For the last year, I've had a nagging feeling that the investment universe was changing in fundamentally significant ways that could have a profound impact on our clients' lives. I'm now convinced that it has, and it will. Although a long-term proponent of the application of mathematics to our work as financial planners, I think we've succumbed to sexy but simpleminded, psuedosophisticated analysis. In the process, we devote endless hours to touting (and inaccurately extolling) secondary issues and techniques.

At the same time, we ignore fundamental issues critical to our clients' well-being. The classic example of "rearranging the deck chairs on the Titanic" is our current fascination with Monte Carlo simulation, while we ignore the "iceberg" on the horizon. The real threat to our clients' well-being is not point-estimation; it's the current "assumption set" (e.g., expected return estimates for bonds and stocks) we use in the investment- planning process.

We have three choices: Continue to rearrange the chairs. Emulate the voyagers on the Titanic by heading for the lifeboats. Or, figure out how to deal with the iceberg.

It's high time for professionals to reconsider their assumption set (AS) and, if changes are necessary, develop strategies for dealing with the implications for our clients' futures. At the risk of disappointing the reader, a brief warning is in order. Although I do believe we need to make changes that will fundamentally affect our clients, I don't yet have a comfort level with what those changes should be. My goal in writing this piece is to refocus the profession's interest from the chairs to the iceberg, so that together we can reduce ambiguity and more effectively manage risk.

Monte Carlo Simulation

I don't pick on Monte Carlo simulation because I think it's a bad tool. It's a wonderful tool. Monte Carlo simulation is an effective way of educating people regarding the uncertainty of risks. Unfortunately, it's not nearly the panacea that is suggested by some commentators. Rather than reducing uncertainty, Monte Carlo simulation increases the guesswork manyfold. A point estimate requires a single guess. Monte Carlo requires three additional estimates for each point estimate: the shape of the distribution (a normal bell distribution is not a given) and the high and the low range for the distribution.

The problem is the confusion of risk with uncertainty. Risk assumes knowledge of the distribution of future outcomes (i.e., the input to the Monte Carlo simulation). Uncertainty or ambiguity describes a world (our world) in which the shape and location of the distribution is open to question. Contrary to academic orthodoxy, the distribution of U.S. stock market returns is far from normal. We need to recognize that in planning our clients' future, we're dealing with ambiguity, not risk. That leads me to believe our current AS is an iceberg of monumental proportions.

The Iceberg

The fundamental investment recommendation a financial advisor makes is the allocation between bonds and stocks. This allocation drives recommendations regarding savings and spending. In fact, it is one of the most significant influences on the quality of our clients' financial future. Whether we arrive at our recommendation for allocation by using a mathematical optimizer, capital-needs programs, a 12-C and/or darts, practitioners begin with an AS.

A practitioner's AS, at a minimum, includes estimates regarding the expected returns on bonds and stocks. They may be based on a total return or real-return estimates and may vary depending on the time frame of the analysis. In our practice, we use different forward-looking, real-return assumption sets for policy development than we do for capital-needs analysis (see sidebar).

As in most distributions, there are a few outliers, so advisors and academics tend to use an AS that reflects the long-term historical results recorded in the Ibbotson data. Total equity return expectations cluster around 11% and bonds, 5%. For those working with real returns, the related AS would be 8% for stocks and 2% for bonds.

As a profession, we are so confident about our assumptions that it is an axiom, repeated by all advisors (large and small, fee and commission, independent and captive), that "stocks are safe in the long run." The almost universally held belief is that, given a 20-year or longer investment horizon, the probability is effectively 100% that an investment in stock will outperform an investment in bonds.

Since the inception of financial planning as a profession, this assumption has served our clients well. We also religiously repeat, even if we don't believe it and without recognizing the inherent conflict, the mantra "past performance is no guarantee of future performance." It's time we reconsider what we tell clients, or we may not continue to service them well in the future.

Not What It Used To Be

"The future ain't what it used to be" is one of my favorite expressions. It's a catchy way to get the attention of those too comfortable with the status quo. It's also handy as a reminder that every professional who eschews ouji boards and follows rigorous econometric procedures can still be woefully off target when looking into the future. The Ibbotson/Sinquefield data is so important to our profession because it serves as the basis for our fundamental assumptions. So it's instructive to see, in retrospect, how successful Ibbotson/Sinquefield has been in predicting the future. Below, I've reproduced a table prepared by Professor Jeremy Siegel of Wharton in the Fall 1999 Journal of Portfolio Management. First, however, it's important to preface Professor Siegel's table with the caveat he included in his paper.

