DeployQuantQID trading strategies, backtested
ProShares UltraShort QQQ: -2x daily Nasdaq-100. Every DeployQuant template run on QID over 5.7 years of minute data, same engine, same window, sorted by return.
QID is ProShares UltraShort QQQ, a fund that targets minus twice the daily move of the Nasdaq-100. From 2021-01-04 to 2026-10-02 holding it lost 29.5% a year. A $10,000 start ended at $1,348, a loss of 86.5%, with a maximum drawdown of 87.3% that began on 2021-03-08 and had not recovered by the end of the window. The Sharpe ratio was negative 0.64.
All 12 DeployQuant templates finished ahead of holding on return, and 11 had a shallower drawdown. None of them made money. The best result was the RSI(2) snapback at negative 1.9% a year, and the worst was the monthly cycle at negative 28.8%, only a little better than holding. The question this page answers is therefore not which rule beat the fund, but which rules lost the least and why a short-horizon rule did so much better than a trend rule.
QID rose 66.4% in 2022 when the Nasdaq fell and lost between 34.0% and 57.3% in every other calendar year. Any rule that held it through the five other years paid for the one good year several times over. The test uses daily decisions with fills on minute bars, and the headline runs carry no fees or slippage. The fund moves opposite to QQQ, with a correlation of negative 1.00, and its siblings SQQQ and SDS behave the same way at different leverage.
| Strategy | CAGR | max DD | Sharpe | trades | win rate | final value |
|---|---|---|---|---|---|---|
| RSI(2) Dip Snapback | −1.9% | −59.0% | 0.06 | 172 | 56% (+1 open) | $8,983 |
| 20-Day Momentum + Trailing Stop | −3.1% | −45.7% | 0.00 | 33 | 39% | $8,354 |
| Golden Cross (SMA 50/200) | −6.0% | −48.4% | -0.11 | 2 | 0% | $6,998 |
| SMA-200 Trend + 15% Trailing Stop | −7.6% | −46.1% | -0.13 | 18 | 28% | $6,354 |
| 200-Day SMA Regime Filter | −10.5% | −55.1% | -0.24 | 17 | 18% | $5,303 |
| SMA 10/50 Trend | −13.1% | −63.2% | -0.28 | 18 | 22% | $4,474 |
| EMA 12/26 Trend | −13.6% | −61.6% | -0.30 | 24 | 17% | $4,312 |
| RSI(14) Mean Reversion | −14.2% | −68.0% | -0.37 | 16 | 44% (+1 open) | $4,157 |
| 3-Month Momentum Switch | −17.2% | −71.8% | -0.49 | 22 | 14% | $3,389 |
| Weekly Entry + 7% Target | −26.2% | −86.1% | -0.65 | 227 | 26% | $1,742 |
| Drawdown Dip Buyer + 8% Target | −27.1% | −86.6% | -0.56 | 3 | 100% (+1 open) | $1,627 |
| First-to-Last Day of Month | −28.8% | −88.0% | -0.59 | 69 | 33% (+1 open) | $1,422 |
| Buy & hold QID | −29.5% | −87.3% | -0.64 | – | – | $1,348 |
How the 12 templates ranked on QID
The RSI(2) snapback lost 1.9% a year with a 59.0% drawdown, 172 round trips and a win rate of 56%. It ended at $8,983. Its profit factor was 0.98, so gross wins were almost level with gross losses. The average win was 3.17% and the average loss was 3.86%. The rule made 95.5% in 2022, 5.4% in 2021, and lost 37.9% in 2023, 13.6% in 2024, 2.3% in 2025 and 16.8% in 2026. It beat holding in all six years.
The momentum breakout lost 3.1% a year with a 45.7% drawdown, 33 round trips and a 39% win rate. It was invested only 16.3% of the time and held for 10.5 days on average. Its best trade ran from 2025-03-25 to 2025-04-07 for a gain of 30.7%. It made 34.2% in 2022 and 21.5% in 2025.
The golden cross lost 6.0% on 2 trades, and both lost money: negative 5.0% over 332 days in 2022 and early 2023, and negative 27.0% over 57 days in 2025. Its exposure was 18.4%. The trend with a trailing stop lost 7.6% with 18 trades and a 28% win rate. The 200-day regime filter lost 10.5% with an 18% win rate and a profit factor of 0.12. The SMA 10/50 trend lost 13.1% and the EMA 12/26 trend lost 13.6%, each winning under a quarter of its trades.
