DeployQuantFirst-to-Last Day of Month on SPY
SPDR S&P 500 ETF Trust: the most-traded S&P 500 ETF and a common benchmark for US large caps. Backtest 2021-01-04 to 2026-10-02, $10,000 starting capital, computed by the same engine that runs live DeployQuant strategies.
The monthly cycle on SPY buys at the open of the first session of each month and sells at the open of the last session. It has no indicator and no stop. From 2021-01-04 to 2026-10-02 it made 69 closed round trips, won 48 of them for a win rate of 70%, and turned $10,000 into $20,007. Buy-and-hold turned the same amount into $21,820. The CAGR was 12.84% against 14.56%, and the maximum drawdown was 22.4% against 24.0%.
The rule held a position at the close on 95.2% of trading days, so its result is close to holding, less the days it skips. It skips the last session of each month and the overnight gap into the next month's first open. The 1.72 point gap in CAGR is the difference over the window. The drawdown was 1.63 points shallower than holding, which is a small difference for the amount of return given up.
On SPY the template ranked third of 12 by CAGR, behind the weekly 7% target at 13.88% and RSI(2) snapback at 13.17%. Across the 59 funds, SPY was its seventeenth best result, and the template's median CAGR across all 59 was 5.46%. The median across the broad index funds was 12.83%, so SPY sits at the median of its group.
The interesting part of the result is the spread inside the 69 trades. A 70% win rate comes with an average win of 3.27% and an average loss of 3.83%, and the profit factor was 2.00. The numbers cover one window of 5.74 years.
Year by year
| Year | monthly cycle | buy & hold |
|---|---|---|
| 2021 | 26.4% | 29.6% |
| 2022 | −19.6% | −17.8% |
| 2023 | 22.4% | 25.5% |
| 2024 | 25.4% | 24.3% |
| 2025 | 18.4% | 17.4% |
| 2026 | 10.1% | 13.6% |
Year by year against holding
The year table shows the rule within a few points of holding in every year, ahead in two and behind in four.
2021: the rule returned 26.4% against 29.6%, a gap of 3.2 points. Eleven of its 12 round trips that year won. The only losing month was September at negative 3.2%. The other eleven months ranged from 0.35% to 5.5%, and the fund's own calendar-year return was 30.46%.
2022: the rule returned negative 19.6% against negative 17.8%, a gap of 1.8 points. Only 5 of 12 round trips won. The rule lost 7.2% in January, 3.6% in February, 6.2% in April, 8.6% in June, 2.4% in August, 7.1% in September, and 6.0% in December, and it gained in March, May, July, October, and November. A rule that holds almost all of the time cannot avoid those months. The fund's calendar-year loss was 18.16%, so the rule did slightly worse than holding in the year of the bear market, by a small amount.
2023: 22.4% against 25.5%, a gap of 3.1 points, with 8 of 12 round trips winning. The losing months were February, August, September, and October, with September the worst at negative 4.3%. November returned 8.5% and December 5.0%, a strong close to the year.
2024: 25.4% against 24.3%, ahead by 1.1 points, with 10 of 12 round trips winning. The losing months were April at negative 2.8% and December at negative 1.8%. This is one of the two years in which the rule beat holding.
2025: 18.4% against 17.4%, ahead by 1.0 point, with 8 of 12 winning. The rule's March was negative 7.2%, the same size as its January 2022. The rest of the year was positive, from May to October, with 4.8%, 5.0%, 3.5%, 3.4%, 4.3%, and 3.2%.
2026 to date: 10.1% against 13.6%, behind by 3.5 points, the largest gap of the six years. Six of 9 round trips won. April returned 8.6% and March negative 5.4%.
The pattern is that the rule gave up a few points in rising years and held up about as well as the fund in falling ones. The two years in which it was ahead were 2024 and 2025.
