DeployQuantFXE trading strategies, backtested
Invesco CurrencyShares Euro Trust: the euro versus the US dollar. Every DeployQuant template run on FXE over 5.7 years of minute data, same engine, same window, sorted by return.
FXE is the euro against the US dollar in a fund wrapper, and it did very little in this window. Between 2021-01-04 and 2026-10-02 buy-and-hold lost 0.9% a year, ended $10,000 at $9,493 and had a maximum drawdown of 22.86%. The drawdown began on 2021-01-06 and the fund took until 2026-01-27 to recover, which is about 1,218 days. Annualized volatility was 7.44%.
All 12 templates finished ahead of holding on CAGR and all 12 had a shallower drawdown. That headline needs a careful reading. Two of the 12, the dip buyer and the 20-day momentum breakout, made no trades at all, so each ended at exactly $10,000 and beat a losing buy-and-hold by sitting in cash. Seven templates had a positive CAGR. The best was the SMA-200 trend plus 15% trailing stop at 2.14% a year with a 7.91% drawdown and a Sharpe of 0.41, and it too made no completed trade.
The pages for volatile funds describe what a rule did with large swings. On FXE the swings were small, and the interesting question is what a trend, mean reversion or calendar rule can extract from a fund that moves 0.39% on an average day and spent years going nowhere. The answer from this test is a small gain from the slower rules and a small loss from the rules that trade often.
The test is one window with no fees or slippage in the headline run, $10,000 starting cash, no margin and daily decisions. The differences between templates on this page are a few points of CAGR at most, so the cost runs and trade counts carry as much weight as the ranking.
| Strategy | CAGR | max DD | Sharpe | trades | win rate | final value |
|---|---|---|---|---|---|---|
| SMA-200 Trend + 15% Trailing Stop | 2.1% | −7.9% | 0.41 | 0 | – (+1 open) | $11,290 |
| SMA 10/50 Trend | 1.6% | −8.2% | 0.33 | 19 | 37% | $10,921 |
| EMA 12/26 Trend | 1.3% | −11.6% | 0.29 | 21 | 29% | $10,785 |
| 200-Day SMA Regime Filter | 1.3% | −6.4% | 0.29 | 17 | 29% | $10,766 |
| 3-Month Momentum Switch | 1.0% | −8.1% | 0.28 | 4 | 50% | $10,576 |
| Golden Cross (SMA 50/200) | 0.4% | −8.7% | 0.10 | 4 | 50% | $10,214 |
| RSI(14) Mean Reversion | 0.3% | −16.2% | 0.07 | 16 | 56% (+1 open) | $10,148 |
| RSI(2) Dip Snapback | 0.0% | −12.8% | 0.02 | 170 | 56% (+1 open) | $9,998 |
| Drawdown Dip Buyer + 8% Target | 0.0% | −0.0% | 0.00 | 0 | – | $10,000 |
| 20-Day Momentum + Trailing Stop | 0.0% | −0.0% | 0.00 | 0 | – | $10,000 |
| First-to-Last Day of Month | −0.3% | −22.4% | -0.00 | 69 | 52% (+1 open) | $9,815 |
| Weekly Entry + 7% Target | −0.4% | −20.6% | -0.01 | 63 | 8% | $9,761 |
| Buy & hold FXE | −0.9% | −22.9% | -0.08 | – | – | $9,493 |
The SMA-200 plus trailing stop returned 2.14% by being invested for one stretch. It entered on 2022-12-02 at an adjusted price of 91.65 and was still holding at the end of the window with a gain of 13.28%. The 15% stop was never hit, which fits a fund whose worst drawdown inside that hold was 7.91%. Exposure was 66.6%, because the 200-day average kept the template out for the first part of the window, including the fall to 2022-09-27. Its yearly returns were 0% in 2021, 2.1% in 2022, 4.7% in 2023, −4.1% in 2024, 14.1% in 2025 and −3.5% in 2026 so far. From 2023 on those match holding, which is what a position held throughout looks like. All the edge is in the first two years, when it avoided the fall. The trailing stop page shows the single open position.
