I've heard the number 7% thrown around a bunch of times, but I am struggling to find reputable data to back it up.

All I found was a quote from Warren Buffet that I couldn't find an original source for and data assembled by individuals on their personal websites. Is there something from an organization such as Market Watch making a statement about this?

I'm hoping to specifically find what the 10 or 20 year return is for investing in the S&P 500 index fund or another diversified portfolio option.

  • 2
    Have you tried looking at index funds that have existed for a long period of time to see what their average returns are? Mar 22, 2017 at 10:35
  • 3
    Just remember that past performance doesn't predict future results. Mar 22, 2017 at 14:22
  • In the UK, the marketing material issued some finance companies (e.g. life insurance) is heavily regulated. The good news is that it prevents the companies cherry-picking the data they use, but the bad news is that every company uses the same arbitrary (and not particularly accurate) assumptions. So those particular "original sources" are worthless. US regulations may be different, of course.
    – alephzero
    Mar 22, 2017 at 14:43
  • Robert Shiller's spreadsheet on the US stock market runs from 1871 to the present day and includes monthly prices, dividends, earnings, CPI (cost of living), and long-term risk-free interest rates: econ.yale.edu/~shiller/data.htm Mar 1, 2020 at 15:36

4 Answers 4


The oddly named MoneyChimp offers a great look at the numbers for the S&P.

enter image description here

You can enter a date as early as 1871 and any ending date you wish. The last 100 years did, in fact, show a CAGR of 10.06%. Take off a bit, .05 - .10 for expenses if you want to see the return for an ETF or mutual fund.


One way would be to take 1st March 2017 S&P 500 index value at close and take 1st March 2007 S&P 500 index value at close and calculate CAGR. Once such calculator is here. However, this will not be the exactly same as the returns of index funds for this period, since index funds can be managed passively or actively. For specific funds you probably have to look at the CAGR for the period for that fund.

Hope this helps

  • 7
    Make sure you account for dividends in the source where you get your numbers, otherwise your calculation will be 'leaking' returns that an investor would have realized if they held shares over that period. Mar 22, 2017 at 13:32

I have recently written a blog post analyzing the historical annualized returns with the S&P 500 all the way from 1928 - 2018. You can check it out here - https://finsers.com/index.php/2019/02/28/long-term-returns-of-the-sp-500/. My analysis shows that the annualized returns one has enjoyed has varied significantly based on the time in history they were investing in - even when annualized over long periods of time such as 30 years! The extent of variation blew my mind away, really.

The median annualized returns were 7%. However, if one were to plan their finances and financial goals assuming a 7% return from the market, they would miss their goal 50% of the times. Many a time they would miss it by a big margin!

  • I'm not sure the median is a relevant statistic here. Yes over one year you would gain less than 7% half of the time (and thus the S&P is a risky investment for a 1-year horizon), but the times you don't make 7% are made up for by the years you make more than 7%. The geometric average is more relevant when analyzing compounding returns over long periods. Also I don't think your return figures include dividends - when looking at total return 30-year CAGR is typically closer to 10%.
    – D Stanley
    May 10, 2019 at 0:58
  • I think the median is the appropriate statistic here. When I say median I am talking about the median of annualized returns. There were many 30-year periods between 1928 - 2018. In my blog post find the CAGR (pls note a slightly modified variant of CAGR in the blog post) for each of those 30-year periods, and 7% happens to be the median of that. And my analysis includes dividends.
    – akhil28288
    May 10, 2019 at 2:49
  • OK I see what you mean now. I thought you were looking at the median of annual returns.
    – D Stanley
    May 10, 2019 at 13:26
  • Was the URL incorrectly typed or is the site broken? It doesn't appear on Wayback Machine either. Mar 1, 2020 at 5:06
  • The link is broken.
    – Flux
    Oct 16, 2020 at 3:10

First, your question is what a mathematician would call "ill-posed." Second, it doesn't matter unless it is for historical curiosity since you should be concerned with future returns rather than historical ones.

It is ill-posed because investments include land, where returns are locally determined rather than nationally or internationally, and private investments alongside public ones. Second, returns would be gross of taxes. If people are intelligent, then they choose their investments, in part, based on their tax situation and so many investments would be irrelevant to you as their after-tax return for you would be poor, but good for someone else.

On the flip side, returns are a function of interest rates and price to income. Interest rates are near enough to historic lows that non-bond investments can pay ridiculously low returns and be competitive. That was not the case across history. For example, you could purchase 18% per year bonds in the '70s and '80s. As a result, equity securities were discounted way way way below that level.

The starting point matters. Had you invested in the Dow on January 1, 1929, reinvesting dividends, and a friend invested in 90 day bills on the same date, it would have taken until 1964 for you to break even with your friend. Within three years you would have shot past him or her by a huge amount, but you would have been worse off in the interim.

Based on the current price to earnings ratio, for US stocks, as well as the dividend yield, and reasonable inflation expectations, my highly educated guess, based on decades of statistical research on the topic by me, is that forward returns before taxes will be six percent, or in that neighborhood.

The personal finance pages do not allow for mathematical notation, so there isn't a simple way to justify it.

Bonds rates could very well be representative of long term bond rates, but my guess is that rates will rise due to serious missteps being made by the Congress and the current president. If you are looking at real property investments, look at the purchase price to rental price ratio. Pittsburgh is among the best values and San Francisco among the worst. Price to income is what matters, and a lot of value gets sunk in land in the US.

The question should be what return should I make from a particular price to income stream. The answer to that is to use standard time value of money discounting methods.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .