DeployQuantSMA 10/50 Trend on RINF
ProShares Inflation Expectations ETF: tracks 30-year breakeven inflation expectations. Backtest 2021-01-04 to 2026-10-02, $10,000 starting capital, computed by the same engine that runs live DeployQuant strategies.
Every calendar year in this test was positive for RINF, and the SMA 10/50 trend rule still lost money on it. Starting at $10,000 on 2021-01-04, the account closed at $8,844 on 2026-10-02: -11.56% in total, -2.12% a year. Holding the fund over the same days produced $14,460, a CAGR of 6.63% and a maximum drawdown of 13.32%. The rule's maximum drawdown was 23.92% and its Sharpe ratio was -0.19, against 0.59 for holding.
The fund is the ProShares Inflation Expectations ETF, built on 30-year breakeven inflation expectations, and it is calm: 12.41% annualized volatility, a 0.61% average daily range, and one fall of 10% or more from a 20-day high over the whole sample. A fast moving-average crossover needs trends to ride, and this fund moved up in small steps with pullbacks that were too small to profit from and large enough to trigger the exit.
The rule made 25 closed round trips plus one open position, for 51 fills. Only nine of the 25 closed higher (36%). Winners averaged 1.95% and losers -1.87%. Those two averages are close in size, and the shortfall comes from the count: more losers than winners. The profit factor was 0.57. The rule was invested on 60.4% of trading days and ranked 11th of 12 templates on this fund. Across all 59 funds, RINF ranks 44th for this rule.
Scope: a single 5.74 year window, daily decisions, no fees in the headline run. The numbers record the past of this rule and fund and carry no view on either.
Year by year
| Year | SMA 10/50 trend | buy & hold |
|---|---|---|
| 2021 | 3.8% | 14.7% |
| 2022 | −14.0% | 8.8% |
| 2023 | 1.5% | 0.2% |
| 2024 | 0.5% | 9.7% |
| 2025 | −2.9% | 1.5% |
| 2026 | 0.0% | 3.7% |
Where the gap to holding opened, year by year
Rule returns by year ran 3.8%, -14%, 1.5%, 0.5% and -2.9% for 2021 to 2025, then 0% for 2026 to date. The fund, held, gave 14.7%, 8.8%, 0.2%, 9.7%, 1.5% and 3.7%. Only 2023 favoured the rule; it fell behind in the other five years.
The widest gap is 2022: the held fund added 8.8% while the rule dropped 14%. This was the year of the rule's six round trips with one win, and the sequence tells the story. The entry on 2022-01-04 at $26.31 sold ten days later at $25.10 for -4.6%. The entry on 2022-03-04 at $26.10 ran 83 days to 2022-05-26 at $27.10 and made 3.83%, the second best trade of the test. Then came the 2022-06-09 entry at $28.99, sold 19 days later at $27.99 for -3.45%, and the 2022-08-23 entry at $28.53, sold on 2022-09-30 at $26.50 for -7.12%, the worst trade of the record. Two more entries in October and December lost 1.2% and 2.36%, the second of them in a single day, from 2022-12-06 to 2022-12-07.
The monthly table for 2022 shows why. January lost 4.5%, May lost 5.4%, June 3.4%, September 5.7% and November 5.6%, while March, April and October gained 4.3%, 5.2% and 4.7%. The fund ended the year up, so the gains and losses inside it were a series of swings, and an average pair that lags by weeks tends to buy after a rise and sell after a fall. The rule was in cash in February and July and gave up those months too.
2021 was the rule's best year at 3.8%, and all three of its exits won, at 1.73%, 0.4% and 1.73%. Holding made 14.7% that year, so the rule captured a small part of it. The first entry, on 2021-03-22 at $24.34, did not come until the fast average crossed the slow one, and January and February sat at 0%.
2023 is the one year in front. The rule made 1.5% against 0.2% for holding, a year in which the fund spent the months until October recovering from its drawdown. One trade did the work, the 189-day hold from 2023-05-17 at $28.11 to 2023-11-22 at $29.70, a return of 5.66% that tops every other trade here. The other two trades that closed in 2023 lost 3% and 1%.
