DeployQuantRSI(2) Dip Snapback on QQQ
Invesco QQQ Trust: tracks the Nasdaq-100, a tech-heavy growth index. Backtest 2021-01-04 to 2026-10-02, $10,000 starting capital, computed by the same engine that runs live DeployQuant strategies.
The RSI(2) Dip Snapback bought QQQ 164 times between 2021-01-04 and 2026-10-02 and made 14.2% a year, which turned $10,000 into $21,396. Buying and holding the same fund made 16.7% a year and ended at $24,265. The strategy therefore finished $2,869 behind on return. It also took a maximum drawdown of 14.3% where holding took 34.2%, and its Sharpe ratio was 0.98 against 0.83.
That is the trade the template makes on a fund like this: a lower return for a much smaller loss in the bad stretch. The strategy was invested 36.1% of the time. Of its 164 round trips, 113 won, a win rate of 69%, and the average trade held for 4.7 days. The wins averaged 1.53% and the losses averaged 1.78%, so the high win rate does the work and the loss size is the weak point. Of the 12 templates run on QQQ it ranked third on return, and QQQ ranked 13th of 59 funds for this template.
The headline run has no fees or slippage. A strategy with 328 fills in under six years is sensitive to costs, and the cost runs below show that a large part of the return is gone at 10 basis points. The RSI(2) strategy page has the same rules on all 59 funds, and the QQQ page compares all 12 templates on this fund.
Year by year
| Year | RSI(2) snapback | buy & hold |
|---|---|---|
| 2021 | 26.7% | 28.4% |
| 2022 | −5.7% | −31.7% |
| 2023 | 13.8% | 53.0% |
| 2024 | 15.2% | 25.1% |
| 2025 | 18.8% | 20.4% |
| 2026 | 15.0% | 22.1% |
How each calendar year played out
The strategy made 26.7% in 2021 against 28.4% for holding. It was close to the market in a year with few pullbacks, and it won 21 of its 26 round trips closed that year. The small gap is the cost of being out of the fund for most days in a year when QQQ rose steadily.
2022 is the year that matters. Holding lost 31.7%. The strategy lost 5.7%, which beat holding by 26 points on the yearly table. Only 14 of its 31 round trips closed in 2022 won, the lowest win rate of any year, and the January 2022 month alone lost 6.5%. The rule kept buying washouts as QQQ fell, and each losing trade was small enough that the damage stayed limited. This is the only calendar year in which the strategy finished ahead of holding.
In 2023 holding made 53.0% and the strategy made 13.8%. That was the largest gap in the six years, 39.2 points. QQQ rose in a sharp trend with few washouts, so RSI(2) rarely dropped below 10 and the strategy sat in cash. It closed 29 round trips and won 20. In 2024 the strategy made 15.2% against 25.1%, with 30 round trips and 21 wins. In 2025 it made 18.8% against 20.4%, with 24 round trips and 19 wins. In the partial year 2026 it made 15.0% against 22.1%, with 24 round trips and 17 wins.
The pattern across the years is steady. In every year except 2022 the strategy trailed holding, and the shortfall was largest when the fund trended hardest. The strategy does not add return in an up-trending year, and it can protect capital in a down-trending one. The tables of the other templates on QQQ, such as the weekly 7% target at 16.5% and the monthly cycle at 14.7%, show the same shape from different rules.
Month by month
| Year | Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2021 | 1.8% | 1.6% | 4.9% | 1.7% | 3.5% | 0.8% | 1.4% | 3.7% | −1.1% | 3.5% | 0.6% | 1.6% |
| 2022 | −6.5% | 0.9% | 2.2% | −1.3% | −2.9% | −5.7% | 7.9% | −1.0% | −4.5% | 7.0% | 2.7% | −3.3% |
| 2023 | 4.8% | 0.3% | 4.6% | 1.4% | 1.5% | 1.6% | 2.1% | −1.6% | −2.4% | −2.0% | 2.5% | 0.5% |
| 2024 | −0.0% | 6.6% | 2.0% | −0.8% | 3.5% | 0.8% | −1.9% | 5.5% | −0.8% | 0.8% | −1.4% | 0.5% |
| 2025 | 5.6% | −3.8% | −5.5% | 12.8% | 3.5% | 0.9% | 0.8% | −0.8% | 3.5% | 3.9% | −2.0% | −0.3% |
| 2026 | 0.2% | 0.1% | 0.6% | 2.7% | 1.4% | 5.3% | 0.6% | 0.7% | 2.7% | 0.0% | – | – |
Which months carried the result
The monthly table shows how small the typical month is. Most months fall between negative 3% and 5%. The best month was April 2025 at 12.81%, which came right after the fund's worst days, 2025-04-03 and 2025-04-04. The strategy bought into the selloff on 2025-04-07 and sold on 2025-04-10 for a 11.45% gain on the trade, and it was in the position for 2025-04-09, the fund's best day at 11.75%. That one trade is the largest in the whole backtest.
