DeployQuantTECL trading strategies, backtested
Direxion Daily Technology Bull 3X Shares: 3x daily leveraged technology sector. Every DeployQuant template run on TECL over 5.7 years of minute data, same engine, same window, sorted by return.
TECL is a 3x daily leveraged fund on the technology sector, and its numbers in this window are large in both directions. Buy-and-hold returned 38.2% a year from 2021-01-04 to 2026-10-02 and turned $10,000 into $63,981, with a Sharpe ratio of 0.82. The maximum drawdown was 77.2%. Annualized volatility of the fund's daily returns was 75.1%, and its worst single day was a loss of 20.4% on 2025-04-03.
We ran all 12 DeployQuant templates on TECL with $10,000 each, no margin, and no fees or slippage in the headline numbers. None beat holding on CAGR. The monthly cycle finished first at 36.0% and the weekly 7% target second at 35.5%. The weakest, 3-month momentum, returned 7.6%. Eleven of the 12 had a shallower maximum drawdown than holding. The monthly cycle did not: its drawdown of 78.0% was a little deeper than the 77.2% for holding.
The spread between the templates is wider here than on any unleveraged fund in this set. The same rules run on TECL, which tracks XLK at three times the daily move, so each exit lands before or after a drop that is three times as large as it would be on the sector fund. The gaps between templates and holding on TECL ran from 2.2 percentage points for the monthly cycle to 30.5 for 3-month momentum.
The window has one deep leveraged drawdown, in 2022, and the rest of the period was mostly up. That mix shapes every result on the page.
| Strategy | CAGR | max DD | Sharpe | trades | win rate | final value |
|---|---|---|---|---|---|---|
| First-to-Last Day of Month | 36.0% | −78.0% | 0.80 | 69 | 52% (+1 open) | $58,352 |
| Weekly Entry + 7% Target | 35.5% | −43.2% | 0.85 | 259 | 51% | $57,202 |
| RSI(2) Dip Snapback | 30.2% | −51.0% | 0.77 | 162 | 65% | $45,495 |
| Drawdown Dip Buyer + 8% Target | 24.7% | −76.3% | 0.67 | 18 | 100% (+1 open) | $35,528 |
| RSI(14) Mean Reversion | 23.9% | −64.1% | 0.67 | 18 | 72% | $34,212 |
| EMA 12/26 Trend | 19.8% | −53.5% | 0.62 | 26 | 38% (+1 open) | $28,253 |
| SMA 10/50 Trend | 18.0% | −62.4% | 0.59 | 17 | 47% (+1 open) | $25,826 |
| SMA-200 Trend + 15% Trailing Stop | 14.4% | −55.8% | 0.52 | 47 | 45% (+1 open) | $21,653 |
| 200-Day SMA Regime Filter | 11.5% | −64.5% | 0.47 | 20 | 20% (+1 open) | $18,706 |
| 20-Day Momentum + Trailing Stop | 10.2% | −48.5% | 0.45 | 64 | 42% (+1 open) | $17,504 |
| Golden Cross (SMA 50/200) | 10.0% | −58.3% | 0.45 | 4 | 25% (+1 open) | $17,300 |
| 3-Month Momentum Switch | 7.6% | −69.9% | 0.40 | 26 | 38% (+1 open) | $15,251 |
| Buy & hold TECL | 38.2% | −77.2% | 0.82 | – | – | $63,981 |
Which templates kept up with a 3x fund, and which gave back most of it
The two calendar-style rules came closest because they stayed invested. The monthly cycle was in the market 95.2% of the time and made 69 round trips with a win rate of 52%. Its winners averaged 19.54% and its losers averaged a loss of 12.15%, and its worst month-long trade lost 30.14% (June 2022). The best gained 63.97% in April 2026. Its drawdown of 78.0% began on 2021-12-27 and was not recovered until 2024-06-12.
