DeployQuantGolden Cross (SMA 50/200)
Hold while the 50-day average is above the 200-day; step aside when it crosses below (the death cross).
The golden cross strategy holds an ETF while its 50-day simple moving average sits above its 200-day average, and moves to cash when the 50-day falls below the 200-day. The median result was a 2.05% annualized return with a median maximum drawdown of 21.84% and a median Sharpe ratio of 0.33. The rule beat buy-and-hold of the same fund in 27 of the 59 cases, finished with a positive return in 41, and had a shallower maximum drawdown than holding in 56.
Those three counts describe a rule that does one job. It cuts drawdowns almost everywhere and it lowers returns wherever the fund rose. The median fund took 3 round trips in 5.7 years and held a position on 59.3% of trading days. Most of the 27 wins against buy-and-hold come from funds that fell during the window, where sitting in cash beat owning them.
The 200-day regime filter is the closest relative and gives a useful comparison, with a median return of 1.84% across the same 59 funds. The SMA 10/50 trend has a median of 2.12%, and the EMA 12/26 trend has 2.76%. All figures come from one window of data that contains one long bear market, so the table describes this window and makes no claim about others.
The rules
- WHEN the market opens · IF not invested AND SMA(50) > SMA(200) · THEN buy with 98% of the sleeve
- WHEN the market opens · IF invested AND SMA(50) < SMA(200) · THEN sell the whole position
A widely used trend filter. When the 50-day simple moving average is above the 200-day, the asset is in a long-term uptrend and the strategy holds. When it crosses below, the strategy moves to cash. It trades rarely, with a handful of signals per decade on an index. Its use is skipping the deepest bear markets, and it will lag some rallies.
Good for: long-horizon investors who want to hold trends but sidestep multi-year bear markets.
Watch out: crosses lag at turning points: the strategy gives back the first leg of a crash and misses the first leg of a recovery, and choppy sideways markets whipsaw it.
How the rule behaves
The entry fires at the open when the sleeve is flat and the 50-day average is above the 200-day average. A buy uses 98% of the sleeve. The exit fires at the open when the sleeve holds a position and the 50-day has fallen below the 200-day. Both conditions use averages computed from earlier closes, so a cross that forms in the afternoon trades the next morning.
Two moving averages crossing is a slower signal than one price crossing one average. The 50-day average must fall below the 200-day, which takes weeks of decline in most funds. By then a good part of a drop has already happened. The template description says as much: the strategy gives back the first leg of a crash and misses the first leg of a recovery. The result is a small number of trades. The most any fund reached was 6 round trips, on SH, and SGOV made none. SGOV's exposure was 86.1% and its return matched holding at 3.22%, so on that fund the rule changed nothing.
The 200-day average needs 200 sessions of history, and the test starts on 2021-01-04. Every fund therefore spent much of 2021 flat while the average formed. This affects the table in a specific way. The exposure column and the year-one return both reflect a start in cash, and the strategy's 2021 median return of 0.2% against a buy-and-hold median of 4.1% is partly that effect.
The rule has no stop, no position sizing and no view on volatility. A fund that falls 5% in a day and a fund that drifts down 5% over a month are treated alike until the averages cross. This is why the leveraged and inverse funds, which move fast, appear at both ends of the table.
Build it from blocks (or type it in English), backtest it on 5.7 years of minute data in seconds, tweak any parameter, then paper trade it on live data. No card, no broker needed to start.
