CLSE trading strategies, backtested

Convergence Long/Short Equity ETF: an active long/short US equity portfolio. Every DeployQuant template run on CLSE over 5.7 years of minute data, same engine, same window, sorted by return. CLSE data starts 2022-02-22, so its results cover a shorter span.

CLSE is an actively managed long/short US equity fund, and it is the youngest fund in this test. Its data starts on 2022-02-22, not on 2021-01-04, so the page covers 1,149 sessions. Buy-and-hold returned 19.49% a year over that stretch, turned $10,000 into $22,713 and had a maximum drawdown of 16.05% with a Sharpe of 1.41. The fund's own close to close return was 132.05%, and its annualized volatility was 13.85%.

All 12 templates had a positive CAGR. Two of the 12 beat holding: the weekly 7% target at 21.66% and the 200-day regime filter at 20.38%. Ten of 12 had a shallower drawdown than holding. The ranking after those two runs from 18.85% for monthly cycle and 18.72% for the trend plus trailing stop down to 3.35% for the dip buyer, so the spread between templates is large even though none lost money.

A fund that rose in most years leaves little room for a rule to add return, and the usual job of a template becomes keeping the return while cutting the drawdown. On CLSE the rules that did that best were the slow trend filters and the weekly target, and the rules that gave up the most were those with low exposure, such as the RSI mean reversion at 3.77% with 21.1% of sessions invested.

The results are from one short window with no fees or slippage in the headline run, $10,000 starting cash, no margin and daily decisions. The window contains a mild 2022, a strong 2024 and a 2025 pullback in April, so it says little about how the fund behaves in a long decline.

The weekly 7% target returned 21.66% with a 12.45% drawdown and a Sharpe of 1.62. It made 50 round trips with a 38% win rate, an average win of 7.27% against an average loss of 1.44%, and a profit factor of 3.42. It was invested 90% of the time. Its best trades were 8.07% from 2026-03-09 to 2026-04-08 and 7.6% from 2024-01-08 to 2024-01-31. The worst was −4.86% in three days from 2025-03-03. Against holding it was ahead in 2022 by 5.4 points, in 2025 by 8.2 and in 2026 by 0.2, and behind in 2023 and 2024. The 7% target was reached often enough to carry the result. Costs barely changed it: 21.66% became 22.44% at 5 basis points and 21.51% at 10.

The 200-day regime filter returned 20.38% with an 8.7% drawdown, a Sharpe of 1.81 and a profit factor of 25.22. It made 8 round trips and one trade did nearly all of the work: 52.85% over 719 days from 2023-03-17 to 2025-03-05. After that it re-entered on 2025-05-13 at an adjusted price of 21.94 and the position was up 56.11% at the end of the window. The other trades were small, between −1.55% and 1.39%. The drawdown of 8.7% began on 2025-01-23, the same peak as the fund's own, and was shallower than the fund's 16.05%. Its beat over holding was 0.89 points of CAGR.

Monthly cycle returned 18.85% on 55 round trips with a 64% win rate and a profit factor of 3.56. Its best month was April 2026 at 10.54% and its worst was June 2022 at −6.93%. It trailed holding by 0.64 points and had a drawdown of 17.45%, slightly deeper than the fund's. The trend plus 15% trailing stop returned 18.72% with a single completed trade of 40.96% from 2022-12-22 to 2025-04-04 over 834 days, then an open position from 2025-05-08 at 21.6 that was up 58.56%. The stop fired on 2025-04-04, the trough of the fund's April 2025 fall, which means the trade exited at the low and the re-entry on 2025-05-08 came after a rebound. See the trailing stop page for CLSE.

Golden cross returned 17.46% from two completed trades, 43.53% over 767 days from 2023-03-06 to 2025-04-11 and 0.21% over 55 days, then an open position up 47.88%. SMA 10/50 returned 16.61% with 12 round trips, a 75% win rate and a profit factor of 10.16. The RSI(2) snapback returned 16.58% over 142 round trips with a 75% win rate and the best Sharpe of the group at 1.95, using only 32.4% exposure. EMA 12/26 returned 14.01% and the 3-month momentum switch 11.46%.

The three lowest are all low-exposure templates. The momentum breakout returned 10.1% with no completed trade and an open position from 2025-05-06 at 21.82, up 56.97%, and a drawdown of 4.74%. It was invested 30.9% of sessions. RSI mean reversion returned 3.77% on 8 round trips, and the dip buyer returned 3.35% on 2 trades, each near 8%, with 25.9% exposure. These templates were right when invested and invested too rarely, which on a fund with a 19.49% buy-and-hold CAGR costs a lot of return.