2024 gave the rule 0.5% against 9.7% for holding, on four round trips with two wins. The 2024-02-16 entry returned 1.78% over 122 days, and the 2024-09-26 entry returned 1.5% over 71 days. Two short trades lost 1.4% and 1.1%. In 2025 the rule lost 2.9% against a gain of 1.5%, with six trades and one win. Four of those six lasted 22 days or less. 2026 runs to 2026-10-02, with the rule at 0% and the fund at 3.7%. The open trade began on 2026-07-30 at $32.43 and was up 0.71% at the last close.
Month by month
| Year | Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2021 | 0.0% | 0.0% | 0.6% | −0.4% | 2.3% | −0.8% | 0.0% | 0.0% | −0.5% | 0.2% | 2.5% | −0.2% |
| 2022 | −4.5% | 0.0% | 4.3% | 5.2% | −5.4% | −3.4% | 0.0% | −1.3% | −5.7% | 4.7% | −5.6% | −2.3% |
| 2023 | 0.0% | 1.0% | −3.9% | −0.9% | −0.2% | 1.2% | 1.5% | 0.0% | 3.7% | 3.3% | −3.9% | 0.0% |
| 2024 | −1.1% | 0.2% | 0.1% | 2.8% | 0.5% | −2.0% | −1.1% | 0.0% | 0.1% | 3.0% | −1.4% | −0.5% |
| 2025 | 1.2% | −2.3% | 0.0% | 0.0% | 1.9% | −0.5% | 0.2% | −0.8% | −1.5% | −1.2% | −0.5% | 0.7% |
| 2026 | 0.3% | −0.7% | 0.0% | 0.7% | 1.0% | −1.6% | −0.1% | −0.2% | 1.0% | −0.3% | – | – |
The monthly table
April 2022 was the rule's top month at 5.18%, then October 2022 at 4.7%, March 2022 at 4.3% and September 2023 at 3.7%. The worst was September 2022 at -5.74%, followed by May 2022 at -5.4%, November 2022 at -5.6% and January 2022 at -4.5%. For the held fund, October 2022 was best at 13.45% and September 2022 worst at -6.11%.
The rule missed that October 2022 jump by a wide margin, gaining 4.7% while the fund gained 13.45%. The fast average had crossed under the slow one in September, and the re-entry on 2022-10-19 at $28.64 came after much of the rise. That entry was sold on 2022-11-30 at $28.30 for -1.2%, so the rule paid for the late entry twice, by missing the move and by holding into the November pullback.
Many months sit at exactly 0.0%, which marks a month in cash. There are more of them in 2021 and early 2023, and the exposure of 60.4% fits the count. The table also shows many small negative months, such as -0.5%, -0.8% and -1.1%, and these are the whipsaw losses the template warns about. In 2025 the rule had six months below zero and four above, with none larger than 2.3% in either direction.
The fund's own month-of-year averages are on the profile and describe a different pattern. October averaged 4.1%, March 1.61%, February 1.29% and April 1.19%, with November at -1.5% and June at -1.23%. Five or six years sit behind each average, so they describe only this stretch of data and say nothing about what the rule would do.
Every trade
SMA 10/50 trend on RINF made 25 closed round trips and one position still open at the end of the test, an average hold of 48 days, an average winner of 1.95%, an average loser of −1.87%, a profit factor of 0.57, a longest losing streak of 6. It held a position at the close on 60.4% of trading days.
