Written by: Clow

In January 2024, Grayscale's GBTC launched its ETF conversion, leading to a net outflow of over 6 billion dollars in two months. Panic spread, but Bitcoin rose from 42,000 to 73,000.

The reason is simple: the outflowing money was absorbed by the real buying power of the new ETFs.

More than two years later, the script flips.

BlackRock, Fidelity, and Bitwise are lined up to obtain approval for spot Bitcoin ETFs, and Wall Street has declared a "institutional bull market".

In the first week of August 2026, the net inflow into spot Bitcoin ETFs over five trading days was approximately 930 million dollars, marking the best performance in months. On-chain whale addresses increased their holdings by about 1.2 billion dollars during the same period.

With a 2.1 billion dollar book buy order dropped, Bitcoin couldn't hold even 65,000, dipping below 63,000 at one point.

The money has come in, but it is not here to "buy" Bitcoin.

If nearly 1 billion dollars in returning funds can't move the price, the narrative that "institutionalization means always rising" has developed cracks.

Those who bought Bitcoin are actually shorting it

Looking at the holdings structure of IBIT, one discovers a disquieting fact: a substantial proportion of the funds are arbitrage chips from hedge funds.

The operation is simple. Buy IBIT spot with one hand, short an equivalent amount of futures on the CME with the other. As long as a futures premium exists, locking both ends results in a risk-free annual return of 5% to 15%. In extreme situations, the interest spread can even exceed 40%.

This is "spot-futures arbitrage". The funds flow into the ETF, Bitcoin is indeed purchased, but equal-sized short positions are generated simultaneously in the derivatives market. In and out, the net exposure is zero.

Quantitative analysis shows that even when engaging in cross-market arbitrage with IBIT options' implied forward prices against CME futures, there remains a 2.58% annualized interest spread after excluding fees. As long as this window remains open, arbitrage funds will continuously pour in.

The inflow numbers for the ETF look good on paper. But for the price, this money is as good as not having come at all.

Even more hidden is the behavior of authorized participants. These large investment banks and market makers primarily work to capture the slight price discrepancies between ETF prices and net asset values.

When funds flow in, they do indeed purchase Bitcoin in the spot market, but every purchase is accompanied by dynamic hedging. Once prices fluctuate, risk control models immediately act in the opposite direction.

The mechanism originally used to smooth price differences turns into an amplifier when liquidity is thin. A few percentage points of normal corrections can be magnified into a small crash by collective rebalancing.

The risks during weekends and after-hours trading are especially high. AP's risk control models do not distinguish between rational adjustments and panic selling; they execute mechanically once threshold values are triggered.

Miners are offloading, MicroStrategy is tightrope walking

The ETF struggles to maintain the decorum of inflows, but on the other side, the selling side features a cast that is despairingly luxurious.

Mining companies are at the forefront. After the halving in April 2024, block rewards were halved, yet the overall network hash rate continued to rise. Hash prices have been hammered down to historical 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 dollars.

At a market price of 65,000, most miners incur a loss of one Bitcoin for each one they mine.

It presents an absurd scene: the world's largest hash rate network produces goods sold below cost, using record levels of electricity.

They lack the capital to hold on. Bills for electricity need to be paid, debt interest must be serviced, and capital expenditures for transitioning to AI data centers must be arranged. On-chain data indicates that every time BTC rebounds above 60,000, miners' daily transfers to exchanges soar into the thousands of coins.

It's not that they want to sell; not selling means death.

Deeper pressure comes from strategic transformation. Leading mining companies like Core Scientific and TeraWulf are turning their mining sites into AI data centers, often signing billion-dollar hash rate contracts.

Transforming cooling and power infrastructure requires a vast amount of capital; Bitcoin reserves have become a cash machine. This kind of selling does not look at market conditions, only at project deadlines.

Now, consider MicroStrategy. This company holds over 840,000 Bitcoins, accounting for about 4% of the total supply. It sounds like the most steadfast bull, but it has over 8 billion dollars in convertible bonds on its balance sheet, maturing between 2028 and 2032.

The financing model is quite clever: taking advantage of the high premium of its stock relative to Bitcoin's net asset value, it continuously issues zero-coupon convertible bonds to raise cash for buying more coins. The premise for the flywheel to keep turning is that Bitcoin prices must keep rising while the premium remains high.

Once the sideways movement continues, and the premium shrinks, the financing channel could be cut off. The probability of being forced to sell coins to repay debts as maturity dates coincide with cash flow exhaustion is not zero.

Market makers are already pricing in this tail risk.

The on-chain competition for chips is equally brutal. The average cost for short-term holders is at 69,000, putting them in an overall losing position. Every time the price rebounds to this level, the liquidation orders pour in like floodgates opening.

Long-term holders, while not panicking, steadily sell about 12,800 Bitcoins each week to secure profits.

Above are the eager liquidation orders; below are the collected profit takers. With two ceilings stacked on top of each other, the price cannot rise.

A sponge that absorbs all momentum

The Coinbase premium index has been negatively impacted for nearly 80 consecutive days, setting a record for the longest duration.

This indicator measures the active buying of domestic institutions and high-net-worth clients in the U.S. Nearly three months of negativity implies that although ETFs are flowing in, the actual spot demand in the U.S. has fallen to cyclical lows.

It's not just the U.S. The world's largest offshore liquidity pool, Binance, continues to see a downward trend in the spot accumulated transaction volume differential, while the high-frequency deposit and withdrawal activities from large accounts suggest that selling intent outweighs building positions.

The macro environment is also working against it. U.S. inflation has proven to be stickier than expected, with market expectations for the Federal Reserve shifting from "when will rates be cut" to "how long will high rates last".

High risk-free yields significantly reduce the attractiveness of Bitcoin, which does not generate interest, in institutional portfolios. JPMorgan estimates the global recession probability for 2026 at 35%, as risk-off sentiment drives hedge funds to reduce exposure to crypto.

The Gamma exposure of options market makers controls the bounds of price volatility. In the range of 62,000 to 65,000 dollars, market makers hold a large number of positive Gamma positions. Every upward price probe results in mechanical selling of spot to maintain Delta neutrality.

This high-frequency hedging acts like a huge sponge, absorbing all the already thin bullish momentum in the market.

More dangerously below. In the range of 57,000 to 58,000 dollars, a considerable amount of negative Gamma exposure is piled up. Once the price dips into this area, the hedging direction for market makers reverses, turning into short chasing, which may trigger accelerated declines.

25-Delta options skew continues to tilt bearishly. In simpler terms: even those holding spot are spending a lot on insurance against a crash.

The spot market has already surrendered pricing power. Prices are not determined by the negotiation between buyers and sellers but are pinned down by the hedging algorithms of derivative market makers.

The nearly 1 billion dollars in ETF inflow is a perfect mask.

Behind the mask, arbitragers use financial engineering to annihilate bullish momentum, miners offload at highs, and MicroStrategy's debt maturity wall looms in 2028.

The 69,000 line of short-term holder costs acts like a wall blocking above, while the market makers' Gamma exposure pins the price within the range.

The most likely outcome for the market is neither a breakout nor a crash but a prolonged exhaustion. Buyers and sellers are evenly matched, and neither side has the incremental ammunition to break through the other's defenses.

Unless there is a substantive turnaround in the macro situation, a sudden drop in inflation triggers a reset in rate cut expectations, or spot prices forcefully breach the 69,000 cost line flipping losses to gains. Until then, the range of 60,000 to 70,000 is Bitcoin's cage.

Bitcoin does not lack inflows; what it lacks is genuine directional betting money.