How each strategy traded CLSE

The RSI(2) snapback is the template where costs matter. It returned 16.58% in the headline run, 13.12% at 5 basis points and 9.77% at 10, and its ending value fell from $20,278 to $17,651 and $15,362. It made 142 round trips with an average win of 1.04% and an average loss of 0.95%, a thin edge per trade. The weekly 7% target, with 50 trades but a 7.27% average win, was far less affected. Monthly cycle went from 18.85% to 17.47% and 16.1%. SMA 10/50 went from 16.61% to 16.31% and 16.01%. The slowest templates were almost unchanged: the trailing stop 18.72%, 18.69% and 18.66%, and golden cross 17.46%, 17.4% and 17.36%.

Average hold times show how different the templates are. RSI(2) snapback held 3.9 days. The weekly target held 26.4 days on average with a median of 17. The regime filter had an average hold of 93.5 days but a median of 3 days, because it made many one-day trades around its single long hold. The trailing stop's hold was 834 days.

The fund's liquidity is modest. Average daily dollar volume in the data was $2,009,643 and the median minute volume was 433 shares. The cost runs matter most for the 142 round trips of RSI(2) snapback, and not at all for the templates with a handful of trades.

Several drawdowns share the same dates. The fund's own worst, 16.05% from 2025-01-23 to 2025-04-04, shows up in almost every template: the trailing stop 14.54%, golden cross 16.06%, monthly cycle 17.45%, the regime filter 8.7%. The second shared episode ran from 2024-07-10 to 2024-08-05, 7.12% for the fund. In 2022 the fund had its decline from 2022-04-20 to 2022-09-30 at 13.86%, which the templates that were invested then, such as monthly cycle at 11.76% and EMA 12/26 at 10.18%, partly shared. Templates that were flat during 2022, such as the trailing stop, golden cross and the regime filter, show returns of −1.4%, −0.1% and −0.5% for that partial year.

How the trend rules traded CLSE

The two moving-average rules on CLSE took the same three big trades, a day or two apart. The SMA 10/50 rule entered on 2025-05-06 at 21.82 and exited on 2026-03-13 at 27.9 for 27.86% over 311 days. The EMA 12/26 rule entered on 2025-05-01 at 21.64 and exited on 2026-03-09 at 27.39 for 26.57% over 312 days. The second pair ran from November 2023: 21.16% over 176 days for SMA 10/50 and 22.85% over 172 days for EMA 12/26. The third pair started in April 2026, with returns of 18.33% and 20.23%.

The losses on those two rules also line up. SMA 10/50 lost 5.71% from 2025-01-23, the fund's own peak, to 2025-02-25. EMA 12/26 lost 4.3% from 2025-01-17 to 2025-02-24. Both bought just before the first quarter of 2025 drawdown and sold into it. EMA 12/26 also had a 7-day trade in June 2022 that lost 5.26%, from 15.22 to 14.42. It made 21 round trips with a 48% win rate, and its median trade lost 0.27% while its average win was 9.95%. A few large winners paid for many small losers, and the profit factor of 4.28 reflects that. SMA 10/50 had a median return of 1.89% and a median hold of 80 days, and it ended the window with an open position from 2026-09-11 that was up 1.21%.

The 3-month momentum rule made 7 round trips and had a median return of 4.66%. Its best trade was 21.85% over 298 days from 2025-06-06, and its worst was a loss of 4.68% from 2022-12-02 to 2023-01-25. It exited on 2026-03-31 and came back on 2026-04-09 at 29.94, after the largest monthly gain of the window had started. That is the cost of a rule that updates once a month: the fund's best month in the buy-and-hold run was April 2026 at 10.98%, and this rule caught only part of it.

The low-exposure rules traded for short stretches. The RSI mean reversion rule had 8 trades with a median return of 3.96%, the best being 5.56% over 64 days from 2022-06-23 and the worst a loss of 4.08% from 2025-02-26 to 2025-04-28, which sat inside the 2025 drawdown. The dip buyer had only 2 trades, 8.13% over 370 days from 2022-06-24 and 7.99% over 70 days from 2025-03-11. Both were bought after a fall of at least 8% and sold at the target, so the rule worked each time it fired. It fired twice in 1,149 sessions.

How CLSE behaved

Calendar years

Biggest single days

Average return by calendar month

Most and least correlated funds

The fund's return pattern is distinctive. Overnight log return was 114.42% and intraday log return was −31.19%, so the overnight share was 137.47% and the intraday share −37.47%. All of the gain came between the close and the next open, and the regular session subtracted. The average overnight gap was 0.5% and the average intraday range 0.86%. A template that is flat overnight would have missed that component, and RSI(2) snapback, invested 32.4% of sessions, held for 3.9 days on average.

The fund returned −3.66% in the partial 2022, 17.93% in 2023, 35.36% in 2024, 20.53% in 2025 and 25.18% in 2026 so far. It closed up on 54.97% of days with an average up day of 0.67% and an average down day of −0.68%. The maximum drawdown was 16.27%, from 2025-01-23 to 2025-04-04, with recovery on 2025-07-31 and a longest drawdown of 301 sessions. It was above its 200-day average on 90.74% of sessions and crossed it 17 times, a pattern that explains the high exposure of every trend template and the success of the regime filter.

