DeployQuantSMA 10/50 Trend
A faster moving-average crossover: hold while the 10-day average is above the 50-day.
SMA 10/50 Trend holds an ETF while its 10-day simple moving average sits above its 50-day average, and sits in cash otherwise. The test ran the same two rules on 59 ETFs from 2021-01-04 to 2026-10-02, starting with $10,000 per ETF. The median CAGR across the 59 was 2.1%, the median max drawdown was 23.0%, and the median Sharpe was 0.35. A typical ETF ran 17 round trips and was invested 57.6% of the time.
The strategy beat buy-and-hold of the same ETF in 21 of 59 cases. It finished with a positive CAGR on 41 of them. It had a shallower drawdown than holding on 52 of 59. That split tells you what the rules do: they cut the depth of a loss on most funds, and they give up return on most funds that rose.
The headline run has no fees or slippage, and each ETF is one path through one window. For the slower relative of this rule set see golden cross, and for a different kind of filter see the 200-day regime filter.
The rules
- WHEN the market opens · IF not invested AND SMA(10) > SMA(50) · THEN buy with 98% of the sleeve
- WHEN the market opens · IF invested AND SMA(10) < SMA(50) · THEN sell the whole position
A faster version of the golden cross. The 10- and 50-day averages catch intermediate trends measured in weeks rather than years. It enters recoveries earlier and exits breakdowns earlier, with more whipsaw trades in sideways markets. Useful for seeing how signal speed changes a strategy's results.
Good for: trending assets with multi-week swings, such as leveraged index ETFs.
Watch out: several false signals a year is normal; each whipsaw costs a small loss and they add up in flat markets.
What the two rules do on a daily chart
Both rules are evaluated when the market opens, using daily closes. If the strategy holds nothing and SMA(10) is above SMA(50), it buys with 98% of the sleeve. If it holds the position and SMA(10) drops below SMA(50), it sells everything. There is no stop, no profit target and no short side. The other 2% of the sleeve stays in cash.
A 10-day average over a 50-day average reacts to a move that lasts a few weeks. The 10-day line crosses above the 50-day line after a rebound has already carried the price up for a couple of weeks, so the entry is late by that amount. The cross back down happens after the price has already slipped for a similar stretch, so the exit is late too. Every round trip therefore starts after some of the gain is gone and ends after some of the loss has arrived. When a trend runs for months, that delay is small compared with the move. When the price oscillates inside a range, the two averages weave around each other and the rules buy and sell repeatedly at slightly worse prices each time.
The median ETF made 17 round trips in 5.74 years, which is about three a year. The ETFs at the top of the trade counts were RINF with 25 round trips, VXZ with 25, VIXM with 24, EEV with 24 and EEM with 23. The lowest was SGOV with 6, a fund that barely moves, followed by IEF, UST and SDS with 14. High counts lined up with weak results: the funds with the most trades had win rates between 17% and 36%.
Exposure explains much of the difference between funds. The median ETF was invested 57.6% of the days. SGOV was invested 93.5% of the time, VOO 69.5%, and TQQQ 58.8%. Funds that trend downward most of the time, such as the inverse ETFs, spent far less time invested: SQQQ 28.2%, SOXS 27.3% and UVXY 19.7%. The rule can only buy when the 10-day average is above the 50-day, so a fund in a long decline keeps it out of the market, which lowers the loss but also means each short rally is bought late.
