Gold IRA vs Stocks and Index Funds
In the monthly data we reviewed, U.S. stocks with dividends grew about 10.5% a year from 1972 to 2022, and gold's price grew about 7.6% a year, with larger swings. Which did better depends on the dates chosen. Both can lose money. This is history, not a forecast or advice.
Is gold or a stock fund the better place for retirement savings? Nobody can answer that for you in advance. What we can do is show what the numbers looked like in the past, how we calculated them, and which rules differ when gold sits in an IRA. This page is part of our guide to Gold IRA vs Other Retirement Investments. It is general education. It does not rank anything, predict prices or tell you what to buy.
How has gold compared with stocks historically?
It depends heavily on the start and end dates. Over the 51 calendar years from 1972 to 2022 in our calculation, stocks with dividends reinvested ended well ahead, but gold led in some long stretches and stocks led in others.
| Measure (1972 to 2022, 51 years) | S&P 500, with dividends | Gold, price only |
|---|---|---|
| Average yearly growth (compound) | about 10.5% | about 7.6% |
| Value of $1 at the end of 2022 (starting from end of 1971) | about $163 | about $42 |
| Years with a loss | 12 of 51 | 20 of 51 |
| Worst calendar year | -39.2% (2008) | -27.5% (2013) |
| Best calendar year | +38.4% (1995) | +118.8% (1979) |
| Typical size of yearly swings (standard deviation) | about 17% | about 26% |
Source: our calculations from the two datasets described below. Not adjusted for inflation. Figures are rounded and approximate.
In plain terms:
- Stocks grew faster over the full period, and gold's yearly results were more spread out.
- Gold gained more than stocks in 22 of the 51 years.
- In 8 of those years, stocks fell while gold rose (1973, 1974, 1977, 1987, 2001, 2002, 2008 and 2022). In 4 years both fell. In 16 years gold fell while stocks rose.
- The yearly returns of the two moved modestly in opposite directions (a correlation of about -0.2, where 0 means no link). That is a statement about this period only.
Does the answer change with the time period?
Yes. Here are the average yearly results, by period, from the same calculation.
| Period | S&P 500, with dividends (per year) | Gold, price only (per year) |
|---|---|---|
| 1972 to 1979 | about +5.3% | about +34.3% |
| 1980 to 1989 | about +17.3% | about -1.1% |
| 1990 to 1999 | about +18.0% | about -3.6% |
| 2000 to 2009 | about -0.7% | about +14.9% |
| 2010 to 2019 | about +13.3% | about +2.7% |
| 2020 to 2022 | about +8.9% | about +6.7% |
Source: our calculations. Each period is compounded from calendar-year returns.
The pattern is that leadership changed hands. A saver who started in 1980 and one who started in 2000 would have seen very different results. That is why a single "winner" claim should make you cautious.
We also checked more recent price-only changes, without dividends for stocks, through September 2026. These use monthly averages and so are rough.
| Period (price change only) | S&P 500 index (per year) | Gold (per year) |
|---|---|---|
| Dec 2000 to Sep 2026 | about +7.0% | about +11.4% |
| Dec 2010 to Sep 2026 | about +12.3% | about +7.5% |
Stock figures in this last table leave out dividends, so they understate stocks' total return. Over long periods dividends can add a lot. Do not compare this table with the first one.
How did we calculate these numbers?
We downloaded two open datasets on 2026-10-08.
- Stocks. The datasets/s-and-p-500 file on GitHub. Its README says it is a tidied copy of data from economist Robert Shiller, with monthly S&P 500 price, dividend and earnings data from 1871, extended from mid-2023 with index prices from the Federal Reserve Bank of St. Louis (FRED). For months from July 2023 on, dividends are blank (zero) in the file, so we calculated total return only through December 2022.
- Gold. The datasets/gold-prices file, which says data from 1960 on comes from the World Bank's Commodity Markets "Pink Sheet." This is the same file used on our gold price volatility page.
- Method. For each year we compared December's monthly average with the prior December's. For stocks, we built a chain-linked monthly total-return index. Each month's index value equals the prior month's value multiplied by (this month's price plus one-twelfth of the annual dividend) divided by the prior month's price. This approximates reinvesting dividends each month. We then compared December's index value with the prior December's. Gold has price change only. "Average yearly growth" is the compound rate.
Limits you should know about:
- We could not open the original Shiller, World Bank, NYU Stern or S&P Dow Jones Indices pages from our work environment, so we could not check these mirrors against the originals. Treat all figures as approximate.
- Shiller's stock prices are monthly averages of daily closes, and his monthly dividends are interpolated. Published index total returns, which use daily or year-end data, will differ somewhat from ours.
- Monthly averages smooth out sharp moves, so real peaks and troughs were larger.
- Results are before inflation, taxes, fund fees, and the costs of buying and selling gold.