The weekly 7% target took a different route and ended almost level. It made 259 round trips, held for about 3.5 days on average, and was invested 55.4% of the time. Its maximum drawdown was 43.2%, against 77.2% for holding. In 2022 it returned 32.4% while the fund lost 73.5%, a gap of 105.9 points in its favour. The weekly entry and 7% target produced short holds in a year when the fund fell hard, and its best trade of the window gained 16.73% from 2022-12-12 to 2022-12-13. The same design gave up ground in the strong years: it trailed holding by 117.1 points in 2023 and by 86.1 points in 2021. Its worst year was 2024, at negative 16.4%, and its profit factor was 1.26. Win rate was 51%, with an average win of 7.49% and an average loss of 5.80%, so the edge per trade is thin and rests on the volume of trades.
RSI(2) snapback returned 30.2% with a 51.0% drawdown. It made 162 round trips, won 65% of them, and held for 4.7 days. In 2021 it returned 176.7%, which was ahead of holding by 55.2 points, and in 2022 it lost 34.0% against 73.5%. After that it trailed every year: 2023 by 112.8 points and 2026 by 80.4. Its worst trade was a loss of 28.72% from 2025-02-21 to 2025-03-06, entered on an oversold reading and held through the first leg of the 2025 decline. A 65% win rate with an average win of 4.78% and an average loss of 5.46% gives a profit factor of 1.40, which is thinner than the same template shows on unleveraged funds.
The dip buyer made 18 round trips and won all 18. It returned 24.7% and stayed invested 87.5% of the time, because it took long holds. One of its trades ran 912 days. Its drawdown of 76.3% was close to holding, and the position opened on 2026-06-03 was down 9.93% when the data ended. A perfect win rate on 18 trades with a target of 8% shows that the rule exits winners early and carries losers until they recover, and the drawdown measures the cost of that carry.
RSI(14) mean reversion returned 23.9% with a 72% win rate and a profit factor of 2.82. Its three round trips in 2022 all lost, and the worst lost 35.35% over 83 days from 2022-04-19 to 2022-07-11. In 2024 it returned 100.6% against 35.9% for holding, which was its best year. Whether the rule works on TECL depends heavily on which of the two years you look at.
The trend rules lost the most. EMA 12/26 returned 19.8% with 26 round trips and a win rate of 38%, and SMA 10/50 returned 18.0% with 17 round trips. Both had long winners (108.17% over 195 days for EMA, 84.89% over 192 days for SMA 10/50, both entered in May 2025) and lost heavily in 2022 trades of 21% to 26% over two weeks. The trailing-stop trend rule made 47 round trips and returned 14.4%, and it made 11 round trips in 2024, a year in which it lost 16.8%.
The 200-day filter returned 11.5% with 20 round trips and a win rate of 20%. Its profit factor was 0.88, the only one below 1.0 in the table. Its 4 wins averaged 26.98% and its 16 losses averaged 6.31%. The open position from 2026-04-14 was up 113.33% at the end of the data. In 2024 and 2023 it trailed holding by 52.0 and 114.9 points. The cross-fund view is on the 200-day filter page for all funds.
Golden cross made 4 round trips and returned 10.0%. One trade gained 144.38% over 558 days, from 2023-03-24 to 2024-10-02. The other three lost 24.04%, 29.71%, and 4.98%. The rule had a win rate of 25% and a drawdown of 58.3%. The golden cross page for TECL lists the entries. With four trades, the 10.0% is mostly one event.
