Seasonality in Bitcoin Intraday Trend Trading

As our readers are aware, futures trend trading, particularly at higher frequencies, represents a core area of Concretum’s expertise, with a meaningful share of our trading risk allocated to this family of models.

Over recent years, we have also published several papers presenting simple and accessible variations of trend-following strategies, all of which have been well received.

Building on this foundation, we recently extended our research into the cryptocurrency space. Motivated by our earlier work, “Catching Crypto Trends: A Tactical Approach for Bitcoin and Altcoins” (available on SSRN), which documents clear multi-day trend-following tendencies in Bitcoin, we investigated whether similar dynamics are exploitable at intraday frequencies.

Put simply: does intraday Trend-Following work on Bitcoin? 

To address this question, we constructed a long–short portfolio based on an ensemble of multiple high-frequency trend-following models. This approach allows us to isolate Bitcoin’s response to trend signals while minimizing model-specific risk. As in our other research, portfolio exposure is volatility-scaled to target an annualized volatility of 20%. The resulting portfolio serves as a robust high-frequency trend-following benchmark.

Over the 2018–2025 period, this benchmark produced compelling results, achieving a gross-of-fees Sharpe ratio of approximately 1.6. For comparison, a volatility-targeted long-only Bitcoin portfolio (20% target volatility) delivered a Sharpe ratio just below 0.8, roughly half that of the actively managed trend-following approach.

Given that Bitcoin trades continuously, without a formal market open or close, we next examined whether the performance of our high-frequency trend benchmark displays intraweek seasonality, and whether such patterns may be linked to the operating hours of major global financial centers.

Importantly, this analysis does not concern the intraweek seasonality of passive Bitcoin returns, a topic already addressed in the literature. Instead, we focus on the seasonality of intraday trends on Bitcoin. When the benchmark rises, it indicates the presence of persistent and exploitable trends (long or short); when it declines, it highlights periods characterized by weaker trends and more mean-reverting price behavior.

Figure 1 – Intraweek seasonality of our Bitcoin high-frequency trend benchmark. The strategy takes both long and short positions with leverage adjusted via volatility targeting. Black dashed lines mark US equity market open (9:30 AM) and close (4:00 PM). Red dotted lines indicate Tokyo Stock Exchange open at 19:00 ET (winter) and 20:00 ET (summer), reflecting the US daylight saving shift (Japan does not observe DST).

After a quick inspection, a clear pattern emerges: our intraday trend benchmark delivers strongly positive returns starting on Sunday at around 7:00 PM New York time, with performance remaining elevated for roughly the next 24 hours into Monday. Notably, this upswing closely aligns with the Monday open of Asian cash equity markets (see the Tokyo Stock Exchange opening hours highlighted in the figure). In contrast, US Sunday morning is associated with negative benchmark performance, consistent with choppier price action, weaker trend persistence, and a higher propensity for mean reversion.

Internally, we refer to this pattern as the “Monday Asia Open Effect.” One plausible explanation is that fresh liquidity and positioning flows from Asian market participants at the start of the week help establish directional price moves that persist over subsequent hours, creating conditions favorable to intraday trend-following strategies. 

This interpretation is consistent with mechanisms observed in equity markets. Fundamental news and macro developments primarily affect desired portfolio allocations rather than prices directly. The resulting rebalancing trades are not executed instantaneously; due to liquidity constraints and optimal execution considerations, they are typically fragmented and carried out over time. This process generates persistence in order flow, inducing positive serial correlation in intraday returns and giving rise to short-term price trends.

When splitting the sample into pre- and post-mid-2020 periods, the “Monday Asia Open Effect” becomes substantially more pronounced in the latter subsample. This timing coincides with the well-documented entry of larger institutional participants into the cryptocurrency ecosystem, further supporting the interpretation that capital reallocation and execution dynamics play a central role.

Figure 2 – Bitcoin high-frequency trend benchmark intraweek seasonality by sample period: 2018-2020H1 (blue) versus 2020H2-2025 (red). Black dashed lines mark US equity open/close, red dotted lines indicate Tokyo open (19:00-20:00 ET depending on DST). Notice how the “Monday Asia Open Effect” effectively appears after the second half of 2020.

Of course, this remains only one plausible explanation. Market phenomena are rarely attributable to a single cause; rather, they typically arise from the interaction of multiple structural and behavioral forces.

We welcome readers to share alternative hypotheses or related empirical observations from their own trading or research. A deeper exploration of how to systematically exploit intraweek seasonality in Bitcoin trend dynamics may be the subject of a future dedicated study.

For questions or feedback, feel free to reach out at info@concretumgroup.com