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 |
|---|---|---|---|---|---|---|
| QLD | 18.8% | 23.6% | −32.3% | 0.76 | 19 | 37% (+1 open) |
| SOXL | 18.2% | 33.3% | −72.2% | 0.60 | 17 | 47% (+1 open) |
| TECL | 18.0% | 38.2% | −62.4% | 0.59 | 17 | 47% (+1 open) |
| TQQQ | 17.3% | 25.4% | −53.3% | 0.60 | 20 | 45% (+1 open) |
| CLSE | 16.6% | 19.5% | −9.3% | 1.55 | 12 | 75% (+1 open) |
| ROM | 14.5% | 30.2% | −43.5% | 0.58 | 16 | 44% (+1 open) |
| SPUU | 11.2% | 22.5% | −31.4% | 0.62 | 14 | 57% (+1 open) |
| IAU | 10.8% | 13.7% | −23.4% | 0.78 | 16 | 44% |
| XLK | 9.5% | 22.1% | −25.6% | 0.63 | 18 | 44% (+1 open) |
| IOO | 9.2% | 16.7% | −22.0% | 0.87 | 17 | 53% (+1 open) |
| QQQM | 9.1% | 16.7% | −20.9% | 0.68 | 18 | 39% (+1 open) |
| QQQ | 9.0% | 16.7% | −20.8% | 0.68 | 18 | 39% (+1 open) |
| VOOG | 8.9% | 15.7% | −18.0% | 0.72 | 16 | 50% (+1 open) |
| SSO | 8.4% | 21.9% | −32.1% | 0.50 | 15 | 60% (+1 open) |
| VV | 8.2% | 14.3% | −12.6% | 0.79 | 15 | 53% (+1 open) |
| VOO | 8.0% | 14.4% | −12.7% | 0.79 | 17 | 59% (+1 open) |
| SPY | 7.9% | 14.6% | −12.7% | 0.79 | 17 | 59% (+1 open) |
| SOXX | 7.7% | 31.1% | −40.3% | 0.42 | 17 | 53% (+1 open) |
| VTV | 7.4% | 13.4% | −11.6% | 0.83 | 17 | 65% |
| TBF | 6.7% | 12.2% | −13.9% | 0.62 | 17 | 59% (+1 open) |
| VOOV | 6.3% | 12.6% | −14.8% | 0.70 | 16 | 50% |
| VOX | 6.1% | 9.1% | −19.6% | 0.52 | 18 | 56% (+1 open) |
| FAS | 4.6% | 18.3% | −57.0% | 0.30 | 19 | 58% |
| XLF | 4.0% | 12.6% | −27.2% | 0.39 | 19 | 47% |
| CTA | 3.9% | 8.9% | −17.1% | 0.34 | 16 | 50% (+1 open) |
| USDU | 3.5% | 5.2% | −7.4% | 0.74 | 15 | 53% (+1 open) |
| SGOV | 3.2% | 3.2% | −0.1% | 13.91 | 6 | 33% (+1 open) |
| KMLM | 3.1% | 7.1% | −21.3% | 0.35 | 19 | 32% (+1 open) |
| ALTY | 2.9% | 7.6% | −15.0% | 0.46 | 17 | 41% |
| QAI | 2.1% | 3.9% | −8.1% | 0.52 | 18 | 39% |
| IGIB | 1.7% | 0.2% | −6.6% | 0.44 | 16 | 44% |
| FXE | 1.6% | −0.9% | −8.2% | 0.33 | 19 | 37% |
| UDN | 1.5% | −1.1% | −6.0% | 0.35 | 17 | 35% |
| IEF | 1.3% | −2.3% | −5.7% | 0.31 | 14 | 43% |
| IEI | 1.2% | −0.4% | −3.2% | 0.44 | 16 | 50% |
| XLP | 1.2% | 5.8% | −15.6% | 0.17 | 19 | 37% |
| QQQE | 1.2% | 9.8% | −25.6% | 0.16 | 21 | 43% (+1 open) |
| AGG | 0.8% | −0.8% | −6.5% | 0.23 | 18 | 44% |
| UST | 0.7% | −8.5% | −10.6% | 0.12 | 14 | 36% |
| BND | 0.5% | −0.8% | −6.6% | 0.16 | 19 | 42% |
| XLY | 0.3% | 6.3% | −27.5% | 0.10 | 22 | 41% |
| TLT | −0.8% | −8.2% | −14.2% | -0.04 | 17 | 35% |
| IWM | −1.2% | 7.5% | −32.9% | -0.01 | 21 | 38% |
| RINF | −2.1% | 6.6% | −23.9% | -0.19 | 25 | 36% (+1 open) |
| SPDN | −2.6% | −9.7% | −23.0% | -0.16 | 17 | 24% |
| SH | −2.7% | −10.0% | −22.8% | -0.17 | 15 | 20% |
| EEM | −3.9% | 6.6% | −41.2% | -0.23 | 23 | 26% (+1 open) |
| PSQ | −6.3% | −13.8% | −37.5% | -0.32 | 18 | 33% |
| TMF | −10.7% | −31.2% | −57.6% | -0.30 | 18 | 22% |
| SDS | −11.6% | −21.6% | −58.0% | -0.42 | 14 | 21% |
| VXZ | −12.1% | −15.1% | −55.0% | -0.55 | 25 | 24% |