Results on every ETF
| ETF | CAGR | buy & hold | max DD | Sharpe | trades | win rate |
|---|---|---|---|---|---|---|
| SOXL | 45.9% | 33.3% | −69.3% | 0.88 | 3 | 67% (+1 open) |
| SOXX | 28.2% | 31.1% | −28.8% | 1.02 | 3 | 100% (+1 open) |
| CLSE | 17.5% | 19.5% | −16.1% | 1.47 | 2 | 100% (+1 open) |
| XLK | 15.3% | 22.1% | −19.6% | 0.87 | 2 | 50% (+1 open) |
| QQQM | 13.8% | 16.7% | −22.4% | 0.89 | 2 | 50% (+1 open) |
| QQQ | 13.6% | 16.7% | −22.2% | 0.88 | 2 | 50% (+1 open) |
| SPUU | 13.2% | 22.5% | −34.7% | 0.65 | 3 | 67% (+1 open) |
| IOO | 12.4% | 16.7% | −18.8% | 0.98 | 2 | 100% (+1 open) |
| ROM | 12.1% | 30.2% | −35.1% | 0.50 | 3 | 33% (+1 open) |
| QLD | 12.0% | 23.6% | −39.9% | 0.53 | 3 | 67% (+1 open) |
| VOX | 11.2% | 9.1% | −20.9% | 0.84 | 3 | 67% |
| IAU | 10.8% | 13.7% | −25.9% | 0.74 | 4 | 50% |
| TECL | 10.0% | 38.2% | −58.3% | 0.45 | 4 | 25% (+1 open) |
| VOOG | 9.0% | 15.7% | −21.8% | 0.65 | 3 | 67% (+1 open) |
| SSO | 8.9% | 21.9% | −36.0% | 0.50 | 2 | 50% (+1 open) |
| VOO | 8.8% | 14.4% | −18.1% | 0.77 | 2 | 50% (+1 open) |
| SPY | 8.7% | 14.6% | −18.1% | 0.77 | 2 | 50% (+1 open) |
| TBF | 8.5% | 12.2% | −17.5% | 0.69 | 4 | 75% (+1 open) |
| VV | 8.4% | 14.3% | −18.6% | 0.73 | 2 | 50% (+1 open) |
| EEM | 7.8% | 6.6% | −20.6% | 0.58 | 2 | 50% (+1 open) |
| TQQQ | 7.3% | 25.4% | −57.4% | 0.38 | 3 | 67% (+1 open) |
| QQQE | 7.0% | 9.8% | −21.0% | 0.55 | 2 | 50% (+1 open) |
| VOOV | 6.8% | 12.6% | −17.1% | 0.66 | 2 | 50% (+1 open) |
| VTV | 5.0% | 13.4% | −17.8% | 0.51 | 3 | 67% (+1 open) |
| QAI | 4.3% | 3.9% | −7.6% | 0.89 | 1 | 100% (+1 open) |
| USDU | 3.7% | 5.2% | −8.3% | 0.71 | 2 | 100% (+1 open) |
| XLF | 3.6% | 12.6% | −17.9% | 0.33 | 4 | 25% (+1 open) |
| SGOV | 3.2% | 3.2% | −0.0% | 14.13 | 0 | – (+1 open) |
| RINF | 2.4% | 6.6% | −13.3% | 0.28 | 4 | 50% (+1 open) |
| XLP | 2.0% | 5.8% | −14.3% | 0.24 | 3 | 67% (+1 open) |
| IGIB | 1.8% | 0.2% | −5.6% | 0.47 | 3 | 67% |
| ALTY | 1.6% | 7.6% | −13.1% | 0.26 | 3 | 33% (+1 open) |
| XLY | 1.4% | 6.3% | −25.6% | 0.16 | 5 | 60% |
| CTA | 1.4% | 8.9% | −20.1% | 0.17 | 5 | 40% |
| IEI | 1.1% | −0.4% | −5.5% | 0.40 | 2 | 50% |
| KMLM | 1.1% | 7.1% | −25.8% | 0.17 | 3 | 33% (+1 open) |
| IWM | 1.1% | 7.5% | −31.0% | 0.14 | 4 | 25% (+1 open) |
| AGG | 0.8% | −0.8% | −7.1% | 0.22 | 5 | 40% |
| BND | 0.7% | −0.8% | −6.9% | 0.21 | 5 | 40% |
| EEV | 0.7% | −16.1% | −46.9% | 0.15 | 3 | 33% |
| FXE | 0.4% | −0.9% | −8.7% | 0.10 | 4 | 50% |
| IEF | −0.3% | −2.3% | −8.5% | -0.05 | 4 | 50% |
| UDN | −1.1% | −1.1% | −11.1% | -0.22 | 4 | 25% |
| VXZ | −1.1% | −15.1% | −19.9% | 0.00 | 3 | 33% |
| VIXM | −1.8% | −16.1% | −24.8% | -0.04 | 3 | 33% |
| SPDN | −2.3% | −9.7% | −28.4% | -0.16 | 4 | 25% |
| SH | −2.3% | −10.0% | −28.6% | -0.17 | 6 | 33% |
| SDS | −2.3% | −21.6% | −33.4% | -0.04 | 2 | 50% |
| UST | −2.4% | −8.5% | −22.4% | -0.33 | 5 | 40% |
| TLT | −2.5% | −8.2% | −20.6% | -0.30 | 5 | 0% |
| PSQ | −4.3% | −13.8% | −34.9% | -0.24 | 3 | 33% |
| FAS | −5.5% | 18.3% | −49.0% | 0.04 | 5 | 20% (+1 open) |
| QID | −6.0% | −29.5% | −48.4% | -0.11 | 2 | 0% |