The RSI(14) mean reversion lost 14.2% with a 68.0% drawdown. Its two worst trades lost 37.7% over 203 days from 2025-04-30 and 34.7% over 206 days from 2023-01-24. The rule buys an oversold reading and waits, and on an inverse fund that decays, waiting is the cost. The 3-month momentum switch lost 17.2% and won 14% of its 22 trades, with a profit factor of 0.08.
The weekly 7% target lost 26.2% with 227 round trips, an 86.1% drawdown and a losing streak of 20 trades. It won 26% of its trades, with an average win of 7.33% and an average loss of 3.50%. Its trades are short, at 5.1 days, but it was invested 67.4% of the time, so it kept paying the daily decay. The dip buyer lost 27.1% and ended at $1,627 with an open position entered on 2022-10-13 at an adjusted price of 121.99 and down 89.5% by the end of the test. The monthly cycle lost 28.8%.
The ordering follows exposure. The rules that held QID for the smallest share of the window lost the least, and the rules that stayed in it lost close to what holding lost. On an asset with negative drift, time in the position is the cost, and the strategies that were short on exposure were also short on losses.
Drawdowns and the one good year
The 2022 rally is the only period when a position in QID paid. Holding gained 63.3% in 2022 on the equity curve, and the weekly 7% target made 50.1%, the RSI(2) snapback 95.5%, the RSI(14) mean reversion 36.1%, and the monthly cycle 81.2%. Five of the six trend and breakout rules gained between 10.8% and 34.2% that year. The 200-day regime filter lost 25.6% in 2022, since the fund was below its 200-day average for most of the year and the rule only entered on the occasional rally.
The rules that beat holding in 2022 did not carry that into later years. Holding lost 55.6% in 2023, and the RSI(2) snapback lost 37.9%, the momentum breakout 17.8%, the golden cross 23.4% and the 3-month momentum switch 11.7%. Those losses are the price of the rules entering after the fund had started to decline.
The drawdowns show the same split. The momentum breakout had a 45.7% drawdown beginning 2022-10-14, and the golden cross had 48.4% from the same date. The same date marks the end of the 2022 rally in the fund, so every rule that held into the peak went on to a long decline. The RSI(2) snapback's drawdown of 59.0% began on 2023-01-03 and had not recovered by 2026-06-02. The 3-month momentum switch had 71.8%, the monthly cycle 88.0% and the weekly 7% target 86.1%.
One point deserves attention in the table. The dip buyer and the monthly cycle finished close to holding because both stayed invested through almost the whole window: 98.1% and 95.2% exposure. The dip buyer made only 3 round trips and its fourth trade stayed open from 2022-10-13 to the end of the test. A rule that buys a decline and holds for a target never reaches the target on a fund that keeps falling.
Best and worst months, streaks and the position still open
April 2022 was the best month for holding QID at 28.51%, and it was also the best month for the golden cross at 28.96%, the dip buyer at 28.58%, the RSI(14) rule at 22.93%, the RSI(2) snapback at 22.53% and the 3-month momentum switch at 21.64%. September 2022 was the best month for the momentum breakout at 22.36%, the trailing stop strategy at 22.57%, the 200-day filter at 22.46%, the SMA 10/50 trend at 25.73%, the EMA 12/26 trend at 24.36% and the weekly target at 22.48%. Every best month was in 2022, and each is a single month of a fund that rose in a falling market.
April 2026 was the worst month for holding at negative 22.08%, the monthly cycle at negative 22.8%, the weekly target at negative 19.33%, the dip buyer at negative 22.03% and the SMA 10/50 trend at negative 15.44%. The RSI(2) snapback's worst trade also fell there: it bought on 2026-04-02, sold on 2026-04-22 and lost 21.77%. July 2022 was the worst month for the golden cross at negative 21.97%, the trailing stop strategy at negative 22.03%, the 200-day filter at negative 21.93% and the 3-month momentum switch at negative 17.75%.
The streaks are long. The weekly target lost 20 round trips in a row at its longest and won at most 3 in a row. The 3-month momentum switch lost 10 in a row, the SMA 10/50 trend and the EMA 12/26 trend lost 9 in a row each, and the momentum breakout lost 7. The RSI(2) snapback won 8 in a row and lost 5. A losing run of 20 trades at 3.5% average loss is the mechanism behind the weekly target's 86.08% drawdown.
