5-Yr Projection of Mean Reversion for S&P 500 Price, GAAP Earnings and Dividends

One approach to seeking fair value, or simply what is most likely, over the intermediate-term to long-term is the assumption of mean reversion.  That approach basically assumes that long-term means act a bit like gravity or a magnetic attraction that pulls on prices, earnings or dividends to return to the mean growth level.

It may take a lot of patience for that approach, and if there is a permanent structural change in markets, the former mean may not continue to serve as gravitational or magnetic attraction.   On the other hand structural changes are few and far between, making mean reversion a pretty good bet more often that not.

So, lets discover the long-term means for S&P 500 prices, GAAP earnings and dividends; and then apply those means to make reasonable 5-year projections of those dimensions into the future.

Let’s use the S&P monthly data available from Professor Shiller at Yale (you can download the data for yourself to do various studies on your own).  That data begins in the 1800’s using precursors to the S&P 500, such as the Dow Industrials and other data, but we will confine our study to actual S&P 500 data from its inception in 1957 (more than 57 years of data).


The long-term compound growth rate of the price of the S&P 500 is 6.88%.  Continuing that growth curve out 5 years comes up with a future  price projection in the vicinity of 2700.  That is around 36% to 37% above the current level.



The long-term compound growth rate of the GAAP earnings for the S&P 500 is 5.90%.  Continuing that growth out 5 years comes up with a future earnings level projection in the vicinity of $115 per index share. That is around 14% above the most recent 2-month trailing earnings levels.



The long-term compound growth rate of dividends for the S&P 500 is 5.37%.  Continuing that growth out 5 years comes up with a future earnings level projection in the vicinity of $45 per index share.  That is around 20% above the most recent 12-month  trailing dividend level.



Now let’s look at the growth trends since the effective beginning of the internet era in 1995, and also use the long-term growth factors from that beginning to see where that ends up.

In 1995, HTML had been around for a few years, and eventually it included ways to add images, not just text. That image capability was added and expanded to browser capability when Mark Andreesen released the Netscape browser in November 1994.  The internet rapidly transformed from a Geeky academic and military technology, to a personal and commercial technology.  It has been disruptive and transformative in business and for stocks ever since.

Keep in mind that if January of 1995 was particularly high or low relative to fair value, the projection would be comparably high or low.  For reference and to make your own judgment about that, here are the multiples at as of December 30, 1994:

  • GAAP earnings yield 5.95% (P/E ratio 18.81)
  • Dividend yield 2.90%
  • Moody’s long-term Baa corporate bonds 9.1%
  • 30-yr Treasuries 7.89%
  • 10-yr Treasuries 7.81%
  • 5-year Treasuries 7.81%
  • 2-yr Treasures 7.69%
  • 3-mo Treasuries 5.6%

The yield curve was essentially flat, while today it is steep. and the Fed was not such an intrusive player.


These charts have two 5-year projections.  The black dotted plot is a simple exponential trendline on the data from the beginning of 1995 projected out 5 years.  The red solid line is a projection from the beginning of 1995 at the long-term growth rate of the price of the S&P 500.

The exponential trend line is heavily influenced by the extremes of the dot-com bubble, while the long-term growth rate projection is impacted by the degree of “normalcy” of the starting point multiple.

The exponential projection suggests there is little if any price growth in the 5-year future of the S&P 500.  The long-term growth rate applied to the internet era suggests a price level 5 years out in the vicinity of 2900, or about 31% to 32% cumulative price change over the next 5 years.

At a minimum, the explosive growth of the past few years driven by both earnings recovery from the deep 2008 crisis, and by P/E multiple expansion.  Those forces are behind us, not in front of us.


There are many optimists among the analyst community, and there are long-term optimists who expect a correction, but there is a much more limited number of outright long-term pessimists.

Two prominent pessimists are Bill Gross of PIMCO and Jeremy Grantham of GMO.

  • PIMCO, as you may know, is predicting stock price appreciation in the 3% to 5% range over the next several years.
  • GMO is another important voice expressing concerns. Their quarterly asset class 7-year forecast is for negative 1.7% rate of return for US large-cap stocks (essentially the S&P 500), negative 5.2% return for small-cap stocks (essentially the Russell 2000), and positive 2.8% for high quality stocks. They see positive 3.6% returns for emerging market stocks,

As to High Quality Stocks, we suspect for the most part GMO was thinking about large-cap stocks that have not participated fully in this momentum market, and which are particularly strong otherwise – such as with consistent but modest revenue and dividend growth, low leverage and strong Balance Sheets, and wide business moats.  See our post of high quality stocks.)


The exponential trendline from the beginning of 1995 suggests earnings are already too high by about $10 per index share ($100 versus a 5-year projection of about $90).  Let’s hope that is not a good forecast, because the absolute level of earning and the associated negative growth rate would probably crush the stock market.

We do have historically high profit margins, historically low corporate borrowing costs, stagnant wages, above historical earnings growth rates, above average P/E multiples, and stock market capitalization that is at an historically high ratio to US GDP.  Those are key fundamental risk factors that should be evaluated.

The long-tern earnings growth rate applied to the earnings at the beginning of 1995 projected out 5 years, suggests earnings of about $122 per index share, or about 21% above current levels.

Standard & Poor‘s forecasts 2014 GAAP earnings of $119.87 (basically the 5-yr projection level at the long-term growth rate); and $136.39 for 2015.  The 3-year period leading up to the end of 2015 would have a compound earnings growth rate of 16.39% (about 2.5 times the long-term earnings growth rate).  They may be right, but if they are way wrong, there will be major market problems.



The exponential 5-year projection for S&P 500 dividends goes to about $40 per index share (about a cumulative 7% growth).  The long-term dividend growth rate applied from the beginning of 1995 suggests a 5-year projected cumulative growth of about 22% to about $46 per index share.


On balance, we have a greater faith in the growth of dividends than in the growth of either earnings or index price.  We also share GMO’s prediction that high quality stocks will do better over the next few years than large-caps in general.  Additionally, our client basis is at or near the end of the accumulation stage of their financial lives, where assets cannot be replaced easily or at all from new wages or business profits.  That means asset preservation is key, further pushing us in the direction of dividend and high quality bias within the stocks allocation, along with some amount of allocation to partially market neutral assets, such as long/short funds or positions.


Comments are closed.