DeployQuantEEM trading strategies, backtested
iShares MSCI Emerging Markets ETF: emerging-market equities, on a different cycle than US indexes. Every DeployQuant template run on EEM over 5.7 years of minute data, same engine, same window, sorted by return.
EEM holds emerging-market stocks, and over 2021-01-04 to 2026-10-02 it did not behave like a US index fund. Buy-and-hold returned 6.58% a year, ended a $10,000 start at $14,416, and had a maximum drawdown of 39.09%. The fund peaked on 2021-02-17, bottomed on 2022-10-24, and did not recover that peak until 2025-09-15.
That weak, slow window is why EEM is the one broad index fund in the set where most templates beat holding. Six of the 12 templates finished with a higher CAGR than buy-and-hold, ten had a shallower drawdown, and ten had a positive CAGR. The weekly 7% target led at 11.62%, followed by RSI(14) mean reversion at 9.51% and the trend plus trailing stop template at 8.85%.
The two templates that lost money were 3-month momentum at negative 1.62% and SMA 10/50 at negative 3.89%. Every number here comes from one window with no fees or slippage in the headline run. The cost runs reorder the table.
| Strategy | CAGR | max DD | Sharpe | trades | win rate | final value |
|---|---|---|---|---|---|---|
| Weekly Entry + 7% Target | 11.6% | −27.4% | 0.72 | 101 | 26% (+1 open) | $18,802 |
| RSI(14) Mean Reversion | 9.5% | −14.0% | 0.80 | 18 | 83% | $16,851 |
| SMA-200 Trend + 15% Trailing Stop | 8.8% | −16.3% | 0.65 | 2 | 100% (+1 open) | $16,276 |
| Golden Cross (SMA 50/200) | 7.8% | −20.6% | 0.58 | 2 | 50% (+1 open) | $15,361 |
| 200-Day SMA Regime Filter | 7.8% | −17.7% | 0.60 | 15 | 27% (+1 open) | $15,362 |
| Drawdown Dip Buyer + 8% Target | 7.1% | −26.2% | 0.56 | 5 | 100% | $14,854 |
| RSI(2) Dip Snapback | 3.9% | −15.9% | 0.36 | 162 | 60% (+1 open) | $12,437 |
| First-to-Last Day of Month | 3.1% | −44.4% | 0.27 | 69 | 54% (+1 open) | $11,943 |
| 20-Day Momentum + Trailing Stop | 2.4% | −24.4% | 0.27 | 5 | 60% | $11,434 |
| EMA 12/26 Trend | 0.3% | −31.8% | 0.09 | 25 | 32% (+1 open) | $10,167 |
| 3-Month Momentum Switch | −1.6% | −32.7% | -0.06 | 13 | 23% | $9,106 |
| SMA 10/50 Trend | −3.9% | −41.2% | -0.23 | 23 | 26% (+1 open) | $7,962 |
| Buy & hold EEM | 6.6% | −39.1% | 0.44 | – | – | $14,416 |
Which templates beat holding EEM, and why
The six templates ahead of holding were the weekly 7% target, RSI(14) mean reversion, the trailing-stop template, the golden cross, the 200-day regime filter and the dip buyer. The golden cross and the 200-day filter beat holding by the same margin on CAGR, both at 7.76% against 6.58%. They took different routes to get there.
Almost all of the advantage came from 2021 and 2022, when EEM lost 4.2% and then 20.2% held. Three of the six winners spent those two years in cash. The golden cross, the 200-day filter and the trailing-stop template all show 0% for 2021 and 2022 because they had no position. In the yearly gaps against holding, the trend templates were ahead by 4.2 points in 2021 and 20.2 points in 2022, and behind in 2023 and 2025. The lead was made in the first two years and then partly given back in the rebound. EEM returned 33.9% in calendar 2025 and 24.4% in 2026 through 2026-10-02, and the slow templates entered those moves late. The golden cross made 27.1% in 2025 against 33.1% for holding.