None of the templates recovered from its worst drawdown inside the window, and holding's 87.33% drawdown from 2021-03-08 was still open on the last day. Two templates ended the window holding a losing position. The dip buyer bought on 2022-10-13 at an adjusted 121.99 and was marked 89.46% below entry at the last close. The RSI(14) rule entered on 2026-08-18 at 13.86 and was down 7.22%.
The yearly gaps to holding show the same split by rule type. In 2022 the 200-day filter trailed holding by 88.9 points, the 3-month momentum switch by 93.6 and the trailing stop strategy by 52.5. In most other years those rules beat holding by wide gaps: the 3-month switch by 39.3 points in 2021 and 43.9 in 2023, the 200-day filter by 45.2 and 39.9. The RSI(2) snapback beat holding in all six years, by 50.6 points in 2021 and 12.7 in 2026. It was the only rule that did, and it still lost money overall.
How each strategy traded QID
| Strategy | Time in market | Avg hold (days) | Best trade | Worst trade | Profit factor | With 10 bps slippage |
|---|---|---|---|---|---|---|
| RSI(2) snapback | 47.2% | 6 | 11.7% | −21.8% | 0.98 | −7.5% |
| momentum breakout | 16.3% | 11 | 30.7% | −9.9% | 0.85 | −4.2% |
| golden cross | 18.4% | 195 | −5.0% | −27.0% | – | −6.1% |
| trend + trailing stop | 20.1% | 23 | 21.8% | −19.3% | 0.62 | −8.2% |
| 200-day regime filter | 17.2% | 22 | 5.4% | −10.4% | 0.12 | −11.0% |
| SMA 10/50 trend | 30.1% | 35 | 10.5% | −16.4% | 0.30 | −13.6% |
| EMA 12/26 trend | 29.2% | 26 | 14.4% | −10.7% | 0.31 | −14.3% |
| RSI mean reversion | 68.8% | 87 | 21.6% | −37.7% | 0.55 | −14.7% |
| 3-month momentum | 21.8% | 21 | 4.4% | −16.3% | 0.08 | −17.8% |
| weekly 7% target | 67.4% | 5 | 12.7% | −20.3% | 0.76 | −30.8% |
| dip buyer | 98.1% | 203 | 9.1% | 8.0% | – | −27.3% |
| monthly cycle | 95.2% | 28 | 25.6% | −23.3% | 0.61 | −30.5% |
What the trades and cost runs show
The two best-ranked rules traded in brief bursts. The RSI(2) snapback held for 5.7 days on average and its best trade lasted 2 days, from 2025-04-02 to 2025-04-04, for a gain of 11.7%. That trade caught the fund's best day, 2025-04-04 at 12.12%. Its worst trade lost 21.8% over 20 days from 2026-04-02 to 2026-04-22, and its longest losing streak was 5 trades.
A second look at the weekly 7% target shows why a profit target does not rescue an inverse fund. Its best trade gained 12.7% in 2 days from 2022-02-22, but its worst lost 20.3% in 3 days from 2025-04-07, because the rule had bought into the rebound that followed the April 2025 selloff. Wins were capped near 7% and losses were not capped.
Trading costs hit the fast rules hardest. At 5 basis points the RSI(2) snapback went from negative 1.9% to negative 4.7%, and at 10 basis points to negative 7.5%, with a drawdown of 66.2%. The weekly 7% target fell from negative 26.2% to negative 28.3% and then negative 30.8%, which is a worse result than holding. The slow rules moved little: the golden cross went from negative 6.0% to negative 6.1% and the 3-month momentum switch from negative 17.2% to negative 17.8%.
At 10 basis points the RSI(2) snapback is still the best result on the page, but the lead over the momentum breakout, at negative 4.1%, is smaller than in the headline run. The headline order is the best case for each rule.
QID's liquidity is high. It trades about $228,567,798 a day and the median minute bar has 5,631 shares. The cost runs use a flat figure and do not model market impact for a large order.