The worst month was January 2022 at negative 6.5%, followed by June 2022 at negative 5.72%, March 2025 at negative 5.54% and September 2022 at negative 4.51%. The five months below negative 3.5% all fell in 2022 or early 2025, which are the two periods of sustained decline in the window.
The months with large gains follow washouts. July 2022 made 7.94% and October 2022 made 7.03%, both right after the lows of June and September. February 2024 made 6.56% and January 2025 made 5.62%. These are the months where the rule's logic worked as designed: it bought a sharp drop and sold the rebound within days.
The longest run of losing months was in late 2022 and 2023, with September 2022 and December 2022 negative and August to October 2023 negative three months running. The strategy's losing streak in trades never went beyond 3, and its longest winning streak was 8.
A single month says little on its own. April 2025 and the two rebound months of 2022 sit well above the rest of the table, so a few months carry a visible share of the total.
Every trade
RSI(2) snapback on QQQ made 164 closed round trips, an average hold of 5 days, an average winner of 1.53%, an average loser of −1.78%, a profit factor of 1.95, a longest losing streak of 3. It held a position at the close on 36.1% of trading days.
Best 10 round trips
| Entry | Entry price | Exit | Exit price | Return | Days held |
|---|---|---|---|---|---|
| 2025-04-07 | $405.22 | 2025-04-10 | $451.60 | 11.4% | 3 |
| 2024-08-05 | $419.79 | 2024-08-09 | $441.95 | 5.3% | 4 |
| 2022-03-14 | $313.82 | 2022-03-17 | $327.94 | 4.5% | 3 |
| 2023-03-02 | $282.72 | 2023-03-06 | $294.91 | 4.3% | 4 |
| 2021-05-19 | $306.34 | 2021-05-21 | $319.42 | 4.3% | 2 |
| 2022-07-26 | $290.59 | 2022-07-29 | $302.83 | 4.2% | 3 |
| 2022-11-29 | $276.17 | 2022-12-01 | $286.33 | 3.7% | 2 |
| 2024-02-21 | $418.29 | 2024-02-23 | $433.57 | 3.6% | 2 |
| 2022-10-03 | $263.00 | 2022-10-05 | $272.15 | 3.5% | 2 |
| 2025-10-23 | $602.26 | 2025-10-27 | $622.31 | 3.3% | 4 |
Worst 10 round trips
| Entry | Entry price | Exit | Exit price | Return | Days held |
|---|---|---|---|---|---|
| 2025-02-21 | $533.20 | 2025-03-06 | $488.74 | −8.3% | 13 |
| 2022-06-10 | $285.32 | 2022-06-16 | $267.52 | −6.2% | 6 |
| 2022-01-20 | $359.14 | 2022-01-31 | $342.83 | −4.5% | 11 |
| 2022-09-19 | $279.84 | 2022-09-29 | $269.59 | −3.7% | 10 |
| 2022-08-18 | $319.74 | 2022-08-25 | $308.31 | −3.6% | 7 |
| 2022-04-22 | $325.04 | 2022-04-29 | $314.14 | −3.4% | 7 |
| 2022-08-30 | $297.88 | 2022-09-08 | $288.10 | −3.3% | 9 |
| 2021-12-30 | $389.50 | 2022-01-12 | $376.73 | −3.3% | 13 |
| 2023-09-19 | $362.74 | 2023-09-26 | $351.23 | −3.2% | 7 |
| 2021-04-29 | $331.21 | 2021-05-10 | $321.03 | −3.1% | 11 |
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.
What the 164 trades looked like
The median trade returned 0.64% and held for 4 days. The shortest hold was 1 day and the longest was 19. The first trade ran from 2021-01-13 to 2021-01-14 for 0.58%, and the last, from 2026-09-25 to 2026-09-28, lost 0.35%. No position was open at the end of the test.
The five best trades were all short, between 2 and 4 days. The best was 2025-04-07 to 2025-04-10 for 11.45%, with entry at an adjusted price of 405.22 and exit at 451.60. The second was 2024-08-05 to 2024-08-09 for 5.28%. The third was 2022-03-14 to 2022-03-17 for 4.50%, the fourth was 2023-03-02 to 2023-03-06 for 4.31% and the fifth was 2021-05-19 to 2021-05-21 for 4.27%.
