DeployQuant200-Day SMA Regime Filter
Own the asset when price closes above its 200-day average; hold cash when it closes below.
The 200-day regime filter has two rules and one parameter. It buys at the open when yesterday's close was above the 200-day simple moving average, and it sells the whole position at the open when yesterday's close was below it. Nothing else decides anything: no stop, no target, no second indicator. We ran it on 59 ETFs from 2021-01-04 to 2026-10-02, each with $10,000, no margin, and no fees or slippage in the headline run.
The median result across the 59 funds was a CAGR of 1.84% and a maximum drawdown of 22.16%. The median Sharpe ratio was 0.36, the median fund made 18 round trips, and the median time invested was 57.8%. It beat buy-and-hold of the same fund on 26 of 59 ETFs. It had a shallower drawdown than buy-and-hold on 50 of 59, and 37 of the 59 ended with a positive CAGR.
Those figures describe a rule that sits out part of the time, and the pattern shows in every table below. Where holding lost money, the filter often lost less. Where holding made the most, the filter gave up a large share of it. The per-fund pages carry the full trade lists. For the other end of the spectrum, see RSI mean reversion, which buys weakness instead of strength, and the golden cross, which also uses the 200-day line but only with a second average.
The rules
- WHEN the market opens · IF not invested AND yesterday's close > SMA(200) · THEN buy with 98% of the sleeve
- WHEN the market opens · IF invested AND yesterday's close < SMA(200) · THEN sell the whole position
One rule and one number. Price above the 200-day moving average has historically coincided with better returns and lower volatility than price below it. This template uses no crossovers and no oscillators, only which side of the long-term average the price is on.
Good for: a first systematic strategy, simple enough to audit every trade.
Watch out: price whips around the 200-day line during volatile bottoms, generating clusters of buy-sell pairs. Some traders add a small buffer band to reduce churn.
What one moving average does to a position
A 200-day average moves slowly. Each day it adds the newest close and drops the close from 200 days earlier, so it only changes direction after a long run of closes on one side. Price crossing it is therefore a fairly rare event for a calm fund, and a frequent one for a volatile fund that is chopping around the line.
The rules here act on the day after the cross. The check uses yesterday's close, and the order goes in at the next open. A fund that closes just under the average on a Tuesday is sold on Wednesday morning, even if it recovers during Wednesday. The mirror image holds on the way up. This one-day lag is built into every trade in the tables, and it is the reason a rule this simple still produces whipsaws: the exit price can be several percent away from the close that triggered it.
The buy side sizes the position at 98% of the sleeve. The remaining 2% stays in cash as a buffer for rounding. Because the test starts each fund with $10,000 and holds no margin, the cash left after a sale sits idle. The headline runs do not credit that cash with interest. That matters for the bond and currency results below, where idle cash is a large part of the comparison.
The median time invested was 57.8%. The filter was out of the market for roughly two fifths of the window for the median fund, and that is the source of most of the shallower drawdowns. A fund cannot fall while it is not held. The same fact explains most of the missed gains.
The strategy page for the monthly cycle uses a calendar instead of a price filter, and its median CAGR across the same universe was 5.46%. The 200-day filter has no calendar element. It trades only when the price crosses the line, which is why trade counts differ so much between funds: 5 round trips on SGOV, 48 on TLT.
