As William Faulkner wrote, "It's not even past."

This month's headline comes from a William Faulkner novel, Requiem for a Nun. According to an authoritative Faulkner Web site, the line refers to the way that people's present actions tends to mirror their past behavior.

I think it provides an answer to the question Mark Spangler posed at a recent NAPFA session about hedge funds. He asked, "Why do investors continue to believe in a simplistic investment model that has produced such discouraging results?" (He was referring to the traditional approach to diversifying portfolios by allocating resources between publicly owned stocks and bonds.) Spangler thought "wishful thinking" might explain why three years of pain has not changed behaviors; a 60/40-stock/bond portfolio worked for two decades, so advisors hope it will continue to work. Faulkner's notion that most folks simply do not change the way they behave may seem a little dark, but it does have a ring of truth about it.

Indeed the 60/40 club, to which most of us have belonged at some time in our careers, still seems to boast a very large and active roster. I dropped out and formed a 30/70 splinter group some years ago; but while this change has kept our clients in the black during the ferocious bear market, it has not provided the kind of absolute returns that retirees require.

As I contemplate the ways in which the world has changed and continues to change, it occurs to me that investment success in the future may not depend simply on our arriving at an appropriate stock/bond mix. Rather, to provide truly profitable advice, we may have to abandon the entire stock/bond allocation paradigm to which we have become accustomed.

Underlying Assumptions

Two fundamental assumptions that we rarely think about are imbedded in the concept of providing diversification by diluting stock market exposure with a modest dose of bonds. First is the assumption that over a long enough period of time, stocks will provide the patient investor roughly twice the returns on bonds, about 10% versus 5%, right? Therefore, all things being equal, stocks are the asset class of choice. But, of course, as history teaches us, a stock investor may from time to time have to endure relatively brief periods of negative returns, a.k.a. losses, and sometimes rather serious losses, as recent events have impressed upon our collective consciousness. This can be vexing, particularly to the retiree who is making regular withdrawals from his or her portfolio. Hence, the species Homo sapiens investori evolved the concept of blending the asset class of choice with bonds, a non-correlated, lower-return investment class that tends to reduce portfolio volatility. Enter the second assumption underlying the traditional approach to asset allocation: that interest-bearing debt securities will be stable in price and rarely experience a year of loss.

Armed with these assumptions, a mainstream advisor is expected to examine historic periods of stock market decline, ask his client to declare the degree of loss he could endure without leaping from the Talahatchee bridge, and then mix in enough low return, low-volatility bonds to seriously reduce the risk of such a tragedy. But how trustworthy is this traditional approach? And how valid are its underlying assumptions?

Does an examination of historic returns adequately describe stock market risk? Or opportunity? Perhaps we can best answer with another question: Did history prepare advisors for a third down year in the averages? Or, what experience suggested to advisors in 1966 that 1,000 in the DJIA would prove virtually impenetrable for 16 years? Did anyone forecast that for 13 years from 1990 through 2002 stocks would return less than long Treasuries (9.7% versus 10.1%)? Or that for the 15 years through 1979 stocks would earn less than inflation (5.6% vs. 6.2%)?

How dependable is the assumption that stocks of the publicly traded variety will provide higher returns than bonds; enough higher to compensate investors for the risk of ownership? (In recent weeks WorldCom bondholders emerged from that company's ashes; shareholders did not.) Such highly regarded investment thinkers as John Bogle, Bill Gross and Peter Bernstein believe that for the next decade stock and bond returns may be nip and tuck. Now, a ten-year outlook may not matter to the Coca-Cola pension fund or Harvard Endowment, but it could be enormously important to the 68-year-old retiree whose equity portfolio, having shrunk 40%, still looks pricey at 32 times trailing 12-month earnings.

And what about that assumption regarding the low volatility of bonds? And just what bonds are we talking about? Junk bonds, always volatile, are a much bigger part of the bond pie than they have ever been. And an investment-grade rating doesn't offer the assurance we once ascribed to it; witness radical downgrades pursuant to headlined accounting scandals.

Just how much portfolio protection can we expect from a 2.8% coupon on a 5-year Treasury bond? If rates rise to a rather normal-sounding 4.8%, the price of that bond could tumble 8%. TIPS, a new entrant in recent years, have exhibited heroic upside volatility as inflation fears subsided in 2002-03; where there is upside, there is downside, n'est ce pas?

Most bond investors fear the effect of rising interest rates once the economy gains traction. The usual reaction is to shift the portfolio mix away from bonds and more toward stocks. But what if interest rates should rise, not because of increased corporate demand for loans in an improving economy, but because the Fed puts the money supply on steroids as a way of heading off deflation? Couldn't this sink the dollar, drive out foreign savers and stimulate inflation fears. In such circumstances, one could easily imagine higher interest rates, stagflation, falling corporate profits and falling stock prices even as bond prices were being pummeled. Advisors who assumed that stock and bond returns are not correlated could be in for a shock. Regardless of whether they had positioned retirement portfolios as 60/40s or 30/70s, many of their clients would be updating their resumes, looking for ways to cut their living expenses ... and possibly interviewing advisors as well.

Can We Trust The Facts?

