DeployQuantPSQ trading strategies, backtested
ProShares Short QQQ: -1x daily Nasdaq-100, an unleveraged tech hedge. Every DeployQuant template run on PSQ over 5.7 years of minute data, same engine, same window, sorted by return.
PSQ is ProShares Short QQQ, a fund that targets the opposite of the Nasdaq-100's daily return, with no leverage. All 12 DeployQuant templates run on it from 2021-01-04 to 2026-10-02, starting each with $10,000. Buying PSQ and holding it turned $10,000 into $4,275, a 13.76% annual loss, with a 59.64% maximum drawdown from 2022-10-14 to the end of the data on 2026-10-02. The fund had not recovered from that peak when the test ended.
Against that, all 12 templates beat buy-and-hold on return, and 11 had a shallower drawdown. Only one template finished with a positive annualized return, the RSI(2) snapback at 0.13%, which ended at $10,077. The other 11 lost money.
The Nasdaq-100 itself returned 151.07% over the window, and PSQ's total return was negative 59.3%. A fund that moves against a rising index gives a rule little to work with. PSQ closed above its 200-day average on only 23.29% of sessions. The text below goes through the ranking, the single year that rewarded the fund, what slippage did, and the numbers that explain why the mean-reversion rules found no edge. The window is 5.7 years, and one fund in one regime is a narrow sample.
| Strategy | CAGR | max DD | Sharpe | trades | win rate | final value |
|---|---|---|---|---|---|---|
| RSI(2) Dip Snapback | 0.1% | −33.8% | 0.08 | 172 | 57% (+1 open) | $10,077 |
| 200-Day SMA Regime Filter | −2.3% | −20.1% | -0.08 | 15 | 20% | $8,732 |
| 3-Month Momentum Switch | −4.3% | −35.3% | -0.22 | 10 | 20% | $7,769 |
| Golden Cross (SMA 50/200) | −4.3% | −34.9% | -0.24 | 3 | 33% | $7,750 |
| SMA-200 Trend + 15% Trailing Stop | −5.5% | −38.2% | -0.28 | 8 | 38% | $7,213 |
| RSI(14) Mean Reversion | −5.7% | −40.9% | -0.33 | 16 | 50% (+1 open) | $7,131 |
| 20-Day Momentum + Trailing Stop | −5.7% | −33.1% | -0.47 | 8 | 38% | $7,123 |
| SMA 10/50 Trend | −6.3% | −37.5% | -0.32 | 18 | 33% | $6,890 |
| EMA 12/26 Trend | −6.7% | −39.8% | -0.34 | 21 | 14% | $6,711 |
| Weekly Entry + 7% Target | −6.8% | −46.4% | -0.27 | 165 | 17% | $6,692 |
| Drawdown Dip Buyer + 8% Target | −9.3% | −55.4% | -0.41 | 3 | 100% (+1 open) | $5,704 |
| First-to-Last Day of Month | −12.7% | −60.3% | -0.54 | 69 | 36% (+1 open) | $4,594 |
| Buy & hold PSQ | −13.8% | −59.6% | -0.59 | – | – | $4,275 |
The ranking by loss
The ranking runs from a 0.13% gain to a 12.67% annual loss. After the RSI(2) snapback come the 200-day regime filter at negative 2.33%, the 3-month momentum switch at negative 4.3%, the golden cross at negative 4.34% and the trend plus trailing stop at negative 5.53%. Then the RSI(14) mean reversion at negative 5.72%, the momentum breakout at negative 5.74%, the SMA 10/50 trend at negative 6.28%, the EMA 12/26 trend at negative 6.71%, the weekly 7% target at negative 6.76%, the dip buyer at negative 9.32% and the monthly cycle at negative 12.67%.
The most invested rules finished at the bottom. The monthly cycle held a position 95.2% of the time and lost 12.67% a year, ending at $4,594 with a 60.31% drawdown, the only template deeper than holding's 59.64%. The dip buyer was in the market 87.9% of the time and the weekly target 76.2%. The less invested rules finished at the top: the 200-day filter 20.1%, the 3-month switch 21.5% and the golden cross 21.7%. The pattern is not perfect, since the momentum breakout held a position only 13.2% of days and still lost 5.74% a year, but on a fund that fell in five of six calendar years, more time in the market meant more exposure to the decline.
