DeployQuantEEV trading strategies, backtested
ProShares UltraShort MSCI Emerging Markets: -2x daily emerging markets. Every DeployQuant template run on EEV over 5.7 years of minute data, same engine, same window, sorted by return.
EEV is a fund built to return twice the opposite of the daily move in emerging market stocks, as tracked by EEM. Over the window from 2021-01-04 to 2026-10-02 emerging markets rose and EEV fell. Buy-and-hold lost 16.07% a year, turned $10,000 into $3,656 and had a maximum drawdown of 80.06%. The Sharpe was −0.26.
Eleven of the 12 templates finished ahead of holding on CAGR, and 11 had a shallower drawdown. Only 2 made money. RSI mean reversion returned 5% a year with a 56.79% drawdown, and the golden cross returned 0.71% with a 46.86% drawdown. The SMA-200 trailing stop was close to flat at −0.86%. At the other end, SMA 10/50 trend lost 24.43% a year and ended at $2,003, worse than holding, with an 81.25% drawdown.
An inverse leveraged fund in a rising market is a test of whether a rule can keep out of a steady loss without giving up the bursts. This fund had the bursts: it gained 37.31% in 2022, and its best single days were 12.76% on 2026-06-05 and 11.32% on 2026-06-23. The rest of the time it bled. The templates that earned anything did so from a handful of trades in 2022 and in the sell-offs of April 2025.
The results are a backtest on one window with no fees or slippage in the headline run, $10,000 starting cash, no margin and daily decisions. The fund trades about $241,075 a day, so fills assumed here are generous for the templates with many trades.
| Strategy | CAGR | max DD | Sharpe | trades | win rate | final value |
|---|---|---|---|---|---|---|
| RSI(14) Mean Reversion | 5.0% | −56.8% | 0.31 | 17 | 65% (+1 open) | $13,229 |
| Golden Cross (SMA 50/200) | 0.7% | −46.9% | 0.15 | 3 | 33% | $10,416 |
| SMA-200 Trend + 15% Trailing Stop | −0.9% | −54.6% | 0.09 | 11 | 36% | $9,518 |
| 200-Day SMA Regime Filter | −5.6% | −59.4% | -0.13 | 22 | 14% | $7,184 |
| RSI(2) Dip Snapback | −5.9% | −59.8% | -0.12 | 162 | 53% | $7,054 |
| Weekly Entry + 7% Target | −6.2% | −61.5% | -0.03 | 188 | 30% (+1 open) | $6,928 |
| 3-Month Momentum Switch | −9.5% | −55.3% | -0.32 | 17 | 18% | $5,636 |
| Drawdown Dip Buyer + 8% Target | −9.9% | −77.2% | -0.13 | 9 | 100% (+1 open) | $5,490 |
| First-to-Last Day of Month | −10.0% | −75.0% | -0.09 | 69 | 42% (+1 open) | $5,474 |
| EMA 12/26 Trend | −12.7% | −56.9% | -0.37 | 23 | 26% | $4,573 |
| 20-Day Momentum + Trailing Stop | −14.8% | −61.9% | -0.61 | 25 | 24% | $3,983 |
| SMA 10/50 Trend | −24.4% | −81.3% | -0.93 | 24 | 17% | $2,003 |
| Buy & hold EEV | −16.1% | −80.1% | -0.26 | – | – | $3,656 |
RSI mean reversion made 17 round trips with a 65% win rate, a profit factor of 1.22, an average win of 12.47% and an average loss of 13.89%. Its best trades were 29.22% from 2025-02-18 to 2025-04-07 and 23.64% from 2022-04-04 to 2022-04-25, both inside the 2022 and early 2025 moves in the fund. It lost 31.57% on a 209-day trade from 2025-04-29 to 2025-11-24, entered after the fund's April 2025 spike, and that trade accounts for the drawdown of 56.79% that began on 2025-04-29. By year it returned 62.5% in 2022 and 29.1% in 2021, against holding at 36.5% and −3.6%, and lost 22.5% in 2026 against −40.8%.
The golden cross made 3 trades. The first, from 2021-10-20 to 2023-01-17, gained 29.42% over 454 days and ended on 2023-01-17, after the fund's 2022-10-24 peak. The second lost 7.6% and the third lost 12.89% from 2025-01-28 to 2025-03-18. That is a 33% win rate and a CAGR of 0.71% for a template that spent 29% of sessions invested. The profit comes from 2022 and the cost from 2023, when it lost 27.6%.