After three years of crushing stock price declines, it would be easy to assume that U.S. equities must be on sale. We derive a lot of conviction from the historic facts; since the end of 1925 stocks have returned an average of 10.2% a year (Ibbotson), eclipsing the 5.9% average for long-term corporate bonds. Surely we can build our strategies on this historic fact. After all, the 77-year period included a worldwide depression, two world wars and the recent ruthless three-year bear market. The record suggests that now is not an appropriate time for changing investment theories; those who stay the course and refuse to panic will surely be rewarded as they have been in the past.

William Faulkner was no economist (for which he was probably most grateful), and was making a very general observation about life when he said, "I don't care much for facts; am not much interested in them. You can't stand a fact up; you've got to prop it up. When you move to one side and look at it from that angle, it's not thick enough to cast a shadow in that direction." I respect Faulkner as a canny observer of human nature, so this got me to thinking about the facts of stock and bond returns and the approach to portfolio diversification that we have built upon those facts.

The 10.2% and 5.9% figures for 77 years present a rather compelling case in favor of stocks as the long-term asset class of choice. And the 38% plunge in large stocks the past three years is adequate evidence of their short-term risks. These are the facts that support the 60/40 approach to portfolio diversification. But if we look at them from a different perspective we may identify with Faulkner's skepticism. For example: from the end of 1980 through 2002, a period of 22 years or roughly a generation, returns from stocks edged out bonds by a measly 0.3% per year (12.2% vs. 11.9%).

What if we exclude from consideration the anomalous data from the'20s bubble and the Great Depression? That leaves us the latest 60 years of performance information. In the first 30 (1943-1972) stocks returned +13.5% a year, overwhelming bonds (+2.7%) by a whopping 1,080 basis points a year. Now that is casting a big shadow. But in the second 30-year stretch (1973-2002) the story was remarkably different, with +10.7% for equities and +9.3% for Treasury debt, a spread of only 140 basis points.

When the folks at Bank Credit Analyst looked over the historic data recently, they concluded that the only great era for stocks was the 1950s and 1960s. And that, they said, was because the period started with cheap stocks (seven times earnings) and expensive bonds (2% yields). Today one could argue that both asset classes are expensive; it seems appropriate to wonder how that will influence stock/bond relationships in the years ahead.

Changes Galore

Well, if historic stock and bond relationships provide little guidance for our investment strategy, what other fundamentals are up for grabs? Plenty, it turns out: inflation, earnings, productivity, demographics and foreign trade, to name a few.

Since Paul Volcker grabbed inflation by the lapels more than 20 years ago, keeping the general level of prices subdued has been the Federal Reserve Bank's primary mission. But in the last six months, despite a CPI that seems to be hovering in the safe/normal 2% to 3% range, no less than three prominent Fed members have announced plans for dealing with deflation. In an effort to explain what is spooking the Fed, the April 25 issue of Grant's Interest Rate Observer highlighted a chart produced by Dresdner RCM Global Investors showing 45 years of a "non-financial corporate price deflator." Except for the OPEC-dominated 1970s, this inflation index has been largely contained between 0% and 4%. In the past two years, however, it has ranged from 0% to -3%. I hope you noticed the minus sign. Apparently the Fed has.

And then there are earnings. Since stocks represent ownership of businesses, the profit that those businesses are capable of generating has always seemed to have at least some relevance to the value of stocks. Over many, many moons, price/earnings ratios for U.S. publicly traded stocks have ranged widely and erratically between, say, 7X and 20X, averaging something like 15X. At this writing, the S&P 500 trades at 32 times the latest 12 months' earnings. If that were all the information we had, one would feel justified in suggesting the market looks rather expensive in terms of its historic experience.

As we know, earnings have been depressed of late, what with a recession, a war and a scandal or three. Because earnings have always bounced back in our resilient economy and gone on to unprecedented heights, many among us believe that an optimistic 12-month forward estimate of "recovery earnings" is the appropriate denominator. Calculated this way, I am told, the market P/E is a more typical 16X, not anything to be afraid of. Compared with the ten-year Treasury yield (the so-called Fed Model) some say it is a bargain.

But I have a problem developing confidence in the earnings numbers, both reported earnings and, of course, estimated earnings. It's not just the constant restatements, the admitted frauds, the understatement of options expenses, and the fact that pensions are becoming a drag after some years of contributing to the bottom line. My bigger problem is with the working assumption that when revenues do begin to recover, profit margins will zip back to the levels we knew in the late 1990s. A Wells Fargo chart of profit margins over the years suggests that the run-up in margins in the last decade was an aberration, and that what look like depressed profit margins today are actually "high normal" by longer-term standards. A snapback is far from guaranteed.

Space constraints prevent even a cursory treatment of productivity, demographic and trade concerns. I'll just mention that Mr. Greenspan's vaunted productivity figures may be unrealistic (Grant's, April 25, page 10), aging populations and projected smaller work forces in most developed countries put social contracts at risk, and the splintering of traditional alliances over the Iraqi war cast a pall over global trade prospects.

The U.S. and world economies seem to be struggling simultaneously with cyclical and long-range shifts of serious magnitude. The adaptation process could alter many financial and economic relationships that we have considered "normal," so that they will no longer serve as reliable guides to appropriate investment behavior. Though we are understandably inclined to repeat what has worked for us in the past, our survival as financial advisors may require substantive change.

In future columns I will offer some ideas for a new, different way of allocating portfolio resources to achieve both diversification and reasonable returns.

J. Michael Martin, JD, CFP, is president of Financial Advantage in Columbia, Md.