The question
Do US and euro-area government-bond yields move as one, and if so, on what timescale? When the US Treasury 10-year yield moves on a given day, does the euro-area AAA sovereign 10-year yield move with it the same day, the next day, or not in any measurable way? The honest version is narrower than “does the US lead Europe.” At daily close-to-close sampling the two markets close about four hours apart, and that gap alone can manufacture a one-day lag out of nothing but shared global news. So the question we can actually answer is how strongly the two are coupled — and, just as importantly, which parts of an apparent lead we are and are not entitled to claim.
Data
Two first-party official sources, no scrapers or redistributors. The US leg is the U.S. Treasury Daily Par Yield Curve — the 10-year constant-maturity point, struck from roughly 3:30 PM ET New York Fed quotations. The euro leg is the European Central Bank’s euro-area AAA sovereign yield curve — the 10-year Svensson spot point. Both are daily, and each is the primary publisher of its own series, not a mirror or an aggregator. The overlap runs from September 2004, when the ECB series begins, to August 2026: about 5,400 aligned trading days after inner-joining the two holiday calendars. The euro series spends a long stretch of 2019–2022 below zero — down to about −0.82% — which the method has to handle rather than discard.
Method
The registered quantity is the lag-1 cross-correlation, ρ = corr(ΔY_US(t), ΔY_euro(t+1)) — how strongly today’s US yield change lines up with tomorrow’s euro change. To keep the engine’s log-based estimator well-defined through the negative-yield era, each yield Y is carried as a synthetic price exp(Y/100), a rescaling that leaves the correlation untouched. Everything was pre-registered and frozen before the analysis window opened: the statistic, the two null models, the materiality floor (|ρ| ≥ 0.10), the block boundaries, and every threshold. The window is split into three consecutive seven-year forward blocks — 2007–2013, 2014–2020, 2020–2026 — with a 2004–2006 calibration block set aside to check the settings and excluded from every verdict. A result has to clear two null models to count: an ordinary value-permutation, and a more conservative circular-block permutation that preserves volatility clustering (which shuffling would otherwise break, making the ordinary null too easy to beat). And it has to repeat in at least two of the three blocks.
Findings
Contemporaneous co-movement is the dominant relationship. The strongest link between the two markets is same-day. The cross-correlation curve peaks at zero lag, with a same-day correlation of roughly 0.44 to 0.57 across the three blocks. Whatever else is true, US and euro-area 10-year yields overwhelmingly move together on the day.
Lag-1 coupling is real and replicates — but it is the secondary piece. The registered one-day-lag correlation is +0.27, +0.24 and +0.20 across the three blocks: material (all above the 0.10 floor), well above both null models (which sit near 0.06), and replicated in all three. It is a genuine, recurring coupling. The steady decline across the blocks — 0.27, then 0.24, then 0.20 — is consistent with stronger global coupling during the 2007–2013 financial-crisis and euro-crisis years. But the lag-1 figure is much smaller than the same-day one, and — this is the load-bearing caveat — it cannot be separated from the mechanics of the close. The euro curve is struck about four hours before the US close, so any shared global move that lands in those four hours enters the US change today and the euro change tomorrow, producing a one-day lag from timing alone, with no transmission content. From daily close-to-close data, that geometric artifact and any genuine next-day transmission are not distinguishable. We therefore report a coupling measurement, not a claim that one market leads the other.
The coupling is not driven by a few extreme days. Removing the most volatile 5% of days by |ΔY_US| in each block — 87, 80 and 76 days — lowers the lag-1 correlation only modestly, to +0.22, +0.19 and +0.16, with every block still material and still clear of both nulls. A ~20% drop from removing 5% of days says those days carry more than their share, as crisis-period co-movement should; that the result does not collapse says the bulk of the coupling lives in ordinary days too.
The spread does not mean-revert faster than chance — a null. As a secondary, descriptive test we asked whether the US-minus-euro yield spread returns toward its own recent local level faster than a random walk. It does not: the measured reversion rate (0.293) sits below the surrogate’s 99th percentile (0.344). At the registered thresholds the spread’s day-to-day behavior is not distinguishable from a random walk — reported here as a clean null, not dressed up.
What this means
Two large government-bond markets on opposite sides of the Atlantic are strongly coupled — mostly on the same day, with a smaller one-day-lag component that is real and repeats, but that honest measurement cannot attribute to leadership rather than to the four-hour gap between their closes. The value of the exercise is exactly that discipline: a robustly replicated, outlier-resistant coupling, reported as coupling, with the parts we cannot identify named as such rather than sold as a lead.
Two caveats belong in plain sight. The US leg is a par constant-maturity yield and the euro leg is a Svensson spot yield, so their level difference carries a construction wedge and is not a clean sovereign basis — which is why the spread test above is framed as spread behavior, not as a sovereign-basis result. And the pre-2007 calibration years were unusually quiet: the spread threw only 15 reversion events there, fewer than a random walk would, a feature of that low-volatility regime. The 2007–2026 window, with 232 events, is what the settings were sized for.