| QID | −13.1% | −29.5% | −63.2% | -0.28 | 18 | 22% |
| VIXM | −13.1% | −16.1% | −57.0% | -0.52 | 24 | 21% |
| REW | −19.8% | −36.1% | −77.0% | -0.42 | 21 | 19% |
| SQQQ | −21.1% | −42.3% | −79.9% | -0.27 | 18 | 17% |
| EEV | −24.4% | −16.1% | −81.3% | -0.93 | 24 | 17% |
| TECS | −27.1% | −46.7% | −87.2% | -0.38 | 21 | 24% |
| UVXY | −38.0% | −48.7% | −94.0% | -0.44 | 16 | 6% |
| SOXS | −49.0% | −48.3% | −97.9% | -0.65 | 19 | 11% |
Results by fund type
The 59 ETFs fall into nine fund types in the data, and the split by type shows more than the overall median.
Broad index ETFs, 12 of them, had a median CAGR of 8.0% against a buy-and-hold median of 14.4%. The median max drawdown was 20.8%. None of the 12 beat holding. The rule kept some of the return and cut some of the pain, but across this window the broad index funds rose often enough that the exits cost more than they saved. SPY returned 7.9% against 14.6% for holding, with a 12.7% drawdown, and QQQ returned 9.0% against 16.7% with a 20.8% drawdown. VOO came out at 8.0% with a 12.7% drawdown and a 59% win rate, close to SPY on all three.
Sector ETFs, 6 of them, had a median of 6.1% against 12.6% for holding, a median drawdown of 27.2%, and 0 of 6 beat holding. SOXX had the widest gap in the whole table: 7.7% for the strategy and 31.1% for holding. A volatile semiconductor fund that rose in strong bursts is the hardest case for a late entry.
Leveraged ETFs, 10 of them, had the highest median CAGR of any type at 14.5%, with buy-and-hold at 23.6%. The median drawdown was 53.3%, the deepest of the positive-return types. Only 2 of 10 beat holding, and those two were UST and TMF, where holding lost money. The top of the full ranking was leveraged: QLD at 18.8%, SOXL at 18.2%, TECL at 18.0% and TQQQ at 17.3%. Their drawdowns were 32.3%, 72.2%, 62.4% and 53.3%. A high CAGR on SOXL came with a drawdown over 70% in this test, so the ranking by return and the ranking by risk differ a lot.
Bond ETFs, 7 of them, had a median CAGR of 1.2% against a buy-and-hold median of negative 0.8%, and a median drawdown of 6.5%. The rule beat holding on 6 of 7. This was the period in which long bond funds fell, and the strategy sat out much of it. TLT lost 0.85% a year with the rules and 8.2% a year held. The one bond fund that did not beat holding was SGOV, which ties it at 3.2% with a 0.05% drawdown.
Inverse ETFs, 11 of them, had a median CAGR of negative 13.1% against negative 21.6% for holding, and the rules beat holding on 8 of 11. The three that did worse than holding were TBF, EEV and SOXS. Losing less than holding is a low bar here. The median drawdown was 63.2%. The volatility products, 3 of them, lost a median 13.1% a year against 16.1% held, with a median drawdown of 57.0%. SOXS had the worst CAGR at negative 48.95% and a 97.93% drawdown, and UVXY lost 38.04% a year with a 93.96% drawdown.