The SMA-200 plus trailing stop made 11 round trips. The 2022 trades were the strong ones: 30.74% over 160 days from 2022-05-31 to 2022-11-07, 28.27% over 142 days from 2021-10-25 to 2022-03-16 and 7.35% over 71 days. The stop then fired on 2022-11-07 and the next trade lost 14.58% in 8 days, inside the drawdown from 2022-10-24 that reached 54.63%. In 2023 the template lost 29.7%, in 2025 14.5%. The result was a CAGR of −0.86%, a profit factor of 0.95 and 36% winners.
The dip buyer needs its own paragraph. It made 9 completed trades and all 9 won, each near 8%, with the best at 14.46%, giving a 100% win rate. Then it bought on 2022-11-07 at an adjusted price of 44.71 and never sold. By the end of the window that open position was down 76.87%. Its drawdown was 77.21%, its exposure 87.6% and its CAGR −9.92%. A win rate of 100% on 9 closed trades coexists with a loss of nearly all the capital because one trade carried the whole loss. See the dip buyer page for EEV.
Below those, the 200-day regime filter lost 5.6% a year with a 14% win rate and a longest losing streak of 17 trades. RSI(2) snapback returned −5.9% over 162 round trips, and the weekly 7% target returned −6.19% over 188. The 3-month momentum switch returned −9.5% and monthly cycle −9.96%. EMA 12/26 lost 12.74%, the 20-day breakout lost 14.82% and SMA 10/50 lost 24.43%. SMA 10/50's worst trade was a 25-day loss of 26.13% from 2025-04-07 to 2025-05-02, a buy signal that came just after the sell-off ended.
The momentum-style templates fit this fund badly. The 3-month momentum switch won 18% of 17 trades, and its best was 28.46% over 138 days from 2022-01-20 to 2022-06-07. Its worst was a 7-day loss of 15.85% from 2025-04-07 to 2025-04-14, taken the day after the fund's 2025-04-04 jump of 10.78%. EMA 12/26 won 26% of 23 trades and its profit factor was 0.3. The 20-day breakout won 24% of 25 trades with a profit factor of 0.25. Rules that buy strength in the fund after it has risen bought into the reversals, which were fast: the fund's worst day was −16.51% on 2022-03-16 and its second worst −14.02% on 2025-04-09.
How each strategy traded EEV
| Strategy | Time in market | Avg hold (days) | Best trade | Worst trade | Profit factor | With 10 bps slippage |
|---|---|---|---|---|---|---|
| RSI mean reversion | 62.4% | 75 | 29.2% | −31.6% | 1.22 | 4.4% |
| golden cross | 29.0% | 203 | 29.4% | −12.9% | 1.17 | 0.6% |
| trend + trailing stop | 39.0% | 75 | 30.7% | −14.6% | 0.95 | −1.2% |
| 200-day regime filter | 28.2% | 27 | 31.1% | −11.6% | 0.53 | −6.3% |
| RSI(2) snapback | 45.5% | 6 | 13.1% | −22.4% | 0.89 | −10.9% |
| weekly 7% target | 73.2% | 7 | 13.0% | −12.9% | 0.93 | −12.7% |
| 3-month momentum | 29.5% | 36 | 28.5% | −15.8% | 0.38 | −10.0% |
| dip buyer | 87.6% | 47 | 14.5% | 8.0% | – | −9.4% |
| monthly cycle | 95.2% | 28 | 24.7% | −20.7% | 0.87 | −12.1% |
| EMA 12/26 trend | 41.2% | 37 | 15.2% | −12.2% | 0.30 | −13.4% |
| momentum breakout | 25.6% | 21 | 10.5% | −10.6% | 0.25 | −15.6% |
| SMA 10/50 trend | 41.3% | 36 | 6.5% | −26.1% | 0.12 | −25.1% |
Time in market and the sign of the return line up in a way that is specific to inverse funds. The two positive templates were invested 62.4% and 29% of sessions. The momentum breakout, which lost the second most, was invested 25.6% of the time, so a low exposure did not protect it. What mattered was when the exposure fell. The profitable templates were invested in 2022 and mostly out in 2023, 2024 and 2025. The unprofitable ones had trades that lost in each of 2023, 2024 and 2025.
