The Authority Post on Stock Market Returns

Stock MarketsI’ve had enough. It seems every day I turn around and somebody else is spouting off the same sentence, “Stock markets return __% per year over the last __ years, on average.” I wouldn’t have such a problem with this, if not for the huge variance in the number that fills in that blank. Suze Orman says 6%, Dave Ramsey says 12%, most financial analysts model for 8%, many others say 10%. It’s crazy. The way I see it, this is all widely available public data, so why isn’t there just one right answer?

First off, you might be asking yourself, “Seven percent, or eleven percent–what does it really matter?”

Well, it matters a great deal, actually. A difference of a few percent in annual returns has not only monstrous effects over long periods of time (you’ll see in a chart at the end), but even over relatively shorter time periods. Let’s say I’m trying to save $50,000 over the next 20 years to pay for my kid’s college education. If I think I’ll get 6% returns, then I need to sock away $114 a month. However, if I think I’ll get 11% returns, I can reach that $50k with only saving $65 a month. So, it’s important to understand what sort of returns you can reasonably expect.

Alas, I decided to get to the bottom of this myself using a huge pool of market returns data compiled by Robert Shiller. Who’s right, who’s wrong? And what did I find? The actual average return on stocks is…

It depends.

Oh man. Even for this seemingly simple exercise there’s no clear answer. And, here’s why.

Let’s break the sentence up into parts:

“Stock Markets” : What market are you talking about? You could use the total market index, using the Wilshire 5000, which gets you the cap-weighted return of all U.S. listed stocks. Or, you could use the Dow Jones Industrial Average, a price-weighted selection of 30 blue chips. Or, you could use the S&P 500, the 500 largest stocks with Headquarters in the United States. The general consensus is to use the S&P 500. A larger sample (like the Wilshire) isn’t really necessary, as the smaller components are essentially irrelevant. Take JDS Uniphase (JDSU), one of the smallest companies in the S&P 500—if JDSU doubled today, it would only move the S&P 500 by 0.34 points (the S&P 500 closed today at 2,068.59). As such, I believe the S&P500 is a large enough sample to use as a proxy for overall “Stock Markets.”

“Return” : What kind of return? Price return is commonly used, which is just “The S&P is at 2,068 today. It was at 1,500 in 2007. Therefore, the S&P has gone up 37% in 8 years.” That is technically correct, but greatly flawed because it neglects dividends, which juice returns each year. And even then, if dividends are reinvested back into the market each year, those reinvested dividends further bolster annual returns. Price returns are pretty irrelevant because every investor who holds a basket of stocks in “the market” gets dividends for holding said stocks. Investors who are withdrawing their dividends (like retirees) are well suited to use a “With Dividends Return.” Investors who are still accumulating assets and reinvesting dividends (working folks) are best suited to use a “With Dividends Reinvested” return. I present all three below.

Also, are we talking about nominal or real (inflation adjusted) returns? I present both cases below, but nominal returns is the only one I see as relevant. I sympathize with the reasoning behind using real returns: pricing power decreases over time, and if markets are going up simply because prices are going up, the returns are artificial. However, I think this is faulty and is a shortcoming of real returns. Inflation impacts each individual differently. Consumers are dynamic creatures, they adjust their basket of purchased goods based on the prices of those goods. Even with today’s multi-year low gas prices, consumers aren’t staying put at their old consumption levels–they’re going out and buying more trucks and SUVs. If I happen to be an average consumer that’s buying a basket of goods each year aligned with the BLS basket, then using the real return is appropriate. However, there is no such thing as the “average consumer.” It makes more sense to use the nominal return and allow each individual to adjust accordingly.

“over the last __ years” : If I look at markets since 2009, my annualized returns will be extremely good (above 20% per year). If I look at the last 15 years, they’ll be terrible, because I’m looking at a period that includes two awful bear markets but ignores the bull market that drove the overvaluation that caused them (or at least caused one of them). It’s easy to cherry pick a start date to support whatever “return” you want to present, so going more years back tends to be a better estimation. I prefer to look at the post-WWII period (last 70 years). This encompasses several recessions, oil crises, booms and busts. Over this period, the equity markets have been more widely used as an investment and financing tool for companies. Markets are more mature and information is abundant. Thus, when I want to make a blanket statement on historical returns, I’ll go back 70 years to 1945. However, I present the last 144 years below.

“on average” : What kind of average? Geometric? Arithmetic? Compound Annual Growth Rate (CAGR)? Anyone worth their salt should be using CAGR—which is what I used, below. This is how it works in real life.