Currency ETFs, 3, had a median of 1.6% and a drawdown of 7.4%. Alternative-strategy ETFs, 6, had a median of 3.1% against 7.6% held, with 0 of 6 beating holding. The single commodity fund, IAU, returned 10.8% against 13.7% held.
The pattern across types is plain. On funds that rose steadily, the strategy lagged. On funds that fell, it lost less. The type where it clearly helped was the one where the underlying asset lost money.
Limits of the test
Every number here comes from one window and one starting date. The headline run has no fees and no slippage. The cost runs on the individual ETF pages add 5 and 10 basis points, and a rule that trades 17 times over 5.74 years is only mildly exposed to them. A fund with 25 round trips pays more than one with 6.
Prices in the trade lists are adjusted for splits and dividends. The strategy decides once a day, so it does not react to a move that happens and reverses inside a session. Orders fill on minute bars. There is no margin and no shorting.
The medians mix funds with very different histories. CLSE, CTA and KMLM have data that starts after the first day of the window, so their runs are shorter. The 59-ETF median treats a $10,000 start in each fund the same way, although the funds differ widely in volatility. The fund-type split is the cleaner read.
The result is a description of what these rules did to these funds in this period. It does not say how the rules will do in the next one.
Results by fund type
| Fund type | ETFs | Median CAGR | Median buy & hold | Median max DD | Beat holding |
|---|---|---|---|---|---|
| Broad index ETFs | 12 | 8.0% | 14.4% | −20.8% | 0 of 12 |
| Sector ETFs | 6 | 6.1% | 12.6% | −27.2% | 0 of 6 |
| Leveraged ETFs | 10 | 14.5% | 23.6% | −53.3% | 2 of 10 |
| Inverse ETFs | 11 | −13.1% | −21.6% | −63.2% | 8 of 11 |
| Bond ETFs | 7 | 1.2% | −0.8% | −6.5% | 6 of 7 |
| Commodity ETFs | 1 | 10.8% | 13.7% | −23.4% | 0 of 1 |
| Currency ETFs | 3 | 1.6% | −0.9% | −7.4% | 2 of 3 |
| Volatility products | 3 | −13.1% | −16.1% | −57.0% | 3 of 3 |
| Alternative-strategy ETFs | 6 | 3.1% | 7.6% | −17.1% | 0 of 6 |
The funds where the rules worked best and worst
The one fund that stands out on risk-adjusted terms is CLSE, an alternative-strategy ETF. It returned 16.6% with a 9.3% max drawdown, a 75% win rate, and 12 round trips, and it was invested 72.5% of the time. Holding it returned 19.5%. The strategy gave up a few points of return for a drawdown under 10%. A fund that already moves in smooth uptrends gives a moving average little to get wrong.
At the other end of the ranking, many funds had a negative CAGR. The worst were SOXS at negative 48.95%, UVXY at negative 38.04%, TECS at negative 27.05%, EEV at negative 24.43%, and SQQQ at negative 21.12%. These are inverse or volatility products that decay over time, so a trend rule has little to hold on to. SOXS had a win rate of 11% and UVXY 6%.
Two funds show a wide margin over holding among the leveraged and inverse types: TMF, where holding lost 31.2% a year and the strategy lost 10.7%, and QID, where holding lost 29.5% and the strategy lost 13.1%. Both are cases where the rule cut a large loss without producing a gain. Losing less than holding on a declining asset is a smaller loss, and the page does not count it as a profit.
The funds where it lagged most in absolute terms were the ones that rose fastest: SOXX, TECL (18.0% against 38.2% held), SOXL (18.2% against 33.3% held), and ROM (14.5% against 30.2%). These are the funds where a strong uptrend arrived in a few sharp legs, and each pullback inside the trend triggered an exit followed by a late re-entry.
For the broad funds, IOO at 9.2% and VV at 8.2% had drawdowns of 22.0% and 12.6%. Both gave a modest return with a moderate drawdown, against a median drawdown of 23.0% for the whole set.