- The S&P 500 is 500 large U.S. companies. It is not "stocks" in general, and one index fund will not match it exactly.
- Gold is measured as the spot price. Coins and bars cost more to buy than spot and sell for less (see below).
What is different about income?
A stock can pay a dividend. Investor.gov says dividend payments come when a company distributes some of its earnings to stockholders, and that investors may also hope for capital appreciation, which "occurs when a stock rises in price."
Our gold figures count price change only. Our comparison therefore treats gold's return as coming entirely from the price rising or falling. We did not find a government source stating that metal pays no income, so we are not making a general claim beyond our own calculation. If you hold metal through an IRA, ask the custodian in writing about any income, fees or charges that apply.
How do risk and loss compare?
Both can lose money. Investor.gov states that stock prices "move down as well as up" and that you can lose money you invest in stocks. It adds that large company stocks as a group have lost money, on average, in about one of every three years. We counted 12 losing years out of 51 for the S&P 500 with dividends, a different index and period, so the two figures are not directly comparable.
For metals, the Commodity Futures Trading Commission (CFTC) says in its 10 Things to Ask Before Buying Physical Gold, Silver, or Other Metals that "precious metals prices can fluctuate just like other investments."
What the data adds is the size and length of declines. In our monthly gold data, the January 1980 peak was not regained until 2006. See Gold IRA price volatility for the full table. Stocks have also had long drawdowns, for example the 2000 to 2009 period in our table above, when the S&P 500 with dividends averaged slightly negative returns.
What rules and costs are different inside an IRA?
This is where the two choices are least alike.
What the IRA may hold. Under the Internal Revenue Code, an IRA's purchase of a "collectible" is treated as a distribution. The IRS page on investments in collectibles in individually directed qualified plan accounts says the acquisition "is treated as an immediate distribution from such account in an amount equal to the cost to the plan." Metals can qualify under exceptions. The page excludes certain gold, silver or platinum coins described in 31 U.S.C. 5112, and gold, silver, platinum or palladium bullion "of a certain fineness," but only if a bank or approved non-bank trustee keeps physical possession. Details are in our collectibles rule guide. Stock and index funds do not raise this issue.
Costs. A dealer "will always sell metal above the spot price and buy it back below the spot price," according to the CFTC, and "each dealer sets its own spread." That gap is a cost that does not show up in our gold return above. Higher spreads and fees raise the price rise needed to break even, and the CFTC says so directly. A Gold IRA can also carry custodian and storage fees, covered in our fees guide. Fund expense ratios are the stock-side cost. We did not research current fee levels for either side on this page, so we make no cost comparison beyond these points.
Who holds the asset. A Gold IRA needs a custodian and storage arrangement. See how a Gold IRA works.
What mistakes do people make with this comparison?
- Picking the period that makes a story work. Both assets have had decades where they looked bad.
- Comparing gold's price change with a stock index's price change and forgetting dividends, or the reverse.
- Ignoring buying and selling costs on physical metal.
- Treating one past correlation as a promise that gold "protects" a portfolio in the next downturn.
- Mixing up gold itself with gold mining stocks or gold ETFs, which are different products. See Gold IRA vs gold ETF.
What related questions should you look at?
- Employer plans: Gold IRA vs 401(k) covers how the account types differ.
- Fixed income: Gold IRA vs Bonds looks at the other major asset class.
- Concentration: allocation and concentration risk explains why putting too much in one asset matters.
What should you ask a professional?
How much of any asset belongs in your plan depends on your age, income, other savings, taxes and goals. That is a question for a fee-only financial planner, ideally one with no ties to a metals dealer, and for a tax professional. Useful questions: What could I lose in a bad five-year stretch? When will I need this money? What are the total costs in writing? We do not make personal recommendations here.
Sources
- Investments in collectibles in individually directed qualified plan accounts, Internal Revenue Service. Accessed Invalid Date.
- Stocks, Investor.gov (U.S. Securities and Exchange Commission). Accessed Invalid Date.
- Customer Advisory: 10 Things to Ask Before Buying Physical Gold, Silver, or Other Metals, Commodity Futures Trading Commission. Accessed Invalid Date.
- S&P 500 index data since 1871 (data.csv and README), datasets/s-and-p-500; based on Robert Shiller's data, extended with FRED, DataHub / Open Knowledge Foundation (GitHub mirror of Robert Shiller data). Accessed Invalid Date.
- Monthly gold prices in USD since 1833 (monthly.csv and README), datasets/gold-prices; 1960 onward sourced from World Bank Commodity Markets (Pink Sheet), DataHub / Open Knowledge Foundation (GitHub mirror of World Bank data). Accessed Invalid Date.
This guide is general education, not personalized financial, tax or legal advice. See our financial disclaimer, editorial policy and advertising disclosure.