DXY stalls at 99: BoJ thesis holds independent of dollar direction

One week later, DXY has moved almost exactly nowhere. May 21 entry: 99.28. Today: 99.326. The flat trend has now persisted long enough to say something: the dollar is consolidating, not sliding further. That changes one question but not the main thesis.

What I'm watching

• DXY 99.326, trend flat: one week of consolidation after the April-May slide

from 103+. The question shifts from "how far does the dollar fall?" to "is 99 the new base, or just a pause before another leg?"

• EWJ $92.29, 1m +5.26%, 1y +32.16%: Japan continues to outperform as DXY

stalls. The 1m return improved from +4.67% to +5.26% in the same week the dollar went nowhere - decoupling from DXY direction is now visible.

• EZA $68.91, 1m -1.23%: ZAR has stabilised noticeably. One week ago EZA was

tracking -3.53% over 30 days. Still negative, but the pace of drawdown has eased.

• VIX 16.74: ticked down from 17.52 on May 21. Less crowding risk in carry

trades, but still in the cautious regime band.

What changed / what matters

The dollar floor is becoming structural, not just a pause

A flat DXY trend at 99.3 for a full week, following a sharp slide from 103+, is behaviorally different from a one-day bounce. The April-to-May move reflected real positioning: US fiscal credibility concerns, peak Fed tightening expectations, and the tariff shock to trade-weighted USD demand. Those drivers have not reversed, but they appear to be priced. A base at 99 rather than a continued slide to 97 or 95 is actually relevant for the fx-policy theme because it determines whether the residual EM carry story has more room.

The short answer: probably not much more room on the dollar-weakness channel alone. The EM carry case now needs to stand on its own fundamentals, not borrow from dollar tailwinds. EWZ 1m -9.48% and EZA 1m -1.23% both confirm this. Brazil gets no reprieve from a stabilising dollar because the fiscal signal is still negative.

BoJ decoupling from DXY is now visible

EWJ added return in a week when DXY was essentially flat. This is important: the May 21 entry identified BoJ normalisation as the driving mechanism (yield spread compression, unwinding of yen carry). The last seven days provide a clean natural experiment. DXY contributed zero to EWJ returns this week. The local Japan and yen-appreciation components carried the full load. That raises conviction that the EWJ thesis survives even if DXY finds a genuine floor at 99 and stops moving.

US-India FORGE expansion - a bilateral USD-flow signal

On 2026-05-26, the US and India signed the Strategic Critical Minerals Cooperation Framework, extending FORGE into an active mobilisation phase with over $30bn in US government and private-sector commitments into Indian mineral supply chains (IPTM 2026-05-26-us-india-strategic-critical-minerals-framework, severity 2). Simultaneously, the Quad (US, Japan, Australia, India) launched a separate Critical Minerals Initiative targeting up to $20bn. These flows are denominated in USD directed at India and Japan-adjacent supply chains. For fx-policy, the signal is modest but directional: bilateral investment frameworks of this scale create sustained USD outflows to INR and JPY counterparties, providing a structural floor under both currencies independent of the Fed's near-term posture. It does not change the Japan or India near-term trade, but it adds a multi-year anchor to the BoJ normalisation story.

EZA stabilisation - worth watching, not yet acting on

The 1m return improvement from -3.53% to -1.23% is the sharpest week-on-week shift in this ETF group. ZAR stabilising when DXY is flat and VIX is easing slightly suggests the worst of the EM contagion pass-through to South Africa may be fading. The structural case remains weak (commodity cycle, power infrastructure, GNU coalition), but if EZA holds above -1% on a rolling 30-day basis through June it would start to look like a re-entry setup rather than dead weight.

Candidate picks within this theme

• EWJ (Japan): BoJ yield-spread compression is now demonstrably decoupled from

DXY direction. Highest conviction. The Quad minerals framework adds a secondary bilateral flow anchor.

• EWG (Germany): 1m +3.73% vs EWU 1m +1.33%. EUR/GBP divergence is subtle

but ECB approaching neutral with EUR stable is a marginally better setup than a BoE cutting into a flat-trend GBP. Low conviction but favour Germany over UK at the margin.

• EWU (UK): BoE cutting gradually, 1y +23.19% reflects genuine earnings

improvement. Medium hold, low excitement.

• EZA (South Africa): watch. The -1.23% 30-day figure is better than last week.

Not a buy yet, but the rate of deterioration has stopped. Revisit in the June fx-policy entry.

• EWZ (Brazil): still omit. 1m -9.48% and no improvement week-on-week. BRL

fiscal stress has not abated. Dollar stabilisation at 99 removes the residual external tailwind argument.

What I'd revise if I saw

• DXY breaking below 98 convincingly: would reopen the generic EM carry thesis

and bring EWZ back into scope despite fiscal stress.

• BoJ June meeting (scheduled late June) turning dovish with a hold and JPY

weakening toward 155+: would cut EWJ conviction sharply and force a rethink of the normalisation timeline.

• EZA posting a positive 1m return by the June fx-policy entry: would move

South Africa from "watch" to a low-conviction candidate pick.

Cross-references

• Previous entry on this theme: docs/thinking/2026-05-21-fx-policy-boj-carries-dollar-floor.md
• US-India FORGE expansion (bilateral USD flow anchor): docs/iptm/actions/2026-05-26-us-india-strategic-critical-minerals-framework.md
• Relevant macro routes: /api/country/JP, /api/country/ZA, /api/country/BR, /api/country/DE
• Regime context: /regime