DeployQuantXLF trading strategies, backtested
Financial Select Sector SPDR Fund: S&P 500 banks, insurers and asset managers. Every DeployQuant template run on XLF over 5.7 years of minute data, same engine, same window, sorted by return.
XLF is the Financial Select Sector SPDR Fund, which holds the banks, insurers and asset managers inside the S&P 500. From 2021-01-04 to 2026-10-02 it returned 102.88% in total, a 13.11% CAGR, with annualized volatility of 18.43%. A $10,000 buy-and-hold position ended at $19,742, a 12.58% CAGR, after a maximum drawdown of 25.47%.
None of the 12 templates beat that CAGR. Twelve of 12 finished positive, and 9 had a shallower maximum drawdown than holding. The top two came close: RSI mean reversion at 11.82% and the monthly cycle at 11.63%. The weakest was the 200-day regime filter at 2.06%.
XLF is a case where holding was hard to improve on. The fund rose strongly in 2021 and 2024, and the two years that hurt, 2022 and early 2026, were shallow compared with the gains around them. A rule that sits out part of the time gives up the strong years and has to earn that back in the weak ones.
| Strategy | CAGR | max DD | Sharpe | trades | win rate | final value |
|---|---|---|---|---|---|---|
| RSI(14) Mean Reversion | 11.8% | −18.7% | 0.92 | 20 | 85% (+1 open) | $18,993 |
| First-to-Last Day of Month | 11.6% | −23.5% | 0.73 | 69 | 58% (+1 open) | $18,809 |
| RSI(2) Dip Snapback | 9.7% | −14.6% | 0.82 | 159 | 64% (+1 open) | $17,033 |
| Weekly Entry + 7% Target | 7.1% | −31.6% | 0.51 | 104 | 25% | $14,847 |
| 20-Day Momentum + Trailing Stop | 7.0% | −17.6% | 0.60 | 5 | 60% | $14,738 |
| Drawdown Dip Buyer + 8% Target | 6.8% | −16.0% | 0.61 | 5 | 100% | $14,572 |
| EMA 12/26 Trend | 5.6% | −24.0% | 0.52 | 25 | 56% | $13,697 |
| SMA 10/50 Trend | 4.0% | −27.2% | 0.39 | 19 | 47% | $12,541 |
| Golden Cross (SMA 50/200) | 3.6% | −17.9% | 0.33 | 4 | 25% (+1 open) | $12,249 |
| 3-Month Momentum Switch | 2.7% | −23.0% | 0.30 | 15 | 60% | $11,682 |
| SMA-200 Trend + 15% Trailing Stop | 2.6% | −27.9% | 0.26 | 4 | 50% (+1 open) | $11,602 |
| 200-Day SMA Regime Filter | 2.1% | −22.0% | 0.23 | 19 | 21% (+1 open) | $11,243 |
| Buy & hold XLF | 12.6% | −25.5% | 0.76 | – | – | $19,742 |
The ranking and what separated the top from the bottom
The full order by CAGR is RSI mean reversion at 11.82%, monthly cycle at 11.63%, RSI(2) snapback at 9.72%, weekly 7% target at 7.13%, momentum breakout at 6.99%, dip buyer at 6.78%, EMA 12/26 at 5.63%, SMA 10/50 at 4.02%, golden cross at 3.6%, 3-month momentum at 2.74%, trend plus trailing stop at 2.62%, and the 200-day filter at 2.06%.
RSI mean reversion bought deep oversold readings and was invested only 37.2% of the time. It made 20 round trips and won 17 of them, an 85% win rate. The average win was 5.53% and the average loss was 7.38%, so the template wins often and loses big when it loses. The profit factor was 3.82. Its maximum drawdown was 18.73%, against 25.47% for holding. What stands out is 2022: it returned 12.3% in a year when XLF lost 10.4%. The template made five round trips that year and won four. The losing one entered on 2022-04-14 and exited on 2022-08-04, a loss of 9.4% over 112 days. Its best month was October 2022, at 11.68%.