The RSI(2) snapback is the exception worth reading. It made 172 round trips, won 57% of them, and held for an average of 5.5 days with exposure of 46.3%. Its average winner was 1.57% and its average loser 1.99%, and its profit factor was 1.01, which is roughly break-even. It beat holding in all six calendar years. Its drawdown was 33.82%, running from 2023-01-03 to 2026-06-02, and its worst trade was an 11.44% loss over 20 days from 2026-04-02. Its Sharpe ratio was 0.08 against negative 0.59 for the holder.
The dip buyer has a 100% win rate and the second-worst return, which needs explaining. It made 3 round trips, all winners near 8%: one held 409 days from 2021-04-06 to 2022-05-20, one held 46 days from 2022-08-01 to 2022-09-16, and one held 38 days from 2022-11-14 to 2022-12-22. A fourth position opened on 2023-01-27 at $55.53 and was still open at the end of the test, marked 55.85% below entry. The win rate counts closed trades only. The open loss is the reason the rule shows a 55.38% drawdown and a final value of $5,704.
The trend and breakout rules share a pattern of small wins and larger losses. The EMA 12/26 trend made 21 round trips and won 3, with an average winner of 7.24% and an average loser of 3.36%, and its best trade was 9.48% from 2022-04-18 to 2022-07-20. The SMA 10/50 trend made 18 round trips and won 6. The trend plus trailing stop strategy made 8 round trips and won 3, with an average loser of 9.07% and a worst trade of 14.38% over 120 days from 2024-08-06. The momentum breakout made 8 round trips, won 3, and lost 10.03% in 10 days from 2024-08-05. All four have profit factors between 0.2 and 0.35.
The golden cross shows the opposite failure. It made 3 round trips and won one. The winner returned 2.62% over 361 days, and the losers were 18.28% and 8.14%. Its profit factor was 0.1. The average loser of 13.21% was much larger than the average winner of 2.62%.
The one year that paid
2022 is the only calendar year in which PSQ rose, and it gained 35.3% on the sleeve basis. The dip buyer returned 55.2% that year. The RSI(2) snapback returned 42.7%, the monthly cycle 41.7%, the weekly target 26.3%, the RSI(14) rule 18.5%, the golden cross 15.4%, the SMA 10/50 trend 15.1%, the EMA 12/26 trend 13.5% and the trailing stop strategy 13.1%. The 3-month momentum switch returned 4.2% and the 200-day filter 3.7%. The momentum breakout lost 9.6%, the only template with a loss in 2022.
Three templates beat holding in 2022: the dip buyer by 19.9 points, the RSI(2) snapback by 7.4 and the monthly cycle by 6.4. The trend rules gave up 2022 relative to holding. The 200-day filter trailed by 31.6 points, the golden cross by 19.9 and the momentum breakout by 44.9. The 200-day filter's first trade of the year started on 2022-01-24 and lost 5.31% in 8 days. Its next two trades, from 2022-04-07 to 2022-08-11 and from 2022-08-22 to 2023-01-24, gained 3.11% and 8.81%.
In 2023 PSQ lost 32.02% and the trend rules did best against holding. The 200-day filter lost 7% and beat holding by 24.3 points. The momentum breakout lost 7.7% and the golden cross 11.1%. A year in which the Nasdaq-100 returned 54.81% was the year when staying out of PSQ was worth the most.
The RSI(2) snapback gained 0.6% in 2025, when PSQ lost 14.9%. The dip buyer's open position kept it level with holding in 2024, 2025 and 2026, each within 0.1 points.
April 2025 and April 2026, the two months that hurt
April was PSQ's best month and its worst, in different years. April 2022 was the best month for holding at 13.94% and for most of the templates: the weekly target made 14.13%, the 3-month momentum switch 14%, the golden cross 14.01%, the dip buyer 13.99% and the 200-day filter 11.26%. April 2026 was the worst month for holding at negative 12.8%, and also for the monthly cycle at negative 11.94%, the dip buyer at negative 12.89%, the RSI(2) snapback at negative 11.38% and the trailing stop strategy at negative 11.35%. The fund fell as the Nasdaq-100 rose, and long positions in PSQ lost with it.
