A myth about long run stock market returns exists. A powerful, but maybe dangerous, myth.
That myth goes like this: Over the long run, because much of the up and down choppiness evens out, you can count on a pretty good, average-ish outcome if you’re investing for decades.
Restated another way, the myth says that if you or I can stay in the market for decades, we can probably get that nice average long run stock market return everybody seems to know about. That 6% or 7% figure. Or whatever.
Unfortunately, this myth about the reliability of long run stock market returns misleads. Long run stock market returns absolutely are not reliable. And you and I need to plan accordingly.
My Crude Line Chart of Long Run Stock Returns
Let me show you a simple line chart that visually depicts this reality. (If the image doesn’t render well, click it to display a larger version.)
The crude little line chart shown above depicts the worst case, average, and best case returns someone saving $10,000 annually enjoyed if they saved over 5, 10, 15, 20, 25 or 30 years using a typical 75% stocks and 25% bonds portfolio.
The average return, for the record, quickly converges to around 6%. The orange line shows this value.
But here’s the thing to note: Though the range from worst to best narrows over time, results still vary widely even when you look at long run stock market returns.
At year 30, for example, the best case scenario (shown as a blue line) delivers a 10% real rate of return on your savings. And the worst case scenario (shown as a gray line) delivers no real rate of return–so zero percent–on your savings.
You see the reality, right? Sure. Maybe long run stock market returns average out to a healthy number. A typical retirement savings portfolio that emphasizes stocks maybe averages out to 6%.
But wide, wide variability in results occurs.
So here’s the point of of the post: Yes, the myth whispers you can count on average-like long run stock market returns if you just show enough patience. But history says something different. History says you better not.
How I Calculated Long Run Stock Market Returns
Let me quickly explain how I got the numbers plotted in the line chart. (If you don’t care about the math or numbers are not your friend, skip ahead to the next section.)
The data comes out of the cFIREsim online calculator, which uses a data set that starts in 1872.
The asset allocation equals 75 percent stocks and 25 percent bonds. So not 100% stocks but rather a more typical stocks and bonds blend.
The scenarios I modeled say some investor saves $10,000 annually for a given number of years: 5 years, 10 years, 15 years and so on at five year increments right up to 30 years.
I used the cFIREsim default .18% expense ratio and default annual rebalancing.
Finally, I calculated the average annual return using Microsoft Excel’s RATE function using as the function inputs that $10,000 annual savings amount, the number of years of saving, and then the cFIREsim-calculated future value for the average, best case and worst case scenarios.
And now back to the message from the chart…
Why My Tail is Fatter than Other Charts
The line chart above (and others like it) show you bear more risk than you realize.
Now I can guess you’re slightly disoriented by this statement and by the way it contrasts with what you thought. So let me try to explain the apparent differences between my cFIREsim-derived results and the information you’ve seen elsewhere.
Four factors probably explain most or all of the apparent disconnect.
First, and very frankly, you possibly just missed the reality discussed here. Regularly, authors don’t highlight the variability. Instead, they focus on the fact that the range of possible returns (that “tail”) narrows over time. This is true as the chart above shows. But the range doesn’t compress to a single value like 6%.
Note: If you’ve read and still own books written by John Bogle or Burton Malkiel, look for their charts that shown the range of returns on stocks over time. You’ll now spot the 3% or 4% or more difference between the worst case and best case returns.
Second, I’m calculating the future value of an annuity, $10,000 a year for up to 30 years. I’m not calculating the future value of an initial one-time deposit. This change from a one-time deposit to an annuity fattens the tail by nearly 70%. You do want, however, to calculate returns using an annuity since you save for retirement by making regular contributions and not a one-time deposit.
Third, some of the data people use to make the calculations and create the charts you’ve seen before don’t go back very far. To name just two examples, both the PortfolioCharts.com website and the PortfolioVisualizer’s back test asset allocation tool start their data in 1970s. That start date means they miss some really bad and some really good patches in investing. This change seems to fatten the tail by another 25% or so.
Note: Please don’t construe my statements as criticisms of the Portfolio Charts or Portfolio Visualizer websites or tools. I love both of them for what they do. But their data sets aren’t big enough to highlight what I’m talking about here.
A fourth explanation for a disconnect–at least visually: Some of line charts that show the range of long run stock market returns scrub the data of the extreme results. (Portfolio Visualizer does this, for example, showing not the worst and best cases but the 25th percentile and 75th percentile cases.) Scrubbing the data makes the line charts more usable and readable. (I’d do the same thing if I was them.) But in this particular comparison, that scrubbing hides the variability in returns.
If the stuff discussed here is all new to you, you’ve got enough new stuff in your head for now. No reason to drag this on further.
Let me, though, close with this point: If this wide variability in returns is new to you, you probably have misunderstood the financial risks you’re bearing in your retirement savings program. And given that, you probably want to do some thinking about that risk and how you want to deal with it.
I’ve got some other blog posts that may help you do this: