DeployQuant200-Day SMA Regime Filter on TQQQ
ProShares UltraPro QQQ: 3x daily leveraged Nasdaq-100, with large trends and deep drawdowns. Backtest 2021-01-04 to 2026-10-02, $10,000 starting capital, computed by the same engine that runs live DeployQuant strategies.
The 200-day regime filter holds TQQQ while yesterday's close is above its 200-day average and sits in cash while it is below. TQQQ is a 3x daily leveraged Nasdaq-100 fund. From 2021-01-04 to 2026-10-02 the rule turned $10,000 into $35,918, a CAGR of 24.95%. Buy-and-hold of the fund made 25.35% and ended at $36,585, so the rule finished $667 behind. The drawdown is where the two separate. The rule's maximum was 36.54%, and holding fell 80.77%.
This is the best of the 59 funds for this strategy. The median CAGR for the 200-day filter across all 59 was 1.84%, and the median across leveraged funds was 12.4%, so TQQQ at 24.95% is well above both. On TQQQ itself the rule ranked fourth of 12 templates, behind RSI(2) snapback at 39.31%, weekly 7% target at 29.44% and trend plus trailing stop at 25.03%.
The rule made 14 closed round trips and held one more at the end, with a win rate of 43%. Six winners averaged 28.31% and the losers averaged 4.01%, a profit factor of 4.11. Most of the result sits in five winning trades of 17.64% or more. The window is one stretch of 5.7 years with one deep bear market, the headline run has no fees, and the result describes this window only.
Year by year
| Year | 200-day regime filter | buy & hold |
|---|---|---|
| 2021 | 17.5% | 88.9% |
| 2022 | −18.3% | −78.1% |
| 2023 | 83.3% | 187.6% |
| 2024 | 43.0% | 57.1% |
| 2025 | 17.5% | 33.9% |
| 2026 | 21.5% | 54.1% |
Calendar years against holding
Holding TQQQ gained 88.9% in 2021, lost 78.1% in 2022, then gained 187.6% in 2023, 57.1% in 2024, 33.9% in 2025 and 54.1% in 2026 so far. The rule beat that in exactly one year, 2022, and trailed in the other five. That is the shape of a trend filter on a fund that mostly rose: it pays for protection in every year the fund rises.
2021 returned 17.5% against 88.9%, a gap of 71.4 points, the second widest. The rule did not own the fund until 2021-10-19, so January to September show 0.0% in the month table. October gained 10.02%, November 5.02% and December 1.72%.
2022 is the year that defines the result. The rule lost 18.3% against 78.1%, 59.8 points ahead. The whole loss is January 2022, minus 18.29%, and the sell came on 2022-01-19 at an adjusted price of 32.11. The trade itself, from 2021-10-19 at 33.47, lost only 4.06%, which shows the position had gained more than that before the sell and given it back. After January the rule held cash through the end of the year and into 2023, while holding went on to its low on 2022-12-28.
2023 returned 83.3% against 187.6%, 104.3 points behind, the widest gap of any year. The rule re-entered on 2023-02-03 and then lost three small trades in February before buying on 2023-03-17 and holding to 2023-10-27. The rule buys only after yesterday's close is above the average, and the wait for that is part of this gap.
2024 returned 43% against 57.1%. The rule was invested from January until 2024-08-06, then re-entered on 2024-08-09. 2025 returned 17.5% against 33.9%, including a minus 13.89% February while invested and a one-day loss of 9.73% on re-entry in March. 2026 so far shows 21.5% against 54.1%, a gap of 32.6 points, helped by a 22.6% April and 32.61% May and hurt by a minus 25.76% July.
Month by month
| Year | Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2021 | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | 10.0% | 5.0% | 1.7% |
| 2022 | −18.3% | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% |
| 2023 | 0.0% | −6.3% | 13.6% | 0.1% | 22.5% | 18.2% | 10.4% | −6.5% | −15.4% | −11.0% | 32.0% | 15.8% |
| 2024 | 3.8% | 14.6% | 2.1% | −14.1% | 18.4% | 18.0% | −7.2% | −4.9% | 1.6% | −4.1% | 14.7% | −0.6% |
| 2025 | 4.5% | −13.9% | −9.5% | 0.0% | −1.0% | 15.0% | 6.2% | 1.2% | 15.7% | 12.7% | −6.4% | −3.3% |
| 2026 | 2.4% | −8.1% | −10.9% | 22.6% | 32.6% | −4.0% | −25.8% | 11.1% | 8.5% | 3.8% | – | – |
Months in cash and months that hurt
The month table has two long runs of 0.0%. The first covers January to September 2021 and the second runs from February 2022 to January 2023, 12 months. Those 12 months are the protection the rule bought: holding lost 78.1% in 2022 while the rule's equity did not move after the January sell. April 2025 also shows 0.0%, a month the rule spent in cash after selling on 2025-03-04.
The best months were May 2026 at 32.61%, November 2023 at 31.99%, April 2026 at 22.6% and May 2023 at 22.46%. The worst were July 2026 at minus 25.76%, January 2022 at minus 18.29%, September 2023 at minus 15.42% and April 2024 at minus 14.13%. Every one of the worst months came while the rule was fully invested.
Holding's best month was April 2026 at 51.68% and its worst was April 2022 at minus 36.68%. The rule avoided April 2022 entirely, and in April 2026 it entered on 2026-04-14 and so captured part of a month that holding captured in full.
September and October 2023 lost 15.42% and 10.97%, and February and March 2026 lost 8.12% and 10.95%, all while the rule was invested.
The invested months after 2023
Once the rule was in a long hold, its monthly table follows TQQQ closely and with the same size of swing. In 2024 February made 14.61%, May 18.41%, June 18.04% and November 14.7%, and April lost 14.13%, July 7.24% and August 4.89%. In the second half of 2025 June made 14.98%, September 15.65% and October 12.73%, then November lost 6.4% and December 3.26%. The year 2026 opened with January at 2.39%, then February lost 8.12% and March 10.95%, April and May made 22.6% and 32.61%, June lost 3.98% and July lost 25.76%. August, September and October 2026 then made 11.13%, 8.5% and 3.77%. A 200-day filter does not smooth any of these months, because the rule only checks the average once a day and the fund can move 25% inside a month while staying above it.
Every trade
200-day regime filter on TQQQ made 14 closed round trips and one position still open at the end of the test, an average hold of 85 days, an average winner of 28.31%, an average loser of −4.01%, a profit factor of 4.11, a longest losing streak of 4. It held a position at the close on 59.8% of trading days.
| Entry | Entry price | Exit | Exit price | Return | Days held |
|---|---|---|---|---|---|
| 2021-10-19 | $33.47 | 2022-01-19 | $32.11 | −4.1% | 92 |
| 2023-02-03 | $12.10 | 2023-02-06 | $11.92 | −1.5% | 3 |
| 2023-02-08 | $12.48 | 2023-02-09 | $12.40 | −0.6% | 1 |
| 2023-02-15 | $12.01 | 2023-02-17 | $11.48 | −4.4% | 2 |
| 2023-03-17 | $11.93 | 2023-10-27 | $15.33 | 28.5% | 224 |
| 2023-11-01 | $16.00 | 2024-08-06 | $26.70 | 66.9% | 279 |
| 2024-08-09 | $28.19 | 2024-09-09 | $28.58 | 1.4% | 31 |
| 2024-09-11 | $29.77 | 2025-02-28 | $35.02 | 17.6% | 170 |
| 2025-03-03 | $37.50 | 2025-03-04 | $33.85 | −9.7% | 1 |
| 2025-05-19 | $34.09 | 2025-05-22 | $33.74 | −1.0% | 3 |
| 2025-06-03 | $35.68 | 2026-03-04 | $48.86 | 36.9% | 274 |
| 2026-03-05 | $49.44 | 2026-03-09 | $46.37 | −6.2% | 4 |
| 2026-03-10 | $49.45 | 2026-03-13 | $47.21 | −4.5% | 3 |
| 2026-04-14 | $51.36 | 2026-07-30 | $60.89 | 18.6% | 107 |
| 2026-07-31 | $65.76 | open | – | 23.2% | – |
Prices are adjusted for splits and dividends, so they sit below the quotes printed at the time. An open position is marked at the last close.
The 15 trades
The first trade ran from 2021-10-19 to 2022-01-19 and lost 4.06%. After it the rule was out for more than a year.
The next three trades are the whipsaw cluster of February 2023. The rule bought on 2023-02-03 and sold on 2023-02-06 for a loss of 1.49%, bought on 2023-02-08 and sold the next day for 0.64%, and bought on 2023-02-15 and sold on 2023-02-17 for 4.41%. The price hovered around the average and the rule flipped. The month table shows minus 6.31% for February. The cost of the cluster was small because each stay was short.