| Entry | Entry price | Exit | Exit price | Return | Days held |
|---|---|---|---|---|---|
| 2021-03-22 | $24.34 | 2021-06-29 | $24.76 | 1.7% | 99 |
| 2021-09-08 | $24.54 | 2021-09-27 | $24.63 | 0.4% | 19 |
| 2021-09-29 | $24.88 | 2021-12-07 | $25.31 | 1.7% | 69 |
| 2022-01-04 | $26.31 | 2022-01-14 | $25.10 | −4.6% | 10 |
| 2022-03-04 | $26.10 | 2022-05-26 | $27.10 | 3.8% | 83 |
| 2022-06-09 | $28.99 | 2022-06-28 | $27.99 | −3.5% | 19 |
| 2022-08-23 | $28.53 | 2022-09-30 | $26.50 | −7.1% | 38 |
| 2022-10-19 | $28.64 | 2022-11-30 | $28.30 | −1.2% | 42 |
| 2022-12-06 | $28.79 | 2022-12-07 | $28.11 | −2.4% | 1 |
| 2023-02-21 | $28.14 | 2023-03-21 | $27.30 | −3.0% | 28 |
| 2023-04-12 | $27.70 | 2023-05-03 | $27.43 | −1.0% | 21 |
| 2023-05-17 | $28.11 | 2023-11-22 | $29.70 | 5.7% | 189 |
| 2024-01-25 | $29.17 | 2024-02-08 | $28.77 | −1.4% | 14 |
| 2024-02-16 | $29.15 | 2024-06-17 | $29.67 | 1.8% | 122 |
| 2024-07-09 | $30.13 | 2024-07-31 | $29.79 | −1.1% | 22 |
| 2024-09-26 | $29.82 | 2024-12-06 | $30.28 | 1.5% | 71 |
| 2024-12-24 | $31.08 | 2025-02-28 | $30.63 | −1.4% | 66 |
| 2025-05-02 | $30.81 | 2025-08-07 | $31.07 | 0.8% | 97 |
| 2025-09-02 | $31.53 | 2025-09-10 | $31.04 | −1.6% | 8 |
| 2025-10-01 | $31.34 | 2025-10-03 | $31.24 | −0.3% | 2 |
| 2025-10-09 | $31.47 | 2025-10-14 | $31.19 | −0.9% | 5 |
| 2025-11-10 | $31.37 | 2025-12-02 | $31.36 | −0.0% | 22 |
| 2025-12-10 | $31.40 | 2026-02-19 | $31.33 | −0.2% | 71 |
| 2026-04-13 | $31.64 | 2026-06-25 | $31.66 | 0.1% | 73 |
| 2026-07-27 | $32.13 | 2026-07-28 | $32.02 | −0.3% | 1 |
| 2026-07-30 | $32.43 | open | – | 0.7% | – |
Prices are adjusted for splits and dividends, so they sit below the quotes printed at the time. An open position is marked at the last close.
The 25 round trips
The 25 closed trades had a median return of -0.34% and a median hold of 28 days. The average hold was 47.7 days, the longest 189 days and the shortest a single day. Nine won and sixteen lost; the best run of wins was 3 and the worst run of losses was 6. The 26th position, bought on 2026-07-30 at $32.43, remained open.
The winners were long holds. The five best returned 5.66% (189 days), 3.83% (83 days), 1.78% (122 days), 1.73% (99 days) and 1.73% (69 days). Every one of them lasted more than two months, and none returned more than 5.66%. A fund with annualized volatility of 12.41% does not hand out large trend gains, so even the best trade was modest.
The losers were short and larger. The five worst were -7.12%, -4.6%, -3.45%, -2.99% and -2.36%, with holds of 38, 10, 19, 28 and 1 days. The losing trades lasted well under the winners, and the average loser of -1.87% sits next to an average winner of 1.95%. With 16 losers and 9 winners that produces the profit factor of 0.57.
The losing streak of 6 ran from the 2022-06-09 entry to the 2023-04-12 entry, with returns of -3.45%, -7.12%, -1.2%, -2.36%, -2.99% and -1%. A second cluster of five losses came in late 2025, at -1.6%, -0.3%, -0.9%, 0% and -0.2%, with holds of 8, 2, 5, 22 and 71 days. Those were tiny in size, but they show the rule entering and leaving within days as the fund moved sideways. The trades of 2025-10-01 and 2025-10-09 lasted two and five days.
Grouped by exit year, trades and wins were 3 and 3 (2021), 6 and 1 (2022), 3 and 1 (2023), 4 and 2 (2024), 6 and 1 (2025) and 3 and 1 (2026). Only 2021 was clean. Listed prices are adjusted for splits and dividends and read under the quotes of the day, and the open trade is valued at the last close.
