Myth

Does the golden cross mean buy?

VerdictIt dependsChecked against the data on 20 November 2026

Not on its own. In the 12 months after SPY's 15 golden crosses since 1993 the median gain was +17.5%, a little above the +14.8% from buying on any day, but trading the 50-day and 200-day rule grew money more slowly than holding, which is why the verdict is It depends.

The myth, and why people believe it

A moving average is the average closing price over the last so many trading days: the 50-day covers roughly the last ten weeks, the 200-day a little over nine months. A golden cross is the day the 50-day average climbs above the 200-day; a death cross is the reverse. Financial news treats both as events, and they're easy to believe in because they look decisive on a chart and often appear just after a big rebound. The catch is in that word "after": both averages look backwards, so by the time they cross, much of the move has already happened.

The test

How we tested this

DataSPY daily closes from EODHD: plain closing prices for the moving averages; closes adjusted for splits and dividends (total return) for every return.
Period10 April 2007 to 20 November 2026, the longest period every holding has prices for
BasketsSPY 60 / BND 40: BND 40%, SPY 60%, rebalanced yearly; SPY: SPY 100%, held throughout (benchmark)
StudySPY 29 January 1993 to 20 November 2026. Simple 50- and 200-day moving averages of plain closing prices; a golden cross is the first close with the fast average above the slow one, a death cross the reverse. Returns after each cross (3-month, 6-month, 12-month) use dividend-adjusted closes and are compared with the same windows from every trading day. The rule as a strategy holds the fund only while the fast average is above the slow one, otherwise cash at 0%, no costs or taxes.
How we testedA golden cross is the first close where the 50-day average of closing prices is above the 200-day after a close at or below it; a death cross is the reverse. Returns are measured 3, 6 and 12 months (63, 126 and 252 trading days) after each cross and compared with the same returns from every trading day. Strategy: own SPY while the 50-day average was above the 200-day at the previous close, cash otherwise, against buy and hold.
Not includedInterest on cash, trading costs and taxes, all of which would make the strategy look worse. Moving-average pairs other than 50 and 200 days.
Verdict rule, set before lookingBusted if the median 12-month return after a golden cross minus the any-day median is at most +1.0 points and the yearly return of holding only while the fast average is above the slow one, minus buy and hold is below zero; Proven if the median 12-month return after a golden cross minus the any-day median is at least +2.0 points and the share of golden crosses followed by a 12-month gain minus the any-day share is at least +5.0 points and the yearly return of holding only while the fast average is above the slow one, minus buy and hold is above zero; It depends otherwise.

We took SPY, the S&P 500 in one fund, from 29 January 1993 to 20 November 2026, and found every golden and death cross of its 50-day and 200-day averages. For each cross we measured what SPY returned over the next 3, 6 and 12 months, dividends included, and compared that with the same returns from every trading day. Then we ran the rule as a strategy: own SPY while the 50-day average is above the 200-day, hold cash otherwise. The verdict rule was set before we looked at the results.

The result

After a golden cross, SPY did a little better than after any day, with a wide range from one cross to the next:

After a golden crossAny-day median
12-month return of SPY after each golden cross, with the median from any day. Golden crosses: median +17.5%, positive 93.3%; any day: median +14.8%, positive 81.4%-10%0%+10%+20%+30%+40%1994199920042009201220162020202514 September 1994: +27.0%8 December 1998: +20.2%11 November 1999: +2.1%15 May 2003: +17.5%5 November 2004: +6.2%11 September 2006: +15.5%23 June 2009: +24.6%22 October 2010: +6.8%31 January 2012: +17.7%17 December 2015: +12.9%25 April 2016: +16.7%1 April 2019: −8.0%9 July 2020: +40.6%2 February 2023: +20.0%1 July 2025: +21.9%any day: median +14.8%
  • Golden crosses (15): median 12-month return +17.5%, positive 93.3% of the time
  • Any day: median +14.8%, positive 81.4%
CrossType3 months6 months12 months
14 September 1994Golden−2.7%+7.2%+27.0%
8 December 1998Golden+10.7%+11.2%+20.2%
11 November 1999Golden+0.4%+3.6%+2.1%
15 May 2003Golden+4.8%+12.6%+17.5%
5 November 2004Golden+3.2%+1.6%+6.2%
11 September 2006Golden+8.9%+7.9%+15.5%
23 June 2009Golden+20.4%+25.9%+24.6%
22 October 2010Golden+9.7%+14.0%+6.8%
31 January 2012Golden+7.6%+5.9%+17.7%
17 December 2015Golden+1.0%+3.1%+12.9%
25 April 2016Golden+4.4%+3.6%+16.7%
1 April 2019Golden+3.9%+4.8%−8.0%
9 July 2020Golden+8.8%+21.6%+40.6%
2 February 2023Golden−2.4%+8.0%+20.0%
1 July 2025Golden+8.2%+11.9%+21.9%

50/200-day moving averages on plain prices, returns on dividend-adjusted closes. Indices Lab study on SPY, dividends reinvested, 29 January 1993 to 20 November 2026. Data: EODHD. Hypothetical results: past performance does not predict future returns.

