Average Stock Market Return

Measured over many decades, the S&P 500 (500 large US companies, dividends reinvested) has produced an average annual total return of roughly 10% in nominal terms since 1926. The exact number shifts a little depending on the start date and whether you use an arithmetic or a compound (geometric) average, but ~10% is the figure almost every source lands near. It reflects a specific index, over a specific country, over a long and mostly favorable stretch of history.

Two things are easy to miss. First, that figure includes dividends: price appreciation alone has been more like 6 to 7% nominally, and dividends made up the rest. Second, it is a very long-run average. Pull any shorter slice and the number can look wildly different, which is the whole point of the sections below. For the practical mechanics of owning that broad market, see how to invest in the S&P 500.

The ~10% headline is a nominal return, meaning before inflation. What actually grows your purchasing power is the real return, which subtracts inflation. US inflation has averaged roughly 3% per year over the long run, so the market's real return has been closer to 6 to 7% per year. That gap sounds small, but compounded over decades it is the difference between how big your balance looks and how much it can actually buy.

For long-horizon goals like retirement, real return is the honest yardstick. A portfolio that grows 10% in a year when inflation runs 8% has barely moved in real terms, even though the statement looks great. When you see a stock-market average quoted, it is worth asking whether it is nominal or real, because the two answers differ by several percentage points a year.

The single most important thing to understand about the ~10% average is that it is the center of a very wide distribution, not a typical result. Historically, S&P 500 annual returns have landed in a narrow band around the average (say, between +8% and +12%) only a handful of times across the whole record. Far more common are years up 20 to 30% or down 10 to 40%. The market spends most of its time nowhere near its own average.

A famous illustration: 2008 was down about 37%, and 2013 was up about 32%. Neither is close to 10%, yet both are part of the history that produces the 10% average. Averaging those extremes gives a smooth number, but no single year feels smooth. Expecting a steady 10% is how investors get surprised, and often scared out of the market, by perfectly normal volatility. Whether individual skill can consistently beat this index is a separate question, covered in can you beat the S&P 500.

The scatter that dominates a single year shrinks as your holding period lengthens. The table below shows approximate ranges of annualized S&P 500 total return across rolling windows of different lengths. These figures are approximate and drawn from long-run history; they are illustrative, not a forecast.

Read down the table and the pattern is clear: one-year outcomes are almost a coin toss between a big gain and a big loss, while twenty- and thirty-year windows cluster near the long-run average and, historically, twenty-year windows have never been negative. Time does not raise the average, it narrows the range of outcomes around it. That is the statistical case for a long horizon.

Averages assume the order of returns does not matter, and for a buy-and-hold investor who never adds or withdraws, that is roughly true: only the compounded result matters. But once you are contributing or, especially, withdrawing money, the order of returns matters a great deal. This is sequence-of-returns risk.

Two retirees can experience the exact same average return over thirty years and end up in very different places. The one who hits a steep bear market in the first few years of retirement, while withdrawing, sells shares into weakness and may never fully recover, because those early withdrawals lock in losses the average never sees. The one who gets good years early is far safer. This is why the same ~10% average can feel benign to a young saver and dangerous to someone drawing down a portfolio: the average is identical, the lived outcome is not.

None of this is advice, but the history points to a few durable, widely cited ideas. The wide single-year scatter and the narrow long-horizon range together explain why time in the market has historically mattered more than timing it. The nominal-versus-real gap explains why costs and inflation deserve attention: fees and taxes come straight out of that ~10%, and a couple of percentage points compounded over decades is enormous. Sequence risk explains why risk tolerance often shifts as you approach a withdrawal phase.

What you do with those observations depends entirely on your goals, timeline, and temperament, and this page cannot decide it for you. Some people conclude that a broad, low-cost index fund held for a very long time fits them; others want more control or a different mix. If you want to sanity-check the funds that track this market, see the best S&P 500 ETFs, and to check how your own holdings have done against it, see how to compare your portfolio to the S&P 500.

Walnut does not forecast the market or tell you what to buy. It is an AI investing assistant you use on the broker you already own: connect any major US broker (read-only by default, your login stays with your broker), then chat through Claude, ChatGPT, or a built-in assistant to understand what you hold. For a topic like this, Walnut can frame each holding and your whole portfolio against a benchmark such as the S&P 500 over a window you choose, and explain concepts like real return and volatility in plain language. You can build baskets around a thesis and track them against targets, and you approve every trade at your own broker. Walnut does not tell you what to buy, and historical averages are never predictions. A common reference fund for this market is VOO.