FX regime notes — first half of 2026

The dominant feature of the first half of 2026 has been an unusual compression in realised volatility across G7 majors, even as central bank divergence has widened. EUR/CHF has spent the year in one of its tightest six-month ranges in a decade. EUR/USD realised vol has spent most of the year below its long-run median. Implied vol in the front end has compressed alongside realised, suggesting the market is genuinely positioned for the calm rather than disbelieving it.

This is a difficult regime for breakout strategies, which rely on directional follow-through that simply has not been present. It is a friendly regime for carry structures, short-vol structures, and mean-reversion strategies that can pick up the small range expansions that do occur. Our book has tilted accordingly, with the deliberate caveat that vol regimes do not unwind politely.

Selected EM pairs continue to offer a real carry premium that is not fully explained by realised volatility. The pairs we have small, sized positions in are those where the carry differential is large, the realised vol is low for the asset class, and the political and macro context does not suggest an imminent re-pricing.

We are conscious that EM crises are the textbook case of correlated tail events. Positions are sized accordingly — small relative to the rest of the book, with hard stops at levels that respect the typical gap sizes for the pair rather than the typical intraday range. We would rather miss a great year on a carry trade than carry the existential risk of a bad week.

The leading indicator we track most closely is the term structure of implied vol in EUR/USD and USD/JPY. When the front end starts pricing more than realised, and the term structure inverts, we treat that as the first credible signal of a regime shift and begin reducing carry exposure.

Other inputs feed in. The pace of speculative positioning unwinds. Cross-asset confirmation from rates and equity vol. The behaviour of correlations between pairs that have been moving independently during the calm. None of these is a trigger on its own; in combination they are typically enough to motivate a precautionary reduction in carry exposure.

In practice the first half of 2026 has meant fewer breakout entries, more patience around mean-reversion setups, and tighter intraday stops on directional positions to compensate for the lower range expansion. Position sizes on the carry book have been steady rather than increasing — when carry is paying well, the temptation to scale up is real and historically punished.

The single most useful operational discipline in a regime like this is not chasing the strategies that worked best in the previous regime. The cost of that mistake — staying long breakouts into compression, for instance — is much larger than the cost of the rotation we have done.