Price Return

Price Returns

Nominal Returns since WWII: 7.42% CAGR

Real Returns since WWII: 3.53% CAGR

As I said before, price return is largely irrelevant and should only be used by somebody trying to slant the truth to fit their viewpoint. The price return was what was commonly touted when referencing the “Lost Decade” of 2000-2012, because price levels were generally the same at both ends of that time period. However, in reality, nobody actually receives just the price level. If you’re holding stocks, you’re collecting dividends—which are as real to an investor as the price level. Ignoring dividends (i.e. using price level returns) is ignoring reality.

With-Dividends Return

With Dividends

Nominal Returns since WWII: 8.02% CAGR

Real Returns since WWII: 4.09% CAGR

The inclusion of dividend payments in return calculations improves accuracy and enhances returns. These returns would be appropriate for somebody who is investing in stocks and withdrawing the dividends.

With Dividends Reinvested Return

Dividends Reinvested Return

Nominal Returns since WWII: 11.17% CAGR

Real Returns since WWII: 7.14% CAGR

These returns are what investors who are still investing and adding new money in the market should expect. When investors allow dividends to be reinvested, those dividends compound, which multiplies on the CAGR rather than just adding to it. This analysis, indirectly, shows the value of reinvesting dividends as opposed to withdrawing them.

Let’s get back to the rub. You might be asking, “what’s the big difference between using price return, dividends-paid return, and dividends-reinvested return?”

The difference, my friend, is enormous. Using our three nominal returns exhibited since WWII ended, let’s look at the growth of $100 from then until today (2015).

Growth of $100

Pretty impressive, right? So, now you understand that they’re not kidding when they say “90% of market returns are due to dividends.”

In sum, let it be forever known, here are the percentages to use when talking about market returns:

Summary Table

I’m still investing in the stock market, and I reinvest all my dividends into more shares of companies. As such, I can fairly model an 11.17% annual return for the future.

Thanks for reading!



  1. Great article. The main problem is that even with your 70 year “post-WWII” time frame, it’s not at all clear whether the next 30, 50, or 70 years can be assumed to be the same in the future. During the post WWII period, the U.S. population more than doubled, significant economic/industry transformation happened, and it grew to be the largest economy in the world. Maybe 70 years from now the U.S. population will be 700 million and the economy will have grown at the same pace. But that outcome is far from certain and nobody can be expected to predict what will happen with any degree of accuracy. That leaves a huge amount of variation in possible future stock market returns, even if you are confident the U.S. S&P 500 grew at an annual rate of 11% during that 70 year window (with dividends reinvested). Who’s to say the next 70 years that return won’t be 6%, or 8%, or 14%. There are a lot of factors that could cause 2015-2085 to look different in the U.S. than 1945-2015 did.

    • Hi dh,
      There’s no perfect way to predict the future, so your point is valid. There are a lot of things that could result in a paradigm shift in the next 70 years that is either much higher or much lower than the returns of the past 70 years. However, this is simply the best information available. The consensus seems to be that we’re in for perpetually lower returns going forward due to: demographic shifts, slowing global economy, population contraction in many developed markets.

      However, there are a lot of reasons to think the next 70 years will be amazing
      – Much longer lifespans–so more years of health and productivity
      – The “race to zero” in technological hardware
      – Middle class emergence in the world’s biggest populations (China, Africa, India)
      – Virtual Reality, Self-Driving Cars, Personalized Medicine, 3D Printing, Internet of Things are all potentially enormous industries that could shape the world–some won’t mean anything, but some will be in history books
      – So much more. Just look at the last five years. Spotify. Waze, Uber. iPad. Ubiquitous wi-fi. Khan Academy. Coursera. So many things are increasing productivity and expanding knowledge–I actually thing the next 70 years will be insanely better (for life and for markets) than the last 70.

  2. Hi Eric – that’s a very solid analysis of the market returns, thanks for putting this together. I also like to refer to the Vanguard analysis that goes all the way back to 1926 and shows how it changes with a bonds/stock allocation. In any case they find 10.2% for 100% stock, so your numbers make sense.
    I would however be careful to say that you can expect the same kind of returns for the future. That may be true for the future over the same period but I don’t know how many people have a 70+ years of investment horizon.
    Just imagine if you were doing this analysis in 2000, it would be quite deceptive to expect an 11% cagr from then on. I’d be cautious expecting such returns for the future.
    Great analysis anyway, this puts all this information in context.
    Out of curiosity, would the overall cagr return be much different if the end date was in the dip of 2009?
    TheMoneyMine recently posted…Comment on Should you Buy Mortgage Points? by NickMy Profile

Leave a Reply

Your email address will not be published.

CommentLuv badge