DeployQuantREW trading strategies, backtested
ProShares UltraShort Technology: -2x daily technology sector. Every DeployQuant template run on REW over 5.7 years of minute data, same engine, same window, sorted by return.
REW is ProShares UltraShort Technology, a fund that targets minus two times the daily return of a US technology index. The leverage table measures it against XLK, the technology sector fund, and finds a realized beta of 2.03 negative. From 2021-01-04 to 2026-10-02 REW lost 95.01% of its value, a CAGR of 40.68% negative, on annualized volatility of 51.89%. Its max drawdown of 95.02% began on 2021-01-06 and had not recovered by the end of the window.
Holding REW in the strategy engine, with $10,000 and no fees, ended at $763. That is a CAGR of 36.12% negative with a max drawdown of 92.67%. Ten of the twelve rule sets lost less than holding, ten had a shallower drawdown, and none ended with a positive CAGR. The best result was the golden cross at 6.71% negative, and the worst was the monthly cycle at 37.19% negative, which was also the only rule set besides the weekly target to finish behind holding.
The central fact for this fund is that its underlying, XLK, rose 226.13% over the same dates. A fund built to move opposite to technology, with a daily reset, will lose in a market that climbs in most months. The rules on this page do not change that. They differ in how much of the loss they avoid by being out of the fund. Rules that keep exposure low and make few trades come out ahead on this fund, and the rules that hold it for long stretches come out behind.
The momentum breakout and trend plus trailing stop templates are backtest-only for now, so their rows describe simulated history. Each run uses daily decisions, minute-bar fills, no margin, and no fees or slippage in the headline figures.
| Strategy | CAGR | max DD | Sharpe | trades | win rate | final value |
|---|---|---|---|---|---|---|
| Golden Cross (SMA 50/200) | −6.7% | −58.5% | -0.08 | 2 | 50% | $6,713 |
| SMA-200 Trend + 15% Trailing Stop | −7.1% | −50.2% | -0.08 | 20 | 20% | $6,544 |
| 20-Day Momentum + Trailing Stop | −8.1% | −62.8% | -0.13 | 44 | 34% | $6,149 |
| RSI(2) Dip Snapback | −10.9% | −72.4% | -0.21 | 166 | 55% (+1 open) | $5,138 |
| 200-Day SMA Regime Filter | −13.0% | −61.3% | -0.28 | 19 | 16% | $4,499 |
| 3-Month Momentum Switch | −15.7% | −70.8% | -0.34 | 18 | 11% | $3,746 |
| EMA 12/26 Trend | −18.5% | −71.7% | -0.38 | 22 | 18% | $3,097 |
| RSI(14) Mean Reversion | −18.5% | −76.5% | -0.40 | 17 | 47% (+1 open) | $3,091 |
| SMA 10/50 Trend | −19.8% | −77.0% | -0.42 | 21 | 19% | $2,828 |
| Drawdown Dip Buyer + 8% Target | −33.5% | −91.1% | -0.67 | 1 | 100% (+1 open) | $961 |
| First-to-Last Day of Month | −37.2% | −93.4% | -0.70 | 69 | 32% (+1 open) | $692 |
| Weekly Entry + 7% Target | −39.1% | −95.3% | -0.99 | 224 | 29% | $578 |
| Buy & hold REW | −36.1% | −92.7% | -0.73 | – | – | $763 |
How each rule set lost less or more than holding
The golden cross made 2 round trips and was invested 18.4% of the time. It gained 15% from 2022-03-21 to 2023-01-31 over a 316-day hold and then lost 42.41% from 2025-04-04 to 2025-06-13. Its average loss was larger than its average win, which gave a profit factor of 0.31, and its drawdown was 58.47%. The result of 6.71% negative a year rests on two trades, and the winning one fell in 2022, the one year the fund rose, at 65.85%.
The trend plus trailing stop came second at 7.12% negative. It made 20 round trips with a 20% win rate, an average win of 20.73% and an average loss of 6.98%, a profit factor of 0.63. Its best trade gained 31.7% from 2025-02-28 to 2025-04-07 and its drawdown was 50.17%. The momentum breakout lost 8.12% a year on 44 round trips with a 34% win rate. Its best trade, 34.84%, ran from 2025-03-24 to 2025-04-07, and nine of its 18 trades in 2022 were winners.