The SMA 10/50 trend made 19 round trips with a 37% win rate and a profit factor of 1.69, and returned 1.55% with an 8.19% drawdown. Its average win was 3.33% and its average loss 1.1%. The best trade was 12.14% over 170 days from 2025-02-19 to 2025-08-08, and the second best was 7.81% over 113 days starting 2022-10-31. EMA 12/26 returned 1.32% on 21 round trips, with a 29% win rate, a profit factor of 1.57 and a drawdown of 11.59% that ran from 2021-05-25 to 2022-11-03. The 200-day regime filter returned 1.29% with the shallowest drawdown among the traders at 6.39%, a profit factor of 2.52 and a best trade of 7.55% over 365 days from 2025-03-06 to 2026-03-06. Its average loss was 0.41%, which is why a 29% win rate still made money.
The 3-month momentum switch made 4 round trips and returned 0.98%. Golden cross made 4 round trips and returned 0.37%, with its best trade 4.27% from 2025-04-15 to 2026-05-05 and its worst −1.97% from 2023-12-26 to 2024-12-06 over 346 days. Both trade counts are too small to say more than that they were invested less than half the time. RSI mean reversion returned 0.26% and had the deepest drawdown of the active templates at 16.18%, because it bought in 2021 and early 2022 and held through a decline: its trades from 2021-09-23 to 2022-02-17 and from 2022-02-25 to 2022-06-03 lost 3.49% and 4.61% over 147 and 98 days.
RSI(2) snapback made 170 round trips and finished at $9,998, a CAGR of 0%. Its average win was 0.49% and average loss 0.6%, and 56% of trades won. That ratio leaves almost no margin for costs, as the next section shows. Monthly cycle returned −0.32% and the weekly 7% target −0.42%, both with drawdowns of about 20% to 22%, close to holding's 22.86%. The weekly 7% target won 8% of 63 trades, with a longest losing streak of 32. Its average win was 7.25% and its average loss 0.64%, so the few winners nearly paid for the many small losses.
Zero-trade templates are worth a sentence each. The dip buyer enters after a 20-day drawdown worse than 10%, and FXE had no such event in the data, so it never entered. The momentum breakout enters after a 20-day return above 10%, and FXE never produced one. These results are real in the sense that they describe what the rules would have done. They are uninformative about the rules' quality, and the comparison with a losing buy-and-hold flatters them.
How each strategy traded FXE
| Strategy | Time in market | Avg hold (days) | Best trade | Worst trade | Profit factor | With 10 bps slippage |
|---|---|---|---|---|---|---|
| trend + trailing stop | 66.6% | – | – | – | – | 2.1% |
| SMA 10/50 trend | 45.9% | 51 | 12.1% | −1.9% | 1.69 | 0.9% |
| EMA 12/26 trend | 45.8% | 46 | 10.7% | −2.6% | 1.57 | 0.6% |
| 200-day regime filter | 49.5% | 62 | 7.5% | −1.0% | 2.52 | 0.7% |
| 3-month momentum | 23.0% | 121 | 7.6% | −3.3% | 1.96 | 0.8% |
| golden cross | 49.4% | 260 | 4.3% | −2.0% | 1.56 | 0.2% |
| RSI mean reversion | 53.5% | 69 | 3.3% | −4.6% | 1.37 | −0.3% |
| RSI(2) snapback | 45.3% | 5 | 2.4% | −4.3% | 1.03 | −5.6% |
| dip buyer | 0.0% | – | – | – | – | 0.0% |
| momentum breakout | 0.0% | – | – | – | – | 0.0% |
| monthly cycle | 95.2% | 28 | 5.3% | −4.7% | 0.97 | −2.6% |
| weekly 7% target | 91.1% | 29 | 7.5% | −2.9% | 0.93 | −2.6% |
Cost sensitivity is the clearest finding on this page. RSI(2) snapback fell from 0% to −2.87% at 5 basis points and −5.62% at 10 basis points, and its drawdown widened from 12.81% to 17.43% and 28.63%. Its ending value dropped from $9,998 to $8,462 and $7,172. Monthly cycle fell from −0.32% to −1.5% and −2.64%. The weekly 7% target fell from −0.42% to −1.46% and −2.6%. The SMA 10/50 template went from 1.55% to 1.21% and 0.89%, and EMA 12/26 from 1.32% to 0.96% and 0.6%. The slow templates barely moved: the trailing stop went from 2.14% to 2.13% and 2.12%, and golden cross from 0.37% to 0.3% and 0.24%.
