Bitcoin's price rose far faster than the S&P 500 over the decade to October 28, 2026, and it fell far harder along the way. This study measures both from one closing price a day, over four windows that end on that date, and compares returns, volatility, the deepest falls, how closely the two moved and how much return each earned for the risk it carried. The S&P 500 series is the index level without dividends, so it understates what an index fund holder earned; a section below puts a number on the gap. For background on each market, start with the crypto hub or the stocks hub.
| Measure | Value |
|---|---|
| Window | Oct 31, 2016 to Oct 28, 2026 |
| Bitcoin annualized return over the window | 61.0% |
| S&P 500 annualized return over the window (price only) | 13.8% |
| Bitcoin annualized volatility | 68.3% |
| S&P 500 annualized volatility | 18.1% |
| Bitcoin maximum drawdown | -83.4% |
| S&P 500 maximum drawdown | -33.9% |
| Correlation of daily returns | 0.28 |
Key findings
- Returns. From October 31, 2016 to October 28, 2026, bitcoin returned 61.0% a year and the S&P 500 price index 13.8%, a total of 11,557% against 264.2% before dividends.
- Start dates. Bitcoin led over 3 years, 33.8% a year against 23.4%, but trailed over 5 years, 6.0% against 11.0%, and lost 28.0% in the past year while the index gained 12.4%.
- Volatility. Bitcoin's annualized volatility was 3.0 to 3.8 times the index's, from 44.1% against 12.7% over 1 year to 68.3% against 18.1% over 10 years.
- Return per unit of risk. Bitcoin scored higher only over 10 years, 0.86 against 0.62. The index scored higher over 5 years, 3 years and the past year.
- Correlation. Daily returns were loosely linked, 0.28 over 10 years, but more closely since 2020, reaching 0.56 in 2022 and 0.46 in 2026 through October 28.
- Deepest falls. Bitcoin's maximum drawdown was 83.4% over 10 years against 33.9% for the index, and 48.2% against 9.1% over the past year.
How was the study done?
The study takes one closing price a day for each asset and measures four windows, 1, 3, 5 and 10 years long, that end on October 28, 2026. Bitcoin prices are Bitstamp's BTC/USD daily closes, taken at midnight UTC. S&P 500 levels are the ones S&P Dow Jones Indices publishes through FRED as series SP500, set at the 4 p.m. New York close, and the cash rate is the 3-month Treasury constant maturity yield, FRED series DGS3MO.
Each window starts on the last S&P 500 trading day on or before the same date 1, 3, 5 or 10 years earlier. FRED publishes only the latest 10 years of S&P 500 data, so the 10-year window starts on its first available date, October 31, 2016, three days short of a full decade.
| Measure | How it is calculated |
|---|---|
| Total return | Last close in the window divided by the first close, minus 1 |
| Annualized return | Total return converted to a compound yearly rate, counting a year as 365.25 days |
| Annualized volatility | Standard deviation of daily log returns, multiplied by the square root of 365 for bitcoin and of 252 for the S&P 500 |
| Maximum drawdown | Largest fall from a running peak close to a later close inside the window |
| Excess return per unit of volatility | Annualized return minus the window's average 3-month Treasury yield, divided by annualized volatility |
| Correlation | Pearson correlation of daily log returns between consecutive S&P 500 trading days, with bitcoin measured between the same dates |
| Best and worst day | Largest one-day percentage rise and fall from one close to the next |
| Share of up days | Percentage of daily returns above zero |
| Calendar-year return | Change from the previous year's last close to the year's last close |
| Month-end drawdown | Distance below the highest close since October 31, 2016, read on the last S&P 500 trading day of each month |
Returns are price changes only, in US dollars, before fees and taxes.
Why bitcoin's volatility uses 365 days and the index's uses 252
Annualizing turns the spread of daily returns into a yearly figure by multiplying it by the square root of the number of daily returns a year contains. Bitcoin trades around the clock, weekends and holidays included, so a year holds about 365 daily closes. The S&P 500 is calculated only on days US exchanges open: the NYSE's 2026 calendar has 261 weekdays and 10 weekday holidays, leaving 251 sessions, and 252 is the usual round figure. Using each market's own count puts both figures on the same one-year footing, and an index return on a Monday already carries whatever happened over the weekend.
Annualized volatility = standard deviation of daily log returns × √(number of daily returns in a year)
Worked example: One daily swing, annualized two ways
Take an asset whose daily log returns have a standard deviation of 3.0%, a round number chosen for illustration.