RSI(14) mean reversion is the cleanest result. It made 18 round trips, won 83% of them, was invested 35.7% of the time, and had a profit factor of 9.02. Its drawdown of 14.01% was about a third of the 39.09% for holding. It was positive in every calendar year: 9.1% in 2021, 7.9% in 2022, 0.4% in 2023, 2.1% in 2024, 22.9% in 2025 and 13.5% in 2026. Its best trade was entered 2025-04-07 and closed 2025-04-29 for 11.91%, bought during the April 2025 sell-off. Its worst closed trade lost only 2.57%. Eighteen trades is a small sample, and one of them accounts for a large share of the gain, but the shape is consistent with a fund whose RSI(14) dips under 30 tended to bounce: 27 such sessions had a median forward 20-day return of 3.65% against a baseline of 0.54%.
The weekly 7% target did the most with the least discipline. It won 26% of its 101 round trips, had a losing streak of 13, and still returned 11.62%, because its average win was 7.19% and its average loss was 1.56%. It was invested 86.1% of the time, so it behaved mostly like holding with a profit cap and a re-entry rule. It beat holding in 2021, 2022, 2023, 2025 and 2026, and trailed in 2024 by 6.4 points. Its drawdown was 27.40%, shallower than holding but not by the margin of the RSI template. Its best month was November 2022 at 19.72%, in the window when EEM itself returned 14.98%.
The trailing-stop template made 2 closed trades and holds a third entered 2026-07-30. The first ran from 2023-01-05 to 2025-04-07 for 4.69% over 823 days, and the second from 2025-04-25 to 2026-07-29 for 44.84% over 460 days. Two wins and a 100% win rate say little. Most of its CAGR came from the second trade.
At the bottom, SMA 10/50 won 26% of its 23 round trips, had a losing streak of 13 and a profit factor of 0.52. Its 2021 and 2022 results were negative 16.8% and negative 14.6%, so the fast crossover was not even protective in the down years. Its drawdown of 41.17% ran from 2021-04-28 and was still unrecovered at the end of the window. The EMA 12/26 template was similar: 25 round trips, 32% wins, a profit factor of 1.01 and a 31.79% drawdown.
The conclusion for this fund is narrow. On a fund that drifted sideways for years and then rose sharply, the templates that did best either stayed out of the long decline entirely or bought a specific oversold condition.
How each strategy traded EEM
| Strategy | Time in market | Avg hold (days) | Best trade | Worst trade | Profit factor | With 10 bps slippage |
|---|---|---|---|---|---|---|
| weekly 7% target | 86.1% | 17 | 9.8% | −4.7% | 1.63 | 11.8% |
| RSI mean reversion | 35.7% | 41 | 11.9% | −2.6% | 9.02 | 8.8% |
| trend + trailing stop | 64.1% | 642 | 44.8% | 4.7% | – | 8.8% |
| golden cross | 59.0% | 325 | 11.3% | −10.3% | 0.99 | 7.7% |
| 200-day regime filter | 57.4% | 46 | 12.3% | −3.7% | 0.76 | 7.2% |
| dip buyer | 50.7% | 213 | 9.0% | 8.0% | – | 7.1% |
| RSI(2) snapback | 39.0% | 5 | 5.9% | −9.5% | 1.23 | −1.7% |
| monthly cycle | 95.2% | 28 | 11.7% | −10.4% | 1.20 | 0.8% |
| momentum breakout | 30.3% | 128 | 29.8% | −8.8% | 1.91 | 2.2% |
| EMA 12/26 trend | 54.2% | 44 | 22.2% | −6.3% | 1.01 | −0.6% |
| 3-month momentum | 44.7% | 72 | 24.0% | −12.8% | 0.72 | −2.0% |
| SMA 10/50 trend | 53.6% | 47 | 20.4% | −9.7% | 0.52 | −4.7% |
Costs, hold times and the trades that decided each result
The cost runs change the table more on EEM than on VOO. At 10 basis points per fill, RSI(2) snapback fell from 3.87% to negative 1.72%. It made 325 fills and had a profit factor of only 1.23 to begin with, with an average win of 1.27% and an average loss of 1.55%. At 5 basis points it made 1.04%. The monthly cycle, with 139 fills, went from 3.14% to 0.75%. EMA 12/26 dropped from 0.29% to negative 0.58%, and SMA 10/50 from negative 3.89% to negative 4.66%.