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 |
|---|---|---|---|---|---|---|
| TQQQ | 24.9% | 25.4% | −36.5% | 0.76 | 14 | 43% (+1 open) |
| SOXL | 23.4% | 33.3% | −76.7% | 0.66 | 22 | 36% (+1 open) |
| QLD | 21.6% | 23.6% | −28.8% | 0.84 | 12 | 42% (+1 open) |
| CLSE | 20.4% | 19.5% | −8.7% | 1.81 | 8 | 50% (+1 open) |
| SOXX | 19.2% | 31.1% | −39.2% | 0.77 | 19 | 26% (+1 open) |
| ROM | 18.0% | 30.2% | −42.2% | 0.66 | 18 | 33% (+1 open) |
| XLK | 16.8% | 22.1% | −19.7% | 0.98 | 11 | 36% (+1 open) |
| QQQM | 13.0% | 16.7% | −19.8% | 0.93 | 10 | 20% (+1 open) |
| IAU | 13.0% | 13.7% | −20.0% | 0.87 | 22 | 23% |
| QQQ | 12.9% | 16.7% | −19.9% | 0.92 | 10 | 20% (+1 open) |
| SSO | 12.4% | 21.9% | −32.5% | 0.70 | 18 | 33% (+1 open) |
| TECL | 11.5% | 38.2% | −64.5% | 0.47 | 20 | 20% (+1 open) |
| IOO | 11.2% | 16.7% | −16.5% | 1.00 | 15 | 33% (+1 open) |
| VOOG | 11.1% | 15.7% | −21.7% | 0.88 | 16 | 19% (+1 open) |
| SPUU | 10.5% | 22.5% | −37.3% | 0.61 | 17 | 24% (+1 open) |
| VOO | 8.4% | 14.4% | −17.9% | 0.84 | 15 | 40% (+1 open) |
| SPY | 8.3% | 14.6% | −18.0% | 0.84 | 15 | 40% (+1 open) |
| VV | 7.9% | 14.3% | −19.9% | 0.79 | 13 | 31% (+1 open) |
| EEM | 7.8% | 6.6% | −17.7% | 0.60 | 15 | 27% (+1 open) |
| VOX | 6.5% | 9.1% | −20.1% | 0.55 | 17 | 24% |
| QQQE | 5.7% | 9.8% | −15.3% | 0.51 | 15 | 33% (+1 open) |
| IWM | 4.7% | 7.5% | −22.2% | 0.40 | 17 | 41% (+1 open) |
| VOOV | 4.3% | 12.6% | −18.4% | 0.48 | 22 | 23% (+1 open) |
| TBF | 4.3% | 12.2% | −24.0% | 0.39 | 27 | 37% (+1 open) |
| VTV | 4.3% | 13.4% | −23.3% | 0.47 | 21 | 29% (+1 open) |
| QAI | 4.3% | 3.9% | −6.0% | 0.94 | 7 | 29% (+1 open) |
| SGOV | 3.2% | 3.2% | −0.0% | 14.13 | 5 | 40% (+1 open) |
| USDU | 2.7% | 5.2% | −11.2% | 0.54 | 17 | 29% (+1 open) |
| XLF | 2.1% | 12.6% | −22.0% | 0.23 | 19 | 21% (+1 open) |
| ALTY | 1.8% | 7.6% | −12.1% | 0.31 | 15 | 20% |
| IGIB | 1.4% | 0.2% | −5.4% | 0.35 | 26 | 38% |
| FXE | 1.3% | −0.9% | −6.4% | 0.29 | 17 | 29% |
| UDN | 1.3% | −1.1% | −5.4% | 0.30 | 20 | 50% |
| IEI | 1.0% | −0.4% | −4.6% | 0.36 | 21 | 33% |
| BND | 0.4% | −0.8% | −6.5% | 0.12 | 21 | 29% |
| AGG | 0.3% | −0.8% | −6.2% | 0.10 | 25 | 32% |
| KMLM | 0.2% | 7.1% | −28.6% | 0.08 | 21 | 14% (+1 open) |
| XLP | −0.8% | 5.8% | −21.4% | -0.03 | 37 | 30% |
| IEF | −1.5% | −2.3% | −11.5% | -0.33 | 30 | 20% |
| RINF | −2.1% | 6.6% | −21.4% | -0.16 | 37 | 32% (+1 open) |
| XLY | −2.2% | 6.3% | −27.3% | -0.09 | 30 | 23% |
| PSQ | −2.3% | −13.8% | −20.1% | -0.08 | 15 | 20% |
| FAS | −2.7% | 18.3% | −54.8% | 0.08 | 31 | 29% |
| CTA | −3.0% | 8.9% | −22.8% | -0.15 | 25 | 24% (+1 open) |
| SPDN | −4.5% | −9.7% | −29.6% | -0.34 | 24 | 17% |
| SH | −4.9% | −10.0% | −31.2% | -0.38 | 25 | 20% |
| EEV | −5.6% | −16.1% | −59.4% | -0.13 | 22 | 14% |
| UST | −6.7% | −8.5% | −35.6% | -0.91 | 34 | 21% |
| SDS | −7.5% | −21.6% | −44.4% | -0.26 | 17 | 0% |
| TLT | −7.6% | −8.2% | −40.2% | -1.04 | 48 | 15% |
| QID | −10.5% | −29.5% | −55.1% | -0.24 | 17 | 18% |
| TMF | −10.5% | −31.2% | −53.2% | -0.70 | 17 | 0% |
| SQQQ | −10.5% | −42.3% | −63.8% | -0.06 | 13 | 31% |
| VXZ | −11.4% | −15.1% | −53.5% | -0.60 | 31 | 26% |
| VIXM | −12.2% | −16.1% | −53.6% | -0.60 | 30 | 17% |
| REW | −13.0% | −36.1% | −61.3% | -0.28 | 19 | 16% |
| TECS | −13.9% | −46.7% | −62.4% | -0.15 | 17 | 18% |
| SOXS | −19.7% | −48.3% | −72.5% | -0.22 | 14 | 0% |
| UVXY | −29.8% | −48.7% | −86.9% | -0.91 | 12 | 0% |
Where the filter helped and where it cost the most
The top of the table is dominated by leveraged and technology funds, because those had the highest buy-and-hold returns in the window. TQQQ returned 24.9% a year with the filter against 25.4% for holding, with a maximum drawdown of 36.5% and 14 round trips. That is a close result on CAGR with a much shallower drawdown than a leveraged fund usually shows. SOXL returned 23.4% against 33.3% for holding, but its maximum drawdown with the filter was still 76.7%, so the filter did not prevent a very deep fall in that fund. QLD returned 21.6% against 23.6% with a 28.8% drawdown.