The worst trades held longer. The worst was 2025-02-21 to 2025-03-06 for negative 8.34% over 13 days, entered at 533.20 and exited at 488.74. The second was 2022-06-10 to 2022-06-16 for negative 6.24%, the third 2022-01-20 to 2022-01-31 for negative 4.54% over 11 days, and the fourth and fifth were 2022-09-19 and 2022-08-18 entries at negative 3.66% and negative 3.57%. All five worst trades fell within 2022 or the February 2025 slide.
The shape is typical for this rule. Winners are short because the exit fires as soon as RSI(2) rises above 70, which often happens within a few sessions. Losers are long because a trade that does not bounce stays open until the oscillator recovers, and in a fall that can take two weeks. The loser is therefore larger than the winner on average, and the strategy depends on winning about seven trades in ten. The profit factor was 1.95.
The worst trade, 8.34%, is a small fraction of the fund's 34.2% drawdown. No single trade did serious damage, since the rule only ever holds one position of this size, and the exit is tied to the oscillator and not to a price stop. A trade that kept falling would stay open until RSI(2) recovered.
Largest drawdowns
| Peak | Low point | Depth | Days to low | Recovered | Days to recover |
|---|---|---|---|---|---|
| 2021-12-15 | 2022-06-30 | −14.3% | 197 | 2023-02-13 | 228 |
| 2025-02-05 | 2025-04-02 | −11.0% | 56 | 2025-04-09 | 7 |
| 2023-08-07 | 2023-10-20 | −7.6% | 74 | 2024-02-22 | 125 |
Buy-and-hold's deepest drawdown ran from 2021-11-19 to 2022-11-03 and reached −34.2%.
How the drawdowns compare with holding QQQ
The strategy's deepest drawdown was 14.3%, from 2021-12-15 to 2022-06-30, and it recovered on 2023-02-13. That took 197 days down and 228 days to recover, a total of more than a year. Holding QQQ fell 34.2% from 2021-11-19 to 2022-11-03 and recovered on 2023-12-12, which is 349 days down and 404 days to recover.
The second drawdown for the strategy was 11.0%, from 2025-02-05 to 2025-04-02, and it recovered on 2025-04-09, only 7 days later. The sharp rebound that month helped the strategy as it helped the fund. Holding had 22.4% in the same episode, from 2025-02-19 to 2025-04-08. The third was 7.6% in 2023, from 2023-08-07 to 2023-10-20, with recovery on 2024-02-22. Holding had 13.3% from 2024-07-10 to 2024-08-07, a drawdown that the strategy's table does not list in its top three.
The strategy cut each of the fund's three largest drawdowns by roughly half or more. It did that by being out of the fund on most days. The cost was the 2023 rally, where the fund made 53.0% and the strategy 13.8%.
The comparison with the Sharpe ratio is the cleanest summary: 0.98 for the strategy and 0.83 for holding. A lower return with a much lower drawdown gives a better return per unit of risk. That does not mean a higher final value. A holder of QQQ ended with $24,265 and a holder of the strategy with $21,396.
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) | 14.2% | −14.3% | $21,396 | 0.98 |
| 5 basis points | 11.0% | −15.6% | $18,240 | 0.79 |
| 10 basis points | 8.0% | −17.5% | $15,544 | 0.60 |
What trading costs do to a 164-trade strategy
At 5 basis points per trade the strategy made 11.04% a year with a 15.6% drawdown and ended at $18,240. At 10 basis points it made 7.99% with a 17.5% drawdown and ended at $15,544. The Sharpe ratio fell from 0.98 to 0.79 and then 0.60. The headline run's 14.2% is the figure with no cost at all, so it is an upper bound.
Costs come off 328 fills, so the effect adds up quickly. When the average win is 1.53%, a cost of 10 basis points on each of two fills takes a visible share of each trade. QQQ is very liquid, with an average daily dollar volume of $18,772,024,579 and a median minute bar of 77,976 shares, so the real cost for a small account on QQQ may be near the low end of the range. The cost runs are flat figures. At 10 basis points the strategy's return is lower than the 16.7% of holding by a wide margin, and it is lower than the 11.3% of the trend with a trailing stop.
The strategy does not reach holding's return at any cost level shown. The case for it on QQQ rests on drawdown and not on return, and the drawdown figure holds up under costs, as it only rises from 14.3% to 17.5%.