| REW | −6.7% | −36.1% | −58.5% | -0.08 | 2 | 50% |
| SQQQ | −7.4% | −42.3% | −51.5% | -0.03 | 3 | 33% |
| TMF | −8.4% | −31.2% | −50.1% | -0.50 | 4 | 0% |
| UVXY | −10.1% | −48.7% | −51.3% | -0.52 | 1 | 0% |
| TECS | −21.0% | −46.7% | −80.0% | -0.44 | 3 | 0% |
| SOXS | −24.4% | −48.3% | −85.5% | -0.30 | 3 | 0% |
Reading the 59-fund table
The top of the table is led by SOXL, a leveraged semiconductor fund, with a 45.93% annualized return against 33.32% for holding it. The rule beat holding there by 12.61 points a year. The cost was a maximum drawdown of 69.3%, the deepest of any fund that finished with a positive return. The rule made 3 round trips on SOXL and won 67% of them.
Below SOXL the unleveraged funds follow. SOXX returned 28.24% against 31.11% for holding, with a 28.82% drawdown. CLSE returned 17.46% with a 16.06% drawdown. XLK returned 15.33% against 22.07%. The two Nasdaq-100 funds, QQQM and QQQ, returned 13.82% and 13.6%, each with 2 round trips and a drawdown near 22%. SPY returned 8.71% with an 18.13% drawdown against 14.56% for holding.
Broad index funds show the cleanest version of the trade-off. IOO returned 12.39% with an 18.79% drawdown. VOO returned 8.77% with an 18.06% drawdown against 14.42% for holding. In each one the rule lost several points of annual return and kept the drawdown in the high teens. Only one of the 12 broad index funds beat holding.
Wins against buy-and-hold also come from the other direction. VOX returned 11.2% against 9.08%, and EEM returned 7.76% against 6.58%, two funds that went sideways enough for cash to help. IGIB returned 1.84% against 0.21%.
The bottom of the table is the leveraged inverse and volatility funds. SOXS returned negative 24.44% with an 85.51% drawdown, the deepest in the table. TECS returned negative 21.03% with an 80.03% drawdown, and UVXY returned negative 10.09% with a 51.34% drawdown on a single round trip. Those are still far better than holding them, since UVXY lost 48.74% a year on a buy-and-hold basis. The rule limited the damage and still lost money, and the two deepest drawdowns in the table came on SOXS and TECS, funds the rule held on only 10.5% and 12.1% of days. The Sharpe ratios in the table are mostly below 1, with a few exceptions among funds that made 1 or 2 trades.
Leverage and drawdown go together in the table. SOXL's 69.3% drawdown, TECL's 58.25% and TQQQ's 57.35% are the three deepest among funds with a positive return, and EEV is fourth, returning 0.72% with a 46.86% drawdown while holding lost 16.1% a year. The rule does not remove the risk of a leveraged fund. It shortens the time spent in the worst declines.
QLD returned 11.99% against 23.55% for holding with a 39.91% drawdown, and SPUU returned 13.22% against 22.48% with 34.67%. Both are leveraged funds, and both gave up a large part of the holding return in exchange for a drawdown that stayed under 40%.
Fund-by-fund results rest on 1 to 6 trades each. A fund with 2 round trips and a 100% win rate, such as IOO, has two data points. The win rate column is better read as a count than as a rate.