The reason is the size of each trade's edge compared with the cost. On a fund with an average up day of 0.36% and an average down day of −0.35%, a round trip charged at 10 basis points on each side takes a large share of what the fund moves in a day. The templates with 63 to 170 round trips pay that many times. The fund's average daily dollar volume was $9,500,738 and its median minute volume was 299 shares.
Holding times split the field. RSI(2) snapback held 5.4 days on average and the weekly target 29.2 days, though its median hold was 3 days. Golden cross held 259.5 days on average. The regime filter averaged 61.8 days, but its median hold was 11 days because a few long trades pulled up the mean. In 2024 the regime filter made 8 round trips and won 1, then made 7 in 2026 and won 2. A fund that crosses its 200-day average 34 times in 5.7 years produces whipsaws, and the regime filter had 17 round trips to show for that.
The calendar templates reflect the fund's seasonal pattern only weakly. Monthly cycle's best month was April 2025 at 5.25% and its worst April 2022 at −4.67%, and across the year the average return by calendar month ranged from −1.35% in September to 0.87% in November, with 5 or 6 observations per month. Monthly cycle won 10 of 12 trades in 2025 and 4 of 12 in 2021 and 2022. The same template returned −7.6% in 2021, −8.2% in 2022 and 11.8% in 2025.
The yearly table shows where the returns came from. In 2024 holding lost 4%, and EMA 12/26 and RSI(2) snapback ended the year at 0% and 1.5%, while golden cross lost 2.1% and the regime filter lost 3.5%. In 2025 holding gained 14.1% and every active template trailed it: the weekly 7% target returned 13%, monthly cycle 11.8%, SMA 10/50 11.2%, EMA 12/26 9.9%, the regime filter 9%, RSI mean reversion 8.6%, RSI(2) snapback and 3-month momentum 7.5% and golden cross 4.3%. Exit rules that react after a move begins give up part of a sharp rally, and 2025 was the fund's one strong year. In 2021 and 2022, when holding lost 7.9% and 6.4%, templates that stayed flat or invested late, such as the regime filter at 0% and 2.1%, gained the most against it.
What the trade lists show year by year on FXE
The trade lists show when each rule worked on FXE, year by year. The weekly 7% target made 18 round trips in 2021 and won none. It won 1 of 15 in 2022, 1 of 7 in 2023, 1 of 9 in 2024 and 2 of 10 in 2025. A euro fund that moves 0.36% on an average up day rarely climbed 7% in a week, so the target was reached on a minority of the 63 trades, and the average loss was small at 0.64%. The best trade was 7.51% over 150 days from 2022-11-14 at 90.14 to 2023-04-13, held far longer than a weekly rule intends. The worst was a 2.85% loss over 10 days from 2022-04-18, in the month the fund fell 4.58%.
The monthly cycle is a coin flip with a year-to-year pattern. It made 12 round trips in each full year and won 4 in 2021, 4 in 2022, 8 in 2023, 6 in 2024, 10 in 2025 and 4 of 9 in 2026. Its best trade was 5.25% in April 2025, and its worst was a 4.67% loss in April 2022. April gave the best result in 2025 and the worst in 2022. The median trade returned 0.09% and trades lasted 29 days at the median.