How each strategy traded TECL
| Strategy | Time in market | Avg hold (days) | Best trade | Worst trade | Profit factor | With 10 bps slippage |
|---|---|---|---|---|---|---|
| monthly cycle | 95.2% | 28 | 64.0% | −30.1% | 1.66 | 32.8% |
| weekly 7% target | 55.4% | 4 | 16.7% | −20.2% | 1.26 | 22.9% |
| RSI(2) snapback | 36.4% | 5 | 15.6% | −28.7% | 1.40 | 23.2% |
| dip buyer | 87.5% | 95 | 9.3% | 8.0% | – | 23.8% |
| RSI mean reversion | 32.1% | 37 | 49.6% | −35.4% | 2.82 | 23.1% |
| EMA 12/26 trend | 59.8% | 48 | 108.2% | −22.8% | 1.53 | 18.8% |
| SMA 10/50 trend | 57.3% | 69 | 84.9% | −26.3% | 1.75 | 17.3% |
| trend + trailing stop | 57.7% | 25 | 54.1% | −19.6% | 1.21 | 12.6% |
| 200-day regime filter | 58.9% | 54 | 40.3% | −14.5% | 0.88 | 10.8% |
| momentum breakout | 43.4% | 14 | 57.5% | −13.5% | 1.17 | 7.8% |
| golden cross | 59.3% | 275 | 144.4% | −29.7% | 1.29 | 9.8% |
| 3-month momentum | 57.8% | 47 | 70.4% | −24.3% | 1.21 | 6.7% |
Trade sizes, hold lengths and the cost of 10 basis points
A 3x fund turns ordinary swings into large single trades, and the trading table shows the size of that effect. The best trades by template ran from 16.7% for the weekly rule to 64.0% for the monthly cycle, and the worst from a loss of 13.5% to a loss of 35.4%. A unit of risk on TECL is three times a unit on XLK, and the template parameters, set once for all funds, were not adjusted for that.
The weekly rule's 7% target was reached quickly because TECL's average intraday range was 5.28%. The rule's best trade of 16.73% took one day, and its median trade gained 6.99% with a median hold of 3 days. Its win rate was 51%, so about half of its trades reached the target or finished positive.
The cost runs separate the templates by trade count.
The long-hold rules lost very little to 10 basis points. Golden cross went from 10.0% to 9.9%, the SMA 10/50 trend from 18.0% to 17.3%, EMA 12/26 from 19.8% to 18.8%, and the 200-day filter from 11.5% to 10.8%. The dip buyer went from 24.7% to 23.8%.
The weekly 7% target went from 35.5% to 32.2% at 5 basis points and to 22.9% at 10 basis points. It made 518 fills, and the maximum drawdown widened from 43.2% to 45.9%. At 10 basis points it fell from a close second to fifth. The monthly cycle fell from 36.0% to 32.8%, and stayed first.
RSI(2) snapback fell from 30.2% to 26.7% at 5 basis points and 23.2% at 10. The 20-day momentum template, with 64 round trips, went from 10.2% to 9.0% to 7.8%.
The average daily dollar volume of TECL was $122,978,604 and the median minute volume was 2,800 shares. That is much thinner than XLK. The 10 basis point run is a flat assumption and does not model thin minute volume, so the high-turnover results on TECL carry more uncertainty than the table shows.
The profit factors tell the same story. The calendar rule and the 200-day filter bracket the table, at 1.66 and 0.88. The weekly rule's 1.26 and the trailing-stop trend rule's 1.21 mean that gross wins exceeded gross losses by little, so any added cost moves the result a lot. Golden cross at 1.29 and momentum breakout at 1.17 sit in the same range, though for different reasons: golden cross has one large winner, and momentum breakout has 64 round trips with a 42% win rate.