The RSI(2) snapback lost 10.95% a year over 166 round trips with a 55% win rate. Its average win of 3.89% was smaller than its average loss of 5.31%, a profit factor of 0.89. It was the rule with the best year of any strategy in 2022, at 88.4%, and then lost in each year from 2023 to 2026, with 36.1% negative in 2026. It beat holding in all six calendar years, yet its max drawdown of 72.41% began on 2023-03-14 and had no recovery in the window.
The 200-day regime filter lost 12.99% a year and won 16% of 19 trades. The 3-month momentum won 11% of 18 trades and lost 15.72% a year. The EMA 12/26 trend lost 18.47% with an 18% win rate, and the SMA 10/50 trend lost 19.75% with a 19% win rate. Each trend rule won most of its trades in 2022 or none at all. The SMA 10/50 won all 3 of its 2022 trades and none of its 4 trades in 2021.
The RSI mean reversion lost 18.49% a year with 17 round trips and a 47% win rate. Its average loss of 20.01% outweighed its average win of 14.2%. Its worst trade lost 45.8% over the hold from 2025-04-30 to 2025-11-19. The dip buyer won its first trade, 8.07% from 2021-02-02 to 2021-03-04, and then bought again and held to the end of the window, finishing with a loss of 33.5% a year and a 91.1% drawdown.
The monthly cycle lost 37.19% a year, worse than holding, with 69 round trips, a 32% win rate and exposure of 95.2%. The weekly 7% target lost 39.14% a year on 224 round trips with a 29% win rate and a drawdown of 95.26%, also worse than holding. Its worst trade lost 23.95% from 2025-04-07 to 2025-04-10.
Which months paid, how long the losing runs lasted, and what stayed open
April 2022 was REW's best month for holding at 27.51%, and the golden cross made 28.23%, the 3-month momentum switch 28.04% and the dip buyer 27.09% in it. September 2022 was the best month for the trailing stop strategy at 26.92%, the 200-day filter at 26.77%, the weekly target at 26.42%, the momentum breakout at 24.31%, the EMA 12/26 trend at 23.98% and the SMA 10/50 trend at 21.47%. April 2026 was the worst month for holding at negative 25.62% and for the dip buyer at negative 24.29%, the SMA 10/50 trend at negative 19.04% and the 3-month switch at negative 18.02%. The RSI(2) snapback's worst month was also April 2026, at negative 30.21%, and its worst trade, from 2026-04-02 to 2026-04-29, lost 30.86%. The RSI(14) rule's worst month was the following May, at negative 30.35%.
The losing runs are long for rules that exit on a signal. The 3-month momentum switch lost 13 round trips in a row and won 1 in a row at most. The weekly target lost 13 in a row and won at most 5. The 200-day filter, the EMA 12/26 trend, the momentum breakout and the SMA 10/50 trend each lost 9 in a row. The RSI(2) snapback won 8 in a row and lost 6.
The yearly gaps to holding fall into two groups. In 2022, the only year REW rose, every trend and breakout rule trailed holding: the 200-day filter by 95.2 points, the 3-month switch by 92.8, the trailing stop strategy by 71.9, the EMA 12/26 trend by 71.4 and the weekly target by 64.6. In 2021, 2023, 2024, 2025 and 2026 the same rules beat holding, most of them by wide margins, such as the 3-month switch by 54.1 points in 2021 and 51.3 in 2023. The RSI(2) snapback beat holding in all six years, from 2.9 points in 2024 to 42.2 in 2023.
None of the rules recovered from its worst drawdown. Holding's 92.67% drawdown from 2021-01-06 was still open at the end, and the shallowest of the rule drawdowns, the trailing stop strategy's 50.17% from 2022-11-03, had no recovery date either. The dip buyer bought on 2021-03-12 at an adjusted 179.14 and is marked 94.47% below that price at the last close, with its single closed trade made in early 2021. The RSI(14) rule entered on 2026-08-18 at 11.42 and was down 13.22% at the end, and the monthly cycle entered on 2026-10-01 at 10.319 and was down 3.96%. The RSI(2) snapback is the only rule with an open gain, 1.54% from 2026-10-02 at 9.76.