- Scaled by 365 days, as this study scales bitcoin: 3.0% × √365 = 3.0% × 19.10 = 57.3% a year.
- Scaled by 252 days: 3.0% × √252 = 3.0% × 15.87 = 47.6% a year.
The 252-day figure is about 17% lower, because √(252 ÷ 365) = 0.83. Using it for bitcoin would drop 113 days of real trading from the count. Going the other way, scaling the S&P 500 by 365 would add days on which the index does not trade and overstate its volatility by about 20%.
How did bitcoin's returns compare with the S&P 500's?
Bitcoin beat the S&P 500's price index over 3 and 10 years and trailed it over 1 and 5 years, with much larger swings in every window.
| Period | Asset | Total return | Annualized return | Annualized volatility | Maximum drawdown | Excess return per unit of volatility |
|---|---|---|---|---|---|---|
| 1 year | Bitcoin | -28.0% | -28.1% | 44.1% | -48.2% | -0.72 |
| 1 year | S&P 500 | 12.4% | 12.4% | 12.7% | -9.1% | 0.67 |
| 3 years | Bitcoin | 139.6% | 33.8% | 46.7% | -53.1% | 0.63 |
| 3 years | S&P 500 | 88.1% | 23.4% | 14.9% | -18.9% | 1.27 |
| 5 years | Bitcoin | 34.1% | 6.0% | 51.0% | -76.7% | 0.04 |
| 5 years | S&P 500 | 68.5% | 11.0% | 17.0% | -25.4% | 0.41 |
| 10 years | Bitcoin | 11557.0% | 61.0% | 68.3% | -83.4% | 0.86 |
| 10 years | S&P 500 | 264.2% | 13.8% | 18.1% | -33.9% | 0.62 |
Over the full decade the gap is enormous: a total return of 11,557% for bitcoin against 264.2% for the index. Annualizing tames those totals, to 61.0% and 13.8% a year, because each year's gain compounds on the last. Shorter windows give a less one-sided result, and the start date explains most of it.
The 5-year window opens on October 28, 2021, 11 days before bitcoin's November 8, 2021 record close on Bitstamp, so it carries nearly all of the 76.7% fall that followed, which the bitcoin drawdowns study dates from record to record. Bitcoin ends that window with 6.0% a year to the index's 11.0%. Start two years later, on October 27, 2023, and bitcoin leads again, 33.8% a year against 23.4%. The 1-year window began three weeks after bitcoin's October 6, 2025 record close, and bitcoin lost 28.0% over it while the index gained 12.4%.
Worked example: What $10,000 became in each window, before costs
Multiply the starting amount by one plus the total return in the table.
- 10 years: bitcoin, $10,000 × 116.57 = $1,165,700. S&P 500 price index, $10,000 × 3.642 = $36,420.
- 5 years: bitcoin, $10,000 × 1.341 = $13,410. S&P 500, $10,000 × 1.685 = $16,850.
- 3 years: bitcoin, $10,000 × 2.396 = $23,960. S&P 500, $10,000 × 1.881 = $18,810.
- 1 year: bitcoin, $10,000 × 0.720 = $7,200. S&P 500, $10,000 × 1.124 = $11,240.
These amounts leave out trading fees, fund costs and taxes, and the S&P 500 lines leave out dividends, which the next section estimates.
How much would dividends add to the S&P 500's return?
About 1.6 to 1.9 percentage points a year in the periods to September 30, 2026, based on S&P Dow Jones Indices and State Street figures from outside this study.
FRED describes its SP500 series as a price index, not a total return index, and says it contains no dividends. Every S&P 500 figure in the tables above therefore understates what someone holding an index fund and reinvesting the payouts would have earned. Bitcoin pays no dividends or interest, so nothing is missing from its side.
| Annualized return to September 30, 2026 | 3 years | 5 years | 10 years |
|---|---|---|---|
| S&P 500, price only | 21.29% | 12.18% | 13.44% |
| S&P 500, dividends reinvested | 22.89% | 13.79% | 15.33% |
| SPY ETF at net asset value, after costs | 22.73% | 13.65% | 15.18% |
| Dividends reinvested minus price only | 1.60 points | 1.61 points | 1.89 points |
These windows end a month before the study's, so they don't match its figures exactly; they size the dividend gap rather than replace the study's numbers. The index's dividend yield was 1.11% on September 30, 2026, according to S&P Dow Jones Indices. The SPY row shows what one large index fund returned after its 0.0945% expense ratio and other costs: 0.14 to 0.16 points a year less than the index with dividends. The SPY page has the fund's latest price.