"My purpose here is not to highlight errors in Ibbotson's and Sinque-field's past forecasts. Their analysis was state-of the-art, and their data have rightly formed the benchmark for the risk and return estimates used by both professional and academic economists. I bring these forecasts to light to show that even the 50-year history of financial returns available to economists at that time was insufficient to estimate future real fixed-income returns."

If even Ibbotson and Sinquefield have trouble predicting the future from historical data, I believe Exhibit 1 is an effective wake-up call for practitioners who still believe in the sanctity of their interpretation of Ibbotson/Sinquefield data.

No Change

Before moving on to a discussion regarding how the future may look very different from the past, it's worth taking a few paragraphs to consider how we might be misadvising our clients, even if future returns are not too different from the past.

Professors Jones and Wilson in their article in the Spring 2000 issue of The Journal of Private Portfolio Management, "Stock Returns in the 1990s: Implication for the Future," provided a series of empirical probabilities of achieving specified returns for various holding periods, based on data from 1920-1998. Contrary to popular belief, they demonstrate that at 20 years, the probability of an investor earning a compounded rate of at least 7% based on the historical record was only four chances out of five. Even at 40 years, the probability of an investor earning a compounded return of 8% or better was only slightly better than four out of five.

And that's the good news.

Bubble Logic

Last year, Robert Levitt, a friend and professional colleague, provided me with a copy of Bubble Logic, Or How to Learn to Stop Worrying and Love the Bull by Clifford Asness. Although during the last year I'd read numerous articles discussing equity risk premium and the future of stock returns, I'd paid (in hindsight) inadequate attention to the implications of the articles. "Bubble Logic" blew away my indifference. Asness concluded his book with:

"Put simply, there are really three possibilities for the broad market.

1) Investors understand and are now more comfortable with a very low expected return on the stock market going forward.

2) We are in for an exceptionally long period of exceptionally high growth in real earnings that justifies today's market prices.

3) Most investors are not really thinking about either 1) or 2), but are engaged in wishful thinking..."

Asness' powerful, persuasive and often humorous writing convinced me that I fell into No. 3. That started me questioning my most closely held and revered investment assumptions.

It was sobering to realize that I, an "expert," might fit all too neatly into the camp of deluded wishful thinkers. Still, before jettisoning my current assumption set, I recognized that it was prudent to look further than one person's opinion. So, I collected the opinions of others from the journals and professional press and badgered professional money managers I respected for their thoughts. Although the basis for their conclusions varied widely, they were uniformly consistent. The last decade of wonderful equity markets is an anomaly. The expectation for future equity market returns equaling the historic "average" is, at best, grossly optimistic.

Possible Futures

As an introduction to the real new millennium, Jonathan Clements of the Wall Street Journal corralled five market icons and asked them to share their market views. Although their time horizon was five years, their conclusions so closely parallel the range of long-term expectations found in my readings and conversations that I'll use Jonathan's concise survey results as a good overview:

Jeremy Siegel: "My feeling is that we might get a real rate of return between 5% and 8%. If you add inflation, you might get between 8% and 11%." Note that 11% is at the high end of the range.

Peter Bernstein: "...[he] figures stocks will deliver three or four percentage points a year more than inflation. He says that Treasury bond investors will probably earn comparable returns ..."

John Bogle: "He looks for the market's price-earnings multiple to fall below 20 over the next five years ... Results? Investors will collect just 3% per year."

Robert Arnott: "... reaches an even more dire conclusion. 'Our expectation is that the return over the next five years should be zero or less.' "

Robert Shiller: "It's quite possible that we could have negative returns for the next five or even 10 years."

Equity Risk Premium

As reflected in the survey above, there are innumerable opinions regarding the future of the equity risk premium, far too many to address in detail in this article, so I've elected to briefly describe the logic behind two examples that represent the range of opinions.

Siegel represents the "high estimate." Since the publication of his book Stocks For The Long Run, he has been the poster child for equity optimism. Consequently, it's sobering to find an optimist such as Siegel penning an article titled "The Shrinking Equity Premium, Historical Facts and Future Forecasts," that concludes, "the degree of the equity premium calculated from 1926 is unlikely to persist in the future." Agreeing with most other commentators, Siegel adds, "Furthermore, despite the acceleration in earnings growth, the return on equities is likely to fall from its historical level due to the high level of equity returns relative to fundamentals."

Siegel also concludes that the real return on fixed-income assets is likely to be significantly higher than estimated on earlier data. For practitioners, this is as important as the conclusion regarding the equity risk premium, and I'll address the implications later in this article.