The monthly cycle needs little explanation. It was invested 95.2% of the time and tracked the fund: 34% in 2021, -10.9% in 2022, 14.6% in 2025. Its drawdown of 23.47% was only slightly shallower than holding's 25.47%. It beat holding in only one year, 2024, by 0.5 points. The 0.95-point CAGR shortfall is the cost of the 69 round trips and the days when it was in cash between the last session of one month and the first of the next.
The RSI(2) snapback was the most active rule on XLF with 159 round trips, a 64% win rate and an average hold of 4.9 days. Its drawdown of 14.57% was the shallowest of the 12. It returned 9.72%, and 2025 was its best year at 19.5%. In 2021 it returned 16.1% while XLF returned 35.8%, which is why it finished behind. The snapback rule buys short-term weakness and was invested 37% of the time.
The weekly target on a fund that does not move 7% in a week
The weekly 7% target did worse on XLF than it did on most funds in the study. It returned 7.13% with a maximum drawdown of 31.64%, deeper than holding's 25.47%. Its win rate was 25%, from 26 winners in 104 round trips, with a longest losing streak of 10. The average win was 7.17% and the average loss 1.77%, and the profit factor was 1.28.
The yearly numbers show where it lost ground. In 2021 the template returned 33.1% against 35.8% for holding, close. In 2022 it fell 20.6% while XLF fell 10.4%, and the gap was the largest in the table. The 2022 loss came from 27 round trips with only 5 wins, including an 8.37% loss between 2022-05-23 and 2022-06-16. A weekly rule that buys every Monday meets every dip in a declining year, and the Thursday exit then books the loss. June 2022 was its worst month at -17.22%, against -10.64% for holding.
In 2024 the template returned 20.2% against 29.9%, even though XLF was strong. The best trade of the window, 9.48%, ran 51 days from 2024-09-16 to 2024-11-06, which is a long time for a rule built around a one-week cycle. Winners that fail to hit 7% by Thursday are not sold, and XLF's quiet advance meant many trades sat in the green for weeks. The average hold was 16 days. The template beat holding in only one year, 2023, by 2.4 points.
Costs hit this template hard. At 5 basis points per trade the CAGR fell to 3.97%, with a drawdown of 36.46%. At 10 basis points it fell to 2.71%, with a drawdown of 38.55%. It made 208 fills, and each fill pays the charge.
Trend rules, which entered after the move
The trend rules did poorly in the context of a fund that rose 36.61% in 2021 and 30.43% in 2024. Golden cross entered on 2021-10-19, exited on 2022-04-21 for a 2.72% loss, re-entered on 2022-12-20 and exited on 2023-04-24 at a 0.29% loss, bought again on 2023-08-08 and lost 8.76% by 2023-10-30, and then caught the big one: 2023-11-17 to 2026-03-19, 853 days, a gain of 44.33%. That made one winner out of four round trips and a 3.6% CAGR. The trade in progress at the end, opened on 2026-07-06, was down 3.76%.
Trend plus trailing stop had a similar pattern. Its first two trades lost 10.26% and 11.27%, and the third, from 2023-06-16 to 2025-04-04, gained 36.67% over 658 days. The 200-day filter had a 21% win rate with an average win of 9.42% and an average loss of 1.55%, and its seven-trade losing streak showed how often the fund crossed its average without following through. XLF crossed its 200-day average 38 times in the window and sat above it for 71.16% of sessions.
The faster pair did a little better. EMA 12/26 won 56% of its 25 round trips and returned 5.63%, and SMA 10/50 won 47% of 19 and returned 4.02%. Both caught the 2024 advance: EMA 12/26 gained 18.1% that year and SMA 10/50 gained 24.7%. Both lost in 2022 (-12.8% and -18.2%). Their best single trades, 18.76% and 19.52%, both began in November 2023 and ended in April 2024.
The 3-month momentum switch returned 2.74% and never beat holding in any year. Its best trade was 17.02% from 2024-07-16 to 2025-03-05. Momentum breakout was a different case. Five round trips returned 6.99%, and one trade, 2023-11-27 to 2025-03-11, gained 37.1% over 470 days. The other four trades ranged from a 7.6% loss to a 13.16% gain.