Then came the first long hold, from 2023-03-17 at 11.93 to 2023-10-27 at 15.33, a gain of 28.5% over 224 days. The best trade began 2023-11-01 at 16.00 and ended 2024-08-06 at 26.70, a gain of 66.88% over 279 days, the longest hold of the test. After it came a small 1.38% trade in August and September 2024 and a 17.64% trade from 2024-09-11 to 2025-02-28 over 170 days.
The worst trade was 2025-03-03 to 2025-03-04, a loss of 9.73% in one day. Three more short losses came in May 2025 and March 2026, of 1.03%, 6.21% and 4.53%. Another long trade ran from 2025-06-03 to 2026-03-04 and gained 36.94% over 274 days. A trade from 2026-04-14 to 2026-07-30 gained 18.56%. The last position was opened on 2026-07-31 at 65.76 and was up 23.19% at the end.
The median trade lost 0.64% and the median hold was 31 days. The average hold was 85 days. The mean is pulled up by the long holds and the median is set by the many short losers. Five winners of 17.64% or more carry the profit, which is why the profit factor of 4.11 is high while the win rate of 43% is low.
The trade list also shows how the signal works. Whipsaws cluster after a big decline, when price crosses the 200-day line repeatedly during a bottom: February 2023, May 2025 and March 2026 are all examples.
Prices at each exit and re-entry
The price levels show what the cash periods were worth. The rule sold at $32.11 on 2022-01-19 and did not buy again until $12.10 on 2023-02-03. That gap is the 2022 protection in price terms, and it explains why holding fell 80.77% while the rule's largest 2022 loss was the 4.06% trade and the January month.
The short trades of 2025 and 2026 show the opposite side. The rule sold at $35.02 on 2025-02-28 and bought at $37.50 on 2025-03-03, higher than the exit, then sold at $33.85 the next day for the 9.73% loss. It bought at $34.09 on 2025-05-19 and sold at $33.74 on 2025-05-22, and then bought at $35.68 on 2025-06-03 for the trade that returned 36.94%. In March 2026 it sold at $48.86 on 2026-03-04, bought at $49.44 on 2026-03-05, sold at $46.37 on 2026-03-09, bought at $49.45 on 2026-03-10 and sold at $47.21 on 2026-03-13. Three trades inside ten days lost 6.21% and 4.53% in two of them, and the next entry, on 2026-04-14 at $51.36, began the 18.56% trade.
The long winners are separated by short gaps. The sale at $15.33 on 2023-10-27 was followed by a purchase at $16.00 on 2023-11-01, and the sale at $26.70 on 2024-08-06 by a purchase at $28.19 on 2024-08-09. Each re-entry paid a little more than the exit, which is the price of waiting for a close above the average. The longest winning streak was 4 trades.
Largest drawdowns
| Peak | Low point | Depth | Days to low | Recovered | Days to recover |
|---|---|---|---|---|---|
| 2024-07-10 | 2024-08-05 | −36.5% | 26 | 2025-10-02 | 423 |
| 2026-06-02 | 2026-07-29 | −33.2% | 57 | not yet | – |
| 2023-07-18 | 2023-10-26 | −32.9% | 100 | 2023-12-13 | 48 |
Buy-and-hold's deepest drawdown ran from 2021-11-19 to 2022-12-28 and reached −80.8%.
Drawdowns
The maximum drawdown was 36.54%, from 2024-07-10 to 2024-08-05, a fall of 26 days, and it took 423 days to recover, to 2025-10-02. The path back included a minus 13.89% February and a minus 9.53% March in 2025. The second drawdown, 33.21%, ran from 2026-06-02 to 2026-07-29 and had not recovered by 2026-10-02. July 2026 lost 25.76% on its own. The third was 32.87%, from 2023-07-18 to 2023-10-26, with recovery on 2023-12-13.
Holding had drawdowns of 80.77% from 2021-11-19 to 2022-12-28, 57.4% from 2024-12-16 to 2025-04-08 and 36.73% from 2025-10-29 to 2026-03-30. The rule avoided the first entirely. The rule was in cash for most of the second after selling on 2025-02-28, apart from a one-day trade on 2025-03-03 that lost 9.73%. In the third, the rule's own drawdown in June and July 2026 reached 33.21% against 36.73% for holding.
A 200-day filter removed the largest decline, and the rule still had three drawdowns above 32%.
