After the previous round of Bitcoin price declines, Coinbase's negative premium for Bitcoin significantly widened, which was viewed by many participants as a signal of increased selling pressure from U.S. investors. Jiang Zhuoer, the founder of the Liebit mining pool, publicly judged that there would likely be a net outflow of funds from the Bitcoin spot ETF the next day, while also reminding that ETF data has a time lag and using its trading is akin to "driving while looking in the rearview mirror." However, the data published on September 8 provided a completely different answer: the Bitcoin spot ETF recorded a slight net inflow of 398 BTC for that day, amounting to about 31.07 million U.S. dollars, while the Ethereum ETF saw a significant net outflow of 19,667 ETH on the same trading day, approximately 48.34 million U.S. dollars. While Bitcoin continued to attract incremental funds, there was a clear withdrawal of funds on the Ethereum side. Compared to September 7, the rhythm of the two ETFs highlighted a stark contrast: the Bitcoin ETF recorded a large net inflow of 8,937 BTC on September 7, amounting to approximately 698 million U.S. dollars, while the Ethereum ETF also saw a simultaneous net inflow of 15,939 ETH, about 39.18 million U.S. dollars, indicating a collective inflow of funds; by September 8, the Bitcoin ETF's cumulative net inflow over two days reached 9,335 BTC, whereas the Ethereum ETF shifted sharply from net inflow to net outflow, completing a drastic reversal. The differentiation in capital flow between Bitcoin and Ethereum sharply contrasted with Jiang Zhuoer's net outflow prediction based on Coinbase's negative premium, providing a clear example of the limitations of observing a single indicator for predicting ETF fund behavior.
Prediction Error: Coinbase Signal Did Not Translate into ETF Selling Pressure
Jiang Zhuoer's reasoning was not complicated: Coinbase’s premium is viewed as an important indicator of U.S. investor demand for Bitcoin. When the negative premium widens, it usually implies that the U.S. market is willing to sell at a discount to local prices. Following a round of declines in Bitcoin prices, the widening negative premium, in his view, is direct evidence of increased selling pressure stateside. Based on this signal, he inferred a continuation effect at the capital level—if there was increased selling pressure in the U.S. spot market, then it is highly likely that the Bitcoin spot ETF, reflecting the willingness of U.S. institutions and compliant fund allocation, would register a net outflow of funds the following day.
However, the ETF data released that day presented a different picture. According to the disclosed results, on September 8, the Bitcoin spot ETF not only did not experience the anticipated net redemption, but instead recorded a slight net inflow of 398 BTC, forming a continuous absorption pattern alongside the previous day's large net inflow of 8,937 BTC, directly negating the judgment that "selling pressure would rapidly materialize at the ETF level." Jiang Zhuoer himself pointed out that the fund flows in ETFs reflect changes from the previous trading day, and “driving while looking in the rearview mirror” inherently has a time lag, while the Coinbase premium is an immediate snapshot of the trading sentiment. This time, the real-time signal of a widened negative premium did not translate into a net outflow of the ETF the next day, highlighting the misalignment between sentiment indicators and capital outcome metrics in terms of dimensions and rhythms.
Changes in the Rhythm of Bitcoin ETF Absorption of Over 9,300 BTC in Two Days
From the data rhythm, the Bitcoin ETF showed a "rapid start followed by a slowdown" trend from September 7 to September 8. On September 7, the Bitcoin ETF recorded a net inflow of 8,937 BTC, which, when calculated for that day, amounted to approximately 698 million U.S. dollars, representing a typical large single-day capital absorption; immediately following on September 8, the net inflow narrowed to 398 BTC, about 31.07 million U.S. dollars, which, in absolute terms, represents only a slight continuation. However, the total net purchase of 9,335 BTC over the two days still constituted a substantial increment of capital within such a short time window, and since its launch, the Bitcoin spot ETF has been seen as an important price driver. Regardless of how the rhythm changes, this scale of continuous inflow signifies that mainstream funds were still increasing their exposure during this phase rather than reducing their holdings.
Even more noteworthy is that this capital absorption rhythm occurred against the backdrop of Bitcoin having recently undergone a decline and Coinbase's negative premium widening, both of which are typically regarded as signaling selling pressure; nevertheless, funds did not reflect net outflows at the ETF level. Considering that ETF data reflects capital changes from the previous trading day, this can be understood as follows: short-term trading sentiment leaned toward risk aversion or even reduction in the spot and on-market premiums, whereas the allocation side in ETFs still opted to progressively increase holdings during price declines as U.S. investors deepened their discounts. The result of absorbing over 9,300 BTC in two days, coupled with the significant reversal in fund direction of Ethereum ETFs within the same window, illustrates Bitcoin's relative resilience at the institutional allocation level and the structural divergence in the time dimension and behavioral motives between sentiment indicators and funds inflow data.