How each strategy traded XLF
| Strategy | Time in market | Avg hold (days) | Best trade | Worst trade | Profit factor | With 10 bps slippage |
|---|---|---|---|---|---|---|
| RSI mean reversion | 37.2% | 39 | 12.2% | −9.4% | 3.82 | 11.1% |
| monthly cycle | 95.2% | 28 | 12.5% | −11.3% | 1.58 | 9.0% |
| RSI(2) snapback | 37.0% | 5 | 8.9% | −8.5% | 1.77 | 3.9% |
| weekly 7% target | 84.9% | 16 | 9.5% | −8.4% | 1.28 | 2.7% |
| momentum breakout | 60.0% | 253 | 37.1% | −7.6% | 5.99 | 6.8% |
| dip buyer | 30.3% | 127 | 8.0% | 8.0% | – | 5.4% |
| EMA 12/26 trend | 62.5% | 53 | 18.8% | −8.8% | 2.12 | 4.8% |
| SMA 10/50 trend | 62.5% | 69 | 19.5% | −6.2% | 1.75 | 3.4% |
| golden cross | 63.4% | 311 | 44.3% | −8.8% | 3.41 | 3.5% |
| 3-month momentum | 55.5% | 78 | 17.0% | −8.3% | 1.53 | 2.2% |
| trend + trailing stop | 70.6% | 330 | 36.7% | −11.3% | 1.66 | 2.5% |
| 200-day regime filter | 61.3% | 62 | 29.2% | −5.9% | 1.49 | 1.4% |
How long each rule held and what costs did
Time in market ran from 30.3% for the dip buyer to 95.2% for the monthly cycle. The dip buyer made five trades, won all five, and closed each near 8%: 2022-03-08 to 2022-03-18, 2022-04-27 to 2023-12-13, and three in April and May 2025. The 2022-04-27 entry was held 595 days before the target filled on 2023-12-13, and its 15.99% drawdown, from 2023-02-07 to 2023-03-17, fell inside that hold. Three of its five trades started within four days of each other, 2025-04-07 to 2025-04-10, around the worst and best days of the window.
The cost runs sort the templates by trading frequency. The RSI(2) snapback went from 9.72% with no costs to 6.78% at 5 basis points and 3.92% at 10. The monthly cycle went from 11.63% to 10.32% and 9.04%. The weekly target fell to 3.97% and 2.71%. RSI mean reversion barely moved: 11.82%, 11.44%, 11.06%. Golden cross, trend plus trailing stop and momentum breakout lost under 0.2 of a point at 10 basis points.
At 10 basis points, RSI mean reversion at 11.06% leads, the monthly cycle at 9.04% is second, and momentum breakout at 6.83% is third. The RSI(2) snapback at 3.92% falls below the dip buyer at 5.39% and EMA 12/26 at 4.75%. The facts support the usual pattern: slow rules keep their results and fast rules lose them, and a fund with $1,682,204,673 in average daily dollar volume and a median minute volume of 62,901 shares is a liquid fund. The test does not measure real fills.
Long trades and the April 2025 low
Golden cross on XLF is one trade. It gained 44.33% over 853 days from 2023-11-17 to 2026-03-19, and the other three trades were small or negative, so the 3.6% CAGR rests on a single holding period. The 200-day filter won 21% of 19 trades with an average win of 9.42% and an average loss of 1.55%, and its best trade ran 518 days from 2023-11-06 for 29.19%. Its worst month, 2025-04 at -15.49%, was the largest monthly loss of any template, so the filter's trades were long when the fund fell in the April 2025 low.
RSI mean reversion won 85% of 20 trades and had a profit factor of 3.82. Its best was 12.2% over 23 days from 2025-04-07, the day before the 2025-04-08 low, and its worst was a loss of 9.4% over 112 days from 2022-04-14. The average win was 5.53% and the average loss 7.38%, so a single bad trade is as large as more than one typical win. It was in the fund 37.2% of the time, and the run ends with an open trade from 2026-09-24 at -2.23%.
The dip buyer made 5 trades and won all of them, each at about 8%. One was a 10-day trade from 2022-03-08, and another was a 1-day trade from 2025-04-07 to 2025-04-08. One was held 595 days from 2022-04-27 to 2023-12-13 before it reached the target, which is a long time to wait for 8%. Its best month was 2025-04 at 21.55%, and its drawdown of 15.99% started on 2023-02-07.
