- Treasury balance growth is a weak and unreliable predictor of Bitcoin returns. While Strategy's balance changes show a statistically significant positive correlation with 1-week forward Bitcoin returns, this signal diminishes materially once serial dependence and overlapping observations are accounted for via permutation testing. Treasury balance dynamics may be more appropriately viewed as a secondary, confirmatory indicator - not a standalone forecasting tool.
- Strategy's purchase announcements follow a "buy the rumour, sell the news" pattern. Across 100 events since August 2020, Bitcoin peaks on average two hours before the announcement and drifts lower in the hours that follow. Large tranches - the top 10% by BTC volume - exhibit the most pronounced post-announcement weakness, with Bitcoin declining to 99.14 by t+120. The announcement itself carries little incremental bullish information.
- Smaller, quieter purchase announcements have historically been associated with more sustained positive intraday price action. The bottom 10% of tranches show essentially flat price action ahead of the announcement, followed by a steady grind higher to 100.51 by t+120. This suggests that low-profile accumulation - less susceptible to front-running - may carry a more genuine and sustained demand signal than headline-grabbing large purchase announcements.
Treasury Companies
Treasury companies have been one of the major sources of demand over past 2 years. The pace of buying has significantly accelerated after Trump took office in November 2024 and promised to promote the US to become the ‘crypto capital of the world'.
Last year alone, bitcoin treasury companies accounted for the vast majority of institutional demand for bitcoin even outsizing acquisitions by bitcoin ETPs. On aggregate, treasury companies already control 5.5% of the supply with Strategy (MSTR) – formerly known as Microstrategy - already accounting for more than 3% of the total supply based on Glassnode data.
The primary aim for a treasury company is to accumulate and hold bitcoins for their shareholders, ideally maximising bitcoin-per-share holdings.
Yet, treasury companies are a fairly recent phenomenon in the bitcoin ecosystem – the first company transitioned to become a dedicated treasury company as late as 2020.
However, given their significant (and growing) size, this research piece aims to analyse treasury company purchase (announcements) and their effect on the performance of bitcoin with a specific focus on Strategy (MSTR).
To begin our investigation, we assessed whether changes in reported corporate Bitcoin treasury balances provide any useful signal for future Bitcoin price performance. The analysis focused on three balance-based series:
- Total Balance
- Strategy Balance
- Total Balance minus Strategy
This segmentation allows us to isolate the impact of Strategy, the dominant treasury company by balance sheet, and uate whether broader corporate treasury activity contains incremental information once its influence is removed.
We first examine the relationship (180d rolling correlation) between 14-day percentage changes in both treasury holdings and the corresponding Bitcoin price.
Changes in reported Strategy balances exhibited the strongest, though still modest, positive correlation with price movements. The aggregated treasury balance series largely moved in lockstep with Strategy, reflecting its dominant contribution to overall treasury flows, while the remaining treasury companies appeared to move more independently.
We next uated the predictive relationship between treasury balance growth and subsequent Bitcoin performance by measuring the correlation between 14-day balance changes and forward Bitcoin returns over 1-week, 2-week, and 4-week horizons.
Across the sample, predictive correlations were generally modest in magnitude. Strategy balances produced the strongest results, with positive correlations at the 1- and 2-week horizons, suggesting some short-term alignment between large treasury accumulation and subsequent price performance.
In comparison, both the aggregated corporate balance series and the Strategy-excluded cohort exhibited weaker correlations with forward returns, although all subsets remained statistically significant.
Source: Glassnode, Bitwise Europe.
A key challenge in uating this relationship is that both the predictor (treasury balance growth) and the target (forward Bitcoin returns) are constructed using rolling windows. This mechanically introduces serial dependence through persistence and overlapping observations, which reduces the effective number of independent data points and can make weak relationships appear more statistically compelling than they are.
To address this issue, we implemented a stricter testing framework. Observations were sampled at non-overlapping intervals aligned with each forward return horizon, reducing mechanical dependence between adjacent data points. We then applied a circular permutation test, which preserves the internal time-series structure of both treasury balance changes and Bitcoin returns while repeatedly shifting their alignment. This approach allows us to assess whether the observed correlations are stronger than what would be expected from two persistent but otherwise unrelated market series.
Under this stricter framework, the apparent predictive signal weakened materially. Only Strategy balances at the 1-week horizon remained statistically significant, while results across other horizons and balance cohorts became inconsistent and largely inconclusive.
However, the modest magnitude of Strategy's 1-week correlation suggests that even this signal offers limited predictive power in practice, implying that treasury balance dynamics are better viewed as a secondary or confirmatory indicator of short-term Bitcoin price moves.
Source: Glassnode, Bitwise Europe.
An additional structural limitation relates to the timing of balance updates. Treasury datasets likely capture when balances were updated or reported, rather than when Bitcoin purchases actually occurred. This introduces the possibility of delayed information and weakens the link between observed balance changes and actual market transactions. In some cases, treasury balance increases may also reflect consolidation between existing holders or institutional transfers rather than genuine open-market accumulation.
