The long-term average return from the stock market is 10.4%. As the earliest baby boomers...

... are now beginning to retire, they will be relying upon their investments for income. The latest boomers have two more decades to compound their savings into a retirement payload. At 10%, boomers young and old--so to speak--have a good chance of a secure retirement. Yet, from 2005, what length of time is needed to assure the long-term average return?

NEVER--investors from today will never achieve the long-term average return. Not in ten years, twenty years, fifty years, or the seventy-nine years that represent the most recognized long-term average return.

According to the 2005 Yearbook published by Ibbotson Associates, the long-term average return from the stock market is 10.4% (pg. 29). Ibbotson starts their long-term series of financial data in 1926 (pgs. 27, 201). Eight decades is a long, seemingly credible period of time--why shouldn't today's investors reasonably expect a similar return over the next one, two, or eight decades?

There are only three components to stock market returns: earnings growth, valuation-level changes (i.e. the change in the P/E ratio), and dividend yields. A discussion of these three components will confirm that a reasonable future return assumption is less than two-thirds of the long-term average.

Before we look forward, let's look backwards for insights. Let's use the certainty of history to explain the contribution of each of the components to the long-term average of 10.4%. According to Ibbotson, earnings growth contributed 5.0% to the long-term average (pgs. 178, 180). Since P/E ratios were 10.2 in 1926, the effect of the increase to 20.7 at the end of 2004 provided 0.9% to the long-term average (pg. 179). Finally, partially related to the starting and average P/E ratios, the dividend yield averaged 4.5% over Ibbotson's period of choice (pg. 180). Combined together, the compounded total return (before transaction costs, fees, expenses, etc.) averaged 10.4%.

So looking forward, from conditions that exist at the starting point of 2005, what are reasonable assumptions for the three factors over the next few decades? To assist in the discussion, concepts and data from the book Unexpected Returns: Understanding Secular Stock Market Cycles will be referenced.

First and foremost, we can eliminate the impact of significantly higher P/Es--the level of valuation cannot be reasonably expected to double to 41 over the next seventy-nine years. Given that we are near historical highs for the P/E ratio (excluding the two bubbles during the past century), any further material increase in P/Es in unrealistic. Past bull markets peaked with P/Es in the low to mid 20s; there are financial reasons, explained in Unexpected Returns (pg. 155-161), that P/E ratios cannot be sustained above the mid-20s. Therefore, if P/Es can at least be maintained at currently high levels, the best-case long-term return is 9.5%, the long-term average of 10.4% less the 0.9% impact of P/E expansion.

The second component, earnings growth, is closely tied to economic growth. Over the past decades and century, as discussed in chapter 7 of Unexpected Returns, earnings growth is closing related to Gross Domestic Product ("GDP"). GDP growth is comprised of real growth in GDP plus inflation. Today, inflation is being tightly controlled by the Federal Reserve Bank and is running below the historical average. As a result, future nominal earnings would be expected to grow at a slower rate than the historical past. Although it may not be much of a change, a 1% slower nominal growth rate shaves almost another 1% off of the potential return provided by earnings growth. Please keep in mind that if inflation does increase, the resulting decline in P/E ratios will more than offset the benefit to earnings growth. So with the more optimistic low-inflation scenario, we're down to a best-case long-term return of 8.5%.


The final component, dividend yield, is directly and mathematically related to the starting level of valuation--the P/E ratio (Unexpected Returns, pg. 103-105). In 1926, when the P/E ratio was close to 10, the dividend yield was approximately 5%. At the current P/E of 20, the normalized dividend yield drops to near 2.5%. The dividend policy and payout rates for companies do not change as the result of the level of its P/E ratio. A company that generates $2 per share will typically pay out $1 per share in dividends regardless of whether its stock price is $20 or $40 (i.e. 10x P/E or 20x P/E). Yet the dividend yield when the P/E is 10 will be 5% ($1 dividend on a $20 price), while the dividend yield at a P/E of 20 will be 2.5% ($2 dividend on a $40 price). The effect of today's valuation levels, P/E near 20, reduces the expected yield by more than 2% versus the historical dividend yield. As a result, our best-case future long-term return approaches 6%.

Of our three components in the future, two of them--earnings growth and dividend yield--are good soldiers that will provide a fairly predictable contribution to total return near 6%. The third component--changes in the P/E ratio--will determine whether realized returns are near 6% or are much less. The trend in P/E ratios significantly impacts multi-year returns. During periods when the P/E increases, earnings growth is multiplied; whereas, periods of P/E declines mitigate EPS growth. The result is periods known as secular stock market cycles. From currently high levels, any decline in P/Es will reduce long-term returns below 6%. The magnitude of the shortfall will depend upon whether the decline stops at the historically average level or further declines to typical secular market lows.