The weekly 7% target moved the other way: 11.62%, then 11.75% at 5 basis points and 11.85% at 10. Costs cannot help a strategy, so the small rise is path noise: a different fill price shifts which trades reach the 7% target. Differences of a fraction of a point between runs are not meaningful on a result built from 101 trades with a 26% win rate.
With 10 basis points the order was weekly 7% target at 11.85%, RSI(14) mean reversion at 8.85%, the trailing-stop template at 8.76%, golden cross at 7.67%, the 200-day filter at 7.19%, the dip buyer at 7.09%, and then the negative group. The top of the table barely changed because those templates trade rarely.
Hold times split the templates into two groups. The trend templates held for months: the trailing-stop template averaged 642 days, the golden cross 325 days and the dip buyer 213. The short-term templates held for days: RSI(2) snapback averaged 5 days and the monthly cycle 28. The dip buyer made 5 trades in all and every one hit its 8% target, one of them after 932 days, entered 2022-03-07 and closed 2024-09-24. That trade spans the 2022 drawdown, and it is why the template shows a 26.19% drawdown and 50.7% exposure. A 100% win rate on 5 trades with a 932-day hold is a different thing from a 100% win rate on 100 trades.
The golden cross shows how little a two-trade strategy can say. Its closed trades were a loss of 10.31% from 2023-01-25 to 2023-10-05 and a gain of 11.30% from 2024-01-05 to 2025-02-04. The open position, entered 2025-02-20 at an adjusted 43.69, was up 54.91% at the end of the window. Its profit factor on closed trades was 0.99. The result was set by whether the open position was counted, and the CAGR counts it.
The 200-day regime filter made 15 round trips and won only 4, 27%. Its profit factor was 0.76. Its winners were few and one was large: 12.34% from 2024-02-02 to 2024-12-19. Its current position, entered 2025-04-25 at 42.00, was up 61.14%. A filter that loses on most trades and still beats holding is relying on a few long holds, and it is also the template most exposed to how the window ends.
3-month momentum had 13 round trips and 3 wins. Its worst trade was a 12.81% loss from 2025-03-25 to 2025-04-07, entered one session before the April 2025 low. Its best was 23.99% held 330 days from 2025-05-05 to 2026-03-31. It had a 32.73% drawdown that started 2023-01-26, ended 2025-05-08, and took until 2026-06-18 to recover.
Why the moving-average rules lost on EEM
The two worst templates, SMA 10/50 and 3-month momentum, lost most of their money before the 2025 rally. SMA 10/50 made 5 round trips in 2021, 3 in 2022 and 4 in 2023, and none of those 12 won. It then won 2 of 5 in 2024 and 3 of 3 in 2026. The fund fell from its 2021-02-17 peak for almost two years, and a 10-day average crossing a 50-day average in a falling market produced entries that reversed within days. The 2022-02-11 trade lost 5.58% in 14 days. The best trade, 20.36% over 205 days from 2025-05-05 at 44.15, came after the April 2025 low. The worst, 9.71% from 2025-01-31 to 2025-04-07.
EMA 12/26 followed the same pattern with a faster signal. It won 0 of 6 round trips in 2021 and 0 of 5 in 2022, then 4 of 5 in 2024. Its median trade lost 1.34%, and its largest winner was 22.24% over 205 days from 2025-04-30 at 42.32. The CAGR of 0.29% is the sum of eleven early losses and a handful of large 2024 to 2026 wins. 3-month momentum had 13 round trips with a 23% win rate and a median return of negative 1.86%. Its best trade was 23.99% over 330 days from 2025-05-05, and its worst lost 12.81% over 13 days from 2025-03-25 to 2025-04-07. It won 0 of 3 trades in 2024 and 0 of 2 in 2025.
The golden cross had two completed trades. It entered 2023-01-25 at 38.61 and left 2023-10-05 at 34.63 for negative 10.31% over 253 days, then entered 2024-01-05 at 37.44 and left 2025-02-04 at 41.67 for 11.3% over 396 days. A third trade began on 2025-02-20 at 43.69 and was still open at the end of the window with a 54.91% gain. The CAGR of 7.76% depends mostly on that open position, which is a result of the 2025 rally rather than a closed trade.