How each strategy traded REW
| Strategy | Time in market | Avg hold (days) | Best trade | Worst trade | Profit factor | With 10 bps slippage |
|---|---|---|---|---|---|---|
| golden cross | 18.4% | 193 | 15.0% | −42.4% | 0.31 | −6.8% |
| trend + trailing stop | 19.0% | 20 | 31.7% | −15.2% | 0.63 | −7.8% |
| momentum breakout | 21.3% | 10 | 34.8% | −10.1% | 0.76 | −9.5% |
| RSI(2) snapback | 48.9% | 6 | 14.3% | −30.9% | 0.89 | −15.9% |
| 200-day regime filter | 16.6% | 19 | 5.8% | −12.9% | 0.13 | −13.6% |
| 3-month momentum | 22.8% | 27 | 15.7% | −18.7% | 0.20 | −16.3% |
| EMA 12/26 trend | 29.6% | 29 | 10.7% | −12.3% | 0.16 | −19.1% |
| RSI mean reversion | 67.5% | 81 | 24.3% | −45.8% | 0.55 | −19.1% |
| SMA 10/50 trend | 30.9% | 31 | 24.3% | −15.1% | 0.26 | −20.4% |
| dip buyer | 98.2% | 30 | 8.1% | 8.1% | – | −35.4% |
| monthly cycle | 95.2% | 28 | 24.9% | −29.9% | 0.55 | −38.6% |
| weekly 7% target | 65.7% | 5 | 13.1% | −23.9% | 0.68 | −42.2% |
The single days that decided the trades
A fund with two times inverse exposure turns one violent day into a trade-defining event. REW's worst day was 2025-04-09, when it fell 27.01%, and its best days were 2025-04-03 and 2025-04-04 at 13.79% and 13.27%. The weekly target's worst trade, from 2025-04-07 to 2025-04-10, spans that 27.01% drop. The trend plus trailing stop's best trade ended on 2025-04-07 and the momentum breakout's best ended the same day, both captured the run-up into early April 2025 and exited before the reversal.
A rule that holds through an event like that carries the full loss. The RSI(2) snapback's best trade was 14.34% from 2025-04-02 to 2025-04-04, which caught the two best days, and its worst was 30.86% negative from 2026-04-02 to 2026-04-29. The monthly cycle lost 29.86% in May 2026 and 28.96% in April 2026, which are its two worst trades. Single-trade losses of that size are large compared with the 7% target and the 10% and 15% trailing stops the templates use, and a fund that moves 27% in one day can pass through any of them in a single session.
Median trades split the group. The RSI(2) snapback had a positive median trade of 0.71% over 5 days. The trend rules had median trades of 2.92% to 9.43% negative. The weekly target's median was 2.55% negative over 3 days. That the RSI(2) snapback's median is positive while its CAGR is negative shows the shape of the problem. Many small wins, with an average of 3.89%, were erased by a few large losses.
The cost runs add to that. At 5 basis points the RSI(2) snapback went from 10.95% negative to 13.47% and at 10 basis points to 15.91%. The weekly target went from 39.14% negative to 40.79% and 42.16%. The golden cross barely moved, from 6.71% negative to 6.75% and 6.8%. A further limit applies to REW in particular. The fund trades about $420,530 a day on average, and its median minute volume is 57 shares. Fills on minute bars in a backtest assume the order trades at the bar price, and a fund this thin would not always allow that. The cost runs of 5 and 10 basis points may understate the real cost of trading it.