The best days were 4.85% on 2025-04-09 and 3.12% on 2022-02-25, and the worst were −4.59% on 2025-01-27 and −4.45% on 2025-04-04. The 2025-01-27 drop came 2 sessions after the fund's 2025-01-23 peak, so the largest drawdown of the window started with its single worst day. The 2025-04-09 gain of 4.85% came 3 sessions after the low on 2025-04-04, and the trailing stop template, which exited on 2025-04-04, did not hold the fund through that day. The window holds only one drawdown of that size, so the ranking of templates rests heavily on how each one handled the first quarter of 2025. A fall of 10% or more from a 20-day high happened 4 times across 6 days, which is why the dip buyer traded only twice. Beta to SPY was 0.54 and to QQQ 0.39, with correlations of 0.67 and 0.64. The nearest funds were VOOG at 0.68 and IOO at 0.67, and the most negative were SPDN, SDS and SH at −0.67.

On mean reversion the sample is small. After 7 sessions with RSI(14) below 30 the median 5-day move was 0.95% and the 20-day move −1.88%, against baselines of 0.45% and 1.62%. After 86 sessions with RSI(2) below 10 the medians were 0.91% and 1.5%, the 5-day figure above baseline and the 20-day figure close to it. That matches RSI(2) snapback's good Sharpe on short holds, while RSI mean reversion's 8 trades produced only 3.77%. Monthly averages were highest in November at 3.5%, May at 3.18% and October at 2.9%, with December the only negative at −0.59%, from 4 or 5 observations per month.

Among other alternative-strategy funds, CTA held 8.9% a year and its best template, monthly cycle, returned 13.3%. RINF held 6.63% with RSI(2) snapback at 8.52%. KMLM held 7.07% and QAI held 3.89%, with golden cross at 4.28%. ALTY held 7.61% and the weekly 7% target returned 8.42%. CLSE held 19.49% and the best template returned 21.66%.

Year by year, the templates trailed or matched holding in most years, and the two leaders gained their edge in particular years. In the partial 2022, holding lost 3.6%. The weekly 7% target gained 1.8% and RSI(2) snapback gained 10.3%, while EMA 12/26 lost 6.2% and 3-month momentum lost 5.1%. In 2023 holding gained 17.5%. The trailing stop returned 17.6%, the regime filter 16.4% and golden cross 15.6%, while the weekly target returned 13.7% and 3-month momentum only 3.9%. In 2024 holding gained 34.7%, and the trailing stop, golden cross, regime filter, monthly cycle and weekly target were all within a point of it, at 34.7%, 34.7%, 34.8%, 34.1% and 33.9%. The momentum breakout and the dip buyer returned 0% in 2024 because they held no position.

In 2025 holding returned 20.2%, and the results spread out. The momentum breakout returned 24.9%, the weekly target 28.4% and 3-month momentum 23.4%, while golden cross returned 8.2% and the trailing stop 13.2%, since both had exited on 2025-04-11 and 2025-04-04 and re-entered in May and July. In 2026 so far holding returned 24.9% and five templates were within about a point of it. RSI(2) snapback returned 9.5% and RSI mean reversion 0%. The pattern is that templates which stay invested through trends match holding in strong years and the ones that win in the weak years, such as 2022, are the ones that trade often.

Correlation, weekdays and shared best months

CLSE's link to the equity market is moderate. Its correlation to SPY was 0.67 and to QQQ 0.64, with betas of 0.54 and 0.39, and its correlation to TLT was negative 0.03. The funds with the highest correlation were VOOG at 0.68, IOO at 0.67, SPUU at 0.67 and VOO at 0.67, all long US equity funds. The lowest were inverse funds: SPDN, SDS and SH at negative 0.67, TECS at negative 0.66 and REW at negative 0.64. A long/short fund that tracks the index this closely behaved like a reduced-size index position over this window, and the templates treated it that way.

The day-of-week averages were small and close together: 0.17% on Monday, 0.04% on Tuesday, 0.11% on Wednesday, 0.01% on Thursday and 0.07% on Friday. With about 1,149 sessions spread across five days, those differences are inside normal noise, and no template here uses the weekday. The lag-1 autocorrelation of daily returns was negative 0.03, which also leaves nothing for a rule built on one-day persistence or reversal to exploit.

Several templates share their best month. The 200-day filter, the trailing stop, EMA 12/26, golden cross and the monthly cycle all had April 2026 as their best month, between 10.32% and 10.98%, and the fund's own best month was the same. A single month at the end of the window therefore lifts the headline CAGR of most templates at once. The worst months differ more: the weekly target's was March 2025 at negative 4.36%, the regime filter's February 2025 at negative 3.33%, and the trailing stop's April 2025 at negative 4.89%. The RSI(2) snapback was the exception, with its best month in January 2025 at 6.47% and its worst in March 2025 at negative 2.17%, a smaller range than any trend rule.

Other alternative-strategy 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.