What is the long term S&P growth rate?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
ge1
Posts: 816
Joined: Sat Apr 28, 2012 8:15 pm

What is the long term S&P growth rate?

Post by ge1 »

In order to get a sense if the market is "overvalued" or "cheap" at the current level, I like to discount the current value of the S&P 500 by the normal growth rate and compare the computed value with the actual value.

Example: Using a 6% growth rate, 1820 in 2016 is equivalent to 1360 in 2011 (1360 x 1.06^5 = 1820), 716 in 2000 or 567 in 1996. In January 1996 the S&P was 614, so I get a rough sense that the market is a bit cheaper today than it was in 1996, but more expensive than in 1991 (424 calculated vs 325 actual). I can go back as far as like and if my calculated value is lower than the actual value, that would indicate that the market is relatively cheap, whereas the market would seem expensive if the calculated value exceeds the actual value in most years.

The calculation obviously depends heavily what I use as the "normal" growth rate of the S&P, i.e. nominal growth rate excluding dividends. Using the Shiller data set I can see that the growth rate ranges between 5.8% (going back to 1920) to 7.8% starting in 1980. What do people think is a reasonable number, 6.5%?

thanks
moshe
Posts: 565
Joined: Thu Dec 12, 2013 12:18 pm
Location: Boston, MA

Re: What is the long term S&P growth rate?

Post by moshe »

It is unknowable unless past performance does now affect future performance?

Any other response requires a crystal ball and mine is in the shop for its 1000 year checkup.

We say "Stay the course!"

~Moshe
My money has no emotions. ~Moshe | | I'm the world's greatest expert on my own opinion. ~Bruce Williams
tibbitts
Posts: 23588
Joined: Tue Feb 27, 2007 5:50 pm

Re: What is the long term S&P growth rate?

Post by tibbitts »

ge1 wrote:In order to get a sense if the market is "overvalued" or "cheap" at the current level, I like to discount the current value of the S&P 500 by the normal growth rate and compare the computed value with the actual value.

Example: Using a 6% growth rate, 1820 in 2016 is equivalent to 1360 in 2011 (1360 x 1.06^5 = 1820), 716 in 2000 or 567 in 1996. In January 1996 the S&P was 614, so I get a rough sense that the market is a bit cheaper today than it was in 1996, but more expensive than in 1991 (424 calculated vs 325 actual). I can go back as far as like and if my calculated value is lower than the actual value, that would indicate that the market is relatively cheap, whereas the market would seem expensive if the calculated value exceeds the actual value in most years.

The calculation obviously depends heavily what I use as the "normal" growth rate of the S&P, i.e. nominal growth rate excluding dividends. Using the Shiller data set I can see that the growth rate ranges between 5.8% (going back to 1920) to 7.8% starting in 1980. What do people think is a reasonable number, 6.5%?

thanks
Although I think it's a silly exercise, I'll guess 3.
long_gamma
Posts: 420
Joined: Mon Sep 14, 2015 4:13 pm

Re: What is the long term S&P growth rate?

Post by long_gamma »

"Everyone has a plan 'till they get punched in the mouth." --Mike Tyson
rbaldini
Posts: 1444
Joined: Mon Mar 23, 2015 3:20 pm

Re: What is the long term S&P growth rate?

Post by rbaldini »

ge1 wrote:In order to get a sense if the market is "overvalued" or "cheap" at the current level, I like to discount the current value of the S&P 500 by the normal growth rate and compare the computed value with the actual value.

Example: Using a 6% growth rate, 1820 in 2016 is equivalent to 1360 in 2011 (1360 x 1.06^5 = 1820), 716 in 2000 or 567 in 1996. In January 1996 the S&P was 614, so I get a rough sense that the market is a bit cheaper today than it was in 1996, but more expensive than in 1991 (424 calculated vs 325 actual). I can go back as far as like and if my calculated value is lower than the actual value, that would indicate that the market is relatively cheap, whereas the market would seem expensive if the calculated value exceeds the actual value in most years.

The calculation obviously depends heavily what I use as the "normal" growth rate of the S&P, i.e. nominal growth rate excluding dividends. Using the Shiller data set I can see that the growth rate ranges between 5.8% (going back to 1920) to 7.8% starting in 1980. What do people think is a reasonable number, 6.5%?

thanks
If you take Jan 1 2016's price (1890.28), divide by Jan 1, 1871's price (4.44), and raise it to the power of 1/(2016-1871), you get 1.043. So that implies price growth of 4.3% per year. Go wild. (Found numbers here: http://www.multpl.com/s-p-500-historica ... le/by-year)

I don't think it's a great exercise because you don't know what year to calculate back to for comparison. Also, the number I just gave includes the current price, so it's not really independent of recent market behavior.

Edit: Another issue: suppose you believed that the market is an independent random walk. Suppose also that the price is "low" right now - relative to whatever past price you compare it to, corrected for growth rate. Then according to an independent random walk, this shouldn't matter: there is no tendency for such a model to "catch up" with the expected return. It simply has an expected growth rate with noise. From the data I've seen, this is a pretty good model: the last 1, 2, 3, 4, 5... years of return have virtually no ability to predict next year's return. So, point being, I'm skeptical that this exercise will produce much fruit. You should still do it though.
Topic Author
ge1
Posts: 816
Joined: Sat Apr 28, 2012 8:15 pm

Re: What is the long term S&P growth rate?

