Quick summary
Across 2013 2025, Bitcoin finished eight of thirteen Septembers lower, averaging minus 4.02 percent
Monthly pattern shows early strength, sharp mid month drop, partial recovery in the final week
Weekly averages highlight week three as weakest and week four as strongest rebound period
Three day data shows gains early, deepest losses on days nineteen to twenty one, then recovery
Bitcoin closed lower in eight of the thirteen Septembers from 2013 through 2025. This has given September a reputation as Bitcoin's weakest month. A closer look at the weekly and three day results shows a clearer pattern. Bitcoin has often started the month with some strength, fallen more sharply between September 15 and 21, and recovered part of the loss during the final full week. These averages show what happened in the past, but they do not predict what will happen in any future September.
How The Numbers Are Calculated
Three blocks of time get measured three different ways.
Monthly return: Price change from September 1st to September 30th, open to close.
Weekly and three-day return: Price change from the start to the end of that block, close to close.
Average: Each block's thirteen yearly results are averaged with a simple average, not compounded.
The Monthly Septemebr Bitcoin Read: 2013 to 2025
Year | Open | Close | High | Low | % Change |
2013 | $138 | $132 | $140 | $122 | -4.72% |
2014 | $476 | $387 | $490 | $373 | -18.59% |
2015 | $227 | $236 | $244 | $226 | +3.91% |
2016 | $572 | $608 | $625 | $572 | +6.42% |
2017 | $4,863 | $4,298 | $4,863 | $3,101 | -11.63% |
2018 | $7,049 | $6,573 | $7,374 | $6,191 | -6.75% |
2019 | $9,599 | $8,029 | $10,620 | $8,029 | -16.36% |
2020 | $11,649 | $10,839 | $11,899 | $10,134 | -6.95% |
2021 | $47,335 | $41,681 | $52,740 | $40,554 | -11.95% |
2022 | $20,021 | $19,564 | $22,339 | $18,528 | -2.28% |
2023 | $25,932 | $26,970 | $27,454 | $25,006 | +4.00% |
2024 | $58,967 | $63,243 | $66,439 | $52,690 | +7.25% |
2025 | $108,224 | $114,024 | $117,888 | $107,295 | +5.36% |
Data covers the thirteen completed Septembers from 2013 through 2025, September 2026 had not yet occurred as of this writing and is excluded. Source: DefiLlama.
Average change: negative 4.02 percent.
Median: negative 4.72 percent.
Best year: 2024, up 7.25 percent.
Worst year: 2014, down 18.59 percent, with 2019 close behind at negative 16.36 percent.
The Weekly Read: Averaged Across All 13 Years
Week of September | Avg % Change | Note |
Week 1 (days 1 to 7) | -0.81% | negative week |
Week 2 (days 8 to 14) | -0.95% | negative week |
Week 3 (days 15 to 21) | -2.19% | weakest week historically |
Week 4 (days 22 to 28) | +1.11% | strongest week, often the rebound |
Week 5 (days 29 to 30) | -0.11% | partial two-day period; not directly comparable |
The Three-Day Read: Averaged Across All 13 Years
Period | Days | Avg % Change |
1 | 1 to 3 | +1.10% |
2 | 4 to 6 | -0.91% |
3 | 7 to 9 | -1.38% |
4 | 10 to 12 | +0.21% |
5 | 13 to 15 | -1.45% |
6 | 16 to 18 | +0.68% |
7 | 19 to 21 | -1.98% (weakest block) |
8 | 22 to 24 | +1.36% |
9 | 25 to 27 | +0.61% |
10 | 28 to 30 | 0.00% |
What the finer view adds: The three-day blocks catch two things the weekly numbers alone miss.
Open is actually positive: Days 1 to 3 average a gain of 1.10 percent, even though Week 1 as a whole finishes negative.
Worst stretch is narrower than a week: Days 19 to 21 average a loss of 1.98 percent, the single worst block in the data.
Bitcoin September Seasonality Explained
The pattern is consistent across thirteen years, but the short timeframe means it should be treated with caution.
Thirteen years is a short sample: Thirteen years is enough to identify a pattern worth tracking, but not enough to treat it as a reliable rule. Coinjuice's 900-year review of Bitcoin's odds of lasting explains why longer timeframes provide stronger evidence.
Volatility goes both ways: Bitcoin can move sharply in either direction during this period. In 2017, it rose 11 percent and then fell 7 percent just days later. In 2021, it dropped 11 percent during the same period. In 2022, it gained 2 percent from days 16 to 18 before falling 3 percent from days 19 to 21.
Conclusion: What The Numbers Do Not Say
None of this data can predict what Bitcoin will do in September 2026. It only shows what happened during the thirteen completed Septembers from 2013 through 2025, five of which ended higher.
Market conditions, ETF flows, interest rates, and Bitcoin adoption change every year. History suggests that September has often finished lower, with the greatest weakness appearing in the middle of the month and some recovery during the final week. That pattern provides useful context, but it cannot tell us whether this year will follow it.
—-
Disclosure: AI-based research tools assisted with compiling and analyzing the data.
—-
Further reading: for traders considering position size during a historically choppier part of the calendar, Coinjuice's guide to trading Bitcoin without leverage covers one approach to staying in a position through a drawdown rather than being forced out. The full research library and subscription pricing for deeper coverage are available on Coinjuice's pricing page.
FAQ
Why is September considered one of Bitcoin’s weakest months?
From 2013 through 2025, Bitcoin finished September lower in eight of thirteen years. Its average return was negative 4.02 percent, while the median return was negative 4.72 percent.
What pattern has Bitcoin followed during September?
Bitcoin has often started the month with some strength before weakening. The largest declines have generally occurred between September 15 and 21, followed by a partial recovery during the final full week.
Which parts of September have been the weakest and strongest?
Days 15 through 21 were the weakest week, with an average loss of 2.19 percent. Days 22 through 28 were the strongest week, with an average gain of 1.11 percent. Among the three day periods, days 19 through 21 were the weakest at negative 1.98 percent, while days 22 through 24 were among the strongest at positive 1.36 percent.
Can this data predict how Bitcoin will perform in a future September?
No. The data only describes the thirteen completed Septembers from 2013 through 2025. It shows a historical pattern worth watching, but thirteen years is a small sample and cannot predict what Bitcoin will do in any future September.
Disclaimer
The information provided in this article is for informational purposes only. It is not intended to be, nor should it be construed as, financial advice. We do not make any warranties regarding the completeness, reliability, or accuracy of this information. All investments involve risk, and past performance does not guarantee future results. We recommend consulting a financial advisor before making any investment decisions.
More like this
Written by
Andrew Kamsky
Andrew Kamsky is a Bitcoin analyst. He spent a decade in traditional finance across a Big Four firm and a listed fintech bank before going deep on Bitcoin full-time.
