这周大盘表面上看很平静，比特币在6.39万到6.68万美元之间晃了一圈，周五收在64,964美元，波动率不算大。但如果只看价格，会错过这周真正在发生的事——衍生品市场（杠杆合约交易的地盘）和ETF资金流层面，出现了一个从"多头逆势加仓"到"多头开始松手"的转折，这个转折比价格的涨跌更值得琢磨。
先说表象。周一（7月20日）比特币收在65,096美元，周二冲到全周最高66,264美元（盘中最高触及66,803美元），周三小幅回落到66,004美元，周四跌到65,156美元，周五收在64,964美元。总市值从周初的2.28万亿美元一路涨到周二的2.347万亿美元，随后连续三天缩水到周五的2.291万亿美元。表面上看，这是一次"涨了又跌回来"的行情，很多人会归因于"中东局势""油价"这类新闻标题。这些因素确实存在——周三美国空袭伊朗、胡塞武装威胁沙特航运的消息发酵，加上油价冲高、10年期美债收益率突破4.7%（为2025年1月以来最高水平），传统市场跟着降温，纳指、标普当天分别跌约1.0%和0.38%。但这只是触发因子，真正决定这周涨跌节奏的，是资金流和杠杆结构这两层机制。
第一层机制：ETF资金流的方向切换。比特币现货ETF（一种能在股票交易所买卖、跟踪比特币价格的基金产品）从周初开始连续多个交易日净流入，规模逐步放大：截至周一的净流入是2.268亿美元，截至周二是2.032亿美元，这是这轮反弹里机构买盘最强的两天，也正好对应周二价格冲到全周高点。但到了截至周三的数据，净流入骤降到6910万美元，而截至周四的数据直接转为净流出2.251亿美元，这是这周ETF资金流方向第一次转负。需要说明的是，14天累计净流入仍有5.47亿美元，说明大方向没有崩，只是短期这股买盘明显在减速甚至短暂反向。这意味着什么？ETF是这半年多头行情里最稳定的边际买盘来源，一旦这个买盘变弱，价格就失去了一个重要的"托底"力量，周三到周五的回调和这个转弱几乎同步出现，不是巧合，是因果。
第二层机制：杠杆结构的同步变化。资金费率（合约多头付给空头的费用，正值代表多头情绪更强、愿意付费维持仓位）这周从周二的最高点0.0089%一路降到周四的0.0003%，几乎降到零，未平仓合约量（OI，衡量场内杠杆仓位总规模的指标）也从周二的105,572张降到周四的102,196张。这个组合意味着什么？不是新空头大举进场做空，而是前期的多头主动在减仓——费率降到接近零、OI同步收缩，是"多头怕了先撤"的信号，而不是"空头猛攻"的信号。这个区分很重要：如果是空头进攻，费率应该转负甚至更负；这里费率只是从偏高回落到中性，说明这是一次谨慎的降杠杆，不是恐慌性的多空对决。周五费率和OI又双双小幅回升（费率0.0051%，OI回到104,369张），说明这个减仓过程本身也没有失控，市场在试探性地重新加仓。
反馈循环怎么运转的？这周能看到一个清晰的链条：中东局势升级和油价走高，压制了整体风险偏好→ETF买盘从周二的高点开始退潮（周三到周四由正转负）→场内杠杆多头看到价格滞涨、买盘变弱，主动降低仓位（费率和OI同步走低）→价格失去两个支撑（新增现货买盘+杠杆多头信心），周三到周五连续三天回落。这个循环里，恐惧贪婪指数（衡量市场情绪冷热的指标，0代表极度恐惧，100代表极度贪婪）的走势也印证了同样的节奏：周三（7月22日）是本周情绪最不恐慌的一天，读数33，随后连续两天下滑，周五回到28。这不是简单的"情绪跟着价格跑"，而是情绪指标和资金流、杠杆数据在同一时间窗口共振下滑，说明背后是同一套机制在起作用，不是三个独立的巧合。
但这个循环目前还没有演变成失控的下跌螺旋，原因有两个锚点没有松动。一是稳定币（锚定美元的加密货币，市值变化能反映场外候场资金的多少）总市值这周基本稳定在3,036亿到3,047亿美元之间，波动很小，没有出现大规模资金离场加密生态的迹象，场外的"弹药库"没有明显缩水，只是没有新增。二是比特币在总成交额里的占比从周初56.5%微升到周末56.7%，山寨币（除BTC、ETH外的品种）占比从33.7%小幅降到33.5%，说明这周资金总体是往比特币这个相对确定性更高的品种收缩，而不是集体逃离整个加密市场，这是一种防御性的收缩，而不是撤退。
往前看，如果这个"ETF转弱→杠杆多头减仓→价格承压"的循环继续运转下去，下一步大概率是ETF资金流能不能重新转正——截至周五，过去14个交易日的累计净流入已经从周四的约5.47亿美元迅速降至约4100万美元。接下来只要再出现一个中等规模的净流出日，这项指标就可能重新转负，那时候杠杆端的谨慎情绪会进一步强化，价格可能继续在这周区间的下沿附近寻找支撑。打破这个循环需要两个条件之一：要么ETF买盘重新放量流入，给市场一个"机构还在进场"的确认信号；要么外部风险（中东局势、油价）出现实质性降温，让风险偏好的压制解除，杠杆多头才有理由重新加仓。
下周需要盯紧的时间窗很明确：周三（7月29日）晚间的美联储利率决议（FOMC），和周四（7月30日）的美国个人收入与支出数据（PCE，一项重要的通胀观察指标）。这两个事件会直接决定10年期美债收益率是继续维持高位还是回落，而这又会传导到风险偏好和ETF资金流的方向。如果利率决议偏鹰派、PCE数据显示通胀顽固，前面提到的循环大概率延续，资金费率和OI可能继续维持低位震荡；如果决议偏温和、通胀数据降温，ETF资金流有机会重新转正，杠杆端的谨慎情绪也可能随之松动。这两个数据点，会是判断这周这套"降杠杆"机制是短暂休整还是趋势转折的第一个真正考验。
【副题】Bitcoin traded in a narrow range this week, but underneath the calm price action, ETF buying stalled and leveraged longs began quietly unwinding.