The RSI(2) snapback's worst trade came in that month. It bought on 2026-04-02 and sold on 2026-04-22 for a loss of 11.44%. A rule built to buy a drop kept buying a fund that was still falling. The other large April hit was a year earlier. PSQ's worst day was negative 11.71% on 2025-04-09, and in the days before it the weekly target lost 9.8% on a trade from 2025-04-07 to 2025-04-10 and the golden cross opened a position on 2025-04-08 that closed on 2025-06-20 at negative 18.28%. The RSI(2) snapback had caught the other side of that week: its best trade was 5.87%, from 2025-04-02 to 2025-04-04, the fund's best day at 6.09%.
The streaks make the same point at length. The weekly target lost 24 round trips in a row at its longest and never won more than 2 in a row. The EMA 12/26 trend lost 9 in a row and the SMA 10/50 trend 6. The RSI(2) snapback won 9 in a row and lost 5 in a row. A rule that fires on every weekly rebound in a fund that keeps falling accumulates small losses without a break, and the monthly cycle's longest losing run was 7 months against 3 winning months.
None of the templates recovered from its worst drawdown inside the window. Holding's 59.64% drawdown began on 2022-10-14 and ran to the last day of data. The RSI(2) snapback's 33.82% drawdown ran from 2023-01-03 to 2026-06-02 with no recovery date, the 200-day filter's 20.06% from 2022-06-16 to 2026-04-08, and the weekly target's 46.39% from 2022-07-01 to 2026-06-15.
How each strategy traded PSQ
| Strategy | Time in market | Avg hold (days) | Best trade | Worst trade | Profit factor | With 10 bps slippage |
|---|---|---|---|---|---|---|
| RSI(2) snapback | 46.3% | 6 | 5.9% | −11.4% | 1.01 | −5.6% |
| 200-day regime filter | 20.1% | 28 | 8.8% | −5.3% | 0.45 | −2.8% |
| 3-month momentum | 21.5% | 46 | 5.3% | −7.5% | 0.31 | −4.6% |
| golden cross | 21.7% | 152 | 2.6% | −18.3% | 0.10 | −4.4% |
| trend + trailing stop | 28.3% | 74 | 7.8% | −14.4% | 0.32 | −5.8% |
| RSI mean reversion | 66.5% | 87 | 11.0% | −19.8% | 0.55 | −6.3% |
| momentum breakout | 13.2% | 34 | 4.8% | −10.0% | 0.20 | −6.0% |
| SMA 10/50 trend | 32.5% | 38 | 8.3% | −7.2% | 0.35 | −6.9% |
| EMA 12/26 trend | 31.8% | 32 | 9.5% | −5.2% | 0.35 | −7.4% |
| weekly 7% target | 76.2% | 9 | 9.2% | −9.8% | 0.84 | −11.6% |
| dip buyer | 87.9% | 164 | 8.5% | 8.0% | – | −9.4% |
| monthly cycle | 95.2% | 28 | 12.3% | −12.2% | 0.63 | −14.7% |
Trading costs
Slippage turned the one positive result negative. The RSI(2) snapback went from 0.13% a year to negative 2.77% at 5 basis points and negative 5.6% at 10, with its Sharpe ratio moving from 0.08 to negative 0.144 and then negative 0.365. At 10 basis points its drawdown was 45.6% against 33.82% in the headline run. The rule makes 345 fills in 5.7 years, so a small cost on each one removes the thin edge.
The weekly target is the most cost-sensitive of the rest. It went from negative 6.76% to negative 8.85% and negative 11.55%, and its drawdown from 46.39% to 50.89% and 57.07%. The monthly cycle went from negative 12.67% to negative 13.75% and negative 14.74%. The slow rules barely changed. The golden cross went from negative 4.34% to negative 4.39% and negative 4.44%, and the dip buyer from negative 9.32% to negative 9.35% and negative 9.38%.
PSQ's average daily dollar volume was $206,478,236 and its median minute volume was 5,460 shares, which is liquid for a $10,000 sleeve. The costs in the runs are slippage on each fill and no more.
