Bitcoin's daily price barely tracks Treasury yields even at a 5.3% series high

The rise since spot ETFs opened was 127 basis points in the 10-year nominal yield and 113 in the real yield ^[6]. Working back from Oct. 5, that puts the starting points at about 4.04% and 1.82% on Jan. 10, 2024 ^[19][20]. The gap between the two, roughly what bondholders are paid for expected inflation, widened only from 2.22 to 2.36 percentage points ^[21]. About 89% of the climb was real ^[22]. For an asset with no coupon, the real yield is the relevant benchmark, or rather the one that ought to bite hardest, and it moved more than a full point against bitcoin over a stretch in which the coin rose anyway ^[4].

There are three ways to square that. One is that yields weigh on bitcoin with a lag of months, too slowly to register in matched daily changes. CryptoSlate found the daily link weak across 2,435 days since 2017 and again across the 682 days after the ETFs began trading ^[7][8]. A second is that the positive post-ETF monthly readings reflect something pushing bonds and bitcoin in the same direction ^[10]. The third is noise. CryptoSlate's own caution is that 32 observations do not establish a lasting change ^[11].

I think the third reading fits the evidence best. It leaves the 5.27% yield as a carrying cost to set against the size of a bitcoin position, with little to say about the timing of a sale ^[2]. The case against that view is the full-sample monthly real-yield correlation, the most negative monthly figure CryptoSlate reported, and it comes with real yields at the top of their series ^[9][5]. If bitcoin keeps sliding while real yields hold near 3%, and the post-ETF monthly correlations turn negative as months are added, the yield would be working as a price signal after all.

The comparison pairs Coinbase bitcoin prices with FRED yield data and the ICE dollar index as distributed by Yahoo Finance, matched by date with missing days left out ^[12]. Closing times differ across the series, and some returns span weekends or holidays ^[13]. The work does not isolate any effect of the ETFs ^[14]. S&P Global's research from before the ETFs found that crypto's relationship with interest rates varied over time ^[15].