Dollar carry's longest run since 2008 reads as a crowding signal, not a return

Annualise the January figure and nothing has cooled: 1.3% by late in the month compounds to roughly 17% over twelve months, within a point of what the index actually delivered across all of last year ^[1]. Pace like that does not come out of interest differentials on their own. Cryptobriefing's own account attributes the size of the 2025 return partly to a weakening funding currency ^[12], and a falling dollar is a price that can reverse inside a week, not a coupon that accrues.

Then there is the reference class. The last time the streak ran this long was 2008 ^[2], which is also the first entry on the same article's list of episodes where this trade came apart, alongside the 2013 taper tantrum and the August 2024 yen unwind ^[8]. The record for persistence and the record for damage share a date.

Goldman Sachs widens the problem rather than narrowing it: it put G10 carry conditions at their most favourable since 2000, with yen and Swiss franc funding working as well as the dollar ^[6]. A book borrowing in several currencies and lending into eight looks diversified on a risk report while depending on one variable, realised volatility, staying where it is. The documented behaviour is that gains accumulate over months and disappear in days ^[8], which is a statement about liquidity at the exit rather than about the quality of the analysis at the entry.

The volatility comparison reaches back twenty-six years ^[3], so the sample of similar periods is thin, and the trigger named in the source is procedural rather than dramatic. A hawkish Federal Reserve surprise, or any move in US rates that narrows the yield gap, removes the reason to hold the position ^[10]. Interventions around the yen have already produced fluctuations without breaking the trend ^[9], the kind of near-miss usually filed as resilience. Meanwhile inflows into emerging-market assets are running at their fastest pace since 2019 ^[7]: money that arrived quickly has no record of leaving slowly.

Morgan Stanley and Bank of America both expect conditions to hold through 2026, absent unforeseen macroeconomic shifts ^[11]. That caveat is doing all the work. Carry rarely loses money because the macro view was wrong. It loses money because the view was right up until the day it was not, and by then everyone holding it wants the same exit.