Results by fund type
| Fund type | ETFs | Median CAGR | Median buy & hold | Median max DD | Beat holding |
|---|---|---|---|---|---|
| Broad index ETFs | 12 | 8.7% | 14.4% | −20.6% | 1 of 12 |
| Sector ETFs | 6 | 11.2% | 12.6% | −20.9% | 1 of 6 |
| Leveraged ETFs | 10 | 10.0% | 23.6% | −49.0% | 3 of 10 |
| Inverse ETFs | 11 | −4.3% | −21.6% | −46.9% | 10 of 11 |
| Bond ETFs | 7 | 0.8% | −0.8% | −6.9% | 6 of 7 |
| Commodity ETFs | 1 | 10.8% | 13.7% | −25.9% | 0 of 1 |
| Currency ETFs | 3 | 0.4% | −0.9% | −8.7% | 2 of 3 |
| Volatility products | 3 | −1.8% | −16.1% | −24.8% | 3 of 3 |
| Alternative-strategy ETFs | 6 | 2.4% | 7.6% | −16.1% | 1 of 6 |
Results by fund type
Fund type explains more of the table than any single fund does. Among the 12 broad index funds the median return was 8.71% against a median buy-and-hold return of 14.42%, with a median drawdown of 20.65%. One of the 12 beat holding. For the 6 sector funds the median return was 11.2% against 12.58%, and one beat holding. The sector group is the closest any category came to matching holding.
Leveraged ETFs are the widest case. The median return was 10.02% against 23.55% for holding, a median drawdown of 48.96%, and 3 of 10 beat holding. A leveraged fund can fall fast enough that the 50-day average drops below the 200-day only after much of the decline is done, and it can rise fast enough that the cross back above comes after the first part of the rebound. TECL returned 10.02% against 38.16% for holding. FAS returned negative 5.49% while holding returned 18.34%, with 5 round trips and a 20% win rate. TQQQ returned 7.33% against 25.35%, with a 57.35% drawdown.
Inverse ETFs show the strategy in a different role. The median return was negative 4.34% against negative 21.57% for holding, and 10 of 11 beat holding. An inverse fund loses value when the market rises, and the strategy's exposure to most of them was low: 2.8% on SOXS, 15.9% on SQQQ and 21.7% on PSQ. TBF was the exception, with a position on 66.3% of days and a return of 8.47%. Beating a fund that lost 48.27% a year, as SOXS did, is a low bar. The rule still lost 24.44% a year there, on three round trips with none winning.
Bond ETFs gave the rule its steadiest use. The median return was 0.78% against negative 0.78%, the median drawdown 6.88%, and 6 of 7 beat holding. AGG returned 0.78% against negative 0.78%, and BND returned 0.72% against negative 0.8%. TLT returned negative 2.46% against negative 8.23% with a 0% win rate across 5 round trips, so the rule lost money on every trade there and still beat holding.
The volatility products were 3 of 3 against holding, with a median return of negative 1.81% against negative 16.06%. Currency ETFs had a median of 0.37% and 2 of 3 beat holding. The one commodity fund, IAU, returned 10.79% against 13.66%. Alternative-strategy funds returned a median 2.38% against 7.61% and one of 6 beat holding.
Year by year, median across all ETFs
| Year | golden cross | Buy & hold | ETFs with a gain |
|---|---|---|---|
| 2021 | 0.2% | 4.1% | 31 of 59 |
| 2022 | −3.8% | −12.7% | 15 of 59 |
| 2023 | 0.0% | 8.9% | 26 of 59 |
| 2024 | 8.4% | 9.7% | 41 of 59 |
| 2025 | 1.1% | 11.1% | 35 of 59 |
| 2026 | 0.2% | 3.7% | 30 of 59 |
Year by year across the 59 funds
The year table shows the median across all funds, which hides the dispersion but shows the rhythm of the rule. In 2021 the strategy's median was 0.2% against 4.1% for holding, with 31 funds up. This is the warm-up year for the 200-day average. In 2022 the median was a loss of 3.8% against 12.7% for holding, and only 15 funds finished with a gain. This is the year that justifies the rule, since the median strategy lost much less than the median holder.
2023 shows the other side. The strategy's median return was 0% against 8.9% for holding, with 26 funds up. A cross takes weeks to confirm, and after a bear year, the strategy was still flat when the rebound started. In 2024 the median was 8.4% against 9.7%, with 41 funds up. Here most funds held a position through the year and the two numbers are close.
In 2025 the median was 1.1% against 11.1%, with 35 funds up, and 2026 so far shows 0.2% against 3.7%, with 30 up. The strategy lagged in both. A fast fall followed by a fast recovery is the pattern a slow cross handles worst, because the position closes after the drop and reopens after the rebound, which is consistent with these two years.