Changing the parameters
| Version | CAGR | Max drawdown | Round trips | Win rate | Final value |
|---|---|---|---|---|---|
| Published rules | 14.2% | −14.3% | 164 | 69% | $21,396 |
| RSI(2) < 5 / > 70 | 15.0% | −16.4% | 157 | 69% | $22,286 |
| RSI(2) < 15 / > 70 | 12.2% | −15.6% | 173 | 69% | $19,327 |
| RSI(2) < 10 / > 60 | 14.3% | −12.4% | 170 | 69% | $21,585 |
| RSI(2) < 10 / > 80 | 17.1% | −12.9% | 156 | 69% | $24,716 |
What the four parameter variants show
The base rule buys under 10 and sells over 70. The four variants change one threshold at a time.
Buying under 5 made 14.98% with a 16.4% drawdown and 157 trades, of which 108 won. A stricter entry made slightly more than the base at a slightly deeper drawdown, with a Sharpe ratio of 1.04. Buying under 15 made 12.16% with 173 trades and a Sharpe ratio of 0.85, which is lower. A looser entry took more trades and earned less on each.
The exit changes mattered more. Selling over 60 made 14.34% with a 12.4% drawdown and 170 trades. Selling over 80 made 17.07% with a 12.9% drawdown, 156 trades, 108 wins and a Sharpe ratio of 1.115. That variant ended at $24,716, above the $24,265 of buy-and-hold, with a drawdown of 12.9% against 34.2%. It is the only variant on the page that beat holding on return.
The reason is the holding period. Waiting for RSI(2) to reach 80 instead of 70 keeps the position through more of the rebound. The bounce after a washout on QQQ often continues past 70, and the median 5-day return after an RSI(2) reading under 10 was 0.80% against 0.56% for a normal day.
The four variants range from 12.16% to 17.07% on return, a spread of nearly five points. Choosing the best one afterward is a form of fitting to a single 5.74-year window, and with 156 to 173 trades the result is not a precise estimate. The fair reading is the direction: on QQQ, later exits did better than earlier ones, and looser entries did worse than stricter ones.
How QQQ behaved
| Measure | QQQ |
|---|---|
| Data in this test | 2021-01-04 to 2026-10-02 (1444 sessions) |
| Total return, buy and hold | 151.1% |
| Annualized volatility | 22.4% |
| Deepest drawdown | −35.0% (2021-11-19 to 2022-11-03) |
| Up days | 54.8% |
| Average daily range | 1.58% |
| Average overnight gap | 0.59% |
| Correlation to SPY | 0.94 |
| Correlation to TLT | 0.09 |
| Sessions above the 200-day average | 75.2% |
| Crossings of the 200-day average | 20 |
| Falls of 10% or more from a 20-day high | 21 |
How QQQ behaved and why the rule trailed it
QQQ tracks the Nasdaq-100. Its annualized volatility was 22.4% and its average daily range was 1.58%. Up days were 54.8% of sessions, with an average gain of 1.01% against an average down day of negative 1.06%. The lag-1 autocorrelation was negative 0.04, which is close to none. The beta to SPY was 1.29.
The fund spent 75.2% of its sessions above the 200-day average and crossed it 20 times. That is a trending fund. Calendar years were 29.2% in 2021, negative 32.4% in 2022, 54.8% in 2023, 25.6% in 2024, 20.8% in 2025 and 22.4% for the partial year 2026. Five of six years were strongly positive, and a rule that sits out most days trails in a market like that.
The washouts the rule looks for were frequent enough. RSI(2) fell below 10 on 144 sessions. After those sessions the median 5-day return was 0.80% against 0.56% on any day, and the median 20-day return was 2.18% against 1.74%. The edge is positive and small. The 14-day RSI fell below 30 on 15 sessions, with a median 5-day return of 3.69% and a 20-day return of 1.42%. That is a bigger bounce, but on 15 sessions it is a thin sample.
QQQ earned 64.6% of its log return overnight and 35.4% during the session, with an average overnight gap of 0.59%. The strategy enters at the open, so it collects the overnight gap on the days it holds.
The fund fell 10% or more from a 20-day high on 21 occasions, covering 82 days. The deepest drawdown on daily closes was 35.0% from 2021-11-19 to 2022-11-03. The best day was 2025-04-09 at 11.75%, and the worst were 2025-04-04 at negative 6.10% and 2022-09-13 at negative 5.52%. By calendar month, May averaged 5.19% and November 4.42%, while September averaged negative 1.65%.