The "low estimate" is well-represented by Anthony Brown's "What's Next for the S&P 500?" in the Winter 2000 issue of The Journal of Investing. Brown concludes, "We argue that the S&P 500 is not only unlikely to match its recent 20-year return over the next 20 years, but that it may also very well fall short of the long-term 11% return (since 1926)." What is instructive about Brown's article is the thoughtful and analytical process he uses to reach his conclusion. Decomposing historical return into the components of dividend yield, earnings growth and valuation change, he considers each in detail.

In his evaluation of dividend yield, Brown considers its significant decline vis-a-vis long Treasuries since 1958, the possible reasons for this change and its impact on stock prices. In his evaluation of earnings growth, he acknowledges the possible "catch-up" effect from the dismal performance of the 1970s. He also considers the impact of share repurchase, integrating these influences to quantify an estimate for the future. As for valuation, he utilizes P/E as a valuation proxy and concludes that there is no reason to expect a systematic positive valuation change in the future.

Based on his deconstruction analysis, he concludes (and describes in detail his rationale) that the implied equity risk premium going forward is only 1.7%. Finally, he suggests a reason for Siegel's observation regarding the increasing return on bonds by noting the increasing risk of bonds, particularly since the 1970s, as a result of an inflation risk not being reflected in the long-term data.

What, Me Worry?

Despite this growing consensus by both academics and practitioners that future equity returns may be a tad disappointing, many investors and advisors seem bent on denial.

In Bubble Logic, with specificity and humor, Asness addresses many of the classic denial responses, such as "Equities always win over in the long term," "My estimates of expected stock returns are based on long-term data" and "The long term will be OK, we've entered a period of spectacular growth." One shouldn't forget Paul Samuelson's famous remark that the long-term data for the U.S. stock market is based on a sample of one.

Exhibit 2 from Bubble Logic, demonstrates that by any historical standard, the market today is significantly priced.

Framing this historically unique market valuation from a different perspective, Barr Rosenberg, in a presentation to the AXA Investment Managers Client Conference, noted that the market capitalization of the 500 largest companies in the San Francisco bay area was $3.5 trillion. By comparison, he pointed out that the market capitalization of all of Asia (except Japan), a region of 3.2 billion people, including China and India, was $2.2 trillion.

However one wishes to frame the issue, there is certainly a strong case to be made that equity returns, in the long run, may be lower in the future than the AS currently used by most practitioners.

Don't Forget Bonds

While many academics and practitioners expect the future equity risk premium to be lower, they also expect the bond risk premium to be higher. Going into great detail to defend this expectation is unnecessary. With TIPs currently yielding more than the 1%-2% real return reflected in the traditional AS, there can be little question that our assumptions need revisiting.

If one concludes that the equity risk premium remains stable and the bond premium increases, allocations will shift in favor of bonds. The risk of such a shift is that the stock premium might actually decrease and a shift to bonds will prove counterproductive.

If the conclusion is that the stock premium will decrease and the bond premium increase, we can no longer say with assurance that "stocks will outperform bonds in the long-term." We will also have to revisit the reality of our clients' ability to achieve goals previously believed to be obtainable.

Conclusion

Practitioners need to consider and plan for the impact of the changes in risk premium. I believe that, until recently, I've personally succumbed to wishful thinking. In addition, we must consider investor heuristics. On one hand, a reduced stock return would suggest an increase in the stock allocation. On the other hand, the uncertainty of stocks' long-term superiority vis-a-vis bonds will multiply a loss-averse client's discomfort with an increased stock allocation. Our quandary is further compounded by another behavioral response noted by Rosenberg. Namely, "Investors have forgotten the meaning of risk: They tend to believe that risk guarantees reward: under that interpretation, risk is not risky."

Regarding investment issues, I've revisited my position relative to risk premium and tentatively concluded that the bond premium has increased to 3%-4% and that the stock premium has probably decreased by an unknown amount. I also agree with the conclusions of Professors Chan, Karceski and Lakonishok that recent stock price performance is neither a result of real asset pricing or a new paradigm economy. The explanation is behavioral.

Regarding behavioral considerations, I believe that my clients look to me to use my professional skills and knowledge to assist them in achieving their personal goals, including my ability to balance the expectations for an uncertain future with an understanding of my clients' psychological, emotional and behavioral attributes. They want to know such things as:

How much can I spend?

How much should I save?

How long should I work? How realistic are my retirement plans?

How much can I afford to give my children?

How much do I need, and can I afford long-term-care insurance?

Based on the issues raised in this article, I'm less confident in my answers. Hence, I look forward to learning from the research and experience of my friends and professional colleagues as they address these same issues.