Largest drawdowns
| Peak | Low point | Depth | Days to low | Recovered | Days to recover |
|---|---|---|---|---|---|
| 2022-04-21 | 2023-03-17 | −23.9% | 330 | not yet | – |
| 2021-11-15 | 2022-03-04 | −9.2% | 109 | 2022-04-19 | 46 |
| 2021-10-21 | 2021-11-01 | −4.9% | 11 | 2021-11-10 | 9 |
Buy-and-hold's deepest drawdown ran from 2022-10-24 to 2023-02-03 and reached −13.3%.
A drawdown deeper than the fund's
The worst drawdown began on 2022-04-21, hit 23.92% on 2023-03-17 after 330 days, and was still unrecovered on 2026-10-02. Holding had its worst drawdown from 2022-10-24 to 2023-02-03 at 13.32%, recovered on 2023-10-18. The rule's drawdown is the larger of the two and the only one still open at the end of the data.
The cause is the run of losses through 2022. The trades of 2022-06-09, 2022-08-23, 2022-10-19 and 2022-12-06 lost 3.45%, 7.12%, 1.2% and 2.36%, and the 2023-02-21 trade added -2.99%. A rule with a profit factor of 0.57 can be underwater for years if the winners are small, and the largest winner after the low was 5.66%.
The second drawdown was 9.24%, from 2021-11-15 to 2022-03-04, recovered on 2022-04-19, 46 days later. The third was 4.93%, from 2021-10-21 to 2021-11-01, recovered on 2021-11-10. Those two were the only drawdowns the rule recovered before the main one began.
The usual argument for a trend rule is shallower drawdowns than the fund. That did not hold here. The fund's own drawdowns were 13.32%, 10.05% and 9.46%. A rule that exits after a fall and re-enters after a rise can lose more than a fund that simply held through each dip, and here it did.
With trading costs
The headline run fills at the bar price. These runs charge slippage on every fill.
| Slippage per fill | CAGR | Max drawdown | Final value | Sharpe |
|---|---|---|---|---|
| None (headline) | −2.1% | −23.9% | $8,844 | -0.19 |
| 5 basis points | −2.5% | −24.3% | $8,624 | -0.24 |
| 10 basis points | −3.0% | −24.7% | $8,409 | -0.29 |
Costs on 51 fills
With no charge the run gave -2.12%, a 23.92% drawdown, $8,844 and a Sharpe of -0.19. Slippage of 5 basis points on each fill moved those to -2.5%, 24.3%, $8,624 and -0.24, and 10 basis points moved them to -3%, 24.7%, $8,409 and -0.29.
The rule made 51 fills, so each charge applies 51 times. The average trade returned a few tenths of a percent or a little under two percent, so a charge at both ends removes a visible share of each result. Trades that returned between -0.3% and 0.8%, such as those of 2025-10-01, 2025-11-10 and 2026-04-13, are the ones where costs decide the sign. The cost runs show the direction and the size on a flat charge, and they do not model the quoted spread of a particular minute.
Liquidity deserves a note here. Average daily dollar volume was only $433,360 and median minute volume 300 shares, small beside the biggest funds in the set. A $10,000 account can still trade, but real spreads could be wider than the flat charge in the cost runs.
Changing the parameters
| Version | CAGR | Max drawdown | Round trips | Win rate | Final value |
|---|---|---|---|---|---|
| Published rules | −2.1% | −23.9% | 25 | 36% | $8,844 |
| SMA 5/50 | −2.4% | −22.7% | 30 | 30% | $8,676 |
| SMA 20/50 | −1.2% | −24.5% | 17 | 47% | $9,341 |
| SMA 10/100 | −4.0% | −25.1% | 17 | 24% | $7,914 |
Three other speeds
One average changed per variant. SMA 5/50, with a faster short leg, returned -2.44% with a drawdown of 22.67%, a final value of $8,676 and a Sharpe of -0.234, across 30 trades with 9 wins. SMA 20/50 uses a slower fast leg and returned -1.18% with a drawdown of 24.49%, $9,341, a Sharpe of -0.087 and 17 trades with 8 wins. SMA 10/100 keeps the fast leg and slows the other one, returning -3.99% with a drawdown of 25.14%, $7,914, a Sharpe of -0.381 and 17 trades with 4 wins.
