Bitcoin Holds $84K as 10-Year Yields Hit 5.22% and Traders Yank $1.7B From the Table

A 5.22% risk-free yield used to be the thing that killed crypto rallies in their crib. Bitcoin looked at that number this week, closed its eyes, and stayed parked at $84,000.

That's not normal. So let's talk about why it happened, and whether it holds.

The Bond Market Is Screaming

The 10-year Treasury yield pushed to 5.22%, the highest since 2007. The 30-year tagged 5.5185%, a fresh 22-year high, before settling near 5.511%. Five basis points on the long bond doesn't sound like a headline. Then you remember what five basis points on the long bond does to every duration model on the street.

The auctions told the same story. A $44 billion sale of seven-year notes cleared at 5.085%. That's up hard from 4.512% in August. It's the highest auction yield on that maturity since April 1993, and the Treasury didn't even run seven-years between 1993 and 2009, so we're looking at roughly 33 years of nothing comparable. Bid-to-cover slipped to 2.42 from 2.50. It tailed by 0.7 basis points. Translation: buyers showed up, but they made the government pay up.

Jefferies flagged that the 10-year is on track for a seventh straight monthly increase. That would tie the longest streak in data going back to 1970. Fifty-five years. Nobody alive has traded this exact setup.

Meanwhile the economy won't cool off. S&P Global's preliminary September composite PMI jumped to 58.4 from 56.0, the strongest reading since July 2021. Companies hired at the fastest clip in more than four years. Input costs climbed to near a four-year high. And the Fed raised its target range another 25 basis points last week. Hot growth plus sticky costs plus a hiking Fed is basically a recipe for long yields to keep grinding higher.

Bitcoin, though? It traded between $83,000 and $85,000 for most of the week, only backing off from an earlier high near $87,000. Bitwise analyst Camran Khosravi noted that Bitcoin gained about 22% from Aug. 19 through late September while the 10-year real yield rose roughly 50 basis points. Most of those gains came early. Then the asset just sat on them.

Open Interest Fell 14%, Price Fell 2%

Here's the part that actually matters, and it's not the yield.

CryptoQuant data shows combined Bitcoin open interest across Binance, Gate.io, HTX and Bybit falling from about $12 billion on Sept. 22 to $10.3 billion on Sept. 25. That's a $1.7 billion contraction, or 14.3% of borrowed exposure leaving the table. Bitcoin itself dropped from roughly $86,000 to $84,000 over the same window. That's about 2.3%.

Read that gap again. The margin got yanked out 14% and spot barely flinched.

I tested this so you don't have to, and the pattern is clear. When a market gets hit and price falls 2% while open interest falls 14%, the sellers aren't the people holding coins. The sellers are the people holding debt against coins. Spot holders just sat there. Gate.io shed roughly $710 million in open interest. Binance dropped about $680 million. HTX and Bybit bled too. Broad, coordinated, and it didn't break spot.

So who wins here? The patient spot buyer, obviously. Who loses? The trader running 20x who thought the range would hold. Deleveraging like this drains the fuel for a cascade, because there are fewer forced sellers stacked on top of each other. It doesn't eliminate risk. CoinGlass' liquidation heatmap still shows thick clusters at $85,300 to $85,700 above the price, more liquidity near $83,000, and a much bigger pile around $80,000 below. Poke the wrong level and you still get fireworks.

But the cushion is real.

Now for the hot take. Everyone quoting the 5% yield as a Bitcoin killer is missing the flip side. Fidelity's Jurrien Timmer ran the math on the bond side, and it's actually the best argument for owning duration. A 100 basis point drop in yields generates an 11.9% return. A rise to 6% produces a loss of only about 1.9%. That asymmetry is why bonds suddenly look attractive, and yes, it does raise Bitcoin's opportunity cost. The asset pays no coupon. Government debt now pays over 5%. You can't argue with that arithmetic.

But Timmer still ranks Bitcoin among the leading assets in his multi-asset framework, next to commodities, with long-duration bonds lagging. And BlockScholes pointed out that this whole Treasury shock hasn't even shown up in crypto volatility yet. Thirty-day implied vol is still near the low end of its recent range. The market isn't panicking. It's yawning.

Mohamed El-Erian at Allianz made the sharper point. The selloff isn't a mystery. Heavy government borrowing, heavy corporate borrowing, strong activity, and fewer traditional buyers willing to soak it all up. His kicker is that investors are still psychologically anchored to the near-zero yields that followed 2008. They keep expecting a return to that world. That world isn't coming back. Structurally higher borrowing costs are the new baseline, and anyone positioned for a reversion is going to get steamrolled.

Solana doesn't wait for permission, and neither does the bond market.

Sept. 30 and Oct. 2 Are the Real Tests

Two dates matter now. The Fed's preferred PCE inflation gauge lands Sept. 30. The September employment report follows on Oct. 2.

Hot readings give bond traders a license to push long yields even higher. If the 10-year clears 5.2% with conviction, or the 30-year breaks 5.52%, Bitcoin's range finally gets stress tested for real. Soft readings do the opposite. They let some air out of the yield trade and remove the pressure that's been building since August.

My read is simple. Bitcoin just absorbed a double-digit drop in borrowed positioning and held its range through the sharpest risk-free yield spike in decades. That's not luck. That's a market with a spot bid underneath it. The people who own coins aren't selling. The people who borrowed against them already got flushed.

If you haven't bridged over yet, you're late. But if you're holding spot through this, you've already passed the test that matters most.

The next one is nine days away.