Month by month
| Year | Jan | Feb | Mar | Apr | May | Jun | Jul | Aug | Sep | Oct | Nov | Dec |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2021 | 1.7% | 2.6% | 2.7% | 4.5% | 0.3% | 1.4% | 2.0% | 2.6% | −3.2% | 5.5% | 0.3% | 3.4% |
| 2022 | −7.2% | −3.6% | 5.2% | −6.2% | 0.2% | −8.6% | 7.7% | −2.4% | −7.1% | 6.9% | 1.4% | −6.0% |
| 2023 | 4.3% | −1.9% | 2.7% | 0.7% | 0.7% | 5.8% | 3.1% | −0.9% | −4.3% | −2.4% | 8.5% | 5.0% |
| 2024 | 3.3% | 4.6% | 2.9% | −2.8% | 4.2% | 3.6% | 0.5% | 1.4% | 2.0% | 0.4% | 4.8% | −1.8% |
| 2025 | 3.0% | −1.3% | −7.2% | −1.9% | 4.8% | 5.0% | 3.5% | 3.4% | 4.3% | 3.2% | −0.6% | 1.4% |
| 2026 | 1.0% | −1.0% | −5.4% | 8.6% | 4.6% | −1.5% | 0.1% | 2.3% | 0.9% | 0.7% | – | – |
Which months paid and which did not
A monthly rule makes the month table the trade list. Each cell is one round trip, apart from the open position at the end.
The best month was April 2026 at 8.61%. Buy-and-hold's best month was also April 2026, at 10.33%. The rule's April trade ran from 2026-04-01 to 2026-04-30 and gained 9.09%, the best trade in the whole window. The worst month was June 2022 at negative 8.64%. Buy-and-hold's worst month was September 2022 at negative 9.0%.
September was the weakest calendar month for the fund. The averages in the profile put September at negative 2.21% and November at 4.01%, and the rule's table shows the same shape: September returned negative 3.2% in 2021, negative 7.1% in 2022, negative 4.3% in 2023, and positive 2.0%, 4.3%, and 0.9% in 2024, 2025, and 2026. Each calendar month has only 5 or 6 observations, so this is a description of one window.
The table also shows how bad months cluster. In 2022 there were seven losing months out of 12. In 2025 the sequence of February, March, and April lost 1.3%, 7.2%, and 1.9%. In 2026 February and March lost 1.0% and 5.4%, and then April gained 8.6% and May gained 4.6%. A calendar rule has no way to step aside during a sequence like that, and it also has no way to miss the rebound that follows. The monthly cycle took part in both halves of those swings.
October 2022 returned 6.9% and July 2022 returned 7.7%. Both came in the middle of the bear market. The rule was invested for all of that rebound, and that is the reason its 2022 loss was close to the fund's and not larger.
Every trade
monthly cycle on SPY made 69 closed round trips and one position still open at the end of the test, an average hold of 28 days, an average winner of 3.27%, an average loser of −3.83%, a profit factor of 2.00, a longest losing streak of 3. It held a position at the close on 95.2% of trading days.
Best 10 round trips
| Entry | Entry price | Exit | Exit price | Return | Days held |
|---|---|---|---|---|---|
| 2026-04-01 | $650.88 | 2026-04-30 | $710.06 | 9.1% | 29 |
| 2023-11-01 | $404.10 | 2023-11-30 | $439.10 | 8.7% | 29 |
| 2022-07-01 | $356.69 | 2022-07-29 | $385.39 | 8.1% | 28 |
| 2022-10-03 | $342.88 | 2022-10-31 | $366.97 | 7.0% | 28 |
| 2023-06-01 | $400.05 | 2023-06-30 | $423.94 | 6.0% | 29 |
| 2021-10-01 | $403.50 | 2021-10-29 | $426.48 | 5.7% | 28 |
| 2022-03-01 | $409.10 | 2022-03-31 | $431.11 | 5.4% | 30 |
| 2025-06-02 | $578.87 | 2025-06-30 | $608.92 | 5.2% | 28 |
| 2023-12-01 | $439.34 | 2023-12-29 | $461.58 | 5.1% | 28 |
| 2025-05-01 | $551.15 | 2025-05-30 | $578.49 | 5.0% | 29 |
Worst 10 round trips
| Entry | Entry price | Exit | Exit price | Return | Days held |
|---|---|---|---|---|---|
| 2022-06-01 | $391.26 | 2022-06-30 | $356.09 | −9.0% | 29 |
| 2022-09-01 | $371.85 | 2022-09-30 | $344.19 | −7.4% | 29 |
| 2022-01-03 | $447.19 | 2022-01-31 | $414.04 | −7.4% | 28 |
| 2025-03-03 | $584.36 | 2025-03-31 | $541.12 | −7.4% | 28 |
| 2022-04-01 | $426.55 | 2022-04-29 | $398.70 | −6.5% | 28 |
| 2022-12-01 | $387.70 | 2022-12-30 | $363.24 | −6.3% | 29 |
| 2026-03-02 | $673.89 | 2026-03-31 | $636.36 | −5.6% | 29 |
| 2023-09-01 | $435.54 | 2023-09-29 | $415.83 | −4.5% | 28 |
| 2022-02-01 | $422.51 | 2022-02-28 | $406.47 | −3.8% | 27 |
| 2021-09-01 | $421.71 | 2021-09-30 | $407.85 | −3.3% | 29 |
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.