CLSE is the one fund near the top where the filter ended ahead of holding: 20.4% against 19.5%, with a maximum drawdown of 8.7%. It made only 8 round trips and was invested 74.9% of the time. The sample is small, and the lower drawdown comes with only a slight CAGR edge, so this reads as one fund that trended cleanly rather than a general result.
The worst gap to holding is on TECL, at 11.5% a year against 38.2% for holding. The filter made 20 round trips on TECL with a 20% win rate, and its maximum drawdown was 64.5%. A fund that moves three times as much as its index in a day crosses its own 200-day line often, and each cross costs a gap at the next open. The TECL page for this strategy lists the trades.
SOXX and ROM show a similar pattern at lower leverage. SOXX returned 19.2% against 31.1% for holding, and ROM returned 18.0% against 30.2%. Both gave up a large share of a strong uptrend by sitting out the pullbacks, and in both cases the pullbacks recovered quickly.
The broad index funds sit in the middle. QQQ returned 12.9% against 16.7% with 10 round trips and a 20% win rate, and QQQM, which tracks the same index, returned 13.0% against 16.7%. SPY returned 8.3% against 14.6% and VOO returned 8.4% against 14.4%. The maximum drawdowns with the filter were 19.9% for QQQ and 18.0% for SPY.
The bottom of the table is almost entirely inverse, volatility, and leveraged bond funds. UVXY lost 29.8% a year with the filter, against a loss of 48.7% for holding. REW lost 13.0% against 36.1%. TLT lost 7.6% against 8.2% and made 48 round trips, the most of any fund. A fund in a persistent downtrend rarely closes above its 200-day average for long, so the filter spends most of its time in cash. The few entries it makes tend to be short-lived rallies that reverse, which is why the win rates on these funds run from 0% to 31%.
Results by fund type
| Fund type | ETFs | Median CAGR | Median buy & hold | Median max DD | Beat holding |
|---|---|---|---|---|---|
| Broad index ETFs | 12 | 8.3% | 14.4% | −19.8% | 1 of 12 |
| Sector ETFs | 6 | 6.5% | 12.6% | −22.0% | 0 of 6 |
| Leveraged ETFs | 10 | 12.4% | 23.6% | −42.2% | 2 of 10 |
| Inverse ETFs | 11 | −7.5% | −21.6% | −55.1% | 10 of 11 |
| Bond ETFs | 7 | 0.4% | −0.8% | −6.2% | 6 of 7 |
| Commodity ETFs | 1 | 13.0% | 13.7% | −20.0% | 0 of 1 |
| Currency ETFs | 3 | 1.3% | −0.9% | −6.4% | 2 of 3 |
| Volatility products | 3 | −12.2% | −16.1% | −53.6% | 3 of 3 |
| Alternative-strategy ETFs | 6 | 1.8% | 7.6% | −21.4% | 2 of 6 |
Results by fund type
The fund-type table splits the 59 ETFs into nine groups. The groups with positive buy-and-hold returns were the ones where the filter had the hardest job.
Broad index funds: 12 funds, a median CAGR of 8.3% against 14.4% for holding, and a median maximum drawdown of 19.8%. The filter beat holding on 1 of 12. Sector funds: 6 funds, a median of 6.5% against 12.6%, and the filter beat holding on 0 of 6. Commodity funds had one member, IAU, at 13.0% against 13.7%. For these groups the filter reduced return and left drawdown roughly where it was. Over a window that began in January 2021 and ended in October 2026, most of the equity index funds spent long stretches above their 200-day averages, and the filter was out of the market only for the shorter declines.
Leveraged funds were a different case. The median CAGR was 12.4% against 23.6% for holding, but the median maximum drawdown with the filter was 42.2%. The filter beat holding on 2 of 10. Those leveraged funds are the funds where a rule like this would be expected to matter most, because the drawdowns without a filter are the deepest, and the table shows the drawdown reduction came at the cost of a large share of the return.