Bitcoin swung between $63,900 and $66,800 this week, closing Friday at $64,964 — not a particularly volatile week on the surface. But if you only look at price, you'll miss what was actually happening: on the derivatives side (the realm of leveraged contract trading) and in ETF flows, there was a turning point — from "longs adding against the trend" to "longs starting to let go" — that's worth paying more attention to than the up-and-down of price itself.
First, the surface picture. Bitcoin closed Monday (July 20) at $65,096, spiked to the week's high of $66,264 on Tuesday (intraday touching as high as $66,803), pulled back slightly to $66,004 on Wednesday, fell to $65,156 on Thursday, and closed Friday at $64,964. Total market cap rose from $2.28 trillion at the start of the week to $2.347 trillion on Tuesday, then shrank for three straight days to $2.291 trillion by Friday. On the surface, this looks like a "rally that gave itself back," and a lot of people will pin it on headlines like "Middle East tensions" or "oil prices." Those factors were indeed in play — on Wednesday, news broke of US airstrikes on Iran and Houthi threats against Saudi shipping, oil prices spiked, and the 10-year Treasury yield broke above 4.7% (the highest level since January 2025), which cooled off traditional markets too, with the Nasdaq and S&P falling about 1.0% and 0.38% respectively that day. But these were just the trigger. What actually set the pace of this week's rise and fall were two underlying mechanisms: capital flows and leverage structure.
**The first mechanism: a reversal in ETF flow direction.** Bitcoin spot ETFs (funds tradeable on stock exchanges that track bitcoin's price) saw net inflows for several consecutive trading days starting early in the week, and the size kept growing: net inflows were $226.8 million as of Monday and $203.2 million as of Tuesday — the two strongest days of institutional buying in this rally, which lines up exactly with Tuesday's price spike to the week's high. But by Wednesday's data, net inflows had plunged to $69.1 million, and by Thursday's data, the flow had flipped to a net outflow of $225.1 million — the first time ETF flows turned negative this week. It's worth noting that the 14-day cumulative net inflow still stood at $547 million, so the broader trend hasn't collapsed — it's just that this buying pressure clearly slowed, or briefly reversed, in the short term. What does this mean? ETFs have been the steadiest source of marginal buying in this bull run over the past half year or so, and once that buying weakens, price loses an important source of support. The pullback from Wednesday to Friday appeared almost simultaneously with this weakening — that's not a coincidence, it's causation.