How REW behaved
| Measure | REW |
|---|---|
| Data in this test | 2021-01-04 to 2026-10-02 (1444 sessions) |
| Total return, buy and hold | −95.0% |
| Annualized volatility | 51.9% |
| Deepest drawdown | −95.0% (2021-01-06 to 2026-10-02) |
| Up days | 44.8% |
| Average daily range | 3.16% |
| Average overnight gap | 1.51% |
| Correlation to SPY | -0.90 |
| Correlation to QQQ | -0.96 |
| Correlation to TLT | -0.07 |
| Sessions above the 200-day average | 19.2% |
| Crossings of the 200-day average | 38 |
| Falls of 10% or more from a 20-day high | 73 |
Calendar years
| Year | Return |
|---|---|
| 2021 | −55.5% |
| 2022 | 65.8% |
| 2023 | −61.4% |
| 2024 | −33.5% |
| 2025 | −43.2% |
| 2026 (part) | −53.6% |
Biggest single days
| Best day | Move |
|---|---|
| 2025-04-03 | 13.8% |
| 2025-04-04 | 13.3% |
| 2026-06-05 | 12.5% |
| 2022-09-13 | 11.3% |
| 2022-06-13 | 10.2% |
| Worst day | Move |
|---|---|
| 2025-04-09 | −27.0% |
| 2022-11-10 | −17.6% |
| 2026-07-30 | −10.8% |
| 2022-11-30 | −10.1% |
| 2026-08-04 | −10.1% |
Average return by calendar month
| Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec |
|---|---|---|---|---|---|---|---|---|---|---|---|
| −1.4% | 0.5% | −2.0% | −2.5% | −12.9% | −7.4% | −3.1% | −1.9% | 5.0% | −7.7% | −8.5% | 1.0% |
Most and least correlated funds
| Most correlated | Least correlated | ||
|---|---|---|---|
| TECS | 0.98 | ROM | -0.99 |
| QID | 0.97 | TECL | -0.98 |
| SQQQ | 0.96 | XLK | -0.98 |
| PSQ | 0.96 | TQQQ | -0.97 |
| SDS | 0.89 | QLD | -0.96 |
Leverage and decay against XLK
| Year | REW | XLK | -2× XLK, reset daily |
|---|---|---|---|
| 2021 | −55.5% | 37.0% | −52.4% |
| 2022 | 65.8% | −27.7% | 38.7% |
| 2023 | −61.4% | 56.0% | −63.0% |
| 2024 | −33.5% | 21.6% | −41.7% |
| 2025 | −43.2% | 24.6% | −48.9% |
| 2026 | −53.6% | 39.3% | −56.9% |
The last column compounds -2× XLK'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, REW returned −95.0% while XLK returned 226.1% and a perfect daily-reset -2× version would have returned −96.9%. Its measured daily beta to XLK was -2.03.
How REW behaves, and why it keeps losing
REW rose on 44.84% of sessions. The average up day was 2.52% and the average down day was 2.39% negative. The average intraday range was 3.16% and the average overnight gap was 1.51%. The fund had 73 separate cases of a 10% fall within 20 days, covering 660 trading days.
Calendar-year returns were 55.51% negative in 2021, 65.85% in 2022, 61.4% negative in 2023, 33.55% negative in 2024, 43.16% negative in 2025 and 53.63% negative in 2026 for the part of the year in the window. 2022 is the only positive year. XLK, the underlying, returned 36.99% in 2021, 27.73% negative in 2022, 56.01% in 2023, 21.64% in 2024, 24.61% in 2025 and 39.31% in 2026. In every year XLK rose, REW lost.
The leverage table sets REW beside XLK and beside a frictionless daily-reset −2× version of XLK. Over the whole window that ideal would have lost 96.86%, and REW lost 95.01%, close to it. By year the two tracked each other: 2021 was negative 55.51% for REW against negative 52.36% for the ideal, 2023 was negative 61.4% against negative 63.03%, and 2022 was 65.85% against 38.71%. The 2022 gap is the widest favourable difference in the table. The ideal leaves out fees, borrowing costs and interest on cash, so the data does not say what moved REW off the line. The realized beta was 2.03 negative with an R-squared of 0.97, and the long-run loss came from compounding a daily inverse in a rising market.
Correlation to SPY was 0.9 negative and to QQQ 0.96 negative, with betas of 2.83 negative and 2.24 negative. The funds most correlated with REW are TECS at 0.98, QID at 0.97, SQQQ at 0.96 and PSQ at 0.96. The fund was above its 200-day average on 19.2% of sessions and crossed it 38 times.
The overnight log return was 102.5% negative and the intraday log return was 193.38% negative. The overnight share of the return was 34.64% and the intraday share was 65.36%, so most of the damage happened during the trading day. A rule that holds the fund only overnight would have avoided most of the loss, and a rule that holds it only during the day would have taken most of it.