Ethereum ETF's One-Day Reversal: From Inflow to Nearly 20,000 Outflow
In contrast to the Bitcoin ETF maintaining net inflows for two consecutive days, the Ethereum ETF displayed a much more dramatic directional reversal within the same timeframe. On September 7, the Ethereum ETF recorded a net inflow of 15,939 ETH, with a capital scale of approximately 39.18 million U.S. dollars, and together with Bitcoin ETF's net inflow of 8,937 BTC that day, it formed an overall pattern tilted toward increased positions, showing that the allocation side was synchronously expanding exposure across both mainstream assets.
However, immediately on September 8, the direction of Ethereum ETF funds sharply turned downwards, registering a net outflow of 19,667 ETH, about 48.34 million U.S. dollars, completely erasing the previous day's net inflow and resulting in an even larger outflow. Comparing this with the Bitcoin ETF, which still recorded a slight net inflow of 398 BTC, it is clear to see that funds at the ETF level did not exhibit a comprehensive "risk aversion"; rather, a relative shift occurred from Ethereum to Bitcoin. This differentiation within a single day could reflect institutions reevaluating the long-term narratives and risk-return profiles for the two asset types, also providing a relatively clear observation starting point for discussions on whether sentiment structures and asset preference changes will persist.
Driving While Looking in the Rearview Mirror: How Lagging ETF Data Misleads Trading
Jiang Zhuoer compared making trades based on spot ETF data to "driving while looking in the rearview mirror," with the core emphasis on the time dimension: the daily published inflow and outflow data of ETFs reflects capital flow from the previous trading day, not orders and sentiment at the current moment. On September 7, with the Bitcoin ETF's net inflow of 8,937 BTC and the Ethereum ETF's net inflow of 15,939 ETH, these figures only enter the market's view the following day, representing the allocation results that institutions made the day before. When traders regard such pre-existing fund flows as "the basis for current market conditions," it is akin to relying solely on the road conditions visible in the rearview mirror to decide whether to speed up or hit the brakes; if today's capital behavior significantly deviates from yesterday's, it may lead to incorrect directional decisions near turning points.
In contrast, Coinbase’s premium is more closely aligned with real-time conditions; a widening negative premium is regarded as a signal of increased selling pressure from U.S. investors. It reflects the buying and selling forces present in the current order book, rather than a statistical result from the previous day's settlement. The advantage of base decisions on premiums lies in their speed of reaction, allowing for the capture of sudden emotional shifts and short-term fund defensive and offensive transitions. However, the downside is high noise, leading to potential misguidance by short-term fluctuations; Jiang Zhuoer's prediction of the next day's Bitcoin ETF net outflow did not materialize into consecutive net inflows is a typical example: real-time sentiment and substantial institutional allocations do not always synchronize in time and behavior. Nonetheless, being lagging does not imply that ETF data can be ignored. Since its launch, the Bitcoin spot ETF has been seen as an important force affecting prices; the cumulative net inflow of 9,335 BTC over two days corresponds to genuine large-scale capital behavior, making it suitable as a benchmark for medium-term capital direction and risk preference judgments. A more reasonable approach would be to treat ETF fund flows as the primary anchor for "size and direction," and to combine it with more sensitive real-time indicators like the Coinbase premium to calibrate rhythm and entry points, rather than relying solely on any single data source for decision-making.
The Rationale Behind Capital Differentiation: Cautiously Interpreting Indicators and Sentiment
Returning to this observation period, the Bitcoin ETF recorded a continuous net inflow of 9,335 BTC over two consecutive days from September 7-8, while the Ethereum ETF experienced a net inflow of 15,939 ETH on September 7 and suddenly switched to a net outflow of 19,667 ETH on September 8, illustrating a phase of preference differentiation between the two major assets. Jiang Zhuoer's prediction of a net outflow from the Bitcoin spot ETF the following day based on the signal of an expanded Coinbase negative premium was contradicted by actual data recording net inflows for two consecutive days, highlighting the limitations of capturing complex fund behaviors with a single high-frequency indicator, and affirming his assertion about the "time lag in ETF data." For trading and research, a more reasonable practice would be to regard the ETF fund flows within the specific timeframe of September 7-8 as a central reference for the size and direction of funds, and then employ real-time indicators like the Coinbase premium reflecting U.S. market sentiment, along with the trends of spot prices and structural positions for multi-dimensional cross-validation, thus accommodating both the time dimension and signal source differences when understanding market sentiment and asset preference disparities.
Join our community for discussion and strength improvement!
AiCoin Exclusive Hyperliquid Benefits: https://app.hyperliquid.xyz/join/AICOIN88
AiCoin Exclusive Aster Benefits: https://www.asterdex.com/zh-CN/referral/9C50e2
On-chain Telegram Community: https://t.me/AiCoinWhaleData
On-chain Community: https://www.aicoin.com/link/chat?cid=N6OVMor5g
AiCoin On-chain Twitter: https://x.com/aicoinwhaledata
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