How PSQ behaved
| Measure | PSQ |
|---|---|
| Data in this test | 2021-01-04 to 2026-10-02 (1444 sessions) |
| Total return, buy and hold | −59.3% |
| Annualized volatility | 22.4% |
| Deepest drawdown | −61.0% (2022-10-14 to 2026-10-02) |
| Up days | 45.5% |
| Average daily range | 1.58% |
| Average overnight gap | 0.60% |
| Correlation to SPY | -0.94 |
| Correlation to QQQ | -1.00 |
| Correlation to TLT | -0.09 |
| Sessions above the 200-day average | 23.3% |
| Crossings of the 200-day average | 30 |
| Falls of 10% or more from a 20-day high | 19 |
Calendar years
| Year | Return |
|---|---|
| 2021 | −25.9% |
| 2022 | 36.4% |
| 2023 | −32.0% |
| 2024 | −15.7% |
| 2025 | −15.5% |
| 2026 (part) | −16.8% |
Biggest single days
| Best day | Move |
|---|---|
| 2025-04-04 | 6.1% |
| 2022-09-13 | 5.5% |
| 2025-04-03 | 5.4% |
| 2022-05-05 | 5.0% |
| 2022-05-18 | 4.9% |
| Worst day | Move |
|---|---|
| 2025-04-09 | −11.7% |
| 2022-11-10 | −7.4% |
| 2022-11-30 | −4.5% |
| 2022-07-27 | −4.2% |
| 2025-05-12 | −4.0% |
Average return by calendar month
| Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec |
|---|---|---|---|---|---|---|---|---|---|---|---|
| −0.9% | 0.9% | −0.5% | −0.6% | −4.8% | −2.4% | −1.7% | −0.3% | 2.3% | −2.5% | −4.0% | 1.0% |
Most and least correlated funds
| Most correlated | Least correlated | ||
|---|---|---|---|
| SQQQ | 1.00 | TQQQ | -1.00 |
| QID | 1.00 | QLD | -1.00 |
| TECS | 0.97 | QQQ | -1.00 |
| REW | 0.96 | QQQM | -1.00 |
| SDS | 0.94 | VOOG | -0.98 |
Leverage and decay against QQQ
| Year | PSQ | QQQ | -1× QQQ, reset daily |
|---|---|---|---|
| 2021 | −25.9% | 29.2% | −25.1% |
| 2022 | 36.4% | −32.4% | 33.3% |
| 2023 | −32.0% | 54.8% | −37.5% |
| 2024 | −15.7% | 25.6% | −22.9% |
| 2025 | −15.5% | 20.8% | −21.6% |
| 2026 | −16.8% | 22.4% | −20.9% |
The last column compounds -1× 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 -1× the underlying's return over a year.
Over 2021-01-04 to 2026-10-02, PSQ returned −59.3% while QQQ returned 151.1% and a perfect daily-reset -1× version would have returned −70.1%. Its measured daily beta to QQQ was -1.00.
What PSQ's own numbers say
PSQ returned negative 59.3% over 1444 sessions, with annualized volatility of 22.35%. The deepest drawdown was 60.96%, from 2022-10-14 to 2026-10-02, and the longest drawdown lasted 994 sessions with no recovery. The calendar years read negative 25.93% in 2021, 36.42% in 2022, negative 32.02% in 2023, negative 15.68% in 2024, negative 15.51% in 2025 and negative 16.84% for 2026 so far. The fund was up on 45.53% of days. Its beta to QQQ was negative 1.0 with a correlation of negative 1.0, and beta to SPY was negative 1.28 with a correlation of negative 0.94.
A fund with a daily target of negative 1x does not return the negative of the index over a longer period. Over this window QQQ returned 151.07% and PSQ returned negative 59.3%. The leverage table sets PSQ beside a frictionless daily-reset ideal of negative 70.14% for the whole window, so the fund lost less than the ideal. In 2023 QQQ returned 54.81% and PSQ lost 32.02%, against negative 37.48% for the ideal. In 2022 QQQ lost 32.39% and PSQ gained 36.42%, against 33.34% for the ideal. The ideal leaves out fees, borrowing costs and interest on cash, so the table describes the window and does not say which of those moved PSQ above the line.
The RSI figures explain why mean reversion found nothing. The 2-period RSI fell below 10 on 241 sessions, and the median 5-day forward return after those sessions was negative 0.49%, the same as the baseline median of negative 0.49%. The 20-day median was negative 1.65% against a baseline of negative 1.55%. The 14-period RSI fell below 30 on 107 sessions, with a median 5-day return of 0.04% and a median 20-day return of negative 1.33%. A washout in PSQ did not predict a bounce, which fits the RSI(2) snapback's profit factor of 1.01. The RSI(14) rule held a position 66.5% of the time and lost 5.72% a year, with an average loser of 8.64% against an average winner of 5.17%, including a 19.82% loss over 203 days from 2025-04-30 and a 17.33% loss over 206 days from 2023-01-24.