The groups where the filter looks best are the ones where holding lost money. Inverse funds: 11 funds, a median CAGR of negative 7.5% against negative 21.6% for holding, and the filter beat holding on 10 of 11. Volatility products: 3 funds, a median of negative 12.2% against negative 16.1%, and the filter beat holding on 3 of 3. Bond funds: 7 funds, a median of 0.4% against negative 0.8%, with a median maximum drawdown of 6.2%, and the filter beat holding on 6 of 7. Currency funds: 3 funds, 1.3% against negative 0.9%, and 2 of 3 ahead.
These are relative wins. A fund that loses 21.6% a year and a filter that loses 7.5% are both losses. The inverse funds also decay through daily resets, which the SPDN page covers, and a fund with a built-in negative drift gives a trend filter an easy comparison. The positive reading is limited to the statement that the filter shortened the time spent in funds that were falling.
Alternative-strategy funds: 6 funds, a median of 1.8% against 7.6%, with 2 of 6 ahead, CLSE and QAI. The group also includes KMLM at 0.2% against 7.1% and CTA at negative 3.0% against 8.9%, where the filter lagged holding. CLSE, CTA and KMLM have data that starts after the first day of the window, so their runs are shorter.
Year by year, median across all ETFs
| Year | 200-day regime filter | Buy & hold | ETFs with a gain |
|---|---|---|---|
| 2021 | 0.0% | 4.1% | 25 of 59 |
| 2022 | −12.4% | −12.7% | 7 of 59 |
| 2023 | 2.8% | 8.9% | 30 of 59 |
| 2024 | 1.5% | 9.7% | 34 of 59 |
| 2025 | 4.3% | 11.1% | 34 of 59 |
| 2026 | 1.5% | 3.7% | 30 of 59 |
Year by year across the 59 funds
The year table reports the median across all 59 funds for each calendar year, along with the number of funds that finished the year with a gain.
In 2021 the median filter result was 0.0%. 25 of 59 funds had a gain, and the median buy-and-hold result was positive.
2022 was the one year where the filter's median came close to protecting capital. The median filter result was negative 12.4% against negative 12.7% for holding, and only 7 of 59 funds finished with a gain. The filter was a little better than holding at the median, but a loss of 12.4% in a year is not much protection. The one-day lag and the slow average are the likely reasons: the 200-day line turns slowly, so exits come after part of a decline has happened. The data here does not show the exact entry dates inside the year, so this explanation is a reading of how the rule works.
The next three years show the cost of that protection. In 2023 the median filter result was 2.8% against 8.9% for holding, with 30 of 59 funds up. In 2024 it was 1.5% against 9.7%, with 34 of 59 up. In 2025 it was 4.3% against 11.1%, with 34 of 59 up. The filter stayed in the black each year but captured only part of the market's gain. The reason is mostly the time spent in cash early in each recovery. After the 2022 decline, price had to rise back above the 200-day line before the filter bought, and by then a good part of the rebound had passed.
2026 runs through 2026-10-02 only. The median filter result was 1.5% against 3.7% for holding, with 30 of 59 up. This is a partial year, so the figures are year to date.
Over the five full-ish years, the filter was ahead of holding at the median in 2022 only. It trailed in 2023, 2024, 2025, and 2026. That tells the same story as the fund-type table: the rule gave up return in uptrends and recovered a small part of it in the one down year in the window. One bear year is a short record. The test cannot say how the rule would behave through a longer bear market, a flash crash that reverses inside a day, or a stretch of sideways price action with many crosses.
Changing the parameters
| Version | Median CAGR | Median max DD | Median round trips |
|---|---|---|---|
| Published rules | 1.8% | −22.2% | 18 |
| 100-day SMA | 1.7% | −24.3% | 30 |
| 150-day SMA | 1.4% | −24.8% | 24 |
| 250-day SMA | 1.6% | −24.6% | 16 |
What changing the window did
The published rule uses 200 days. We also ran 100, 150, and 250 days across the same 59 funds to see whether the choice of 200 was doing any work.
The published rule had a median CAGR of 1.8%, a median maximum drawdown of 22.2%, and a median of 18 round trips. The 100-day version had a median CAGR of 1.7%, a median drawdown of 24.3%, and a median of 30 round trips. The 150-day version came in at 1.4% with a drawdown of 24.8% and 24 round trips. The 250-day version had 1.6%, a drawdown of 24.6%, and 16 round trips.