3 months 6 months 12 months 12 months, share that gained
After a golden cross +4.8% +7.9% +17.5% 93.3%
After a death cross +6.7% +11.5% +20.4% 80.0%
Any day +3.8% +7.1% +14.8% 81.4%

All figures are medians, the middle result, so one huge year doesn't skew them. Over 12 months the median after a golden cross beat any day by +2.8 points, from only 15 crosses in more than three decades. The year after a cross ranged from a loss after the April 2019 cross, a year before the Covid crash, to the biggest gain after the July 2020 one, and the November 1999 cross, months before the dot-com peak, barely gained. Every cross is in the table under the chart.

The supposed bad sign wasn't a reliable sell either. The median year after a death cross gained more than the year after a golden cross, because most death crosses came after much of a fall had already happened:

After a golden crossAfter a death crossAny-day median
12-month return of SPY after each golden cross and death cross, with the median from any day. Golden crosses: median +17.5%, positive 93.3%; any day: median +14.8%, positive 81.4%-40%-20%0%+20%+40%+60%199419982000200420072010201520162020202319 April 1994: +17.0%14 September 1994: +27.0%29 September 1998: +22.3%8 December 1998: +20.2%4 November 1999: +5.6%11 November 1999: +2.1%30 October 2000: −20.3%15 May 2003: +17.5%18 August 2004: +13.3%5 November 2004: +6.2%19 July 2006: +24.4%11 September 2006: +15.5%21 December 2007: −39.8%23 June 2009: +24.6%6 July 2010: +32.7%22 October 2010: +6.8%12 August 2011: +21.7%31 January 2012: +17.7%28 August 2015: +12.0%17 December 2015: +12.9%11 January 2016: +20.4%25 April 2016: +16.7%7 December 2018: +21.5%1 April 2019: −8.0%30 March 2020: +53.3%9 July 2020: +40.6%14 March 2022: −5.1%2 February 2023: +20.0%14 April 2025: +31.6%1 July 2025: +21.9%any day: median +14.8%
  • Golden crosses (15): median 12-month return +17.5%, positive 93.3% of the time
  • Death crosses (15): median 12-month return +20.4%, positive 80.0% of the time
  • Any day: median +14.8%, positive 81.4%
CrossType3 months6 months12 months
19 April 1994Death+2.5%+6.9%+17.0%
14 September 1994Golden−2.7%+7.2%+27.0%
29 September 1998Death+18.9%+23.0%+22.3%
8 December 1998Golden+10.7%+11.2%+20.2%
4 November 1999Death+4.7%+5.7%+5.6%
11 November 1999Golden+0.4%+3.6%+2.1%
30 October 2000Death−2.2%−9.2%−20.3%
15 May 2003Golden+4.8%+12.6%+17.5%
18 August 2004Death+7.9%+11.5%+13.3%
5 November 2004Golden+3.2%+1.6%+6.2%
19 July 2006Death+9.0%+14.8%+24.4%
11 September 2006Golden+8.9%+7.9%+15.5%
21 December 2007Death−9.6%−10.6%−39.8%
23 June 2009Golden+20.4%+25.9%+24.6%
6 July 2010Death+11.2%+24.8%+32.7%
22 October 2010Golden+9.7%+14.0%+6.8%
12 August 2011Death+5.8%+15.9%+21.7%
31 January 2012Golden+7.6%+5.9%+17.7%
28 August 2015Death+5.7%+0.5%+12.0%
17 December 2015Golden+1.0%+3.1%+12.9%
11 January 2016Death+7.7%+13.1%+20.4%
25 April 2016Golden+4.4%+3.6%+16.7%
7 December 2018Death+6.7%+10.7%+21.5%
1 April 2019Golden+3.9%+4.8%−8.0%
30 March 2020Death+16.9%+28.8%+53.3%
9 July 2020Golden+8.8%+21.6%+40.6%
14 March 2022Death−9.8%−5.0%−5.1%
2 February 2023Golden−2.4%+8.0%+20.0%
14 April 2025Death+16.1%+23.5%+31.6%
1 July 2025Golden+8.2%+11.9%+21.9%

50/200-day moving averages on plain prices, returns on dividend-adjusted closes. Indices Lab study on SPY, dividends reinvested, 29 January 1993 to 20 November 2026. Data: EODHD. Hypothetical results: past performance does not predict future returns.