The split between overnight and intraday returns shows the same drift. PSQ's overnight log return was negative 36.52% and its intraday log return negative 51.53%, so 41.48% of the decline happened between the close and the next open and 58.52% during the session. The average overnight gap was 0.6% and the average daily range 1.58%. Rules that enter at the open take the overnight move before they act. Because both parts of the day were negative on average, neither the open nor the close offered a better entry in this window.
The trend picture is consistent. PSQ was above its 200-day average on 23.29% of sessions and crossed it 30 times. Trend rules bought a rally that failed more often than it held. The 200-day filter made 15 round trips and won 3, with a profit factor of 0.45, and the EMA 12/26 trend made 21 round trips and won 3.
Month averages are thin. May averaged negative 4.76% and November negative 4.03%, and September averaged 2.25%, each from 5 or 6 years. The fund's best days were 6.09% on 2025-04-04 and 5.51% on 2022-09-13, and its worst were negative 11.71% on 2025-04-09 and negative 7.4% on 2022-11-10.
For comparison, the other inverse funds show a similar pattern. SH, the inverse of the S&P 500, lost 10% a year as a holding and its best template, the RSI(2) snapback, returned negative 0.53%. SPDN lost 9.68% with a best template of negative 1.45%. QID, ProShares UltraShort QQQ, lost 29.46% with a best template of negative 1.85%, and SQQQ lost 42.33% with golden cross at negative 7.41%. Rules that avoid a fund that loses money do better than holding it, and few turn a profit.
The limits are the usual ones: one window, no fees in the headline runs, daily decisions, and a fund whose behaviour depends on the path of daily returns. The results describe PSQ over this period and carry no forecast.
Decay by year and the weekday averages
The leverage table breaks the gap between PSQ and the negative of QQQ into years. The decay gap was 3.25 points in 2021 and 4.02 in 2022, then widened to 22.79 in 2023 when QQQ gained 54.81%. It was 9.91 in 2024, 5.27 in 2025 and 5.6 in 2026. Over the whole window the gap to negative 1x of QQQ was 91.77 points. The years when the index rose by the most are the years when a daily-reset inverse fund falls short of the simple opposite, and that shows in 2023.
The weekday averages are small: Monday negative 0.17%, Tuesday negative 0.01%, Wednesday negative 0.09%, Thursday 0% and Friday 0%. Monday is the only weekday with a visible average, and the lag-1 autocorrelation of negative 0.04 gives a rule no daily pattern to work with. The beta to QQQ was negative 1 and the correlation negative 1, so PSQ was a mirror of QQQ day by day, with a beta of negative 1.28 to SPY.
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 PSQ?
Of the 12 templates tested on PSQ over 2021-01-04 to 2026-10-02, the strongest by CAGR was RSI(2) snapback at 0.1% (max drawdown 33.8%), versus −13.8% 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 PSQ?
12 of 12 templates beat PSQ buy-and-hold (−13.8% CAGR) on this window; 11 of 12 had a shallower maximum drawdown than holding (59.6%).
Which strategy worked best on PSQ?
The RSI(2) snapback was the only template with a positive return, 0.13% a year, ending at $10,077 from $10,000. Its drawdown was 33.82% against 59.64% for holding. At 5 basis points of slippage it turned to a loss of 2.77%.
Did any strategy beat buy-and-hold on PSQ?
All 12 beat holding on return, because holding lost 13.76% a year. Eleven also had a shallower maximum drawdown. The monthly cycle's 60.31% drawdown was deeper than holding's 59.64%.
Why did the dip buyer show a 100% win rate and still lose money?
It closed 3 trades, all winners near 8%, but a fourth position opened on 2023-01-27 stayed open and was marked 55.85% below entry at the end. The win rate counts closed trades only.
Did mean reversion work on PSQ?
Barely. After the 2-period RSI fell below 10, the median 5-day return was negative 0.49%, the same as the baseline. The RSI(2) snapback earned a profit factor of 1.01.
Which year did PSQ strategies make money?
2022, the one year PSQ gained, with a 35.3% return for holding. The dip buyer returned 55.2% and the RSI(2) snapback 42.7%. Every other year PSQ fell, and the templates mostly lost less than the holder.
Do trading costs matter on PSQ?
They matter for the busy rules. At 10 basis points the RSI(2) snapback fell from 0.13% to negative 5.6% a year and the weekly target from negative 6.76% to negative 11.55%.
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.