Written by: Clow
In January 2024, Grayscale's GBTC launched its ETF conversion, and outflows exceeded $6 billion within two months. Panic spread, but Bitcoin rose from $42,000 to $73,000.
The reason is simple: the outflowing money was absorbed by real buying from new ETFs.
More than two years later, the script has flipped.
BlackRock, Fidelity, and Bitwise have all received approvals for spot Bitcoin ETFs, and Wall Street has shouted the slogan of an "institutional bull market."
In the first week of August 2026, spot Bitcoin ETFs saw net inflows of approximately $930 million over five trading days, the best performance in months. On-chain whale addresses increased their holdings by about $1.2 billion during the same period.
With $2.1 billion in paper buying pressure, Bitcoin couldn't even hold $65,000 and briefly fell below $63,000.
Money came in, but it wasn't there to "buy" Bitcoin.
If nearly $1 billion in capital inflows can't push the price up, the narrative that "institutionalization equals perpetual growth" has a crack.
Those who bought Bitcoin are actually shorting it
Opening IBIT's holdings structure reveals a disturbing fact: a significant proportion of the funds are arbitrage chips from hedge funds.
The operation is simple. On the left hand, buy IBIT spot; on the right hand, short an equivalent amount of futures on CME. As long as the futures premium exists, locking both sides yields an annualized risk-free return of 5% to 15%. In extreme market conditions, the spread has even surged past 40%.
This is "cash-and-carry arbitrage." Funds flow into the ETF, and Bitcoin is indeed bought, but in the derivatives market, an equal amount of short positions is generated simultaneously. In and out, net exposure is zero.
Quantitative analysis shows that even with cross-market arbitrage between IBIT options' implied forward prices and CME futures, after fees, there is still an annualized spread of 2.58%. As long as this window remains open, arbitrage funds will continue to pour in.
The paper inflow numbers for ETFs look good. But for the price, it's as if this money never came.
More subtle is the behavior of Authorized Participants (APs). These large investment banks and market makers have the core job of capturing the tiny spread between the ETF price and its net asset value.
When funds flow in, they do buy Bitcoin in the spot market, but every purchase is accompanied by dynamic hedging. Once the price moves abnormally, risk models immediately reverse operations.
The mechanism originally used to smooth price discrepancies becomes an amplifier when liquidity is thin. A normal pullback of a few percentage points can be amplified into a mini-crash by collective position adjustments.
Risk is especially high during weekends and after-hours sessions. APs' risk models do not distinguish between rational adjustments and panic stampedes; they execute mechanically once thresholds are triggered.
Miners are dumping, MicroStrategy is walking a tightrope
ETFs are barely maintaining the dignity of capital inflows, but on the other side, the seller lineup is so strong it's despairing.
Miners bear the brunt. After the halving in April 2024, block rewards were halved, but network hash rate continued to rise. Hash price has been hammered to historic lows.
CoinShares data shows that the weighted average cash cost for listed mining companies to mine one Bitcoin has soared to between $76,000 and $80,000.
At a market price of $65,000, most miners lose money on every Bitcoin they mine.
A bizarre scene: the world's largest hash power network uses record electricity to produce a product sold below cost.
They don't have the capital to hold. Electricity bills must be paid, debt interest must be repaid, and capital expenditures for transitioning to AI data centers must be raised. On-chain data shows that whenever BTC rebounds above $60,000, miners' daily transfers to exchanges surge to tens of thousands of coins.
It's not that they want to sell; it's that if they don't, they die.
Deeper pressure comes from strategic transformation. Top miners like Core Scientific and TeraWulf are converting mining sites into AI data centers, signing computing power contracts worth billions of dollars.
Transforming cooling and power infrastructure requires massive capital, and Bitcoin reserves become an ATM. This kind of selling doesn't look at market conditions, only at project timelines.
Now look at MicroStrategy. This company holds over 840,000 Bitcoins, about 4% of the total network supply. It sounds like the most steadfast bull, but its balance sheet carries over $8 billion in convertible bonds, concentrated maturing between 2028 and 2032.
The financing model is clever: leveraging the high premium of its stock over Bitcoin's net asset value, it continuously issues zero-coupon convertible bonds to raise cash and buy more coins. The flywheel only spins if Bitcoin prices keep rising and the premium stays high.
Once sideways movement persists, the premium shrinks, and the financing channel is cut off. If maturity coincides with liquidity drought, the probability of being forced to sell coins to repay debt is not zero.
Market makers are already pricing in this tail risk.
The game of on-chain chips is equally brutal. The average cost for short-term holders is $69,000, and they are in an overall floating loss. Every time the price rebounds to this level, the trapped sellers release like a floodgate.
Long-term holders are not panicking, but they consistently sell about 12,800 Bitcoins per week to lock in profits.
Above are trapped sellers eager to escape, below are profit-takers calmly harvesting. Two ceilings stacked together, and the price can't lift its head.
A sponge that sucks up all momentum
The Coinbase premium index has been negative for nearly 80 consecutive days, setting a record for the longest streak.
This indicator measures the active buying pressure from US institutional and high-net-worth clients. Being negative for nearly three months means that despite ETF inflows, real US spot demand has fallen to cyclical lows.
It's not just the US. On Binance, the world's largest offshore liquidity pool, the cumulative spot volume differential is also declining, and high-frequency deposits and withdrawals by large players suggest a stronger intention to distribute than to accumulate.
The macro environment is also working against it. US inflation stickiness is far exceeding expectations, and market expectations for the Fed have shifted from "when will they cut rates" to "how long will high rates last."
With high risk-free yields, Bitcoin, which generates no interest, becomes much less attractive in institutional allocations. JPMorgan puts the probability of a global recession in 2026 at 35%, and risk aversion is pushing hedge funds to reduce crypto exposure.
Options market makers' Gamma exposure controls the price's volatility boundaries. In the $62,000 to $65,000 range, market makers hold large positive Gamma positions. Every time the price tests upward, they mechanically sell spot to maintain Delta neutrality.
This high-frequency hedging acts like a giant sponge, absorbing all the already thin bullish momentum in the market.
More dangerous is below. In the $57,000 to $58,000 range, there is a large accumulation of negative Gamma exposure. Once the price falls into this area, market makers' hedging direction reverses, turning into chasing shorts, which could trigger an accelerated decline.
The 25-Delta option skew continues to tilt toward puts. In plain English: even those holding spot are paying big money for insurance against a crash.
The spot market has already surrendered pricing power. The price is not determined by the game between buyers and sellers, but is pinned down by derivatives market makers' hedging algorithms.
Nearly $1 billion in ETF inflows is a perfect mask.
Behind the mask, cash-and-carry arbitrageurs use financial engineering to dissipate bullish momentum, miners dump at every high, and MicroStrategy's debt maturity wall looms in 2028.
The $69,000 cost line for short-term holders is like a wall blocking above, and market makers' Gamma exposure pins the price in a range.
The market's most likely outcome is not a breakout, nor a crash, but a prolonged grind. Buyers and sellers are evenly matched, and neither has incremental ammunition to break through the other's defenses.
Unless there is a substantial macro turnaround, with inflation unexpectedly falling and triggering a reset of rate cut expectations, or spot prices forcibly break through the $69,000 cost line to turn floating losses into gains. Until then, the $60,000 to $70,000 range is Bitcoin's cage.
Bitcoin doesn't lack inflows; it lacks money that truly bets on direction.