Adding dividends would not change the 10-year ranking, since 15.33% a year is still far below bitcoin's 61.0%, though the two cover slightly different dates. It would narrow bitcoin's lead over 3 years, widen the index's lead over 5 years and the past year and, with volatility little changed, raise the index's return per unit of risk in every window.
How much more volatile is bitcoin than the S&P 500?
Between three and four times as volatile: bitcoin's annualized volatility was 68.3% over 10 years against 18.1% for the S&P 500, a ratio of 3.8.
| Period | Start | End | Correlation of daily returns | Bitcoin volatility ÷ S&P 500 volatility | Average 3-month Treasury yield |
|---|---|---|---|---|---|
| 1 year | Oct 28, 2025 | Oct 28, 2026 | 0.47 | 3.50 | 3.8% |
| 3 years | Oct 27, 2023 | Oct 28, 2026 | 0.37 | 3.10 | 4.5% |
| 5 years | Oct 28, 2021 | Oct 28, 2026 | 0.42 | 3.00 | 4.0% |
| 10 years | Oct 31, 2016 | Oct 28, 2026 | 0.28 | 3.80 | 2.5% |
The ratio column divides bitcoin's annualized volatility by the index's. It was lowest over 5 years, at 3.0, and highest over the full decade, at 3.8. Both assets were calmer in the shorter, more recent windows: bitcoin's volatility was 44.1% over the past year and the index's 12.7%, so the multiple stayed in that range even as bitcoin's own swings narrowed. Volatility describes the typical spread of returns, not the worst case; the best and worst days and the drawdown chart further down show the extremes.
The last column is the average 3-month Treasury yield in each window, the cash return that the risk-adjusted score subtracts. It averaged 2.54% over the decade, which included the near-zero rates that followed the Federal Reserve's March 15, 2020 cut, and 4.5% over the past 3 years. What is the risk-free rate explains why Treasury bills serve as the benchmark, and today's T-bill rates shows where they stand now.
Do bitcoin and the S&P 500 rise and fall together?
Only loosely: their daily returns had a correlation of 0.28 over the 10 years, although the link has been closer since 2020 and peaked at 0.56 in 2022.
| Year | Bitcoin | S&P 500 | Correlation of daily returns |
|---|---|---|---|
| 2017 | 1336.4% | 19.4% | 0.05 |
| 2018 | -73.4% | -6.2% | 0.07 |
| 2019 | 94.1% | 28.9% | -0.11 |
| 2020 | 304.5% | 16.3% | 0.45 |
| 2021 | 59.4% | 26.9% | 0.26 |
| 2022 | -64.2% | -19.4% | 0.56 |
| 2023 | 155.7% | 24.2% | 0.16 |
| 2024 | 121.0% | 23.3% | 0.36 |
| 2025 | -6.3% | 16.4% | 0.42 |
| 2026 (to Oct 28, 2026) | -7.1% | 13.1% | 0.46 |
Correlation runs from −1, for two series that always move in opposite directions, to 1, for two that always move together in proportion; 0 means no linear link. From 2017 to 2019 the yearly figures sat close to zero, between −0.11 and 0.07. Since 2020 they have ranged from 0.16 in 2023 to 0.56 in 2022, the year both fell, bitcoin by 64.2% and the index by 19.4%.
The calendar-year columns show how often the two parted ways. Bitcoin returned more than the index in 6 of the 9 full years, 2017, 2019, 2020, 2021, 2023 and 2024, and less in 2018, 2022 and 2025. In 2025 and in 2026 through October 28, bitcoin fell, by 6.3% and 7.1%, while the index rose 16.4% and 13.1%. Even at 0.56, a correlation describes a tendency across many days rather than a rule for any one of them, and it says nothing about the size of the moves, which were several times larger for bitcoin.
Two timing details shape these numbers. Bitstamp's daily close falls at midnight UTC, which is 8 p.m. in New York during daylight saving time and 7 p.m. in winter, 3 to 4 hours after the stock market's 4 p.m. close, so news in those hours lands on different days for the two assets. And because each bitcoin return is measured between the same two dates as the index's, bitcoin's weekend and holiday moves are folded into the next trading day.