None made money. The best was SMA 20/50 at -1.18%, which beat the published rule by trading less, 17 round trips against 25, and winning 8 of them, a rate of 47%. The worst was SMA 10/100, which won 4 of 17. The 5/50 pair made the most trades and still finished behind the published rule.
The pattern is that fewer trades helped a little and a much slower exit did not. A slower slow-average at 100 days kept the rule in positions longer without finding larger trends, and the drawdown was the deepest of the four. The spread across the four, from -1.18% to -3.99%, is small beside the 6.63% the fund made, so the choice of speed was a second-order issue next to the fund's lack of trends. Three alternatives on a single fund do not show that 20/50 would work in other periods.
How RINF behaved
| Measure | RINF |
|---|---|
| Data in this test | 2021-01-04 to 2026-10-02 (1428 sessions) |
| Total return, buy and hold | 45.4% |
| Annualized volatility | 12.4% |
| Deepest drawdown | −13.5% (2022-10-24 to 2023-02-03) |
| Up days | 51.9% |
| Average daily range | 0.61% |
| Average overnight gap | 0.36% |
| Correlation to SPY | 0.13 |
| Correlation to QQQ | 0.09 |
| Correlation to TLT | -0.40 |
| Sessions above the 200-day average | 82.3% |
| Crossings of the 200-day average | 74 |
| Falls of 10% or more from a 20-day high | 1 |
How RINF behaved and why a crossover struggled
Across 1,428 sessions the fund's price profile shows 45.41%, a CAGR of 6.74%. The backtest's buy-and-hold figure of 6.63% is slightly lower because it buys at the first open with 98% of the account. Its calendar-year returns were 15.05%, 8.98%, 0.21%, 9.89% and 1.48% from 2021 to 2025, with 3.78% so far in 2026. The profile's deepest fall was 13.54%, from 2022-10-24 to 2023-02-03, regained on 2023-10-18, and the longest underwater spell was 303 sessions.
RINF closed higher on 51.93% of days, with up days of 0.55% on average and down days of -0.56%. Lag-1 autocorrelation of 0 means yesterday's move told nothing about today's. The best day was 2022-03-02 at 4.35% and the worst 2022-10-25 at -3.16%. The average overnight gap was 0.36% and the overnight log return was 76.2% against -38.62% for the intraday part, so the fund earned its whole gain between the close and the next open. The rule trades at the open, so it holds the overnight gap on every day it is long.
The 200-day average numbers are the key ones for any trend rule. On 82.26% of sessions RINF finished above its 200-day average, yet price crossed that line 74 times. Staying mostly above the line marks a steady uptrend, and 74 crossings in under six years marks many small breaks. A 10/50 crossover turns on those breaks, and each one risks an exit followed by a later, higher re-entry. The 2022-12-06 entry and its exit the next day show the problem in one trade.
RINF is nearly uncorrelated with stocks, at 0.13 to SPY and 0.09 to QQQ, and it is negatively correlated with TLT at -0.4. Its beta to SPY was 0.09. TBF at 0.41, KMLM at 0.23 and XLF at 0.16 correlated most, and TMF and TLT at -0.4 and IEF at -0.37 correlated least. This is a fund that moves for its own reasons, and the rule has no information about those reasons.
The RSI statistics show a mild bounce. RSI(14) closed under 30 just 10 times, and the median return over the next 20 sessions was 2.99% against a 0.57% baseline. RSI(2) under 10 occurred 139 times with a median 20-day return of 1.23%. Short dips were followed by gains more often than usual, which a crossover rule never uses. The mean reversion templates did far better on this fund, as the comparison below shows.
The rules
A faster moving-average crossover: hold while the 10-day average is above the 50-day.
- WHEN the market opens · IF not invested AND SMA(10) > SMA(50) · THEN buy with 98% of the sleeve
- WHEN the market opens · IF invested AND SMA(10) < SMA(50) · THEN sell the whole position
A faster version of the golden cross. The 10- and 50-day averages catch intermediate trends measured in weeks rather than years. It enters recoveries earlier and exits breakdowns earlier, with more whipsaw trades in sideways markets. Useful for seeing how signal speed changes a strategy's results.