Best and worst trades
All 69 closed trades held between 25 and 30 days, with a median of 29. The median trade gained 1.49%. The open position, entered on 2026-10-01 at $763.99, was up 0.74% at the end of the data. Prices are adjusted for splits and dividends.
The best trade was 9.09% from 2026-04-01 to 2026-04-30. The next four were 8.66% in November 2023, 8.05% in July 2022, 7.03% in October 2022, and 5.97% in June 2023. Two of the five best trades came in the 2022 bear market, in the months when the fund rebounded.
The worst trade was a loss of 8.99% from 2022-06-01 to 2022-06-30. The next four were 7.44% in September 2022, 7.41% in January 2022, 7.40% in March 2025, and 6.53% in April 2022. Four of the five worst trades came in 2022. The fifth, March 2025, came just before the April 2025 low on 2025-04-08.
The trade distribution is almost symmetrical. The best trade was 9.09% and the worst was 8.99%. The average win was 3.27% and the average loss was 3.83%, so the average loser was larger than the average winner, and a 70% win rate was needed to produce a profit factor of 2.00. A rule with a 50% win rate and the same sizes would have lost money. The edge is the market's tendency to rise over a month, not anything the rule adds.
The longest winning streak was 8 trades and the longest losing streak was 3. The year-by-year wins show where the losses clustered: 11 of 12 in 2021, 5 of 12 in 2022, 8 of 12 in 2023, 10 of 12 in 2024, 8 of 12 in 2025, and 6 of 9 in 2026. Losing trades clustered in 2022 and 2025.
Largest drawdowns
| Peak | Low point | Depth | Days to low | Recovered | Days to recover |
|---|---|---|---|---|---|
| 2021-12-29 | 2022-10-12 | −22.4% | 287 | 2024-01-19 | 464 |
| 2025-02-19 | 2025-04-08 | −20.3% | 48 | 2025-08-04 | 118 |
| 2024-07-16 | 2024-08-05 | −8.9% | 20 | 2024-09-19 | 45 |
Buy-and-hold's deepest drawdown ran from 2022-01-03 to 2022-10-12 and reached −24.0%.
Drawdowns against holding
The rule's largest drawdown was 22.4%, from the peak on 2021-12-29 to the low on 2022-10-12, which took 287 days. Recovery took until 2024-01-19, another 464 days. Buy-and-hold fell 24.0% from 2022-01-03 to the same low on 2022-10-12 and recovered on 2023-12-13. The rule's peak came a few days before the fund's, its low came on the same day, and its recovery came later than the fund's.
The second drawdown was 20.3%, from 2025-02-19 to 2025-04-08, recovered on 2025-08-04 after 118 days. Holding fell 18.4% over the same period and recovered on 2025-06-26. This is the one drawdown where the rule was deeper than holding. The March 2025 trade lost 7.40%, and the rule re-entered on the first session of April, before the low on 2025-04-08.
The third was 8.9%, from 2024-07-16 to 2024-08-05, a quick drop recovered in 45 days on 2024-09-19.
The summary is that the rule was 1.63 points shallower than holding in the worst case and deeper in the second. A calendar rule is not a risk control. With 95.2% exposure, it takes almost every decline, and the small differences come from the days it skips. The Sharpe ratio was 0.88 against 0.95 for holding.
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) | 12.8% | −22.4% | $20,007 | 0.88 |
| 5 basis points | 11.4% | −23.1% | $18,621 | 0.80 |
| 10 basis points | 10.0% | −24.0% | $17,305 | 0.72 |
139 fills and the cost runs
The rule made 139 fills, two for each round trip and one for the open position. At 5 basis points of slippage per fill the CAGR was 11.44% and the final value was $18,621. At 10 basis points the CAGR was 10.02% and the final value was $17,305, against $20,007 in the headline run. The maximum drawdown went from 22.36% to 23.05% and then 23.98%, and the Sharpe ratio fell from 0.88 to 0.801 and 0.718.
The cost of a calendar rule comes from trading every month for a result that already trails holding by 1.72 points. At 10 basis points the gap to holding widens further. A template that trades once and holds has no such cost.
SPY trades an average of $30,453,859,411 a day with a median minute volume of 107,870 shares, so a $10,000 order is a very small part of the volume. The cost runs apply a flat offset and do not model the spread on specific days.