Which asset earned more per unit of risk?
It depends on the window: bitcoin scored higher over 10 years, 0.86 against 0.62, while the S&P 500 scored higher over 1, 3 and 5 years.
| Window | Bitcoin | S&P 500 | Higher score |
|---|---|---|---|
| 1 year | −0.72 | 0.67 | S&P 500 |
| 3 years | 0.63 | 1.27 | S&P 500 |
| 5 years | 0.04 | 0.41 | S&P 500 |
| 10 years | 0.86 | 0.62 | Bitcoin |
The score takes each asset's annualized return, subtracts what 3-month Treasury bills paid on average over the same window and divides by the asset's annualized volatility. It answers a narrower question than raw return: how much extra return each point of volatility bought.
Excess return per unit of volatility = (annualized return − average 3-month Treasury yield) ÷ annualized volatility
Worked example: The 10-year scores, step by step
The average 3-month Treasury yield from October 31, 2016 to October 28, 2026 was 2.54%.
- Bitcoin: (61.0% − 2.54%) ÷ 68.3% = 58.46 ÷ 68.3 = 0.86
- S&P 500: (13.8% − 2.54%) ÷ 18.1% = 11.26 ÷ 18.1 = 0.62
Over the past year the same arithmetic gives bitcoin (−28.1% − 3.83%) ÷ 44.1% = −0.72. The score turns negative whenever an asset returns less than Treasury bills.
William Sharpe introduced this kind of measure in 1966 as the reward-to-variability ratio, and his 1994 restatement stresses that the result depends on the period over which it is measured, which the four windows here bear out. This version is simpler than the textbook Sharpe ratio, which averages each period's return above the cash rate. A compound return falls further below that simple average the more an asset swings, so the approach used here is tougher on bitcoin than the textbook one would be. The score also judges each asset alone. It ignores how an asset's moves offset or add to the rest of a portfolio, a limit Sharpe also flagged, and it counts a sharp rise as risk just as much as a sharp fall.
What were the best and worst days for each?
Bitcoin's worst day in the 10 years was a 39.0% fall on March 12, 2020; the S&P 500's worst, a 12.0% fall, came four days later, on March 16.
| Period | Asset | Best day | Date | Worst day | Date | Share of up days |
|---|---|---|---|---|---|---|
| 1 year | Bitcoin | 12.3% | Feb 6, 2026 | -14.0% | Feb 5, 2026 | 47.7% |
| 1 year | S&P 500 | 2.9% | Mar 31, 2026 | -2.6% | Jun 5, 2026 | 52.6% |
| 3 years | Bitcoin | 12.3% | Feb 6, 2026 | -14.0% | Feb 5, 2026 | 50.9% |
| 3 years | S&P 500 | 9.5% | Apr 9, 2025 | -6.0% | Apr 4, 2025 | 56.8% |
| 5 years | Bitcoin | 14.5% | Feb 28, 2022 | -15.6% | Jun 13, 2022 | 49.2% |
| 5 years | S&P 500 | 9.5% | Apr 9, 2025 | -6.0% | Apr 4, 2025 | 53.0% |
| 10 years | Bitcoin | 26.9% | Jul 20, 2017 | -39.0% | Mar 12, 2020 | 51.9% |
| 10 years | S&P 500 | 9.5% | Apr 9, 2025 | -12.0% | Mar 16, 2020 | 54.8% |
Both came around the Federal Reserve's Sunday decision of March 15, 2020 to cut its target range for the federal funds rate to 0 to 1/4 percent, a move its statement tied to the coronavirus outbreak. The index's best day of the decade, a 9.5% rise on April 9, 2025, came three trading days after its worst day of the past 5 years, a 6.0% fall on April 4, 2025. April 9 was the day a White House order suspended for 90 days, for every country except China, the higher country-specific tariff rates set on April 2. Bitcoin's best and worst days of the past 3 years came back to back: a 14.0% fall on February 5, 2026 and a 12.3% rise the next day.
The share of up days was close to half for both assets: 51.9% of days for bitcoin and 54.8% for the index over 10 years, and 47.7% for bitcoin over the past year. The gap in long-run returns came from the size of the moves, not from how often prices rose. Each bitcoin return covers one calendar day, while each index return runs from one trading close to the next, so a Monday's move includes the weekend.
How far below their peaks did bitcoin and the S&P 500 fall?