Among the broad index funds, the same rule made 15.9% on VOOG, 15.6% on QQQM, 14.1% on VOO and 13.2% on SPY. QQQ at 14.2% sits in the middle of that group. The median for the category was 13.2% and the median across all 59 funds was 4.7%.
The rules
Buy extreme 2-day RSI washouts under 10, exit as soon as RSI(2) recovers above 70.
- WHEN the market opens · IF not invested AND RSI(2) < 10 · THEN buy with 98% of the sleeve
- WHEN the market opens · IF invested AND RSI(2) > 70 · THEN sell the whole position
A short-horizon mean-reversion template popularized by Larry Connors' RSI-2 research. A 2-period RSI under 10 flags a sharp multi-day selloff. In assets with a persistent upward drift, those selloffs have tended to snap back within days. Trades are frequent and short. This is the highest-turnover template in the library.
Good for: liquid index ETFs with strong long-term drift; turnover is high so per-trade edges are small.
Watch out: high trade counts make results sensitive to execution quality; a crash that keeps crashing will hand this template several losing entries in a row.
How the two rules interact with QQQ
The entry buys with 98% of the sleeve at the open when RSI(2) is below 10 and the strategy is flat. The exit sells the whole position at the open when RSI(2) is above 70. There is no stop and no time limit.
On QQQ the two thresholds work well together for the reasons above. The entry fires on about 144 sessions in the window and the fund tends to bounce afterward by a small margin. The exit at 70 usually fires within a few sessions. The median hold was 4 days and no hold lasted more than 19.
The interaction that costs return is the exit threshold. RSI(2) rises above 70 after one or two strong days, but QQQ in a trending year often continues. The variant that exits above 80 made 17.07% against 14.17% for the base rule, so the early exit left return behind. The interaction that protects capital is the entry. The rule does not buy until RSI(2) is under 10, so it stayed out of QQQ for most of the time and avoided the slow part of the 2022 decline.
For other templates on this fund, the golden cross made 13.6% with a drawdown of 22.2%, and the 200-day regime filter made 12.9% with 19.9%. The RSI(2) snapback had the lowest drawdown of the three at 14.3% and the highest return of the three. The test stops at one window and one fund. The results do not forecast future returns.
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 RSI(2) snapback beat buy-and-hold on QQQ?
Over 2021-01-04 to 2026-10-02, RSI(2) snapback on QQQ returned 14.2% annualized vs 16.7% for buy-and-hold: it trailed buy-and-hold by 2.5% per year, with a maximum drawdown 20.0 points shallower than holding (14.3% vs 34.2%).
How many trades did it make?
164 completed round trips over 5.7 years (328 fills), with 69% of round trips closing profitably.
How often does RSI(2) trade?
Far more than RSI(14), with dozens of round trips per year on a volatile ETF. The backtest table on each page shows the exact count over the test window.
Is RSI(2) too fast for daily bars?
It is designed for daily bars. The 2-day window catches short, sharp washouts rather than long regimes.
How did RSI(2) Dip Snapback do on QQQ?
It made 14.2% a year from 2021-01-04 to 2026-10-02, turning $10,000 into $21,396, with a 14.3% maximum drawdown. Buy-and-hold made 16.7% and ended at $24,265 with a 34.2% drawdown. The strategy won 113 of 164 trades.
How many trades does the RSI(2) strategy make on QQQ?
It made 164 round trips, which is 328 fills. The median hold was 4 days and the longest was 19. The strategy was invested 36.1% of the time.
Does RSI(2) beat buy-and-hold on QQQ?
Not in the base version. It trailed by 2.53 points a year and beat holding only in 2022, when it lost 5.7% against 31.7%. The variant that sells above 80 made 17.07% and ended above holding.
What do trading costs do to RSI(2) on QQQ?
At 5 basis points the return fell to 11.04% a year and at 10 basis points to 7.99%. The ending value fell from $21,396 to $18,240 and then $15,544. The drawdown rose to 17.5%.
What was the best RSI(2) trade on QQQ?
The best trade ran from 2025-04-07 to 2025-04-10 and returned 11.45%, from an adjusted entry price of 405.22 to an exit of 451.60. The worst lost 8.34% over 13 days from 2025-02-21.
Which RSI(2) settings worked best on QQQ?
Buying under 10 and selling over 80 made 17.07% with a 12.9% drawdown and a Sharpe ratio of 1.115. Buying under 5 made 14.98% and buying under 15 made 12.16%. The sample is one window, so the ranking may not hold elsewhere.
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.