Losses were more common after a death cross, though, and the bad years were real bear markets: 2000, 2007 and 2022.

Run as a strategy, the rule cost money:

Since 12 November 1993, $10,000 End value CAGR Worst fall
Own SPY only while 50-day > 200-day $197,960 9.5% 33.7%
Own SPY throughout $266,825 10.5% 55.2%

CAGR is the steady yearly growth rate that links the start and end values; the worst fall is the biggest drop from a high to the following low. The rule was invested on 75.6% of days, and 60.0% of SPY's 20 best days came while it sat in cash: the sharpest rebounds happen while the averages still point down.

Hold only while fast average is above slowBuy and hold
Growth of $10,000 in SPY: holding only while the 50-day average is above the 200-day one ended at $197,960 (9.5% a year, worst fall 33.7%); buy and hold ended at $266,825 (10.5% a year, worst fall 55.2%)$0k$100k$200k$300k199419982002200620102014201820222026
  • Rule $197,960, 9.5% a year, worst fall 33.7%, invested 75.6% of days, 30 switches
  • Buy and hold $266,825, 10.5% a year, worst fall 55.2%
Drawdown from the previous high, rule vs buy and hold-40%-20%0%199419982002200620102014201820222026

First half: rule 11.0% a year vs buy and hold 7.6%; second half: 8.0% vs 13.5%. 60.0% of SPY's 20 best days fell while the rule was in cash.

Rule from 12 November 1993, cash earns nothing, no costs or taxes. Indices Lab study on SPY, dividends reinvested, 29 January 1993 to 20 November 2026. Data: EODHD. Hypothetical results: past performance does not predict future returns.

Build a portfolio you don't need to time Opens the builder with BND 40%, SPY 60% against SPY. Free account, no card.

Where the myth has a point

Two things. The year after a golden cross ended in a gain more often than the year after any day, 93.3% against 81.4%. And the strategy cut the worst fall by a lot. It stepped aside during the slow bear markets of 2000 to 2002 and 2008, and in the first half of the test it beat holding outright: 11.0% a year against 7.6%, with a worst fall of 19.0% against 55.2%.

Its weakness is speed. In 2020 the market fell so fast that the death cross came after the bottom, so the rule took the same worst fall as holding and then sat in cash for the first months of the recovery. In the second half of the test it made 8.0% a year against 13.5% for holding. A signal that only works against slow crashes can't tell you in advance which kind is coming.

What it means for your portfolio

A golden cross tells you the market has already been rising; it doesn't tell you what comes next. If the appeal is avoiding a big fall, there's a steadier way to get most of it: own a mix whose worst fall you could sit through, such as stocks with bonds, and check what that mix did through 2008, 2020 and 2022 before you own it. A backtest shows that drawdown in money, with no signal to watch. For other sayings we've put to the same kind of test, see does gold beat the S&P 500? and are dividend stocks safer in a crash?, and for judging a portfolio's ride, Sharpe vs Sortino vs Calmar.

Backtest your own portfolio Opens the builder with BND 40%, SPY 60% against SPY. Free account, no card.

Questions

What is a golden cross?

The day a stock's or fund's 50-day moving average rises above its 200-day moving average. The opposite, the 50-day falling below the 200-day, is a death cross.

Is the golden cross a buy signal?

On SPY since 1993, the year after a golden cross was a little better than the year after any day, but trading on the crosses grew money more slowly than simply holding. It describes the past few months more than it forecasts the next.

Does a death cross mean sell?

Not reliably. On SPY, the median year after a death cross was a gain, larger than after a golden cross, because most came after much of the fall. Losses were more common than after a golden cross, though, and two of the worst years followed death crosses in 2000 and 2007.

Does the golden cross work on individual stocks or crypto?

This test covers SPY only. Single stocks and crypto swing more and produce more crosses, which also means more false starts. Test the exact asset and rule before trusting it.

Why 50 and 200 days?

Convention: they are the pair most charts and news stories use. Other pairs exist; this test covers 50 and 200 only.

Build your index Opens the builder with BND 40%, SPY 60% against SPY. Free account, no card.

About this page. Figures on this page come from Indices Lab’s own data, powered by EODHD, and every number is recalculated after each market close; the text is reviewed by a person before it is published. Indices Lab is a research and educational tool: everything here is hypothetical and informational, not investment advice and not a recommendation to buy or sell any security. Past performance does not predict future results. See our methodology.