The RSI mean reversion rule made 16 round trips and won 56%. It lost on all 3 of its 2022 trades, including 4.61% over 98 days from 2022-02-25 at 98.65, and lost 3.49% over 147 days from 2021-09-23. In 2026 it won all 3. Its best trade was 3.31% over 41 days from 2023-02-23. The average win of 1.69% and average loss of 1.57% left a profit factor of 1.37, and its drawdown was 16.18%, deeper than any trend template.
Among the slow rules, the golden cross took 4 trades with a median return of 1.86%. Its winners were 4.27% over 385 days from 2025-04-15 and 1.86% over 277 days from 2023-01-06. Its losers were 1.97% over 346 days from 2023-12-26 and 1.84% over 30 days from 2026-05-27. The 3-month momentum rule had 4 trades, winning 7.63% from 2025-03-19 and 4.51% from 2022-12-02, and losing 2.68% and 3.28% in the 2024 stretch. The 200-day regime filter won 2 of 2 trades in 2023, 1 of 8 in 2024 and 2 of 7 in 2026. Its best, 7.55% over 365 days from 2025-03-06, is the one hold the rule kept through a full leg of the 2025 rise.
The slow rules had 4 to 19 trades each, so a single trade moves their results by a point or more, and these figures describe one window.
How FXE behaved
| Measure | FXE |
|---|---|
| Data in this test | 2021-01-04 to 2026-10-02 (1444 sessions) |
| Total return, buy and hold | −4.8% |
| Annualized volatility | 7.4% |
| Deepest drawdown | −23.4% (2021-01-06 to 2022-09-27) |
| Up days | 48.4% |
| Average daily range | 0.39% |
| Average overnight gap | 0.27% |
| Correlation to SPY | 0.25 |
| Correlation to QQQ | 0.23 |
| Correlation to TLT | 0.18 |
| Sessions above the 200-day average | 57.4% |
| Crossings of the 200-day average | 34 |
| Falls of 10% or more from a 20-day high | 0 |
Calendar years
| Year | Return |
|---|---|
| 2021 | −8.1% |
| 2022 | −6.6% |
| 2023 | 4.9% |
| 2024 | −4.2% |
| 2025 | 14.5% |
| 2026 (part) | −3.6% |
Biggest single days
| Best day | Move |
|---|---|
| 2025-04-10 | 2.5% |
| 2022-11-04 | 2.2% |
| 2022-11-10 | 1.9% |
| 2025-03-05 | 1.8% |
| 2023-11-14 | 1.7% |
| Worst day | Move |
|---|---|
| 2024-11-06 | −1.7% |
| 2022-09-23 | −1.6% |
| 2022-07-05 | −1.6% |
| 2025-05-12 | −1.5% |
| 2023-01-03 | −1.5% |
Average return by calendar month
| Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec |
|---|---|---|---|---|---|---|---|---|---|---|---|
| −0.1% | −0.6% | 0.1% | 0.8% | 0.3% | −0.4% | −0.4% | 0.4% | −1.4% | −0.6% | 0.9% | 0.8% |
Most and least correlated funds
| Most correlated | Least correlated | ||
|---|---|---|---|
| UDN | 0.96 | USDU | -0.86 |
| EEM | 0.41 | EEV | -0.41 |
| IAU | 0.40 | KMLM | -0.26 |
| QAI | 0.36 | SDS | -0.25 |
| IGIB | 0.35 | SH | -0.25 |
FXE is a low-volatility fund. Annualized volatility was 7.44%, the average daily range 0.39% and the average overnight gap 0.27%. The fund closed up on 48.37% of days and lag 1 autocorrelation was −0.01. It had no falls of 10% or more from a 20-day high, which is why the dip buyer never traded. The total return was −4.76% over the window.
Calendar years were −8.07% in 2021, −6.57% in 2022, 4.86% in 2023, −4.17% in 2024, 14.5% in 2025 and −3.63% in 2026 so far. The maximum drawdown was 23.45% from 2021-01-06 to 2022-09-27, recovered on 2026-01-27 after a longest drawdown of 1,268 sessions. A second drawdown from that recovery date had reached 5.96% by 2026-10-01 and was still open. For almost the whole window, then, the fund was either falling or climbing back to its January 2021 level, and a template needed to avoid the fall without missing the 2025 rise.