Year by year, median across all ETFs
| Year | SMA 10/50 trend | Buy & hold | ETFs with a gain |
|---|---|---|---|
| 2021 | 1.8% | 4.1% | 33 of 59 |
| 2022 | −9.3% | −12.7% | 16 of 59 |
| 2023 | 7.0% | 8.9% | 42 of 59 |
| 2024 | 1.3% | 9.7% | 33 of 59 |
| 2025 | 5.3% | 11.1% | 39 of 59 |
| 2026 | 2.7% | 3.7% | 33 of 59 |
How each calendar year played out
The table of yearly medians shows the strategy against buy-and-hold across all 59 funds.
In 2022 the median ETF lost 9.3% under the rules and 12.7% when held. Only 16 of the 59 funds finished the year with a gain. This was the year the rule was built for, and it still lost money for the median fund, because the exits came after the first leg of the decline and the re-entries came after short rallies that then failed. The loss was smaller than holding, but it was a loss.
In 2023 the median strategy return was 7.0% against 8.9% for holding, and 42 funds gained. This was one of the closer years to holding. A broad rebound that persisted for months favors a rule that only needs to be right about direction for a few weeks.
In 2024 the strategy returned a median 1.3% while holding returned 9.7%. 33 funds gained. This is one of the widest yearly gaps in the table. The market made progress through steady gains with brief shakes, and the moving averages crossed back and forth around those shakes, so the strategy was out of the market for parts of the rise.
In 2025 the median was 5.3% against 11.1% for holding, with 39 funds positive. In 2026, a partial year through 2026-10-02, the median was 2.7% against 3.7% and 33 funds gained.
In the yearly medians the strategy trailed holding in every year. The advantage in a down year, 2022, was real but small. The cost in up years was larger. That is the cost structure of a trend rule that exits on a 10 against 50 cross: it does not make a profit in a falling year, and it gives up part of every rising year.
The count of funds with a gain also shows how the breadth of the market varied. In 2022 only 16 funds gained, in 2023 42 did, and in the other years the count was between 33 and 39. A strategy that depends on trends does better in years when many funds trend together.
Changing the parameters
| Version | Median CAGR | Median max DD | Median round trips |
|---|---|---|---|
| Published rules | 2.1% | −23.0% | 17 |
| SMA 5/50 | 2.0% | −20.2% | 23 |
| SMA 20/50 | 1.0% | −28.7% | 15 |
| SMA 10/100 | 0.7% | −27.3% | 10 |
Changing the averages
The parameter table shows what happened when one of the two windows changed. The medians below are across the 59 ETFs.
The published rules, SMA 10/50, had a median CAGR of 2.1%, a median max drawdown of 23.0%, and 17 round trips.
SMA 5/50 is a faster entry signal. Its median CAGR was 2.05%, essentially the same as the published rules. The median drawdown was 20.2%, which is shallower, and the trade count rose to 23. The faster average reacted sooner to a drop, so the exit came earlier and the drawdowns were a little smaller. The extra trades did not add return in this window.
SMA 20/50 is slower on the fast leg. The median CAGR fell to 1.03%, the median drawdown rose to 28.72%, and round trips fell to 15. A slower fast line pulls the entry and exit later, and in this window it hurt both the return and the drawdown.
SMA 10/100 changes the slow leg to 100 days. The median CAGR was 0.72%, the median drawdown was 27.28%, and the trade count dropped to 10. Fewer trades did not help. A 100-day average makes the crossover signal later in both directions, so the strategy was in the wrong position for longer stretches.
The best of the four variants on median drawdown was the fastest one. The best on median CAGR was the published rule, narrowly. None of the variants was a clear improvement, and the differences in CAGR are small compared with the spread between funds, which ran from negative 48.95% to 18.76%. Moving the window by a few days does not change the story much. The choice of fund changes it far more.
These are medians over a single window of 5.74 years. A different set of years could rank the four variants differently, and the table does not show any variant that was tuned on one window and tested on another.
How it compares with the other templates
The strategy hub lists the median CAGR across the 59 ETFs for every template in the set. SMA 10/50 Trend, at 2.12%, sits in the middle of the group.
