The peso closed last week near 17.25 per dollar. Iran has held the Strait of Hormuz shut since early March, crude is up more than 50% since the war began on February 28, and the IPC had its worst week since April earlier this month. None of it moved the currency. The peso is up roughly 3.9% against the dollar in 2026, within sight of multi-year highs, and it got there while a war played out in the Gulf and global markets dumped risk.
In February we wrote, in “The Peso Is Holding. That Is the Story Nobody Is Writing,” that the currency’s strength rested on three structural supports: trade flows, sticky foreign investment, and a credible central bank. We said only a USMCA collapse could break it. Four months later the peso has passed a harder test than the one we set, holding through a war and an oil shock we did not think to list. Passing it exposed something the calm tape hides. The support has narrowed to a single load-bearing pillar, and that pillar answers to Washington, not Mexico City. The Fed meets Tuesday and Wednesday. That decision matters more to the peso than anything in the Gulf.
What held, and what it held through
The shock was not small. The war that began on February 28 closed the world’s most important oil chokepoint and pushed Brent more than 50% above its pre-conflict level. The IPC lost 3.57% in a single week this month, its worst stretch since April, as the Wall Street selloff bled into Mexico City. We covered that week in “Washington Wants Fifty Percent,” and why Mexican equities barely track crude in “Mexico’s Oil Paradox” back in March.
The currency did something different from the index. It barely registered, trading around 17.2 to 17.4 through the worst of the headlines while Banxico’s reserves sat at a record near USD 250bn. The emerging-market playbook says a currency should weaken when war triggers a rush into the dollar, all the more so for an economy that imports the fuel it burns. The peso ignored it. The explanation has changed since February.
The carry is doing more of the work
The rate gap still pays. Banxico cut to 6.50% on May 7, a move it called the last of the cycle, and the Fed has held at 3.50% to 3.75% all year. The spread sits near 2.9 percentage points, tighter than the 3.25 we cited in February but still one of the widest positive real yields in emerging markets. A foreign investor in short-dated CETES or MBonos earns a real return for the risk. The carry held through the shock.
What changed is everything around it. In February we argued the peso did not depend on hot money because three structural supports sat underneath. Two of the three have weakened, and not as cyclical noise.
The pillars are thinner than they were
Foreign investment was supposed to be the sticky one. In February we pointed to record 2025 FDI of USD 40.9bn and new investment that had more than doubled to USD 7.4bn, the inflow that does not leave when spreads narrow. The first-quarter data undid that. As we showed in “The Capital Showed Up. The Jobs Did Not,” 94.1% of Q1 2026’s USD 23.6bn was reinvested earnings, and genuine new investment was USD 1.7bn, flat after inflation. The inflow we leaned on pads the headline. It does not bid for pesos.
Remittances are recovering, but barely, and into a new tax. The 2025 total fell 4.6% to USD 61.8bn, the first annual drop in over a decade. 2026 has clawed some back, with the first quarter up 1.4% and April at USD 4.9bn, still over 3% of GDP. A new U.S. tax on transfers took effect this year, migration north has thinned, and BBVA sees full-year growth near 2%, flat in real terms. Not collapsing, but no longer the marginal dollar it was.
Trade flows are the real anchor, and they go to a vote in July. Mexico is still the largest U.S. trading partner, and exporters converting dollars remain the steadiest source of demand for the peso. The USMCA review opens next month; the second round runs this week in Washington over auto rules of origin and a proposed U.S.-content threshold, the fight we detailed in “Washington Wants Fifty Percent.” The anchor holds. It is also the one support with a stress date already on the calendar.
Of the three pillars we named in February, one has hollowed out, one is flat, and one faces a July test. What holds the peso day to day is the carry, the same carry we argued was not the marginal price setter. Four months on, it is.
Why the Fed, not the Gulf, is the threat
A currency held up by a rate gap is only as safe as the gap. Banxico is done cutting: the May statement guided to a prolonged hold, and consensus puts the rate at 6.50% through year-end. The spread cannot widen from the Mexican side. It can only move from the U.S. side, and the U.S. side meets this week.
The Fed walks in with American inflation running hot, pushed in part by the same oil shock that left the peso unmoved. If its statement and projections confirm the cutting cycle is over and the next move could be a hike, the differential the peso leans on stops growing and starts to look capped. In February we flagged a carry unwind through Japan; the Bank of Japan tightened and the peso shrugged. The live risk runs through Washington. We put it fifth on our 2026 watch list in January, peso carry unwind risk, and it is moving from theory toward test.
The bottom line
The peso survived the kind of event that usually breaks an emerging-market currency, which vindicates the February call: do not short it on rate cuts alone. The structural story was real. It is also narrower than it was. The dollar supply from investment and remittances has thinned, the trade anchor faces a July review, and the carry has quietly become the marginal support, the role we said it would not play.
This is not a reversal call. Reserves are at a record, the real yield is positive, and nothing about Wednesday forces a move. But watch the Fed, not the Strait. If the U.S. confirms it has stopped cutting and starts pointing at hikes, the cushion under the peso stops building. It will not break on the news. It will just have less between it and the next shock, and the next one will not give two days’ notice the way this one did.