How TECL behaved
| Measure | TECL |
|---|---|
| Data in this test | 2021-01-04 to 2026-10-02 (1444 sessions) |
| Total return, buy and hold | 592.0% |
| Annualized volatility | 75.1% |
| Deepest drawdown | −78.0% (2021-12-27 to 2022-10-12) |
| Up days | 54.8% |
| Average daily range | 5.28% |
| Average overnight gap | 2.06% |
| Correlation to SPY | 0.91 |
| Correlation to QQQ | 0.97 |
| Correlation to TLT | 0.07 |
| Sessions above the 200-day average | 68.4% |
| Crossings of the 200-day average | 40 |
| Falls of 10% or more from a 20-day high | 68 |
Calendar years
| Year | Return |
|---|---|
| 2021 | 124.3% |
| 2022 | −74.3% |
| 2023 | 202.8% |
| 2024 | 36.4% |
| 2025 | 38.3% |
| 2026 (part) | 110.0% |
Biggest single days
| Best day | Move |
|---|---|
| 2025-04-09 | 39.6% |
| 2022-11-10 | 24.6% |
| 2026-07-30 | 16.2% |
| 2022-11-30 | 14.9% |
| 2026-08-04 | 14.7% |
| Worst day | Move |
|---|---|
| 2025-04-03 | −20.4% |
| 2026-06-05 | −20.1% |
| 2025-04-04 | −19.7% |
| 2022-09-13 | −16.0% |
| 2025-01-27 | −14.7% |
Average return by calendar month
| Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1.1% | −3.7% | 0.6% | 4.5% | 22.4% | 9.8% | 5.5% | −0.4% | −4.7% | 11.1% | 13.4% | −1.5% |
Most and least correlated funds
| Most correlated | Least correlated | ||
|---|---|---|---|
| XLK | 1.00 | TECS | -1.00 |
| ROM | 0.99 | REW | -0.98 |
| TQQQ | 0.97 | QID | -0.97 |
| QLD | 0.97 | SQQQ | -0.97 |
| QQQ | 0.97 | PSQ | -0.97 |
Leverage and decay against XLK
| Year | TECL | XLK | 3× XLK, reset daily |
|---|---|---|---|
| 2021 | 124.3% | 37.0% | 130.0% |
| 2022 | −74.3% | −27.7% | −72.7% |
| 2023 | 202.8% | 56.0% | 242.4% |
| 2024 | 36.4% | 21.6% | 55.0% |
| 2025 | 38.3% | 24.6% | 55.2% |
| 2026 | 110.0% | 39.3% | 126.7% |
The last column compounds 3× 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 3× the underlying's return over a year.
Over 2021-01-04 to 2026-10-02, TECL returned 592.0% while XLK returned 226.1% and a perfect daily-reset 3× version would have returned 1075.0%. Its measured daily beta to XLK was 2.99.
Leverage decay, volatility and what they did to the rules
The fund's measured beta to XLK was 2.99 with an R-squared of 1, so day to day it moved three times its underlying. Over the full window XLK returned 226.13% and TECL returned 591.96%. Three times the XLK return would be 678.38%, which puts TECL 86.42 points under that simple multiple. A daily-rebalanced 3x of XLK's returns, before any fund costs, would have been 1,074.97% in the data, so the fund's result sat well below that ideal too.
The year-by-year table shows the gap changing sign. TECL beat 3x XLK in 2021 by 13.3 points, in 2022 by 8.9 points, and in 2023 by 34.8 points. It fell short in 2024 by 28.5 points, in 2025 by 35.5 points, and in 2026 by 8.0 points. A fund that resets every day compounds each day's move, so over a calendar year its return depends on the path as well as the endpoint. A straight climb pays more than 3x, and a choppy climb pays less. 2023 was the first kind in this data and 2024 and 2025 were the second.
Volatility explains the template results better than the trend does. TECL's annualized volatility was 75.1%. Its average up day was 3.43% and its average down day was negative 3.62%, with 54.8% of days up. Its beta to SPY was 4.16 and to QQQ was 3.24. A 15% trailing stop sits within a few average down days of the fund's high, and the 20-day return thresholds of the momentum templates are crossed often. The fund had 68 falls of 10% or more from a 20-day high, spread over 533 days. The dip buyer's condition was true for a large share of the window, which is why it was invested 87.5% of the time.
The fund closed above its 200-day average on 68.4% of sessions, a lower share than XLK, and crossed the line 40 times. That many crossings is the direct reason the 200-day filter made 20 round trips and won 4. Its 16 losing trades averaged a loss of 6.31%, and the 4 winners averaged 26.98%, which left a profit factor below 1.0.