The trade lists have a common shape. Most of the best trades fall in 2022, March 2025 or March 2026. The weekly 7% target's best trade was 12.99% in two days in February 2022 and its worst was −12.92% in three days in March 2022. Monthly cycle's best month was September 2022 at 24.7% and its worst was November 2022 at −20.71%. The fund's own best month was September 2022 at 26.28% and its worst month was November 2022 at −26.38%. A template that takes calendar months holds both extremes.
Costs matter most for the high-turnover templates. RSI(2) snapback fell from −5.9% to −8.46% and −10.95%, and its ending value dropped from $7,054 to $5,138. The weekly 7% target fell from −6.19% to −9.76% and −12.73%. Monthly cycle fell from −9.96% to −11.02% and −12.06%. The dip buyer went from −9.92% to −9.35% and −9.45%. That is within a few tenths of a point of the headline figure, because it has been in one open position since 2022. RSI mean reversion went from 5% to 4.69% and 4.38%, so it stays positive at 10 basis points.
Liquidity belongs in the interpretation. Average daily dollar volume was $241,075 and the median minute volume 200 shares. A minute bar with 200 shares can hold only a small order before the price moves, so the test's flat basis-point costs understate the cost of the 188 and 162 round trip templates.
Holding periods also tell the story. The shortest average hold was 6 days for RSI(2) snapback and 7 days for the weekly target. Golden cross held longest at 202.7 days, then RSI mean reversion at 75.1 days. The two profitable templates held longer than most, which suits a fund whose gains came in a few extended runs, such as 2022. Trade counts of 3 and 17 are small, so a different start date could change the order.
One more comparison clarifies the year-by-year story. Holding returned −3.6% in 2021, 36.5% in 2022, −13% in 2023, −7.6% in 2024, −42.7% in 2025 and −40.8% in 2026 so far. In 2025 and 2026 holding lost heavily, and every template beat it in both years, mostly by being out of the fund. In 2022, when holding gained 36.5%, four templates beat it: RSI mean reversion at 62.5%, the dip buyer at 50.9%, monthly cycle at 50.1% and the trailing stop at 45.4%. The golden cross matched it at 36.5%. In 2023 holding lost 13%, and the templates split: the dip buyer lost 12.9% and RSI(2) snapback lost 8.3%, while the trailing stop lost 29.7% and golden cross lost 27.6%. The headline rank therefore depends on which year a reader weights.
How the losing rules traded EEV
The four losing trend and momentum rules on EEV made their best trades in 2022 and their worst in 2025 and 2026, with little in between. SMA 10/50 won 1 of 5 round trips in 2021 and 2 of 6 in 2022, then none of 4 in 2023 and none of 4 in 2024. Its best trade was only 6.51% over 58 days from 2022-04-12. Its worst was a 26.13% loss over 25 days from 2025-04-07 at 36.4 to 2025-05-02 at 26.89, entered during the April 2025 spike. A second large loss, 15.67% from 2026-03-13 to 2026-04-15, came in the weeks after the fund's 2026-03-03 gain of 10.11%. The average win of 4.1% against an average loss of 8.48% gave a profit factor of 0.12 and the worst CAGR on the page, negative 24.43%.
EMA 12/26 did slightly better by moving faster. Its best trade was 15.2% over 70 days from 2022-08-31, just before the fund's September 2022 best month. It lost 12.19% over 24 days from 2025-04-04 at 32.82 to 2025-04-28 at 28.82, a trade that spanned the fund's 2025-04-09 drop of 14.02%. It won 0 of 4 in 2021, 3 of 6 in 2022 and 0 of 4 in 2024, and its median trade lost 5.1%. The momentum breakout rule made 10 round trips in 2022 and won 3, then won 1 of 4 in 2025 and none of 3 in 2026. The 3-month momentum rule had a 28.46% winner over 138 days from 2022-01-20 and an 18% win rate. A 7-day trade from 2025-04-07 lost 15.85%.
The weekly 7% target shows how fast an inverse fund moves. It made 188 round trips, held 7 days on average and won 30%. The median trade lost 1.73% and lasted 3 days. Its best trade was 12.99% in 2 days from 2022-02-22 at 29.56 and its worst was a 12.92% loss in 3 days from 2022-03-14. It won 13 of 33 trades in 2021 and only 6 of 34 in 2025, the year the fund lost 43.54%, and its return that year was negative 33.8%.
