Of the four largest US-listed ETFs, with total assets under management totaling over $1.1 trillion, three track the S&P 500 index, and the fourth returns practically the same as S&P 500, as charted below. It is fair to say that the future expected rate of return on an S&P 500 index fund is one of the most important assumptions an investor needs to make when deciding whether to invest in a US large cap index fund versus an alternative, so in this article I will show the math and data behind how I estimate that future rate of return, which for 2021-2031, I expect to average in the low single digits. As the tracker fund for the S&P 500 index, I will be using the Vanguard S&P 500 ETF (NYSEARCA:VOO), rather than the SPDR S&P 500 Trust (SPY) or iShares Core S&P 500 ETF (IVV) for the following reasons:
Although SPY and IVV each have more assets under management than VOO, I consider VOO to be almost interchangeable with the slightly larger Vanguard Total Stock Market ETF (VTI), which has over 80% overlap with VOO's holdings, and as charted below, these two funds' returns have tracked each other within 0.02%/year over the 10 year period ending June 30, 2021. In other words, VOO and VTI combined have more money allocated to the S&P 500 portfolio than either SPY or IVV. Another special detail of VOO is that it is a share class of Vanguard's larger S&P 500 fund, with total assets under management over 3x that of VOO ($754 billion as of the June 2021 fact sheet), again highlighting the massive amounts of capital banking on the S&P 500's future returns.
Past performance is supposed to be no indicator of future results, but the best investors do study history to get an idea of how different environments have driven different returns for stocks. The simplest useful long-term series on US stock returns and underlying fundamentals is available for download on Robert Shiller's website, and includes not only stock returns, but also dividends, earnings, inflation, and bond yields going back to 1871.
Some long-term stock charts, especially those starting in the 1930s or later, can almost look like a nearly straight march up and to the right, but ones going back to 1871 show how much lower the rate of price appreciation in stocks was from 1871-1921 (investors expected most of their returns from dividends back then), and how significant the bubble and crash of the 1920s was. The chart below also overlays plots of earnings and dividends, multiplied by the somewhat arbitrary multiples of 20x earnings and 40x dividends, to show how stock prices do indeed track earnings growth and dividend growth over time. Earnings and dividends simply didn't grow as much from 1871-1951 as they have since 1951, partly because dividend payout rates have fallen from around 65% to 40% over that period, and since then more earnings have been reinvested than paid out.
The above chart becomes more dramatic and timely when we zoom in on the past 50 years since 1971, and plot on a linear scale rather than a logarithmic scale. What this chart shows is how consistently US equities averaged P/E multiples below 20, and dividend yields above 2.5%, in the late 20th century, and how sub-2.5% dividend yields seem to be new to the 21st century. In other words, although long-term earnings growth has continued to drive steady long-term dividend growth, the rise in stock prices overshot these fundamentals in the 1990s, and again over the past 10 years, as we will see in the following table.
The following table calculated from Shiller's data provides a closer decade-by-decade view of periods when stock prices rose faster than dividends and earnings, versus periods when the reverse happened. Although I rarely see tables like this one published, I find these to be one of the most useful ways of breaking down rates of return for any stock portfolio. As obvious as it may sound to anyone who knows the math, these are the two most important observations to take away from the below table, and on which the "Bogle model" is based:
As a few examples:
With those charts and numbers providing some context on what might be reasonable rates of earnings and dividend growth to expect, and how changes in dividend yield and P/E ratios can affect our total rates of return, let's plug these into two modified versions of the "Bogle Model" to see what these give us for expected rates of return for VOO out to 2031.
The simplest explanation of Jack Bogle's model for what a "reasonable expectation" of future rates of return from stocks comes from Bogle himself in this 2017 interview with Morningstar, shortly before he passed away in January 2019, but was also explained in more detail in this article by the unrelated Jack Vogel. The "Bogle Model", if we want to call it that, simply puts together that "reasonable expectation" of future rates of return by adding together three components:
For dividend paying stocks and ETFs, I also use a modified version of this model entirely based on dividends, where I estimate dividend growth instead of earnings growth, and change in dividend yield instead of change in P/E ratio. Dividend growth can be more stable than earnings growth, but that is necessarily offset by dividend yields being more volatile than P/E ratios.
In the short term, that third component of "speculative return" dominates the total rate of return, at least for investors who watch the prices at which their stocks are trading day to day, while longer term, as we have seen, earnings growth is the dominant factor. Although 10 years might seem long term for some investors, we have seen that changes in P/E ratios and yields still have a significant impact on total rates of return over a 10 year horizon, but beyond that I have a hard time keeping most investors' attention. As you will see in the tables below, whether the P/E ratio in 10 years is 20 or 30 makes more than 4%/year difference in your total rate of return over 10 years, but over 50 years it makes less than 1%/year difference.
Of the above three inputs, the only ones we know for sure when we start are the starting dividend yield (#1), and the starting P/E ratio (part of #3 in Bogle's original model, based on earnings). Then we can use the below table to look up what the expected rate of return over the next 10 years would be based on our assumptions of earnings growth and what the P/E ratio will be in 2031. I have somewhat deliberately centered this chart around a 2031 P/E of 22-24, and an earnings growth rate of 4-5%, which would put the nominal rate of return on VOO between 3.7-5.6%. More optimistic estimates would be up and to the right, while more pessimistic ones are down and to the left. Part of my expectation of lower earnings growth over the next 10 years comes from my expectations of higher inflation, along with growth headwinds likely to be faced by the trillion-dollar companies at the top of the index (which would be a whole other article).
The pure dividend version of this table is below, again centered on a modest level of 4-5% annual dividend growth and a dividend yield remaining around 1.3%, which I see as optimistic, but could possibly lead to another decade of 4.5-7.1% total returns. The inflation concern mentioned above means I believe it's more likely than not that we will see higher interest rates, which implies higher dividend yields, and even a rise to a 1.8% dividend yield means dividend growth would have to break 7% for VOO total returns to average 5%. Overall, I would still focus my estimates on the center-top of this chart, which we could widen to an estimated return range of 1.5-4.5%.
Although I left the above models open-ended for you to plug in your own assumptions, I have clearly indicated that the most likely assumptions imply VOO delivers annual rates of return of less than 5% over the next 10 years, and your assumptions would have to be more aggressive than mine to get higher returns. On a portfolio allocation level, I regularly run tables like the ones above across different markets and names, and hopefully this example shows why I don't buy S&P 500 index funds, but instead focus on GARP strategies, selected high quality stocks, and high-yielding emerging and frontier markets like Nigeria instead.
Neither standard nor poor, my analysis takes a big step back from all the noise and focuses on the few key visuals and numbers that identify winning long-term investments. Get inspired with your free trial to Big Picture Growth & Income.
This article was written by
Tariq Dennison TEP runs GFM Asset Management, a registered investment adviser for cross-border families and retirement plans between the US and Asia. His marketplace service "Long Run Income" focuses on dividend growth and dividend alternative ideas for generating investment income over multi-decade time horizons. Tariq is the author of the book "Invest Outside the Box: Understanding Different Asset Classes and Strategies", spends most of his time in Hong Kong and around Asia, and teaches two classes at the Masters in Finance program at ESSEC Business School in Singapore.
Disclosure: I/we have a beneficial short position in the shares of SPY either through stock ownership, options, or other derivatives. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.