How QID behaved
| Measure | QID |
|---|---|
| Data in this test | 2021-01-04 to 2026-10-02 (1444 sessions) |
| Total return, buy and hold | −89.3% |
| Annualized volatility | 44.7% |
| Deepest drawdown | −89.6% (2021-03-08 to 2026-10-02) |
| Up days | 45.1% |
| Average daily range | 3.15% |
| Average overnight gap | 1.19% |
| Correlation to SPY | -0.94 |
| Correlation to QQQ | -1.00 |
| Correlation to TLT | -0.09 |
| Sessions above the 200-day average | 20.0% |
| Crossings of the 200-day average | 34 |
| Falls of 10% or more from a 20-day high | 78 |
Calendar years
| Year | Return |
|---|---|
| 2021 | −46.4% |
| 2022 | 66.4% |
| 2023 | −57.3% |
| 2024 | −34.0% |
| 2025 | −35.0% |
| 2026 (part) | −34.4% |
Biggest single days
| Best day | Move |
|---|---|
| 2025-04-04 | 12.1% |
| 2022-09-13 | 11.0% |
| 2025-04-03 | 10.7% |
| 2022-05-05 | 10.0% |
| 2022-05-18 | 9.9% |
| Worst day | Move |
|---|---|
| 2025-04-09 | −23.5% |
| 2022-11-10 | −14.7% |
| 2022-11-30 | −9.1% |
| 2022-07-27 | −8.5% |
| 2025-05-12 | −8.1% |
Average return by calendar month
| Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec |
|---|---|---|---|---|---|---|---|---|---|---|---|
| −2.1% | 1.2% | −1.5% | −1.3% | −9.8% | −4.9% | −3.6% | −1.1% | 4.3% | −5.3% | −8.4% | 1.6% |
Most and least correlated funds
| Most correlated | Least correlated | ||
|---|---|---|---|
| SQQQ | 1.00 | TQQQ | -1.00 |
| PSQ | 1.00 | QLD | -1.00 |
| TECS | 0.97 | QQQ | -1.00 |
| REW | 0.97 | QQQM | -1.00 |
| SDS | 0.94 | VOOG | -0.98 |
Leverage and decay against QQQ
| Year | QID | QQQ | -2× QQQ, reset daily |
|---|---|---|---|
| 2021 | −46.4% | 29.2% | −45.7% |
| 2022 | 66.4% | −32.4% | 60.3% |
| 2023 | −57.3% | 54.8% | −62.2% |
| 2024 | −34.0% | 25.6% | −42.5% |
| 2025 | −35.0% | 20.8% | −41.9% |
| 2026 | −34.4% | 22.4% | −39.3% |
The last column compounds -2× QQQ's daily return with no fees or financing: the return a perfect daily-reset fund would have had. It leaves out the fund's fees, its borrowing costs and the interest it earns on cash, which put a real fund below or above this line. A daily-reset fund does not aim for -2× the underlying's return over a year.
Over 2021-01-04 to 2026-10-02, QID returned −89.3% while QQQ returned 151.1% and a perfect daily-reset -2× version would have returned −93.3%. Its measured daily beta to QQQ was -2.00.
Why inverse funds lose to time
QID lost money in five of six calendar years even though the Nasdaq-100 was up in five of them. The annual figures were negative 46.4% in 2021, positive 66.4% in 2022, negative 57.3% in 2023, negative 34.0% in 2024, negative 35.0% in 2025 and negative 34.4% for the partial year 2026. Up days were only 45.1% of sessions.
Daily volatility is the same as QLD's: annualized volatility was 44.7% and the average daily range was 3.15%. The average up day was 2.16% and the average down day was negative 2.02%. A fund with these daily swings loses to compounding when the underlying drifts upward. The leverage table shows it. QQQ returned 151.1% over the window. A frictionless daily-reset model of minus 2x gives negative 93.3%, and QID returned negative 89.3%, a slightly smaller loss. In the worst year for the fund, 2023, the QQQ gain of 54.8% produced a QID loss of 57.3%, against negative 62.19% for the ideal. The ideal leaves out fees, borrowing costs and interest on cash.
The fund's longest drawdown lasted 1,400 sessions and was still open on 2026-10-02. It fell 10% or more from a 20-day high on 78 occasions, covering 553 days, and it was above its 200-day average on only 20.0% of sessions, with 34 crossings. A trend rule on QID is therefore mostly flat. The SMA 10/50 trend was invested 30.1% of the time and the 200-day regime filter 17.2%, and both still lost money, because each signal arrived after the fund's short rallies were already underway.
The mean reversion signals gave no edge. After RSI(2) closed under 10, which happened on 248 sessions, the median 5-day return was negative 1.17%, the same as the baseline of negative 1.17%. The median 20-day return was negative 3.93% against a baseline of negative 3.55%. After the 14-day RSI closed under 30, on 116 sessions, the median 5-day return was 0.12% and the 20-day return was negative 3.09%. A washout on this fund is not followed by a bounce that beats a normal day. That is why the RSI(2) snapback did not profit here, and why its relative result came from being in cash.