The best days were 2.47% on 2025-04-10 and 2.17% on 2022-11-04, and the worst were −1.74% on 2024-11-06 and −1.59% on 2022-09-23. Overnight log return was −2.08% and intraday log return −3.26%, so the fund lost in both parts of the day. The RSI statistics show little. After 63 sessions with RSI(14) below 30 the median 5-day move was 0.37% and the 20-day move 0.35%, against baselines of −0.04% and −0.05%. After 202 sessions with RSI(2) below 10 the medians were −0.02% and −0.12%. The RSI(14) signal has a small positive lean and RSI(2) none, matching the weak results of the RSI templates.
The fund was above its 200-day average on 57.43% of sessions and crossed it 34 times. Correlation to SPY was 0.25 and to QQQ 0.23, with a beta to SPY of 0.12. Its closest relative was UDN, the bearish dollar fund, at 0.96. Correlation with USDU, the bullish dollar fund, was −0.86. EEM and IAU were at 0.41 and 0.4. Among the other currency funds, UDN held −1.12% a year and its best template, the trailing stop, returned 1.88%, while USDU held 5.22% and monthly cycle returned 5.4%.
The day-of-week averages were between −0.01% and 0%, which shows no weekday pattern. September was the weakest month at −1.35% and November the strongest at 0.87%. The trailing stop and SMA 10/50 returned the most, and the templates that traded often had to work inside a daily range of 0.39%.
Drawdown tells the same story from the other side. Monthly cycle, invested 95.2% of sessions, had a drawdown of 22.44% from 2021-05-25 to 2022-09-27, almost the fund's own. The weekly 7% target, at 91.1% exposure, had 20.58%. Every template with exposure near half or lower had a drawdown under 13%, apart from RSI mean reversion at 16.18% and its 53.5% exposure. Several drawdown tables end with an open drawdown that began on 2026-01-27, the date the fund recovered its 2021 level: the trailing stop shows 5.97%, SMA 10/50 5.41% and the regime filter 5.42%. The euro slipped back after that date, and each template that was invested took part of that decline.
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 FXE?
Of the 12 templates tested on FXE over 2021-01-04 to 2026-10-02, the strongest by CAGR was trend + trailing stop at 2.1% (max drawdown 7.9%), versus −0.9% 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 FXE?
12 of 12 templates beat FXE buy-and-hold (−0.9% CAGR) on this window; 12 of 12 had a shallower maximum drawdown than holding (22.9%).
Which strategy worked best on FXE?
The SMA-200 plus 15% trailing stop returned 2.14% a year with a 7.91% drawdown, against −0.9% for holding. It held one open position from 2022-12-02 that was up 13.28% at the end. It made no completed trade.
Why did the dip buyer and breakout show 0% on FXE?
Neither made a trade. The dip buyer needs a 20-day drawdown worse than 10%, and FXE had none. The breakout needs a 20-day return above 10% and never got one, so both stayed in cash with $10,000.
Did every strategy beat buy-and-hold on FXE?
On CAGR and on drawdown, yes, all 12 did, against −0.9% and 22.86%. Two of those wins came from making no trades. Seven templates had a positive CAGR.
How much do trading costs matter on FXE?
A lot for frequent traders. RSI(2) snapback went from 0% to −5.62% at 10 basis points over 170 round trips. The trailing stop moved from 2.14% to 2.12%, because it made no completed trade.
Is FXE good for mean reversion?
The evidence is weak. After RSI(14) fell below 30, the median 20-day move was 0.35% against a −0.05% baseline. RSI mean reversion returned 0.26% with a 16.18% drawdown, and RSI(2) snapback returned 0%.
How volatile is FXE?
Annualized volatility was 7.44%, with an average daily range of 0.39%. The best day was 2.47% on 2025-04-10 and the worst was −1.74% on 2024-11-06. It had no falls of 10% or more from a 20-day high.
Other currency 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.