The fund's drawdowns show where those entries sat. Holding fell 25.47% from 2022-01-12 to 2022-10-12 and recovered on 2024-02-22. It fell 15.33% from 2025-02-18 to 2025-04-08, and 14.65% from 2026-01-06 to 2026-03-27. Those later two drops were shorter and shallower than 2022. The 3-month momentum and trailing stop templates show drawdowns beginning on the same dates, so they were in the fund when it fell.
How XLF behaved
| Measure | XLF |
|---|---|
| Data in this test | 2021-01-04 to 2026-10-02 (1444 sessions) |
| Total return, buy and hold | 102.9% |
| Annualized volatility | 18.4% |
| Deepest drawdown | −25.9% (2022-01-12 to 2022-10-12) |
| Up days | 52.9% |
| Average daily range | 1.40% |
| Average overnight gap | 0.49% |
| Correlation to SPY | 0.77 |
| Correlation to QQQ | 0.59 |
| Correlation to TLT | -0.05 |
| Sessions above the 200-day average | 71.2% |
| Crossings of the 200-day average | 38 |
| Falls of 10% or more from a 20-day high | 10 |
Calendar years
| Year | Return |
|---|---|
| 2021 | 36.6% |
| 2022 | −10.6% |
| 2023 | 12.1% |
| 2024 | 30.4% |
| 2025 | 14.9% |
| 2026 (part) | −1.1% |
Biggest single days
| Best day | Move |
|---|---|
| 2025-04-09 | 7.6% |
| 2024-11-06 | 6.1% |
| 2022-11-10 | 5.1% |
| 2021-01-06 | 4.4% |
| 2022-10-13 | 4.2% |
| Worst day | Move |
|---|---|
| 2025-04-04 | −7.4% |
| 2025-04-03 | −5.0% |
| 2023-03-09 | −4.1% |
| 2023-03-13 | −3.9% |
| 2022-09-13 | −3.8% |
Average return by calendar month
| Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 2.8% | 1.6% | −1.1% | −0.1% | 1.6% | −0.1% | 4.0% | 1.6% | −3.4% | 3.3% | 4.8% | 0.3% |
Most and least correlated funds
| Most correlated | Least correlated | ||
|---|---|---|---|
| FAS | 1.00 | SH | -0.78 |
| VTV | 0.89 | SDS | -0.77 |
| VOOV | 0.89 | SPDN | -0.77 |
| SSO | 0.78 | UVXY | -0.64 |
| VOO | 0.78 | VXZ | -0.62 |
XLF year by year, and the drawdowns
XLF returned 36.61% in 2021, -10.55% in 2022, 12.11% in 2023, 30.43% in 2024, 14.86% in 2025 and -1.15% so far in 2026. Two of the six years were over 30%. A rule that was in cash for part of those two years started from behind.
The deepest drawdown ran from 2022-01-12 to 2022-10-12, a fall of 25.87%, and the fund did not recover until 2024-02-22. That is 528 sessions underwater, the longest in the window. The second came after the peak of 2025-02-18 and bottomed on 2025-04-08 at 15.33%, recovering by 2025-06-27. The third began on 2026-01-06, bottomed on 2026-03-27 at 14.65% and recovered on 2026-07-06.
The best day was 2025-04-09 at 7.59%, and the worst was 2025-04-04 at -7.36%, five calendar days apart. The other worst days were 2025-04-03 at -4.97%, 2023-03-09 at -4.07% and 2023-03-13 at -3.94%, and 2022-09-13 at -3.78%. The March 2023 days sit next to each other and point to a short cluster of stress for financial stocks specifically. The other best days were 2024-11-06 at 6.09%, 2022-11-10 at 5.11%, 2021-01-06 at 4.38% and 2022-10-13 at 4.17%.
Of the fund's total return, 62.56% came overnight and 37.44% intraday, with overnight and intraday log returns of 43.2% and 25.86%. Both halves were positive. That differs from some other funds in this study and means a rule that buys at the open is not giving up the whole gain. Average daily range was 1.4% and the average overnight gap was 0.49%. Up days were 52.88% of sessions, with an average up day of 0.86% and an average down day of -0.86%.