The discussion of the components for future returns is complete--all three parts indicate below average returns in the future. Earnings growth will be lower than average, unless inflation increases. Dividend yields will be well below average as a result of current valuation levels. P/Es cannot contribute their past benefits due to their currently high levels. Finally, a decrease in P/Es, due to higher inflation or other factors, would offset the resulting modest gains in earnings growth. In the aggregate, investors can expect that the long-term return, based upon 2005 as the starting point, will be less than two-thirds of the historical average. Once P/Es retreat to average levels, future long-term returns from that point will increase. From now to then, investors would suffer the effects of a P/E decline. And only when the starting point for P/Es is again at 10.2 can investors expect that the historical long-term average return will again be possible.

As a result of the current environment and conditions, investors have two alternatives: reasonable expectations or blind hope. Unfortunately for the boomers, historically average returns are not in the cards.

But What About Markowitz, MPT, & Your Stock Market Investments?

Modern Portfolio Theory ("MPT"), the model that acclaimed a Nobel Prize, should come with a warning label. "Use with caution. It's only as good as your assumptions." What did Harry Markowitz intend to impart with his ground-breaking research and what are the implications given a reasonable view of future long-term returns from today?

Harry Markowitz published his research titled "Portfolio Selection" in The Journal of Finance during 1952. He led with: "The process of selecting a portfolio may be divided into two stages. The first stage starts with observation and experience and ends with beliefs about the future performances of available securities. The second stage starts with the relevant beliefs about future performances and ends with the choice of the portfolio. This paper is concerned with the second stage."

Help! What about the first stage? What do you mean that the assumptions are OUR responsibility?!!

It's been many decades since the article was first published. Many, many 'buy-and-hold" constituents have reiterated their mantra in concert with Dr. Markowitz. But that isn't what he intended. Yes, investors should only be rewarded for taking risks that can't be neutralized. Yes, stocks have more risk than bonds and over time have realized higher returns. BUT, what if your timeframe isn't 75 to 100 years and what if you are starting from a period of relatively high valuations and the expectation of below-average future returns?

Please Dr. Markowitz, help me with my 10 to 20 year investment horizon. For that, we can reflect upon historical 10 to 20 year horizons for your assumptions. That is the first stage to which Markowitz referred--before MPT can be applied to your portfolio.

Since 1900, there have been 86 twenty-year periods, the first was from 1900 to 1919 and eighty-five double decade periods thereafter. The results can be sorted into two groups: those above the average and those below the average. Is there a way to determine whether the next twenty years is likely to be a top half or bottom half period? This would enable us to improve our outlook by using an above-average or below-average return assumption.

One characteristic that is blatantly obvious for the two halves is the starting level of valuation in the market as determined by the price/earnings ratio (P/E). It's the bellwether measure of prices in the stock market. Almost unanimously throughout the past century, when the P/E is above average, subsequent returns are below average. As well, below average P/E's historically delivered above average returns.

So since the current P/E is well above average, shouldn't the assumption for Markowitz's model be below average returns? Wouldn't this be consistent with the assessment of future returns provided earlier?


Markowitz gave us the holy grail to portfolio management; conventional wisdom has forgotten or ignored the need to use appropriate assumptions--the essential "first stage" of portfolio management. As Markowitz emphasizes, it is our responsibility to use "observation and experience" to develop "beliefs about the future performances." Although future performance of the stock market cannot be predicted with certainty, through observation and experience we may be able to at least refine the assumptions into above-average or below-average territory. Based upon current market valuations, it is very likely that we're in the 'below-average' batters box and should include a below-average return assumption for the next twenty years and even longer.

Oh no. Should we hang on to hope that this time will be different? Or should we rationally include a scenario that presents below average assumptions to Dr. Markowitz? Dear Dr. Markowitz, what should we do if the assumptions for stock market returns are below average?

Ed Easterling is President of Crestmont Holdings



Conclusion

I hope you enjoyed this week's Outside the Box. To read more insightful research by Ed Easterling, I highly recommend his book Unexpected Returns (www.amazon.com).

Your not expecting long-term average returns analyst,


John F. Mauldin
johnmauldin@investorsinsight.com

www.frontlinethoughts.com