Good for: trending assets with multi-week swings, such as leveraged index ETFs.
Watch out: several false signals a year is normal; each whipsaw costs a small loss and they add up in flat markets.
How the rule met this fund
The 10-day simple average has to sit above the 50-day for the rule to hold. When flat and SMA(10) rises over SMA(50), it puts 98% of the sleeve in at the open; when long and SMA(10) drops under SMA(50), it sells the lot at the open. The template is pitched at assets with multi-week swings and carries a caution that several false signals a year are normal. This fund produced 16 losing trades in 5.74 years.
Results from the other eleven templates show what suited RINF. At the top, the RSI(2) snapback made 8.52% (drawdown 10.64%), followed by the weekly 7% target at 6.57% and RSI mean reversion at 6.34% (drawdown 9.76%). The monthly cycle managed 5.19% and the trend plus trailing stop 4%. The golden cross returned 2.38%, and the dip buyer 1.32% with a drawdown of just 2.78%. Four templates and this one finished negative: the EMA 12/26 trend at -0.22%, the 3-month momentum rule at -0.38%, the 200-day regime filter at -2.11% and the momentum breakout at -2.27%, next to this rule's -2.12%.
Dip-buying and small-profit rules led on this fund, and rules that follow a trend trailed. The two closest trend relatives, the golden cross and the EMA 12/26, both finished ahead of this rule.
Among alternative-strategy funds, the same rule made 16.61% on CLSE, 3.87% on CTA, 3.13% on KMLM, 2.91% on ALTY and 2.12% on QAI, and lost on RINF at -2.12%. The category median was 3.13%, and the all-fund median for this rule 2.12%. RINF had the most round trips of the six, 25 against between 12 and 19, and the deepest drawdown at 23.92%. See the strategy hub and the RINF page for the full tables.
What the test leaves out
RINF gained in each of the six calendar years covered. A stretch with a sustained move in this fund, up or down, would give a crossover something to follow and could change the result. The rule is long or flat only. The headline run has no fees or margin, and with such thin volume actual fills could differ from the bar prices used. None of this is 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
Did SMA 10/50 trend beat buy-and-hold on RINF?
Over 2021-01-04 to 2026-10-02, SMA 10/50 trend on RINF returned −2.1% annualized vs 6.6% for buy-and-hold: it trailed buy-and-hold by 8.8% per year, with a maximum drawdown of 23.9% (buy-and-hold: 13.3%).
How many trades did it make?
25 completed round trips over 5.7 years (51 fills), with 36% of round trips closing profitably.
Why 10 and 50 days?
A common intermediate-trend pairing. It reacts within weeks and ignores single bad days. Both windows are editable parameters in DeployQuant.
Did SMA 10/50 trend beat buy-and-hold on RINF?
No. It made -2.12% a year to the fund's 6.63%, finishing at $8,844 against $14,460, with a maximum drawdown of 23.92% against 13.32%. 2023 was the only year it led.
How many trades did the rule make?
The count was 25 closed round trips plus one open, or 51 fills. Nine won, which is 36%. Holds averaged 47.7 days and ran from 1 day to 189.
Why did the crossover lose money on a fund that kept rising?
RINF climbed in small steps with frequent pullbacks and crossed its 200-day average 74 times. The crossover sold after dips and bought back after rallies, and 16 of its 25 trades lost. In 2022 the rule fell 14% while the fund rose 8.8%.
Do other moving-average speeds work better on RINF?
None was profitable. SMA 20/50 was closest at -1.18% a year over 17 trades, SMA 5/50 returned -2.44% and SMA 10/100 returned -3.99%. That is a small sample from one fund.
How much do trading costs matter?
A 5 basis point charge per fill took the CAGR to -2.5% and 10 basis points to -3%. With 51 fills and many trades under 1%, the charges bite. Median minute volume of 300 shares also means spreads could be wider in practice.
Which templates worked best on RINF?
RSI(2) snapback led with 8.52% a year, then the weekly 7% target at 6.57% and RSI mean reversion at 6.34% with a 9.76% drawdown. Rules that follow trends sat near the bottom of the ranking.
Related
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.