ETF Outflows $4.5 Billion: Will BTC Drop Another 30% in the Next 3 Months?
- Core View: The Bitcoin market exhibits a divergence of "superficial prosperity, internal hollowness." The number of active on-chain addresses has been contracting for six consecutive months, hitting a five-year low, while trading volume remains stable. This indicates a narrowing breadth of market participation, with activity increasingly concentrated among institutions and off-chain products, providing a crucial framework for judging future trends.
- Key Elements:
- Sustained Weakness in On-Chain Activity: Glassnode data shows the 8-day moving average of active Bitcoin addresses has fallen by approximately 31% from August 2025 to February 2026. CryptoQuant has also flagged low network activity for six consecutive months.
- Divergence Between Trading Volume and Active Addresses: The average daily number of transactions remains stable (around 440,000), but the number of active addresses has decreased, indicating that fewer entities (such as exchanges, large holders) are responsible for the same level of on-chain activity.
- Low Fees Reflect Weak Demand: The network's average transaction fee is at a low level (around $0.24), with block space demand shrinking, consistent with the backdrop of declining participation breadth.
- Macro Environment and ETFs Alter Trading Patterns: Unstable market risk appetite and persistent net outflows from US Bitcoin ETFs have led to more exposure adjustments being completed off-chain, shifting activity away from self-custody wallets.
- Three Possible Future Scenarios: These include continued apathy (sustained low on-chain participation), liquidity thaw (macro improvements driving address recovery), and structural substitution (price increases but persistently low breadth, indicating an evolution in Bitcoin's role).
Trading volume hasn't collapsed, but active addresses have been shrinking for six consecutive months, hitting a five-year low. This divergence of "superficial prosperity with internal hollowness" is a negative signal for the structural health of a bull market.
The article cross-validates using three sets of data from Glassnode, Santiment, and CryptoQuant, proposing three future scenarios, providing a suitable reference framework for judging BTC's current trajectory.
The full text is as follows:
Bitcoin's network activity has been weakening for six consecutive months, but this trend is not reflected in the core metrics that many traders first look at.
The clearer signal is not trading volume—which has remained largely stable—but the breadth of participation. Even as the network continues to process a similar number of transactions, the number of active on-chain addresses has been steadily declining.
In a market where price discovery increasingly occurs on ETFs and derivatives, this split is crucial. It means: Bitcoin's on-chain footprint is narrowing, while market exposure remains active elsewhere.
As the bear market persists, this trend has become increasingly difficult to ignore.
Glassnode data shows that in mid-August 2025, the 8-day moving average of Bitcoin active addresses was approximately 778,680. As of February 23, this number has dropped to about 535,942, a decline of roughly 31%.
CryptoQuant has also flagged low network activity for six consecutive months, describing the current phase as a period of sustained weakness in on-chain participation.

Bitcoin Active Addresses Momentum,Source: CryptoQuant
The last time the market exhibited a similar pattern was in 2024—Bitcoin subsequently experienced a correction of approximately 30%.
This doesn't guarantee a repeat now, but it reinforces a historical pattern: prolonged network weakness often coincides with phases of weakening market confidence.
Breadth Declines, But Throughput Hasn't Collapsed
Bitcoin's transaction count has not declined in sync with the number of active addresses.
In mid-August 2025, the average daily transaction count was about 444,000. Blockchain.com data shows the 30-day average is currently around 439,000.
Intraday data still fluctuates, ranging from about 289,000 to 702,000 transactions, but the overall throughput trend has not collapsed.
This divergence is key to understanding the current situation.
If transaction volume remains stable while active addresses decrease, it indicates that fewer entities are responsible for the same amount of on-chain activity.
This situation has multiple causes, none of which require a retail influx. Exchanges and custodians can batch withdrawals; large holders can consolidate transfers; institutional fund flows can be handled through fewer wallets; operational activities can also cause temporary spikes in transaction counts without representing a genuine return of users.
The result is: the chain still appears busy, but the underlying number of participants is shrinking.
This is why the decline in breadth is more telling than raw throughput. Stable transaction counts may mask a market where activity is increasingly concentrated among repeat traders, large institutions, and operational fund flows.
In this landscape, Bitcoin's chain still functions normally, but the breadth of user participation it represents is becoming less genuine.
Blockchain analytics firm Santiment provides a more direct description from a longer time perspective.
The firm states that since February 2021, the number of unique addresses initiating Bitcoin transactions has decreased by 42%, and the number of new addresses created has decreased by 47%.

Santiment does not characterize this as evidence that crypto is dead or that a multi-year bear market is locked in, but it does describe a bearish divergence that persisted through 2025—market capitalization rose while Bitcoin's utility metrics weakened.
This tension is now evident in the six-month trend. Price and market narratives can continue, but the chain itself is becoming quieter.
Low Fees Point to Shrinking Block Space Demand
Fee data further confirms that Bitcoin's Layer 1 is in a state of weak demand.
Data from mempool.space shows the network's recent average transaction fee is approximately $0.24, equivalent to about 1.8 sats/vB.
For a network that experienced sustained block space competition during peak periods of past cycles, this is a low level. At the current transaction pace, this fee level implies the network's daily fee revenue is less than $100,000.
In contrast, the block subsidy is still approximately 450 BTC per day, with fee revenue constituting a minimal portion.