The calendar averages show why a fixed-date rule does not fit. January averaged negative 5.08%, May negative 4.27% and November negative 4.84%, while September averaged 3.29% and July 1.95%, each from 5 or 6 observations. Wednesday averaged negative 0.25% and Friday 0.13%, which is small against a daily volatility that gives an average up day of 1.87%.
After 61 sessions with RSI(14) below 30, the median 5-day return was 1.24% against a baseline of negative 0.44%, and the median 20-day return was negative 1.06% against negative 1.32%. A bounce lasted about a week and then faded. The 2-period RSI below 10 on 198 sessions gave a median 5-day return of 0% and 20-day return of negative 0.53%. A rule that exits quickly after a washout read the first number correctly. A rule that holds for weeks read the second.
How EEV behaved
| Measure | EEV |
|---|---|
| Data in this test | 2021-01-04 to 2026-10-02 (1430 sessions) |
| Total return, buy and hold | −65.5% |
| Annualized volatility | 40.5% |
| Deepest drawdown | −81.1% (2022-10-24 to 2026-06-22) |
| Up days | 48.6% |
| Average daily range | 1.54% |
| Average overnight gap | 1.54% |
| Correlation to SPY | -0.65 |
| Correlation to QQQ | -0.67 |
| Correlation to TLT | -0.09 |
| Sessions above the 200-day average | 32.7% |
| Crossings of the 200-day average | 44 |
| Falls of 10% or more from a 20-day high | 52 |
Calendar years
| Year | Return |
|---|---|
| 2021 | −3.7% |
| 2022 | 37.3% |
| 2023 | −13.2% |
| 2024 | −7.8% |
| 2025 | −43.5% |
| 2026 (part) | −42.3% |
Biggest single days
| Best day | Move |
|---|---|
| 2026-06-05 | 12.8% |
| 2026-06-23 | 11.3% |
| 2025-04-04 | 10.8% |
| 2026-03-03 | 10.1% |
| 2022-05-05 | 8.6% |
| Worst day | Move |
|---|---|
| 2022-03-16 | −16.5% |
| 2025-04-09 | −14.0% |
| 2026-04-08 | −11.3% |
| 2022-11-04 | −8.6% |
| 2026-06-11 | −8.4% |
Average return by calendar month
| Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec |
|---|---|---|---|---|---|---|---|---|---|---|---|
| −5.1% | 0.5% | 1.1% | −2.2% | −4.3% | −3.5% | 1.9% | −0.4% | 3.3% | 1.7% | −4.8% | −0.8% |
Most and least correlated funds
| Most correlated | Least correlated | ||
|---|---|---|---|
| SOXS | 0.69 | EEM | -0.98 |
| SQQQ | 0.67 | QAI | -0.79 |
| QID | 0.67 | SOXL | -0.69 |
| PSQ | 0.67 | SOXX | -0.69 |
| TECS | 0.66 | QQQE | -0.69 |
Leverage and decay against EEM
| Year | EEV | EEM | -2× EEM, reset daily |
|---|---|---|---|
| 2021 | −3.7% | −4.3% | −0.8% |
| 2022 | 37.3% | −20.6% | 34.1% |
| 2023 | −13.2% | 9.2% | −21.9% |
| 2024 | −7.8% | 6.4% | −17.9% |
| 2025 | −43.5% | 33.9% | −49.3% |
| 2026 | −42.3% | 24.4% | −45.8% |
The last column compounds -2× EEM'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, EEV returned −65.5% while EEM returned 46.9% and a perfect daily-reset -2× version would have returned −76.5%. Its measured daily beta to EEM was -2.00.
EEV returned −65.51% in total over the window with annualized volatility of 40.45%. The maximum drawdown was 81.06% from 2022-10-24 to 2026-06-22, the longest drawdown lasted 975 sessions and the fund never recovered the 2022 peak. Calendar returns were −3.7% in 2021, 37.31% in 2022, −13.24% in 2023, −7.76% in 2024, −43.54% in 2025 and −42.27% in 2026 so far. The fund closed up on 48.64% of days, with an average up day of 1.87% and an average down day of −1.9%.