The best days were 2025-04-04 at 12.12%, 2022-09-13 at 11.00% and 2025-04-03 at 10.69%. The worst was 2025-04-09 at negative 23.53%, then 2022-11-10 at negative 14.71%. By calendar month, May averaged negative 9.75% and November negative 8.36%, while September averaged 4.33%. Each month has 5 or 6 observations.
How QID compares with similar funds
Among the inverse funds in this test, the best template differs by fund. On PSQ the RSI(2) snapback reached 0.1% a year against negative 13.8% for holding. On SH and SPDN, also inverse S&P 500 funds, the same rule returned negative 0.5% and negative 1.5%. On SOXS the trend with a trailing stop made negative 4.9% against negative 48.3% for holding, and on SQQQ golden cross made negative 7.4% against negative 42.3%. On EEV the RSI(14) mean reversion made 5.0% against negative 16.1%, the only positive result among the leveraged inverse funds shown.
The pattern on QID is the common one for inverse funds in this window: the best rule loses a little, the typical rule loses a lot, and holding loses the most. The test stops at one window of 5.74 years in which the Nasdaq rose. A window with a longer decline would rank the rules differently, and none of these figures is a forecast.
Overnight and weekday figures, and what the RSI signals showed
QID's overnight log return was negative 92.57% and its intraday log return negative 127.02%, so the fund lost in both halves of the day, 42.16% of the decline coming overnight and 57.84% in the session. The average overnight gap was 1.19% and the average daily range 3.15%. Beta to QQQ was negative 2 and to SPY negative 2.57, with a correlation of negative 1 to QQQ.
Weekday averages show a Monday figure of negative 0.34%, Wednesday negative 0.18%, Tuesday negative 0.04%, Thursday negative 0.01% and Friday negative 0.02%. September averaged 4.33%, December 1.57% and February 1.24%, and every other calendar month averaged below zero, with 5 or 6 observations per month. With so few observations these averages describe the window and are not a calendar rule.
The RSI(14) fell below 30 on 116 sessions. The median 5-day return after them was 0.12%, but the median 20-day return was negative 3.09%. After the 14-day signal, the 5-day result was near flat and the 20-day result was below zero, so the longer holding period is where buying weakness in QID lost money. The 20-day baseline for all sessions was negative 3.55%, so the signal was a little better than an average day, and still negative. The long Nasdaq-100 fund QLD has the same daily range as QID.
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
What was the best strategy for QID?
Of the 12 templates tested on QID over 2021-01-04 to 2026-10-02, the strongest by CAGR was RSI(2) snapback at −1.9% (max drawdown 59.0%), versus −29.5% for buy-and-hold. The best result in hindsight is not a forecast. Check drawdowns and trade counts before drawing conclusions.
Did any strategy beat buying and holding QID?
12 of 12 templates beat QID buy-and-hold (−29.5% CAGR) on this window; 11 of 12 had a shallower maximum drawdown than holding (87.3%).
Did any strategy make money on QID?
No. All 12 templates had a negative return from 2021-01-04 to 2026-10-02. The best was the RSI(2) snapback at negative 1.9% a year, and the worst was the monthly cycle at negative 28.8%. Holding QID lost 29.5% a year.
Why does QID lose money when QQQ rises?
QID targets minus twice the daily return of QQQ. QQQ returned 151.1% over the window and QID lost 89.3%. Daily rebalancing at 2x leverage compounds against the holder when the underlying drifts upward, and the fund fell in five of six calendar years.
What was QID's maximum drawdown?
Holding QID fell 87.3% from 2021-03-08 to 2026-10-02 and had not recovered by the last day of the test. The shallowest template drawdown was 45.7%, from the momentum breakout.
Does RSI(2) mean reversion work on QID?
It lost 1.9% a year with 172 trades and a 59.0% drawdown, and it won 56% of its trades. It made 95.5% in 2022 and lost money in 2023 to 2026. After costs of 10 basis points the loss was 7.5%.
How volatile is QID?
Annualized volatility was 44.7% and the average daily range was 3.15%. The fund's best day was 12.12% on 2025-04-04 and its worst was negative 23.53% on 2025-04-09.
Which strategy lost the least on QID?
The RSI(2) snapback lost 1.9% a year, then the momentum breakout lost 3.1% and the golden cross 6.0%. The rules that held the fund for the smallest share of the window lost the least.
Other inverse etfs
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.