Bitcoin Average Block Fees,Source: Mempool.space
This is not an immediate security issue, nor does it mean Bitcoin's security model is under near-term pressure.
This is because the block subsidy still dominates miner revenue. But it does point to a long-term reality that Bitcoin has not yet been forced to confront at this stage of the cycle.
The topic of transitioning to a fee-supported security budget resurfaces every cycle, but in the current environment, this transition is not being tested—because fee demand itself is weak.
Practically speaking, the quiet fee market is postponing this discussion further.
The chain is not under pressure from sustained congestion, and users are not fiercely bidding to get their transactions included. This could change rapidly during volatility events, speculative waves, or new demand shocks, but it hasn't happened yet.
For now, block space is clearly in a state of low usage compared to past bull market phases, aligning with the broader context of declining participation breadth.

Bitcoin's Empty Mempool,Source: Mononaut
CryptoQuant's assessment also aligns with this fee environment—low network activity is typically associated with declining market interest in the asset and periods of widespread losses.
When interest wanes, new participants decrease, self-initiated transfers decline, and fee pressure subsides.
Bitcoin can still be actively traded as a financial asset, but the chain itself no longer reflects broad user participation.
Macro Environment and ETF Flows Are Changing How Bitcoin Trades
The macro backdrop helps explain why this trend persists.
Bitcoin is increasingly behaving like a macro-sensitive, high-beta asset, particularly standing out during risk-off periods.
Over the past year, US inflation has cooled, with the January 2026 CPI year-over-year growth rate at 2.4%; the Federal Reserve's target rate range was cited as 3.50% to 3.75% at the end of January.
In a simpler market environment, cooling inflation might support a clearer risk asset rebound.
However, market attention is focused on multiple volatility catalysts—including uncertainty around tariff policies. This factor has driven sharp fluctuations in interest rates and the US dollar, keeping overall risk appetite persistently unstable.
In such an environment, both retail and institutional investors tend to reduce their operational frequency. Retail participation declines, trader turnover decreases. Institutions can maintain exposure but prefer adjusting positions through products that don't require on-chain movement of coins.
This is precisely why spot Bitcoin ETFs have become the protagonist of a key narrative.
Coinperps data shows that US Bitcoin ETFs have seen net outflows for multiple consecutive weeks, with cumulative outflows of approximately $3.8 billion over the past five weeks and about $4.5 billion year-to-date.

2026 US Bitcoin ETF Daily Fund Flows, Source: Coinperps
This shifts activity from self-custody wallets to brokerage accounts.
It also explains how the market can remain active while the chain grows quieter. Exposure is still changing hands, but more of this turnover is happening off-chain.
This is a significant shift in Bitcoin's role. It is increasingly resembling a financial product wrapped in an institutional shell, while Layer 1 is used more selectively for settlement, storage, and periodic transfers.
Meanwhile, the daily trading energy within the crypto space is increasingly concentrated elsewhere, particularly in stablecoins.
Coin Metrics identifies stablecoins as a core driver of on-chain activity, with the total stablecoin supply nearing $300 billion and transaction volumes continuing to climb.
If stablecoin rails on other chains absorb more daily settlement demand, Bitcoin's Layer 1 naturally becomes more functionally singular.
This does not inherently weaken Bitcoin's investment thesis, but it does change its form.
Three Scenarios for the Next Three to Six Months
The current six-month decline in network breadth outlines three possible paths for Bitcoin's future trajectory.
The first is a continuation of apathy, which looks like the baseline scenario in a risk-averse market environment.
In this scenario, active addresses remain low (in the 450k to 600k range), transaction counts remain volatile but don't collapse, fees stay low, and ETF flows remain flat or slightly negative.
Here, Bitcoin could still experience sharp volatility due to macro headlines, but on-chain participation does not confirm a broad recovery. The asset trades more like a macro instrument than a network entering a new expansion phase.
The second is a liquidity thaw, the more optimistic path.
If inflation continues to cool and easing expectations stabilize risk appetite, ETF flows could shift from net outflows to sustained net inflows. In this environment, growth in active addresses would become a key confirmation signal.
A rebound to 650k to 800k active addresses would signify that participation breadth is recovering, not just price momentum returning. This looks more like a classic cyclical recovery—price increases supported by growth in on-chain user participation.
The third is a structural substitution scenario, perhaps the most noteworthy.
In this scenario, Bitcoin's price rises, but on-chain breadth remains persistently weak. ETFs, derivatives, and custodial settlements continue to dominate, while stablecoins absorb more trading demand elsewhere in the crypto space.
Here, Bitcoin increasingly resembles a digital macro asset and settlement layer, rather than a chain with broad daily retail activity.
This scenario would mark an evolution in Bitcoin's role, reflecting the deep changes it has undergone compared to many years ago.