Seasonality, RSI behaviour and relationships
By calendar month the strongest averages were November at 4.85%, July at 4.03%, October at 3.3% and January at 2.77%. September was the weakest at -3.37%, followed by March at -1.11%. With five or six observations per month these describe the window and not a lasting pattern.
After RSI(14) fell below 30, the median forward 5-day return was 0.75%, against a baseline of 0.42%, and the median 20-day return was 3.18% against 1.5%. That was from 26 qualifying observations out of 30 sessions below 30. RSI(2) below 10 occurred on 169 sessions, with a median 5-day return of 0.38% and a 20-day return of 1.89%. The 5-day figure is below the baseline. XLF's lag-1 autocorrelation was 0.02, which is close to zero. The 20-day figure of 1.89% is above the 1.5% baseline, and the 5-day figure is not. Falls of 10% or more from a 20-day high happened in 10 episodes across 45 days, which fits the dip buyer's five trades.
XLF's correlation to SPY was 0.77 with a beta of 0.87, and its correlation to QQQ was 0.59 with a beta of 0.48. Against TLT the correlation was -0.05. The most correlated funds were FAS at 1, VTV and VOOV at 0.89, and SSO and VOO at 0.78. FAS is the 3x leveraged financials fund, so a correlation of 1 is expected. The least correlated were SH at -0.78 and SDS at -0.77.
Among the other sector funds in the study, the best rule varies. XLK had its best result from the monthly cycle at 20.86% against 22.07% for holding, XLY from the weekly target at 10.19% against 6.27%, XLP from the RSI(2) snapback at 8.22% against 5.8%, and VOX from golden cross at 11.2% against 9.08%. SOXX was best with the weekly target at 30.54% against 31.11%.
What this test leaves out
The window is 5.7 years. The headline run has no fees or slippage, uses $10,000 per fund, and has no margin. Decisions are daily and fills occur on minute bars at prices adjusted for splits and dividends. XLF's strong 2021 and 2024 weigh heavily on every comparison with holding, and a different start date would change the order. The ranking is a description of this window.
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 XLF?
Of the 12 templates tested on XLF over 2021-01-04 to 2026-10-02, the strongest by CAGR was RSI mean reversion at 11.8% (max drawdown 18.7%), versus 12.6% 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 XLF?
0 of 12 templates beat XLF buy-and-hold (12.6% CAGR) on this window; 9 of 12 had a shallower maximum drawdown than holding (25.5%).
Did any strategy beat buy-and-hold on XLF?
No. Buy-and-hold returned 12.58% a year, and the best template, RSI mean reversion, returned 11.82%. The monthly cycle came next at 11.63%. Nine of the 12 templates had a shallower maximum drawdown than holding's 25.47%.
Which strategy had the smallest drawdown on XLF?
The RSI(2) snapback had a maximum drawdown of 14.57%, the lowest of the 12, with a CAGR of 9.72%. Momentum breakout had 17.64%, golden cross 17.95% and RSI mean reversion 18.73%. The weekly 7% target, SMA 10/50 and trend plus trailing stop had deeper drawdowns than holding.
How did XLF do in 2022?
XLF lost 10.55% in 2022, with a deepest drawdown of 25.87% from 2022-01-12 to 2022-10-12. RSI mean reversion gained 12.3% that year and the dip buyer gained 6.3%. The weekly 7% target lost 20.6%.
Why did the weekly 7% target struggle on XLF?
XLF rarely moves 7% in a week, and the target filled on only 26 of 104 round trips. The template lost 17.22% in June 2022 and had a 31.64% drawdown. At 10 basis points of slippage its CAGR fell to 2.71%.
How much do trading costs matter for XLF strategies?
Slow rules lose little: golden cross went from 3.6% to 3.45% at 10 basis points. The RSI(2) snapback went from 9.72% to 3.92%, and the weekly target from 7.13% to 2.71%. RSI mean reversion stayed near 11%.
Is XLF correlated with the S&P 500?
XLF had a correlation of 0.77 and a beta of 0.87 to SPY over the window. Its correlation to QQQ was 0.59 and to TLT -0.05. FAS, the 3x financials fund, had a correlation of 1.
Other sector 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.