Post by ge1 »

long_gamma wrote:This might be interest for you.

http://www.philosophicaleconomics.com/2015/03/treps/
Excellent article, this is exactly what I was looking for (no actually more than that!). Thanks again.
User avatar
jimb_fromATL
Posts: 2278
Joined: Sun Nov 10, 2013 11:00 am
Location: Atlanta area & Piedmont Triad NC and Interstate 85 in between.

Re: What is the long term S&P growth rate?

Post by jimb_fromATL »

If you take Jan 1 2016's price (1890.28), divide by Jan 1, 1871's price (4.44), and raise it to the power of 1/(2016-1871), you get 1.043. So that implies price growth of 4.3% per year. Go wild. (Found numbers here: http://www.multpl.com/s-p-500-historica ... le/by-year)
ge1 wrote: The calculation obviously depends heavily what I use as the "normal" growth rate of the S&P, i.e. nominal growth rate excluding dividends.
That's only for the value of the stock, not including dividends paid out or reinvested.
According to http://www.multpl.com/s-p-500-dividend-yield/ the dividend has been that much -- and sometimes a lot more.

Even allowing for the likelihood that historical performance is a better predictor of what might happen in the future than perhaps a crystal ball or casing runes, it doesn't seem that it would make sense to go by the growth rate alone and excluding dividends.

Most folks investing in virtually any kind of account intended for growth for the future do reinvest the dividends. And in retirement plans like 401(k)s and IRAs you defer the taxes on the dividends and reinvest that money too. (And never pay tax on the dividends in Roths.)

Over the life of the S&P 500, the dividends have often been as much or more as the growth in value. Even if the investment is for income instead of growth, the dividends are probably even more important.

For gee-whiz historical data about the S&P 500 MoneyChimp has a calculator using the Shiller data by year for the S&P with dividends reinvested. They also go into a lot of detail about the much misunderstood difference between mathematical averaging and CAGR (Compound Annual Growth Rate) for investments.

They unintentionally provide the easiest source I've found for the yearly shiller data for the S&P 500 with dividends reinvested -- for your own spreadsheet. In their menu for choosing the years, you can “select all” and copy all the years from their menu and then paste it into a spreadsheet. You can convert the percentage columns into a multiplication factor in a new column by dividing by 100 then adding 1, then use the factors with the geomean function in Excel to compute the compound annual growth rate -- the average APY -- for the yearly data – for a lump sum.

For yearly dollar-cost averaging of contributions, you can create another column and use the multiplication factor (or a repetitive calculation of the percentage for the previous year) add contributions each month and calculate the year’s gain. Then use the beginning and end result with the RATE() function. This can be used to do rolling averages, too. Just remember that just like Vanguard, Morningstar, Yahoo, and other sources, the yearly data and summaries are only accurate for a lump sum at the beginning and only if you invest on exactly the same day they use.

Poster #cruncher has some instructions and a sample spreadsheet table for working with yearly data from the sources above in this thread ... and I posted a few comments about manipulating the data there too.

I've also created a sample GEOMEAN & CAGR investment comparison worksheet to show some samples of how to do it:
https://docs.google.com/spreadsheets/d/ ... 1753109071



jimb
Topic Author
ge1
Posts: 816
Joined: Sat Apr 28, 2012 8:15 pm

Re: What is the long term S&P growth rate?

Post by ge1 »

Thanks jimb, lots of good information and useful tools with the links provided. Will check this out.

Completely understand that looking at the price growth only tells you little about the total return - all I am trying to do here is comparing the S&P today with S&P values in the past (which I find useful and others disagree with, no need to discuss that.)
lack_ey
Posts: 6701
Joined: Wed Nov 19, 2014 10:55 pm

Re: What is the long term S&P growth rate?

Post by lack_ey »

IMHO if you want to see if the growth rate is above or below trend and by how much, you should probably correct for both dividends and inflation, as was done in the link long_gamma posted.

Implicitly this still assumes that there is some kind of trend and natural rate of underlying growth, which may or may not be a good assumption.
saurabh
Posts: 319
Joined: Thu Aug 20, 2015 11:56 am

Re: What is the long term S&P growth rate?

Post by saurabh »

In a US economy with 3% real GDP nominal growth S&P 500 earnings have grown about 6% fairly consistently when measured over 10-year or longer time horizons. In a economy that is now growing 2% in a good scenario, and some possibility of profit compression I would say that 4% is a realistic conservative estimate, although it could well be higher.
User avatar
midareff
Posts: 7711
Joined: Mon Nov 29, 2010 9:43 am
Location: Biscayne Bay, South Florida

Re: What is the long term S&P growth rate?

Post by midareff »

From January 1960 to end of December 2013 the S&P returned 7% real. You will get a different number for different time periods but that is a reasonable close 53 year period and may or may not produce any useful information. The market can do gyrations lots longer than you can remain solvent.
Post Reply