Changing the parameters
| Version | CAGR | Max drawdown | Round trips | Win rate | Final value |
|---|---|---|---|---|---|
| Published rules | 12.8% | −22.4% | 69 | 70% | $20,007 |
| Only after a positive month | 5.3% | −27.2% | 46 | 67% | $13,464 |
| Only above the 200-day SMA | 4.4% | −24.6% | 46 | 65% | $12,829 |
| With a 5% profit target | 12.5% | −21.5% | 69 | 71% | $19,673 |
Three attempts to improve the calendar rule
We ran three variants of the monthly cycle on SPY. Two gate the entry, and one adds a profit target.
"Only after a positive month" enters at the first session only if the previous month finished up. It returned 5.32% with a 27.2% drawdown and a Sharpe ratio of 0.531, making 46 round trips with 31 wins. "Only above the 200-day SMA" enters only if the fund is above its 200-day average. It returned 4.43% with a 24.6% drawdown and a Sharpe ratio of 0.473, with 46 round trips and 30 wins. "With a 5% profit target" rests a limit order at 5% above entry and returned 12.51% with a 21.5% drawdown and a Sharpe ratio of 0.894, with 69 round trips and 49 wins.
The gated versions lost most of the return. The published rule returned 12.84%, and the gates returned 5.32% and 4.43%, even though each of them made 46 round trips against 69. The trade lists explain why. Several of the best months came right after bad ones: July 2022 returned 7.7% after June's negative 8.6%, October 2022 returned 6.9% after September's negative 7.1%, and April 2026 returned 8.6% after March's negative 5.4%. A gate that waits for a positive month or a price above the 200-day line skipped those rebounds. The fund was above its 200-day average on 78.0% of sessions and crossed it 30 times, so the trend gate switched on and off often.
The 5% profit target variant is the mild case. It returned 12.51% against 12.84%, with a drawdown of 21.5% against 22.4% and a Sharpe ratio of 0.894 against 0.88. It made the same 69 round trips and won 49 instead of 48. The results are within a few tenths of the published rule, so the target did little. The trade count is the same because the rule re-enters at the next month's start.
The variants test whether a simple filter improves a calendar rule. On SPY in this window, the two filters made it worse, and the profit target made little difference. The 200-day filter on SPY as a standalone rule returned 8.3%, which is also below the plain calendar rule here.
How SPY behaved
| Measure | SPY |
|---|---|
| Data in this test | 2021-01-04 to 2026-10-02 (1444 sessions) |
| Total return, buy and hold | 125.3% |
| Annualized volatility | 16.4% |
| Deepest drawdown | −24.5% (2022-01-03 to 2022-10-12) |
| Up days | 54.4% |
| Average daily range | 1.15% |
| Average overnight gap | 0.44% |
| Correlation to QQQ | 0.94 |
| Correlation to TLT | 0.08 |
| Sessions above the 200-day average | 78.0% |
| Crossings of the 200-day average | 30 |
| Falls of 10% or more from a 20-day high | 11 |
What SPY's behaviour says about a first-to-last-day rule
SPY's annualized volatility was 16.4% and its total return in the window was 125.29%. Up days were 54.4% of sessions, with an average up day of 0.74% and an average down day of negative 0.74%. The overnight session produced 62.46% of the fund's log return and the intraday session 37.54%.
That split bears on this rule. The monthly cycle buys at the first session's open, so it misses the overnight gap into that open, and it sells at the last session's open, so it misses the last session's trading and the next overnight. The fund's overnight share of return was the larger part, and the rule skipped one overnight move a month. The gap to holding of 1.72 points is consistent with those skipped moves.
The fund closed above its 200-day average on 78.0% of sessions. Its deepest drawdown on its own series was 24.51%, from 2022-01-03 to 2022-10-12, with recovery on 2023-12-13 after 488 sessions. It had 11 falls of 10% or more from a 20-day high, over 23 days. The best days were 2025-04-09 at 9.39% and 2022-11-10 at 5.48%, and the worst were 2025-04-04 at negative 5.98% and 2025-04-03 at negative 4.78%. A rule that holds through the month takes both.
On calendar months, the fund averaged 4.01% in November, 3.07% in July, 3.00% in May, and 2.88% in October, and negative 2.21% in September. February averaged 0.06%. Each has 5 or 6 observations. The monthly cycle does not choose among these months, and each one is in the 69 trades.