With trading costs
The headline run fills at the bar price. These runs charge slippage on every fill.
| Slippage per fill | CAGR | Max drawdown | Final value | Sharpe |
|---|---|---|---|---|
| None (headline) | 24.9% | −36.5% | $35,918 | 0.76 |
| 5 basis points | 24.6% | −36.6% | $35,430 | 0.75 |
| 10 basis points | 24.4% | −36.6% | $34,974 | 0.74 |
Execution costs
The cost runs are small. With 5 basis points per fill the CAGR was 24.65% and the final value $35,430. With 10 basis points it was 24.37% and $34,974. The headline run ended at $35,918. The maximum drawdown stayed at 36.56% and 36.57%. The rule made 29 fills over the window. TQQQ is liquid, with $4,150,045,180 of average daily dollar volume and a median minute volume of 285,786.5 shares, and the 10 basis point run is a conservative guide for orders of this kind.
Cost runs and Sharpe ratios
The headline Sharpe ratio was 0.76, against 0.69 for holding. At 5 basis points per fill it was 0.749 and at 10 basis points 0.743. The rule kept its higher ratio over holding in all three runs, so the improvement in risk-adjusted terms does not depend on the cost assumption. The runs end at $35,918, $35,430 and $34,974 against $36,585 for holding. With 29 fills, costs reduce the CAGR by about the same amount per step, 24.95% to 24.65% to 24.37%.
Changing the parameters
| Version | CAGR | Max drawdown | Round trips | Win rate | Final value |
|---|---|---|---|---|---|
| Published rules | 24.9% | −36.5% | 14 | 43% | $35,918 |
| 100-day SMA | 17.7% | −46.9% | 30 | 40% | $25,465 |
| 150-day SMA | 20.6% | −44.5% | 13 | 46% | $29,378 |
| 250-day SMA | 25.1% | −37.1% | 7 | 43% | $36,105 |
Changing the length of the average
Three windows were tested against the published 200 days. A 250-day average returned 25.06% with a drawdown of 37.14%, a Sharpe ratio of 0.764, 7 trades and 3 wins, and ended at $36,105. That is almost the same as the published rule, with fewer trades. A 150-day average returned 20.65% with a drawdown of 44.51% and 13 trades. A 100-day average returned 17.68% with a drawdown of 46.86% and 30 trades, 12 of them winners.
CAGR rises with the length of the window across the four, from 17.68% at 100 days to 25.06% at 250, and the two shortest windows had the deepest drawdowns. The drawdown of 46.86% for the 100-day window compares with 36.54% for the 200-day. A 250-day window did as well as 200, which suggests the benefit on this fund comes from using a slow average and not from the exact value. The comparison is one fund over one window and three alternatives, so it is a description and not a tuning result.
How TQQQ behaved
| Measure | TQQQ |
|---|---|
| Data in this test | 2021-01-04 to 2026-10-02 (1444 sessions) |
| Total return, buy and hold | 291.9% |
| Annualized volatility | 67.1% |
| Deepest drawdown | −81.7% (2021-11-19 to 2022-12-28) |
| Up days | 54.5% |
| Average daily range | 4.74% |
| Average overnight gap | 1.78% |
| Correlation to SPY | 0.94 |
| Correlation to QQQ | 1.00 |
| Correlation to TLT | 0.09 |
| Sessions above the 200-day average | 69.5% |
| Crossings of the 200-day average | 28 |
| Falls of 10% or more from a 20-day high | 56 |
What the 3x structure did
TQQQ's annualized volatility was 67.11% and its beta to QQQ was 3, with a correlation of 1. Its beta to SPY was 3.86. The average up day was 3.01% and the average down day was minus 3.22%. The average intraday range was 4.74% and the average overnight gap was 1.78%. About 74.87% of the log return came overnight.
The fund returned 291.92% over the window against 151.07% for QQQ. Three times the QQQ return would be 453.22%, and a daily-reset 3x fund with no costs would have returned 568.66%. The shortfall to the simple multiple was biggest in 2025, 27.96 points, and 2024, 18.56 points. In 2023 the fund beat the simple multiple by 33.9 points, and in 2022 it lost 79.07% against minus 97.18% for the simple multiple.
The fund spent 69.48% of sessions above its 200-day average and crossed it 28 times. The rule made 15 entries in total. The rule's exposure was 59.8%.
The best day was 2025-04-09 at 35.25% and the worst was 2025-04-04 at minus 18.3%. The fund fell 10% or more from a 20-day high 56 times over 478 days. In the leveraged category the large daily moves are normal, and a rule that exits at the open after a close below the average has to accept the overnight gap. RSI(14) fell under 30 on 18 sessions, with a median forward 5-day return of 10.3% against 1.24% for all days, which fits the strong result of RSI(2) snapback on this fund and means a trend rule that sells declines can miss the rebound.