The momentum breakout rule had 5 trades and 3 winners. The 29.84% winner ran 301 days from 2025-05-06 to 2026-03-03 and carried the 2.36% CAGR. Its two losers were 8.82% from 2024-05-20 and 7.68% from 2024-10-03, so it went 0 for 2 in 2024 and its yearly return that year was negative 15.5%.
The monthly cycle shows how much one calendar month can matter. It made 12 round trips in each full year, and its win count rose from 3 in 2022 to 9 in 2025. Its worst trade lost 10.39% from 2022-09-01 to 2022-09-30, and its best gained 11.67% in November 2022. A 44.39% drawdown, deeper than the fund's own 39.09%, came from holding through both ends of that cycle with no stop. The test cannot say whether a different calendar rule would have avoided September 2022, since this one is fixed in advance and the window has only one such month.
How EEM behaved
| Measure | EEM |
|---|---|
| Data in this test | 2021-01-04 to 2026-10-02 (1444 sessions) |
| Total return, buy and hold | 46.9% |
| Annualized volatility | 19.9% |
| Deepest drawdown | −39.8% (2021-02-17 to 2022-10-24) |
| Up days | 52.0% |
| Average daily range | 1.04% |
| Average overnight gap | 0.75% |
| Correlation to SPY | 0.67 |
| Correlation to QQQ | 0.69 |
| Correlation to TLT | 0.09 |
| Sessions above the 200-day average | 66.7% |
| Crossings of the 200-day average | 31 |
| Falls of 10% or more from a 20-day high | 14 |
Calendar years
| Year | Return |
|---|---|
| 2021 | −4.3% |
| 2022 | −20.6% |
| 2023 | 9.2% |
| 2024 | 6.4% |
| 2025 | 33.9% |
| 2026 (part) | 24.4% |
Biggest single days
| Best day | Move |
|---|---|
| 2022-03-16 | 8.1% |
| 2025-04-09 | 6.9% |
| 2026-04-08 | 5.5% |
| 2026-06-11 | 4.4% |
| 2022-11-04 | 4.2% |
| Worst day | Move |
|---|---|
| 2026-06-05 | −6.5% |
| 2026-06-23 | −5.7% |
| 2025-04-04 | −5.5% |
| 2026-03-03 | −4.9% |
| 2022-10-24 | −3.9% |
Average return by calendar month
| Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 2.9% | 0.0% | −1.1% | 1.1% | 2.2% | 1.7% | −0.9% | 0.2% | −1.0% | −0.7% | 3.0% | 0.6% |
Most and least correlated funds
| Most correlated | Least correlated | ||
|---|---|---|---|
| QAI | 0.81 | EEV | -0.98 |
| SOXX | 0.71 | SOXS | -0.71 |
| QQQE | 0.71 | SQQQ | -0.69 |
| SOXL | 0.71 | QID | -0.69 |
| IOO | 0.70 | PSQ | -0.69 |
How EEM behaved in this window
EEM's annualized volatility was 19.86%, its average intraday range was 1.04% and its average overnight gap was 0.75%. It rose on 51.98% of days, with an average up day of 0.92% and an average down day of negative 0.93%.
The return split is unusual. EEM lost 7.75% in log terms between the close and the next open, and gained 45.12% during the session. Overnight moves were a negative share of the total return, negative 20.76%, and intraday moves were 120.76%. Over this window the fund's gains came entirely while US markets were open, and the overnight gaps subtracted. A rule that buys at the open and sells at the open holds through the overnight leg. The same split for VOO went the other way, with most of its return coming overnight.
The calendar years were negative 4.30% in 2021, negative 20.64% in 2022, 9.17% in 2023, 6.39% in 2024, 33.91% in 2025 and 24.37% in 2026 through 2026-10-02. Half of the six years were under 7%. The drawdown from 2021-02-17 to 2022-10-24 reached 39.82% in the profile data, and the longest drawdown lasted 1,148 sessions. A recent drawdown of 14.01% began 2026-06-22 and was still open at the end of the data.