**The second mechanism: a synchronized shift in leverage structure.** The funding rate (the fee longs pay shorts on perpetual contracts; a positive value means longs are more confident and willing to pay to hold their positions) fell all the way from its peak of 0.0089% on Tuesday to just 0.0003% by Thursday — essentially zero. Open interest (OI, a measure of the total size of leveraged positions in the market) also dropped from 105,572 contracts on Tuesday to 102,196 contracts on Thursday. What does this combination mean? It's not that new short-sellers piled in aggressively — it's that existing longs were actively cutting their positions. A funding rate falling toward zero alongside contracting OI is a signal of "longs getting scared and bailing first," not "shorts attacking hard." This distinction matters: if shorts were on the offensive, the funding rate should have turned negative, or even more negative. Here, the rate only fell from elevated to neutral, which suggests this was a cautious deleveraging, not a panicked long-short showdown. By Friday, both the funding rate and OI ticked back up slightly (rate at 0.0051%, OI back to 104,369 contracts), showing that the deleveraging process itself didn't spiral out of control — the market was tentatively re-adding positions.
How did the feedback loop actually run? This week shows a clear chain: escalating Middle East tensions and rising oil prices suppressed overall risk appetite → ETF buying receded from Tuesday's peak (turning from positive to negative between Wednesday and Thursday) → on-chain leveraged longs, seeing price stall and buying weaken, proactively cut their positions (funding rate and OI fell together) → price lost two pillars of support (fresh spot buying plus leveraged long confidence), leading to three straight days of decline from Wednesday to Friday. The Fear & Greed Index (a gauge of market sentiment, where 0 is extreme fear and 100 is extreme greed) confirms the same rhythm within this loop: Wednesday (July 22) was the least fearful day of the week at a reading of 33, followed by two straight days of decline, dropping back to 28 by Friday. This isn't simply "sentiment following price" — the sentiment index, capital flows, and leverage data all declined in resonance within the same window, which suggests one underlying mechanism was driving all three, not three separate coincidences.
That said, this loop hasn't yet turned into an out-of-control downward spiral, because two anchor points haven't budged. First, stablecoins (crypto assets pegged to the dollar, whose market cap changes can reflect how much sideline capital is waiting outside the market) held roughly steady this week, with total market cap staying between $303.6 billion and $304.7 billion — very little fluctuation, and no sign of large-scale capital leaving the crypto ecosystem. The sideline "ammunition" hasn't visibly shrunk, it just hasn't grown either. Second, bitcoin's share of total trading volume ticked up slightly from 56.5% at the start of the week to 56.7% by the weekend, while altcoins' (everything besides BTC and ETH) share edged down slightly from 33.7% to 33.5%. This shows that capital overall was consolidating into bitcoin — the relatively higher-certainty asset — rather than fleeing the crypto market as a whole. This is a defensive contraction, not a retreat.
Looking ahead, if this loop of "ETF weakening → leveraged longs cutting positions → price under pressure" keeps running, the next thing to watch is most likely whether ETF flows can turn positive again — as of Friday, the trailing 14-trading-day cumulative net inflow had already dropped quickly from roughly $547 million on Thursday to roughly $41 million. If even one more moderate net-outflow day shows up, this metric could flip negative again, which would further reinforce cautious sentiment on the leverage side, and price could continue searching for support near the lower end of this week's range. Breaking this loop requires one of two conditions: either ETF buying picks back up in volume, giving the market confirmation that "institutions are still coming in," or external risks (Middle East tensions, oil prices) cool off substantively enough to lift the pressure on risk appetite, giving leveraged longs a reason to add back positions.
The time windows to watch closely next week are clear: the Fed's interest rate decision (FOMC) on Wednesday evening (July 29), and the US Personal Income and Outlays data (PCE, an important inflation gauge) on Thursday (July 30). These two events will directly determine whether the 10-year Treasury yield stays elevated or pulls back, which in turn will feed through to risk appetite and the direction of ETF flows. If the rate decision leans hawkish and the PCE data shows inflation staying sticky, the loop described above will likely continue, with funding rates and OI probably staying stuck in a low range. If the decision leans dovish and inflation data cools, ETF flows have a chance to turn positive again, and cautious sentiment on the leverage side could loosen up too. These two data points will be the first real test of whether this week's deleveraging is just a brief pause or a genuine trend reversal.