Among the broad index funds, the monthly cycle returned 15.80% on IOO, 14.95% on QQQM, 14.73% on QQQ, 13.91% on VOOG, 12.83% on VOO, and 12.44% on VV. Lower down were QQQE at 7.95%, IWM at 6.48%, and EEM at 3.14%, with drawdowns of 29.11%, 36.88%, and 44.39%. Every fund in the group made 69 round trips, so the differences come from price paths. The Nasdaq-100 funds returned more than the S&P 500 funds, with drawdowns of 36.63% on QQQM and 36.07% on QQQ.
On SPY the best template was the weekly 7% target at 13.88%, followed by RSI(2) snapback at 13.17%, which had a drawdown of 12.69%. The monthly cycle had the third-best CAGR and a drawdown of 22.36%, the second deepest among the top five.
The rules
Buy the first session of each month and sell the last, as a calendar-seasonality test.
- WHEN the first session of the month opens · IF not invested · THEN buy with 98% of the sleeve (once per month)
- WHEN the last session of the month opens · IF invested · THEN sell the whole position
This template uses no indicators. It tests the turn-of-the-month effect by holding from each month's first open to its last. Whatever it earns or loses comes from calendar seasonality alone, so it is a useful comparison against buy-and-hold on the same page.
Good for: measuring how much of an asset's return accrues inside the month versus across month boundaries.
Watch out: this is a research template. It holds ~95% of all sessions, so results usually track buy-and-hold minus the boundary days.
How the two rules handle the month boundary
The buy rule fires on the first session of the month, once per month, if the account is not invested. The sell rule fires on the last session of the month if the account is invested. Both act at the open. The position takes 98% of the sleeve.
The design has one feature that explains the exposure figure. The rule holds from the first session's open to the last session's open, so it is out of the market for one session's trading at the end of every month and for the overnight moves on each side of the boundary. A month with 21 sessions leaves the account invested through 20 of them, so the exposure figure sits near 95%.
The rule has no way to react to what happens in the month. A loss of 8.99% in June 2022 was held to the sale on 2022-06-30, and a gain of 9.09% in April 2026 was held to the sale on 2026-04-30. The template page describes it as a research template, and the results here are consistent with that: it tracks the fund closely and adds nothing beyond what the calendar provides.
The other templates on SPY introduce a reason to be out of the market. The golden cross returned 8.71% and the EMA 12/26 trend returned 10.20% with a drawdown of 11.84%. In this window the calendar rule returned more than either, and the signals gave a shallower drawdown. The comparison is one test on one fund, and says little about how the choice would turn out over a different 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
Did monthly cycle beat buy-and-hold on SPY?
Over 2021-01-04 to 2026-10-02, monthly cycle on SPY returned 12.8% annualized vs 14.6% for buy-and-hold: it trailed buy-and-hold by 1.7% per year, with a maximum drawdown 1.6 points shallower than holding (22.4% vs 24.0%).
How many trades did it make?
69 completed round trips over 5.7 years (139 fills), with 70% of round trips closing profitably.
Is the turn-of-the-month effect real?
It has appeared in long historical studies, but it depends on the market regime and is small. These pages show the recent five and a half years per ETF.
Does buying the first day and selling the last day of the month beat buy-and-hold on SPY?
Not in this test. From 2021-01-04 to 2026-10-02 the monthly cycle returned 12.84% a year against 14.56% for holding. It beat holding in 2024 and 2025 and trailed in the other four years. The maximum drawdown was 22.36% against 23.99%.
How many trades does the monthly cycle make on SPY?
It made 69 closed round trips and had one position open at the end, entered on 2026-10-01. Forty-eight of the 69 won, a win rate of 70%. The average win was 3.27% and the average loss was 3.83%.
What was the worst month for the monthly cycle on SPY?
June 2022 at negative 8.64%. The trade from 2022-06-01 to 2022-06-30 lost 8.99%. Four of the five worst trades came in 2022.
Does a trend or momentum filter improve the monthly cycle on SPY?
Not here. Entering only after a positive month returned 5.32% and entering only above the 200-day average returned 4.43%, against 12.84% for the plain rule. A 5% profit target returned 12.51%.
How much do costs matter for the monthly cycle on SPY?
The rule made 139 fills. The CAGR went from 12.84% to 11.44% at 5 basis points and 10.02% at 10 basis points. The final value fell from $20,007 to $17,305.
How does the monthly cycle rank among strategies on SPY?
Third of 12 by CAGR, behind the weekly 7% target at 13.88% and RSI(2) snapback at 13.17%. Across 59 funds, SPY was its seventeenth best result.
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.