensive. Although it has taken on added functionality since its introduction, MoneyGuidePro has always focused primarily on one thing, which is answering the question, “Can my client get through his lifetime without running out of resources?” Although its methodology is different from that of E$Planner, MoneyGuidePro appeals to a broad spectrum of planners who are more concerned with getting the crux of financial planning right than generating a “comprehensive” plan.

Another criticism voiced about E$Planner is that it works, but not for all clients. Says Scott Neal of D. Scott Neal Inc. in Lexington and Louisville, Ky., “If E$Planner is controversial, it’s because it deviates from our standard stuff, not because it’s fundamentally flawed in some way. Like many tools we employ, it has a place and application for which other tools are inappropriate.”

Neal goes on to say that, when working with a client for the first time, he looks for what might be subtle signs of the client’s strategic goal before getting into more precisely defined goals. “For example, I listen for clues that would help me determine whether the client fits into one of the following four camps: a) client wants to maximize lifetime security, b) client wants to maximize lifetime consumption, c) client seeks optimal lifetime income smoothing, or d) client wants his planning to help him achieve something intangible, like power, leadership, recognition or fame.”

Obviously, says Neal, Dr. Kotlikoff’s software meets the needs of the third type of client. “Most clients come to us either in Camp A or B and are, in fact, often in conflict about which of those they are in. For the few who truly fit in Camp C, E$Planner is the tool of choice.”

Ryan believes the choice of software programs depends on the client’s level of wealth. “E$Planner works great for folks with moderate incomes and savings. It does not work well for very high-income or high-net-worth clients. Suppose a retired client shows up with $10 million in savings. Since E$Planner calculates the highest possible level of spending, it might recommend a level of discretionary spending that’s substantially higher than the client’s actual spending.”

That said, my guess is there’s a work-around for that situation, too. Although this article seeks to discuss E$Planner’s philosophical underpinnings more than its nuts and bolts, this author has tested the program for himself (a pre-retirement “client”) and for one post-retirement client, finding that it handles both scenarios quite well. Yes, there are some bugs (e.g., wrong names sometimes print out on reports) and minor inconveniences (E$Planner doesn’t flag a user’s failure to fill in certain critical assumptions or data and, subsequently, may give the user wildly off-base recommendations necessitating a wild goose chase through the program’s input screens to find the incompletes).

Nonetheless, it does seem likely Dr. Kotlikoff’s planning methodology, if not his software, will continue to attract attention and will very likely impact the financial planning software market within the next few years. Says Miller, “New ideas have trouble gaining traction in the [financial services] market, and we’ve all got our own ways of doing things. Someone could theoretically come out with planning software that did everything and people would still be slow to adopt it.” As for Miller, he says his own business processes are built around E$Planner and he’s sticking with it.

If you want to learn more about dynamic programming or living-standard-centric planning, you’ll find the prolific Dr. Kotlikoff’s many writings online, and his next book — Spend Till the End — will be published by Simon and Schuster in the spring of 2008.

An independent financial advisor since 1981, David J. Drucker, MBA, CFP, has been an industry influential for many years. Learn about his upcoming Technology Tools for Today Conference—the industry’s premier technology conference for financial advisors—at http://www.daviddrucker.com (http://www.daviddrucker.com/).