On oversold readings, RSI(14) fell below 30 on 17 sessions. The median forward return was 11.44% over 5 days and 9.59% over 20 days, against baselines of 1.09% and 3.74%. Those are large edges on a small sample, and they match the RSI mean reversion template's 72% win rate. They also hide the cases where the fund kept dropping, as in the 35.35% loss in 2022. RSI(2) fell below 10 on 144 sessions, with a 5-day median of 1.77% and a 20-day median of 6.58%, against the same baselines. The 20-day figure is above baseline, and the average RSI(2) trade lasted 4.7 days, so the template exits well before that edge arrives.
By calendar month, May averaged 22.39%, November 13.40%, and October 11.14%, while September averaged negative 4.71% and February negative 3.71%. Each month has 5 or 6 observations. The best days were 2025-04-09 at 39.56% and 2022-11-10 at 24.63%. The worst were 2025-04-03 at negative 20.40%, 2026-06-05 at negative 20.07%, and 2025-04-04 at negative 19.69%. Three days in one week of April 2025 decided a lot: the fund lost 20.40% and 19.69% on the first two days and then gained 39.56% on 2025-04-09, and any rule that sold at the lows missed the rebound.
The overnight session produced 59.53% of the fund's log return and the intraday session 40.47%. The average overnight gap was 2.06%. Orders that fill at the next open take that gap in full.
Among the other leveraged funds, buy-and-hold returned 25.35% on TQQQ, 23.55% on QLD, and 33.32% on SOXL. On TQQQ the best template, RSI(2) snapback, returned 39.31% and beat holding. On SOXL golden cross returned 45.93% and beat holding. On TECL none did. ROM, the 2x technology fund, had its best template, the weekly 7% target, at 30.27% against 30.19% for holding. The full group is a small sample, and each fund's result depends on which dates its drawdowns fell on.
The test is one window of 5.74 years, daily decisions, and no fees in the headline run. A 3x fund can lose most of its value in a single drawdown, as TECL did for 77.2%, and the results here describe that window and not a forecast.
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 TECL?
Of the 12 templates tested on TECL over 2021-01-04 to 2026-10-02, the strongest by CAGR was monthly cycle at 36.0% (max drawdown 78.0%), versus 38.2% 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 TECL?
0 of 12 templates beat TECL buy-and-hold (38.2% CAGR) on this window; 11 of 12 had a shallower maximum drawdown than holding (77.2%).
What is the best strategy for TECL?
By CAGR, the monthly cycle at 36.0% from 2021-01-04 to 2026-10-02, against 38.2% for buy-and-hold. The weekly 7% target returned 35.5% with a much shallower drawdown of 43.2%. Neither beat holding on return.
Did any strategy beat buy-and-hold on TECL?
None of the 12 templates did. Holding returned 38.2% a year with a maximum drawdown of 77.2%. Eleven templates had a shallower drawdown. The monthly cycle had 78.0%, slightly deeper.
Why did the 200-day moving average filter do badly on TECL?
It returned 11.5% a year with 20 round trips and a 20% win rate. TECL crossed its 200-day average 40 times in the window. Its profit factor of 0.88 was the only one below 1.0 among the 12 templates.
How much does leverage decay affect TECL?
Over the window TECL returned 591.96% and XLK returned 226.13%. Three times XLK would be 678.38%, so TECL was 86.42 points below it. The gap changed sign by year: TECL was ahead of 3x in 2021 to 2023 and behind in 2024 to 2026.
How bad was the TECL drawdown in 2022?
TECL lost 74.3% in calendar 2022 and had a peak-to-trough drawdown of 78.0% from 2021-12-27 to 2022-10-12. It recovered on 2024-06-11, after 616 sessions in the drawdown. The weekly 7% target gained 32.4% in 2022.
Do trading costs change the ranking on TECL?
For the frequent traders, yes. The weekly 7% target fell from 35.5% to 22.9% at 10 basis points, and RSI(2) snapback fell from 30.2% to 23.2%. Golden cross barely moved, from 10.0% to 9.9%.
Other leveraged 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.