RSI is of limited use here. RSI(14) fell below 30 on 142 sessions, with a median forward 5-day return of 0.04% against a baseline of 1.12% negative, and a median 20-day return of 4.67% negative against a baseline of 4.09% negative. RSI(2) below 10 occurred on 265 sessions, with a median forward 5-day return of 0.93% negative. The 5-day reading after an oversold RSI(14) was better than the baseline, which fits the RSI(2) snapback's positive median trade, but the 20-day readings were not.
Monthly averages show May at 12.89% negative, October at 7.74% negative and November at 8.5% negative, with September at 5.01% positive. With 5 or 6 observations for each month these are descriptions and not seasonal rules.
The other inverse funds in the category show how the same rules did on different underlyings. SOXS was best with the trend plus trailing stop at negative 4.93%, against a buy-and-hold CAGR of negative 48.27%, and EEV was best with the RSI mean reversion at 5%. TBF, a rate-driven fund, had a buy-and-hold CAGR of 12.21%. The funds tied to US equity indexes, such as SH and SPDN, had buy-and-hold CAGRs of 10% negative and 9.68% negative, a smaller loss than REW's 36.12% negative. The same rules therefore met a milder decline on those index-wide inverse funds than on this technology fund, and the best rule on each of them was a short-horizon or trend rule that spent most of its time in cash.
Biggest days and the peer inverse funds
The best days were 13.79% on 2025-04-03, 13.27% on 2025-04-04 and 12.5% on 2026-06-05, and the worst were negative 27.01% on 2025-04-09, negative 17.62% on 2022-11-10 and negative 10.81% on 2026-07-30. With an average intraday range of 3.16% and an average overnight gap of 1.51%, one session can move the fund more than most rules in the table expect to make or lose in a trade. The 2025 pair is the same week that shows up on the QID and PSQ pages, since all three funds are short the Nasdaq-100 or a technology index. The lag-1 autocorrelation was negative 0.04, so one day's move said little about the next.
The peer table gives the holding returns for the other inverse funds. QID held for negative 29.46%, SQQQ for negative 42.33%, TECS for negative 46.67% and SOXS for negative 48.27%, against negative 36.12% for REW. The RSI(2) snapback was the best template on QID, PSQ, SH and SPDN, and the golden cross was best on SQQQ, SDS and REW. Only a few inverse funds in the table had a positive best-template result: PSQ at 0.13% and EEV at 5%.
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 REW?
Of the 12 templates tested on REW over 2021-01-04 to 2026-10-02, the strongest by CAGR was golden cross at −6.7% (max drawdown 58.5%), versus −36.1% 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 REW?
10 of 12 templates beat REW buy-and-hold (−36.1% CAGR) on this window; 10 of 12 had a shallower maximum drawdown than holding (92.7%).
Which strategy did best on REW?
The golden cross, at a CAGR of 6.71% negative against 36.12% negative for buy-and-hold. It made only 2 round trips and was invested 18.4% of the time. None of the twelve rule sets had a positive CAGR.
How much did REW lose from 2021 to 2026?
REW lost 95.01% over 2021-01-04 to 2026-10-02, a CAGR of 40.68% negative. Its max drawdown was 95.02%, and it had not recovered by the end of the window. 2022 was the only calendar year with a gain, at 65.85%.
Why does REW lose when technology rises?
REW targets two times the inverse of the daily move of a technology index. XLK rose 226.13% over the window, and REW's daily reset compounded the losses in each rising year. The fund's realized beta to XLK was 2.03 negative with an R-squared of 0.97.
Did the RSI(2) snapback work on REW?
It lost 10.95% a year, but it beat holding in all six calendar years and had a positive median trade of 0.71%. In 2022 it returned 88.4%. Its max drawdown was 72.41%, and the cost runs lowered its result to 13.47% negative at 5 basis points.
Can these results be traded as shown?
REW's average daily dollar volume was about $420,530 and its median minute volume was 57 shares. The backtest fills orders on minute bars with no slippage in the headline run. A fund this thin may cost more to trade than the 5 and 10 basis point cost runs assume.
Did the weekly 7% target work on REW?
No. It lost 39.14% a year over 224 round trips with a 29% win rate and a drawdown of 95.26%. It finished behind buy-and-hold, along with the monthly cycle.
What does this backtest leave out?
It covers one window from 2021-01-04 to 2026-10-02 with daily-decision rules and no fees in the headline run. It contains one year in which the fund rose. The results are hypothetical and are not a forecast.
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.