What Is the S&P 500? How It Works, Year by Year

Michael Whitman · Trading since 2017, full-time since 2020 · Published 2026-09-25

When someone says “the market was up today”, they usually mean the S&P 500. It is the number on the news, the benchmark most funds are measured against, and the thing an index fund investor actually owns.

How it works

S&P Dow Jones Indices describes it as “the best single gauge of large-cap U.S. equities.” Its own page says the index includes 500 leading companies and covers approximately 80% of available market capitalization (read on spglobal.com, 25 Sep 2026).

Weighting, membership and dividends

It is weighted by market value. A company worth ten times as much as another moves the index ten times as much. So the largest few companies carry a large share of each day’s move, and the smallest members barely register.

A committee, not a formula alone, decides who is in. S&P Dow Jones Indices publishes the rules and announces changes ahead of time. On 4 Sep 2026, for example, it announced that Bloom Energy, Illumina and Everpure were set to join. Companies that shrink or are bought out leave, and the replacement is announced the same way.

It is a price index. The level quoted on the news leaves out dividends. The figures on this page do too, so they understate what an investor who reinvested dividends actually received.

You cannot buy the index itself. You buy something that tracks it: an index fund, an exchange-traded fund such as SPY (see how trading SPY is different), or futures and options on it.

The index launched on 4 Mar 1957. S&P Dow Jones Indices notes that everything before that date is a hypothetical back-test using the method in force at launch. The daily history used below starts in January 1950, so its first seven years are that reconstruction.

A worked example

Take the S&P 500’s close on the last trading day of 2024 and of 2025. The calendar-year change is simply the second divided by the first, minus one. For 2025 that came to +16.4%, from the daily closes file described below.

Now the same arithmetic on 2008. The index closed that year 38.5% below where it closed 2007. On a hypothetical $10,000 held in a fund that tracked the price exactly, that is about $6,150 at the end of the year, before dividends and before the fund’s own costs.

And on 1954, the best year in the file: +45.0%, so the same hypothetical $10,000 would have ended the year near $14,500. Both years are the same index, 54 years apart, which is the point of the next section: the average year is a poor guide to any single one.

The original data

The data: every daily close of the S&P 500 from 3 Jan 1950 to 18 Sep 2026, 19,299 trading days, downloaded from Yahoo Finance (^GSPC), price only. The last close in the file is 7,650.50; the first is 16.66.

Counted by calendar year, 1950 to 2025, the index finished 56 of 76 years higher: 73.7%. It finished 20 lower. The median year was +12.4%. Three years fell by more than 20%, and five rose by 30% or more. 1950 is measured from its first close, because the file starts there; every other year runs from the previous year’s last close. The year-by-year table is published as a CSV of every calendar year.

Time spent below a previous record

The same file shows how much of the time the index spends below a previous high. On 92.0% of those 19,299 trading days, the close was under an earlier record. The deepest fall ran from 9 Oct 2007 to 9 Mar 2009, −56.8%, from a close of 1,565.15 to 676.53, and the index did not close above the old record until 28 Mar 2013, about 5.5 years after the peak. The longest wait of all began at the record of 11 Jan 1973 and ended on 17 Jul 1980: 1,897 trading days, or about 7.5 years.

Falls of 5% or more happened 74 times, a little under one a year on average (0.96).

The question: The S&P 500 went up in 56 of 76 years. Does that mean next year is 74% likely to be up?

No. It means that is how the last 76 years went. The count describes history, including decades that will not repeat, and it is on price alone. What it does support is a smaller, more useful claim: a down year is normal, not an exception, and a plan that cannot survive one is not a plan for this index.

When it fails

The first failure is treating “the market” as diversified by default. Because the index is weighted by value, a handful of the largest companies can drive most of a year’s move. Owning the S&P 500 means owning that concentration.

The second is reading the price index as your return. The quoted level leaves out dividends and the fund’s costs, so it is not the number an investor actually received.

A third is using the average year to plan a single year. The median was +12.4%, but individual years ran from −38.5% to +45.0%.

A fourth is assuming a recovery takes months. After 2007 it took about 5.5 years to regain the record; after 1973, about 7.5.

And a fifth is forgetting the early history is reconstructed. Everything before 4 Mar 1957 is a back-test by the index provider, not a live index.

Related

Index funds explains how most people own the S&P 500 and what that costs. Trading SPY covers the exchange-traded fund that follows it and how trading it differs from trading a single stock. And the Nasdaq 100 is the other index people most often mean by “the market”, with a very different make-up.

This page is educational, not financial advice. Test every idea on your own charts before risking money.