The differences in median CAGR are small, between 1.4% and 1.8%. The median drawdown is shallowest with the published window, 22.2% against 24.3% to 24.8% for the other three. The trade count falls steadily as the window lengthens, from 30 round trips at 100 days to 16 at 250. That gives a practical reading: a shorter average reacts faster and trades more without improving the median result, and a longer average trades less without improving it either.
The 200-day setting came out best on both CAGR and drawdown at the median, but the gap is small enough that we do not read it as evidence that 200 is special. The numbers are medians over 59 funds with very different behaviour, and a different window can win on a single fund. For example, on a leveraged fund the faster 100-day line might have cut a drawdown that the 200-day line did not. The per-fund pages carry the variant results for each case.
The parameter test covers the window length only. We did not vary the entry size, add a buffer band around the line, or add a confirmation period. The caveat on the rule list mentions a buffer band as something some traders add to reduce churn. This test does not measure it. The comparison with SMA 10/50 trend shows a rule with a faster signal and a median CAGR of 2.12%, and the EMA 12/26 trend had 2.76%. Both of those use faster signals than a 200-day line.
How it compares with the other templates
The facts for this page include the median CAGR of the other eleven templates on the same 59 funds. The 200-day filter's median was 1.84%.
The highest medians belong to the weekly 7% target at 6.61%, the monthly cycle at 5.46%, and RSI(2) snapback at 4.74%. RSI mean reversion followed at 2.98%. Among the trend rules, EMA 12/26 trend had 2.76%, trend with a trailing stop had 2.61%, SMA 10/50 had 2.12%, and golden cross had 2.05%. The 200-day filter at 1.84% was below all of those. The dip buyer had 1.33%, and the momentum breakout and the 3-month momentum template had 0%.
A rule that runs less often tends to be less exposed to costs, but the headline run has no fees or slippage, so cost does not explain the gaps. The ranking more likely reflects the window. From 2021 to October 2026, the market spent much of its time rising with short, sharp drops. Rules that bought weakness or took a quick profit did better in that kind of tape than a rule that waited for a long average to confirm direction.
That statement is limited to this window and these funds. The head-to-head pages for each pair show the full comparison, for example the 200-day filter against golden cross and against the weekly 7% target.
Frequently asked questions
What is the 200-day regime filter strategy?
Own the asset when price closes above its 200-day average; hold cash when it closes below. One rule and one number. Price above the 200-day moving average has historically coincided with better returns and lower volatility than price below it. This template uses no crossovers and no oscillators, only which side of the long-term average the price is on.
Does 200-day regime filter beat buy-and-hold?
Across 59 ETFs backtested 2021-01-04 to 2026-10-02, it beat same-ETF buy-and-hold on 26 of 59 (44%). Median CAGR was 1.8% with a median max drawdown of 22.2%. Per-ETF results vary widely; the table lists every one.
Why the 200-day average specifically?
It approximates a year of trading days and has been studied across decades of data. It is not the best window for every asset. The per-ETF backtests here show where it helped and where it didn't.
Does the 200-day moving average strategy beat buy-and-hold?
Across the 59 ETFs tested from 2021-01-04 to 2026-10-02, it beat buy-and-hold of the same fund on 26 of 59. The median CAGR was 1.84%. It beat holding most often on inverse, volatility, and bond funds, where holding lost money, and least often on broad index and sector funds, where it beat holding on 1 of 12 and 0 of 6.
How many trades does a 200-day SMA filter make?
The median fund made 18 round trips in the window of 5.74 years. The count ran from 5 on SGOV to 48 on TLT. Funds that trend cleanly, such as CLSE with 8 round trips, trade rarely, and funds that move around the line trade often.
Does the 200-day filter reduce drawdowns?
In this test it did at the median: the maximum drawdown was 22.16%, and the filter had a shallower drawdown than holding on 50 of 59 funds. The reduction was uneven. SOXL still had a maximum drawdown of 76.7% with the filter, and TECL had 64.5%.
Is 200 days the best window for the moving average?
Of the four windows tested, 200 days had the highest median CAGR at 1.8% and the shallowest median drawdown at 22.2%. The 100, 150, and 250 day versions had median CAGRs from 1.4% to 1.7%. The gaps are small, and the result is a median over many different funds.
What happened to the 200-day filter in 2022?
The median filter result was negative 12.4% against negative 12.7% for buy-and-hold, and 7 of 59 funds finished the year with a gain. The filter lost slightly less at the median. It did not avoid the decline, because the average turns slowly and entries came during bounces.
Does the filter work on leveraged ETFs?
It cut the drawdown on some and gave up a large share of the return on most. TQQQ returned 24.9% a year with the filter against 25.4% for holding. TECL returned 11.5% against 38.2%. The filter beat holding on 2 of 10 leveraged funds.
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.