By month, May averaged 14.75%, November 12.06% and June 7.24%, against minus 5.42% for September and minus 3.82% for February. With five or six years of data these averages describe the window and not a seasonal rule.
Weekdays, RSI(2) and the two-times peer
Monday averaged 0.53% and Wednesday 0.28% for TQQQ, with Tuesday, Thursday and Friday between 0.03% and 0.06%. The lag-1 autocorrelation was -0.04. The rule trades the open after a close below the average, so a Monday gap down is one of the ways it exits, and 74.87% of TQQQ's log return came overnight. RSI(2) fell below 10 on 152 sessions and the median 5-day return afterwards was 1.73% against 1.24% for all days, and 4.54% over 20 days against 3.62%. The edge is much smaller than after the RSI(14) readings under 30, where the 5-day median was 10.3%.
QLD, the 2x fund on the same index, correlates 1 with TQQQ in the facts. The rule returned 21.57% on QLD with a drawdown of 28.75% and 12 round trips, against 24.95% and 36.54% on TQQQ. The shallower drawdown on QLD is the result of lower leverage on the same trend signal, and the two results are close enough in CAGR that the extra leverage on TQQQ added return of a few points at the cost of a deeper fall. That comparison covers one window in which the index trended up for most of the time.
The rules
Own the asset when price closes above its 200-day average; hold cash when it closes below.
- 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.
The rule against the other leveraged funds and templates
The same rule on other leveraged funds returned 23.44% on SOXL with a 76.74% drawdown, 21.57% on QLD with 28.75%, 17.99% on ROM, 12.4% on SSO, 11.53% on TECL, 10.54% on SPUU, and negative results on FAS at minus 2.74%, UST at minus 6.68% and TMF at minus 10.5%, which made 31, 34 and 17 round trips. Among the leveraged equity funds only QLD at 28.75% and SSO at 32.46% had a shallower drawdown than TQQQ's 36.54%.
Among templates on TQQQ, monthly cycle made 21.49% but fell 82.48%, SMA 10/50 made 17.3% and EMA 12/26 made 15.25%, and golden cross made 7.33% with a 57.35% drawdown. The 200-day filter had the shallowest drawdown among the top five templates, and SMA 10/50, EMA 12/26 and golden cross did worse on return. The strategy page shows the rule on every fund and the TQQQ page shows every template.
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.
Frequently asked questions
Did 200-day regime filter beat buy-and-hold on TQQQ?
Over 2021-01-04 to 2026-10-02, 200-day regime filter on TQQQ returned 24.9% annualized vs 25.4% for buy-and-hold: it trailed buy-and-hold by 0.4% per year, with a maximum drawdown 44.2 points shallower than holding (36.5% vs 80.8%).
How many trades did it make?
14 completed round trips over 5.7 years (29 fills), with 43% of round trips closing profitably.
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.
How did the 200-day regime filter do on TQQQ?
It returned 24.95% a year and turned $10,000 into $35,918, against 25.35% and $36,585 for buy-and-hold. The maximum drawdown was 36.54% for the rule and 80.77% for holding.
Did the 200-day filter avoid the 2022 TQQQ crash?
Yes. It sold on 2022-01-19 and held cash until 2023-02-03. The year returned minus 18.3% against minus 78.1% for holding, and all of the rule's loss came in January 2022.
How many trades did the rule make on TQQQ?
14 closed round trips and one open position, 29 fills in all. Six of the 14 won, a win rate of 43%. Five winners of 17.64% or more produced nearly all the gain.
Is a 200-day average the right length for TQQQ?
A 250-day average gave nearly the same result at 25.06% and 7 trades. A 150-day average gave 20.65% and a 100-day average 17.68%, with deeper drawdowns of 44.51% and 46.86%. The 100-day average made 30 trades and did worst in this window.
Why did the rule trail buy-and-hold in 2023?
It returned 83.3% against 187.6%, because it waited for the fund to climb above the average and bought on 2023-02-03. Three small losing trades in February cost 6.31% and the main hold began on 2023-03-17.
Do trading costs matter for the 200-day filter on TQQQ?
Not much. Slippage of 10 basis points per fill moved the CAGR from 24.95% to 24.37% and the final value from $35,918 to $34,974.
Related
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.