The six rows give one win for the strategy in 2022, one close year in 2024 and four years of lag. That is the shape of the median strategy against the median fund. A rule that earns its keep in one year in six has to be judged by the size of that year, and in this window 2022 was a year in which most funds lost money.
Changing the parameters
| Version | Median CAGR | Median max DD | Median round trips |
|---|---|---|---|
| Published rules | 2.0% | −21.8% | 3 |
| SMA 40/200 | 2.3% | −23.0% | 3 |
| SMA 50/150 | 2.5% | −22.2% | 4 |
| SMA 60/250 | 2.1% | −22.8% | 2 |
Changing the averages
Three variants were run on all 59 funds. The published 50/200 pair had a median return of 2.05%, a median drawdown of 21.84% and a median of 3 round trips. SMA 40/200 had a median return of 2.34%, a drawdown of 23.03% and 3 round trips. SMA 50/150 returned 2.53% with a 22.17% drawdown and 4 round trips. SMA 60/250 returned 2.08% with a 22.77% drawdown and 2 round trips.
The differences are small. The median return moves between 2.05% and 2.53% across the four settings, and the median drawdown between 21.84% and 23.03%. The published pair has the shallowest median drawdown of the four. The 50/150 pair has the highest median return and one more round trip than the published rule, which fits a shorter long-term average reacting earlier. The 60/250 pair is slowest and made the fewest trades, 2 at the median.
None of the variants changes the conclusion. Moving the windows by a few weeks shifts individual crossing dates, and with 1 to 6 trades per fund a shift in one date can change a fund's result by a large amount. At the median across 59 funds those shifts average out, and the published setting is neither the best nor the worst. The test does not show that any pair is better than another.
Frequently asked questions
What is the golden cross strategy?
Hold while the 50-day average is above the 200-day; step aside when it crosses below (the death cross). A widely used trend filter. When the 50-day simple moving average is above the 200-day, the asset is in a long-term uptrend and the strategy holds. When it crosses below, the strategy moves to cash. It trades rarely, with a handful of signals per decade on an index. Its use is skipping the deepest bear markets, and it will lag some rallies.
Does golden cross beat buy-and-hold?
Across 59 ETFs backtested 2021-01-04 to 2026-10-02, it beat same-ETF buy-and-hold on 27 of 59 (46%). Median CAGR was 2.0% with a median max drawdown of 21.8%. Per-ETF results vary widely; the table lists every one.
How often does a golden cross happen?
On a broad index, roughly every couple of years. The backtest pages show the exact trade count for each ETF over the 2021 to 2026 window.
Golden cross vs buy and hold: which does better?
It depends on the asset and the window. Every backtest page here shows the same-window buy-and-hold comparison.
What is the golden cross strategy?
It holds an ETF while the 50-day simple moving average is above the 200-day average and moves to cash when the 50-day falls below. Orders are placed at the open after the signal, and a buy uses 98% of the sleeve.
Does the golden cross beat buy-and-hold?
On 27 of 59 ETFs from 2021-01-04 to 2026-10-02. Most of those 27 are funds that fell during the window, such as inverse and volatility products. Among the 12 broad index funds only one beat holding.
How many trades does it make?
The median fund had 3 round trips in 5.7 years and the most was 6, on SH. SGOV made none. The median exposure was 59.3% of trading days.
Which fund did best with the golden cross?
SOXL returned 45.93% a year against 33.32% for holding, with a 69.3% maximum drawdown. Among unleveraged funds, SOXX returned 28.24% and CLSE returned 17.46%.
Does the golden cross reduce drawdowns?
In 56 of 59 cases the maximum drawdown was shallower than holding. The median drawdown was 21.84%. Leveraged funds still had a median drawdown of 48.96% under the rule.
Did changing the moving average lengths help?
Not by much. SMA 50/150 had a median return of 2.53%, SMA 40/200 had 2.34% and SMA 60/250 had 2.08%, against 2.05% for the published 50/200. Medians moved less than a point.
Compare with other strategies
Backtests are hypothetical, computed by DeployQuant's engine on minute-resolution consolidated US market data (2021-01-04 to 2026-10-02, $10,000 starting capital, no margin, no fees or slippage in the headline run; buy-and-hold puts 98% of the account in at the first open, as the templates do) and do not guarantee future results. Nothing on this page is investment advice. Live trading involves risk of loss.