Cryptocurrency Seasonality: Which Months Historically Tend to Rise the Most?
Historically, Bitcoin has performed strongest in October and November, with average returns of about 20%–23% and 30%–46%, respectively; September is often called the "curse month," with an average decline of about 3.77% over the past decade. However, these are statistical patterns, not investment advice — the sample size is small, variance is high, and the risk of blindly basing trading decisions on the month is much higher than you might think.
Prerequisites
Before analyzing seasonal patterns, confirm two things:
You know what you plan to do with this information — whether it's adjusting a DCA (dollar-cost averaging) rhythm, gauging current market sentiment, or simply understanding the historical context. Different purposes call for different uses.
You have a bottom-line rule for position management — no matter what month it is, the rule that a single trade's maximum loss should not exceed 1%–2% of total capital must not be relaxed just because "this month has historically risen a lot."
1. First, Recognize the Year's "Strong vs. Weak" Pattern, Not "Certainty Signals"
What to do: Identify which months have historically been relatively strong and which have been relatively weak, building a general sense of rhythm.
How to do it:
According to historical data from 2013 to 2024, the approximate distribution of Bitcoin's monthly performance is as follows:
Strong months: October (average +20% to +23%), November (average +30% to +46%), February (average +13%), April (average +12%–13%)
Weak months: September (average -3% to -4%, known as the "September curse"), January (elevated risk), August (often shows negative returns)
Highly volatile months: May is called the "devil month" — in 2017 and 2019, May surged over 52%, but in 2021 it fell 35% and in 2022 it fell 15%.
What counts as done: You can name at least two strong months and two weak months, and you understand that these are "historical probabilities" rather than "a pattern that will definitely repeat this year."
Common reason for failure: Treating historical patterns as "trading signals," such as blindly going long every October or liquidating every September. A counterexample: In October 2025, Bitcoin only fell slightly by around 0.76%, failing to replicate the historical average strong performance.
2. Understand the "Holiday Effect" and "Intraweek Patterns" as Supplementary References
What to do: Beyond monthly patterns, there are two seasonal regularities frequently discussed in academia and among institutions.
How to do it:
"Halloween Effect": A 2025 academic study found that during the 2018–2024 period, the returns of major cryptocurrencies like Bitcoin and Ethereum during the Halloween period (October 31 to April 30 of the following year) were significantly higher than during the rest of the year. This strategy only requires two operations per year (entering in November and exiting in May), making it more suitable as a quarterly macro rhythm reference.
Intraweek patterns: Academic research shows that liquidity, trading volume, and volatility in cryptocurrencies peak midweek (Wednesday and Thursday) and decline markedly over the weekend. This leads to a more actionable observation: when liquidity dries up on weekends, prices can be more easily pushed or depressed by smaller amounts of capital.
What counts as done: You know roughly which months the "Halloween Effect" covers and which days of the week the market tends to be relatively active or quiet.
3. Understand Coinbase's Warning: Seasonal Patterns Are Not Statistically Significant
What to do: Understand why institutional investors do not rely solely on calendar months to make decisions.
How to do it:
In 2025, Coinbase's global research department published a detailed report, using five statistical methods to examine Bitcoin's monthly seasonality. The conclusion consistently pointed in one direction: the monthly "calendar month" factor is not a reliable predictive indicator.
There are two core issues:
Sample size is too small: Bitcoin has only 10–12 years of complete monthly data, meaning each month has just over 10 data points. Statistically, such a small sample cannot yield definitive conclusions.
High randomness: Some months that appear "strong" have price ranges that overlap heavily with other months, making it impossible to rule out random variance rather than a genuine calendar effect.
In other words: Historical data does show that certain months have higher average gains, but statistically it cannot be proven that this is "not by chance."
Risk reminder: If you go all-in long at the start of October just because "October has historically risen a lot," and then encounter a mild decline like October 2025, your only basis for the decision is "historical probability" without any stop-loss bottom line. No seasonal pattern can replace position management.
4. Place Seasonal Patterns in the "Decision Reference Layer," Not the "Decision Core Layer"
What to do: Determine the "priority" of seasonal information in your trading decisions — where it should be placed.
How to do it:
Correct usage:
If you are dollar-cost averaging, you could increase the allocation amount in relatively weak months (like September) — but only if you are mentally prepared for further price declines.
If you are making a quarterly macro judgment, the historical strength of October–November can serve as one reference, but not the only one.
If you want to know "why market volatility is so low recently," the weekend liquidity drop could be one reason.
Incorrect usage:
Going to cash in September just because it has historically dropped, and going all-in in October because it has historically risen — such an approach directly relies on a statistically insignificant signal.
Ignoring stop-loss orders just because a certain month "averages strong gains" — black swans can happen in any month.
What counts as done: In your trading process, monthly seasonality is marked as "reference information" rather than a "signal source." You still use technical analysis, on-chain data, or fundamental judgment when making entry decisions; seasonality merely serves as an auxiliary.
Reference: Overview of Bitcoin's Historical Monthly Performance
The following data is based on historical statistics from 2013–2024, for reference only, not constituting trading advice:
| Month | Historical Characteristic | Average/Typical Performance |
|---|---|---|
| January | Historically multiple sharp drops | Average ~+3.8%, but high variance |
| February | Historically multiple strong rallies | Average ~+13% |
| March | Relatively neutral performance | Average ~+12%, moderate volatility |
| April | Overall good but volatile | Average ~+12%–13% |
| May | "Devil month," extremely volatile | Average ~+8%, but single-month drop can exceed 35% |
| June | Near zero, often a turning month | Average ~-0.3% |
| July | Generally positive | Average ~+7.5% |
| August | Often negative returns | Average ~+1.7% |
| September | "September curse," weakest performance | Average ~-3.8% |
| October | One of the strongest months of the year | Average ~+20%–23% |
| November | Strongest month of the year, often sees major bull-market rallies | Average ~+30%–46% |
| December | Relatively stable, influenced by year-end effect | Average ~+5% |
After completing these four steps, you now know: which months Bitcoin has historically performed relatively strongly, and which have been relatively weak; and you also understand that the statistical reliability of these patterns is debated. The next step is not to revise your trading plan based on this information, but to write them into your market notes — under the "background knowledge" column, not the "strategy basis" column. The next time the market swings sharply, the first question you ask yourself should still be "Is my stop-loss still in place?" — not "What month is it?"