Correlation to SPY was 0.67 and to QQQ 0.69, with a beta to SPY of 0.81. Correlation to TLT was 0.09. The funds in the set most correlated with EEM were QAI at 0.81, then SOXX, QQQE and SOXL at 0.71. That is far lower than the 1.00 between VOO and SPY, which is the reason the US index templates and the EEM templates rank so differently. EEM is on its own cycle, which the fund note on the page says.
The biggest days were 2022-03-16 at 8.12%, 2025-04-09 at 6.91%, 2026-04-08 at 5.47%, 2026-06-11 at 4.42% and 2022-11-04 at 4.21%. The worst were 2026-06-05 at negative 6.52%, 2026-06-23 at negative 5.65%, 2025-04-04 at negative 5.53%, 2026-03-03 at negative 4.91% and 2022-10-24 at negative 3.89%. The worst days are concentrated in 2026, which fits the drawdowns that began in February and June of that year. 2022-10-24 was also the day of the bear-market low, the worst day in the 2022 decline fell at the trough.
Monthly averages were highest in November at 2.96%, January at 2.92% and May at 2.18%, and lowest in March at negative 1.06% and September at negative 0.98%. There are 5 or 6 observations per month, so these describe the sample and do not predict anything.
For the signals the templates use, EEM was above its 200-day average in 66.67% of sessions and crossed it 31 times, a lower share and more crossings than VOO, which is what a choppier fund looks like to a trend filter. RSI(14) fell below 30 in 27 sessions and above 70 in 78. After the 27 oversold sessions the median forward 5-day return was 2.94% and the 20-day return was 3.65%, both above the baselines of 0.21% and 0.54%. RSI(2) fell under 10 in 163 sessions, with median forward returns of 0.40% over 5 days and 1.20% over 20 days. That is above the baseline too. The edge was small next to the cost of trading it 162 times, which is why RSI(2) snapback ended at 3.87% before costs and 1.04% after 5 basis points. The first-lag autocorrelation was negative 0.07.
There were 14 falls of 10% or more from a 20-day high, over 34 days. The dip buyer used them, and made 5 trades.
The test stops at one window, daily decisions and no margin. EEM would also show a different template ranking from a window that began in a different year.
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 EEM?
Of the 12 templates tested on EEM over 2021-01-04 to 2026-10-02, the strongest by CAGR was weekly 7% target at 11.6% (max drawdown 27.4%), versus 6.6% 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 EEM?
6 of 12 templates beat EEM buy-and-hold (6.6% CAGR) on this window; 10 of 12 had a shallower maximum drawdown than holding (39.1%).
What was the best strategy for EEM?
The weekly 7% target had the highest CAGR at 11.62% against 6.58% for holding, with a 27.40% drawdown. RSI(14) mean reversion was second at 9.51% with a 14.01% drawdown and an 83% win rate on 18 trades. Both are results from one window.
Did any strategy beat buying and holding EEM?
Six of the 12 templates beat holding on CAGR, and ten had a shallower drawdown than the 39.09% for holding. The templates that lost money were [3-month momentum](/learn/strategies/quarterly-momentum/eem/) and [SMA 10/50](/learn/strategies/sma-10-50-trend/eem/).
Why did trend strategies beat holding EEM?
They were mostly in cash during 2021 and 2022, when EEM lost 4.2% and 20.2% held. The golden cross and the [200-day regime filter](/learn/strategies/sma200-regime/eem/) both returned 7.76%. They gave back part of the lead in 2025, when EEM returned 33.1% held.
How big was the EEM drawdown?
Held, EEM fell 39.09% from 2021-02-17 to 2022-10-24 and recovered on 2025-09-15, which is 1,057 days after the low. The profile table on the page shows 39.82% for the same peak and low.
How much do costs matter for EEM strategies?
They matter for the fast templates. RSI(2) snapback went from 3.87% to negative 1.72% at 10 basis points, and the monthly cycle from 3.14% to 0.75%. The golden cross barely moved, from 7.76% to 7.67%.
Is EEM correlated with US stocks?
Correlation to SPY was 0.67 and to QQQ 0.69 in this window, with a beta of 0.81 to SPY. [VOO](/learn/etf/voo/) is close to 1.00 with SPY. EEM's gains came intraday and its overnight moves were negative.
Other broad index 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.