The leverage table compares EEV with EEM and with a frictionless version of the same target. EEM returned 46.91% over the window. A perfect daily-reset −2× version of EEM would have returned −76.53%, and EEV returned −65.51%, a smaller loss than that ideal. The realized beta was −2 with an R squared of 0.97. By year, the fund and the ideal were close in 2021 (−3.7% against −0.8%) and 2022 (37.31% against 34.06%). In 2025 EEM rose 33.91% and the fund fell 43.54%, against −49.26% for the ideal. The table leaves out fees, borrowing costs and interest on cash, so the data cannot say which of them put EEV above the line.
The fund's return split is unusual. The overnight log return was 39.03% and the intraday log return was −143.49%, so the overnight share was −37.36% and the intraday share 137.36%. All of the loss came during the trading day, and the overnight gaps worked for the fund. The average overnight gap was 1.54% and the average intraday range 1.54%. Lag 1 autocorrelation was −0.08.
Correlations confirm the design. EEV's correlation to EEM was −0.98, to SPY −0.65 and to QQQ −0.67, with a beta to SPY of −1.6. It correlated 0.69 with SOXS and 0.67 with SQQQ, QID and PSQ. The funds it moved against were EEM, QAI at −0.79, and SOXL and SOXX at −0.69. The fund was above its 200-day average on 32.74% of sessions and crossed it 44 times. There were 52 falls of 10% or more from a 20-day high, covering 384 days, which means a dip buyer had signals nearly all the time.
The RSI data show that dips did not mean-revert here. After 61 sessions with RSI(14) below 30, the median 5-day move was 1.24% against a baseline of −0.44%, but the median 20-day move was −1.06% against −1.32%, almost the same as baseline. After 198 sessions with RSI(2) below 10, the medians were 0% and −0.53%. The short-term bounce exists but the longer move does not.
The monthly averages were highest in September at 3.29%, July at 1.95% and October at 1.67%, and lowest in January at −5.08%, November at −4.84% and May at −4.27%, with 5 or 6 observations per month. Among other inverse funds, SQQQ held −42.33% a year and its best template, golden cross, returned −7.41%. SOXS held −48.27% and its best, the trailing stop, returned −4.93%. PSQ held −13.76% and RSI(2) snapback returned 0.13%, and TBF was the one inverse fund that rose, holding 12.21%. The templates that gained on EEV were the ones invested during 2022.
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 EEV?
Of the 12 templates tested on EEV over 2021-01-04 to 2026-10-02, the strongest by CAGR was RSI mean reversion at 5.0% (max drawdown 56.8%), versus −16.1% 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 EEV?
11 of 12 templates beat EEV buy-and-hold (−16.1% CAGR) on this window; 11 of 12 had a shallower maximum drawdown than holding (80.1%).
Which strategy worked best on EEV?
RSI mean reversion returned 5% a year with a 56.79% drawdown, against −16.07% for holding. It made 17 round trips with a 65% win rate. Its biggest trades were 29.22% and 23.64%, and its biggest loss was 31.57%.
Why did the dip buyer show a 100% win rate and still lose money on EEV?
It closed 9 trades at about 8% or better, then entered on 2022-11-07 at 44.71 and never reached its target. That open trade was down 76.87% at the end, giving a drawdown of 77.21% and a CAGR of −9.92%.
How many strategies beat buy-and-hold on EEV?
Eleven of 12 beat holding on CAGR and 11 had a shallower drawdown. Only 2 had a positive CAGR, RSI mean reversion at 5% and golden cross at 0.71%. SMA 10/50 lost more than holding at −24.43%.
Why did EEV lose less than twice EEM's gain?
EEM returned 46.91% over the window and EEV returned −65.51%. A perfect daily-reset −2× version of EEM would have returned −76.53%, so EEV lost less than that ideal. The ideal leaves out fees, borrowing costs and interest on cash.
Is EEV liquid enough to trade?
Average daily dollar volume was $241,075 and the median minute volume was 200 shares. The test applies flat basis-point costs, so high-turnover templates such as RSI(2) snapback at 162 trades are likely to cost more in practice.
Does mean reversion work on EEV?
Partly. After RSI(14) fell below 30 the median 5-day move was 1.24% against a −0.44% baseline, but the 20-day move was close to baseline. RSI mean reversion was the best template and RSI(2) snapback returned −5.9%.
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.