Taken together, the evidence suggests that changes in corporate Bitcoin treasury balances may have a modest short-term correlation with future BTC returns, particularly in the Strategy balance series. However, once overlap and serial dependence are accounted for, the apparent predictive content diminishes materially, leaving little evidence that treasury balance growth constitutes a reliable standalone forecasting signal.
Strategy's Bitcoin Purchases: A Contrarian Intraday Signal?
Our event study analysis of Strategy's Bitcoin purchase announcements reveals a consistent and perhaps counterintuitive intraday pattern: Bitcoin tends to underperform in the two hours following an announcement, not outperform.
Across all 100 announcement events since August 2020, Bitcoin's average indexed performance peaks well before the announcement - at 100.28 two hours prior - and drifts modestly lower through the event window, reaching 99.97 at t+30 and 99.96 at t+60 before partially recovering. The median path tells the same story, suggesting this is not driven by outliers.
Sample: All announcements between August 2020 - March 2026
Gray bands denote min and max values, respectively. Past performance not indicative of future returns
The divergence between large and small purchase tranches is particularly instructive. For the top 10% of purchases by BTC volume - the most newsworthy announcements - Bitcoin averages 100.14 at t−30, consistent with pre-announcement price appreciation, before selling off to 99.82 at t+30 and continuing to weaken to 99.14 by t+120. This pattern is consistent with a classic buy the rumour, sell the news dynamic, where markets partially price in the announcement ahead of time, leaving limited upside once confirmed.
Sample: All announcements between August 2020 - March 2026
Past performance not indicative of future returns
Smaller purchases tell the opposite story. The bottom 10% of tranches show Bitcoin at 99.98 at t−30 - essentially flat - before grinding higher to 100.12 at t+30, 100.17 at t+60, and 100.51 by t+120. Quieter accumulation appears to carry a more sustained positive signal, possibly because it occurs with less front-running and possibly implies higher purchases in the future.
All in all, Strategy announcements are not a reliable short-term bullish catalyst for Bitcoin. If anything, large tranches have historically marked local intraday highs. The signal, to the extent one exists, lies in the accumulation trend itself - not the announcement.
Bottom Line
- Treasury balance growth is a weak and unreliable predictor of Bitcoin returns. While Strategy's balance changes show a statistically significant positive correlation with 1-week forward Bitcoin returns, this signal diminishes materially once serial dependence and overlapping observations are accounted for via permutation testing. Treasury balance dynamics may be more appropriately viewed as a secondary, confirmatory indicator - not a standalone forecasting tool.
- Strategy's purchase announcements follow a "buy the rumour, sell the news" pattern. Across 100 events since August 2020, Bitcoin peaks on average two hours before the announcement and drifts lower in the hours that follow. Large tranches - the top 10% by BTC volume - exhibit the most pronounced post-announcement weakness, with Bitcoin declining to 99.14 by t+120. The announcement itself carries little incremental bullish information.
- Smaller, quieter purchase announcements have historically been associated with more sustained positive intraday price action. The bottom 10% of tranches show essentially flat price action ahead of the announcement, followed by a steady grind higher to 100.51 by t+120. This suggests that low-profile accumulation - less susceptible to front-running - may carry a more genuine and sustained demand signal than headline-grabbing large purchase announcements.
Important Information
This publication constitutes a marketing communication and is provided for informational purposes only. It is intended solely for professional investors and is not suitable for retail distribution and reliance. It does not constitute investment advice, a personal recommendation, or an offer or solicitation to buy or sell any financial instrument.
This document (which may take the form of a presentation, press release, social media post, blog article, broadcast communication or similar instrument – collectively referred to as a “Document”) is issued by Bitwise Europe GmbH (“BEU” or the “Issuer”) and has been prepared in accordance with applicable laws and regulations, including those relating to financial promotions.
Bitwise Europe GmbH, incorporated under the laws of Germany, is the issuer of the Exchange Traded Products (“ETPs”) referenced in this Document under a base prospectus and the applicable final terms, as supplemented from time to time, approved by the German Federal Financial Supervisory Authority (BaFin). The approval of the prospectus by BaFin relates solely to the completeness, coherence and comprehensibility of the prospectus in accordance with the Prospectus Regulation and does not constitute an endorsement, recommendation or assessment of the merits of the products.
Past performance is not a reliable indicator of future results.
Capital at risk. Cryptoassets are highly volatile and involve a high degree of risk. The value of investments in cryptoassets and crypto-linked ETPs may fluctuate significantly, and investors may lose part or all of their invested capital. No capital protection or guaranteed compensation mechanism applies in respect of market losses.
Any investment decision should be made on the basis of the relevant base prospectus, the applicable final terms and the key information document, in particular the section entitled “Risk Warning” The base prospectus, final terms and additional risk information are available at: www.bitwiseinvestments.eu
Access to certain documents may require self-certification regarding your jurisdiction and investor status and may be subject to additional disclaimers and important information.
This document may contain forward-looking statements including statements regarding Bitwise Europe's belief or current expectations with regards to the performance of certain asset classes. Forward-looking statements are subject to certain risks, uncertainties and assumptions, and there can be no assurance that such statements will be accurate and actual results could differ materially. Therefore, you must not place undue reliance on forward-looking statements.
For more details and the full disclaimer visit: https://bitwiseinvestments.eu/disclaimer/