Bitcoin's deepest fall in the 10 years was 83.4%, against 33.9% for the S&P 500; at month ends, the deepest readings were 82.2% and 25.2%.
- Bitcoin
- S&P 500
Each point shows how far an asset stood below the highest close it had reached since October 31, 2016, read on the last S&P 500 trading day of each month. Bitcoin's deepest month-end reading was 82.2% on January 31, 2019. The index's was 25.2% on September 30, 2022, when bitcoin was 71.2% below its peak, and at the end of March 2020 the two stood 66.5% and 23.7% below.
Month-end sampling misses the worst days, so the chart looks shallower than the maximum drawdowns in the main table, which use every daily close. It also starts the clock in October 2016. Bitcoin was still below its December 2013 record on Bitstamp until February 23, 2017, so its first months look closer to a peak than they were.
The two lines rarely moved in step. On June 30, 2026, bitcoin was 53.1% below its peak while the index was 1.5% below; by October 28, 2026, the figures were 34.9% and 1.0%. Bitcoin's slow recoveries, from 2018 to 2020 and from 2022 to 2024, show as wide valleys, and the history of bitcoin's drawdowns measures how long each one lasted.
What are the limits of this comparison?
The main limits are the price-only S&P 500 series, the short shared history, the sensitivity to start and end dates and the use of one exchange's prices.
- No dividends. The index figures leave out dividends, which added 1.60 to 1.89 points a year in the periods to September 30, 2026.
- Start and end dates. A window that opens near a bitcoin peak, as the 5-year one does, or closes in a slump, as all four do, can flip the ranking. On October 28, 2026, bitcoin was 34.9% below its peak and the index 1.0%.
- Ten years at most. FRED carries only the latest 10 years of daily S&P 500 data, so earlier bitcoin history, including the 2013 to 2015 crash, is outside this comparison.
- One exchange, one clock. Bitstamp's midnight UTC close differs a little from other exchanges' prices and closing times, which can move a record or a low by a day and a percentage by a few tenths of a point.
- No costs or taxes. Spreads, trading fees, fund expenses and taxes are not deducted, and buying bitcoin through an app, an exchange or a fund adds costs of its own.
- Volatility is not all of risk. Standard deviation treats rises and falls alike and says nothing about exchange failures, lost keys or rule changes; types of investment risk covers the rest.
- Dollars, before inflation. All figures are nominal US dollar returns. Inflation shrinks them in real terms, as real returns and inflation explains, and an investor who converts from another currency also gains or loses on the exchange rate.
What does this history mean for someone weighing the two?
It describes what each asset did over these dates, not what comes next: past performance never predicts future results, and the ranking changed between windows.
The numbers are most useful for sizing what holding each asset has felt like. Over the decade, bitcoin's higher return came with an 83.4% fall along the way; over the past year, it came with a loss while the index gained. A fall of that size is easier to judge in money than in percentages.
Worked example: The deepest 10-year falls in dollars
Apply each asset's maximum drawdown to $10,000 held at the peak.
- Bitcoin, an 83.4% fall: $10,000 × (1 − 0.834) = $1,660.
- S&P 500 price index, a 33.9% fall: $10,000 × (1 − 0.339) = $6,610.
How much of a portfolio to put in an asset that can fall this far is a personal decision that depends on goals, time horizon and other holdings, and the guide to how much crypto to hold in a portfolio sets out the questions to ask. The DCA calculator backtests a fixed weekly or monthly purchase of bitcoin or the S&P 500 over the past year, and the bitcoin price page shows the latest price. For how the two differ beyond price, including ownership and regulation, see crypto vs stocks.
How to use and cite this data
The data is free to download and reuse under the CC BY 4.0 license, provided the source is credited with a link to this page.
- Download the CSV file, which holds the main results table with one row per window and asset, or use the download box near the top of the page. The other tables can be quoted with the same credit.
- Quote each figure with its window and end date, for example: bitcoin returned 61.0% a year from October 31, 2016 to October 28, 2026.
- Say that the S&P 500 figures are price-only and exclude dividends.
- Suggested credit: Source: Sonic Wallet analysis of Bitstamp BTC/USD daily closes and FRED data (S&P 500, series SP500; 3-month Treasury yield, series DGS3MO), data through October 28, 2026.
Each table shows the date its data runs to. How studies are built, checked and corrected is set out in the editorial policy.