Ahead of it were the weekly 7% target at 6.61%, the monthly cycle at 5.46%, the RSI(2) snapback at 4.74%, RSI mean reversion at 2.98%, EMA 12/26 trend at 2.76%, and trend plus trailing stop at 2.61%. Behind it were the golden cross at 2.05%, the 200-day regime filter at 1.84%, the dip buyer at 1.33%, and the momentum breakout and 3-month momentum at 0%.
The ordering shows that the shorter-horizon rules did better than the longer trend rules in this window. The RSI templates and the weekly and monthly templates finished above the crossovers. The ranking does not claim that mean reversion is better in general. It reports one window, and the shorter-horizon templates did better in it.
Within the crossover family, 10/50 was above the golden cross and just under the EMA 12/26 version. The three are close enough that the median alone does not separate them. A head-to-head page exists for each pair, for example golden cross against SMA 10/50 and SMA 10/50 against the 200-day filter.
Frequently asked questions
What is the SMA 10/50 trend strategy?
A faster moving-average crossover: hold while the 10-day average is above the 50-day. A faster version of the golden cross. The 10- and 50-day averages catch intermediate trends measured in weeks rather than years. It enters recoveries earlier and exits breakdowns earlier, with more whipsaw trades in sideways markets. Useful for seeing how signal speed changes a strategy's results.
Does SMA 10/50 trend beat buy-and-hold?
Across 59 ETFs backtested 2021-01-04 to 2026-10-02, it beat same-ETF buy-and-hold on 21 of 59 (36%). Median CAGR was 2.1% with a median max drawdown of 23.0%. Per-ETF results vary widely; the table lists every one.
Why 10 and 50 days?
A common intermediate-trend pairing. It reacts within weeks and ignores single bad days. Both windows are editable parameters in DeployQuant.
Is the SMA 10/50 strategy profitable?
Across 59 ETFs from 2021-01-04 to 2026-10-02 the median CAGR was 2.1%, and 41 of the 59 had a positive CAGR. The median max drawdown was 23.0%. The result varies a lot by fund, from 18.8% on QLD to negative 48.95% on SOXS, and the headline run has no fees or slippage.
Does SMA 10/50 beat buy and hold?
It beat buy-and-hold of the same ETF in 21 of 59 cases. It beat holding on 0 of 12 broad index ETFs and 0 of 6 sector ETFs, and on 6 of 7 bond ETFs and 8 of 11 inverse ETFs. The wins came mostly where the underlying fund lost money.
What is a golden cross and how is this different?
The [golden cross](/learn/strategies/golden-cross/) uses longer averages. SMA 10/50 uses a 10-day and a 50-day average, so it reacts within weeks and makes more trades. The median was 17 round trips here. Its median CAGR was 2.12% against 2.05% for the golden cross.
Which ETF did SMA 10/50 work best on?
By CAGR the best was [QLD](/learn/strategies/sma-10-50-trend/qld/) at 18.8% with a 32.3% drawdown. By drawdown relative to return it was CLSE, with a 16.6% CAGR, a 9.3% drawdown and a 75% win rate. Both are single-fund results from one window.
What happens if I change the averages to 5/50 or 20/50?
SMA 5/50 had a median CAGR of 2.05% and a median drawdown of 20.2% with 23 round trips. SMA 20/50 had 1.03% and 28.72% with 15 round trips. SMA 10/100 had 0.72% and 27.28% with 10 round trips. None improved clearly on the published rules.
How many trades does it make?
The median ETF made 17 round trips in 5.74 years. The count ranged from 6 on SGOV to 25 on RINF and VXZ. Funds with more trades tended to have lower win rates.
Does it work on leveraged ETFs?
The leveraged ETFs had the highest median CAGR of any fund type at 14.5%, but the median drawdown was 53.3% and only 2 of 10 beat holding. [TQQQ](/learn/strategies/sma-10-50-trend/tqqq/) returned 17.3% with a 53.3% drawdown. The test has no fees, and drawdowns this deep are part of the result.
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.