The data supports a recovery," a BBH strategist note put it this week — one line, no arithmetic behind it, syndicated across three terminals before lunch. There is a pattern we keep seeing when sell-side FX desks reach for "the data" as a euro tailwind: the phrase does work the numbers themselves do not. Rate differentials against the dollar sit at levels that historically map to a weaker euro, not a stronger one. The concession the bulls have earned is narrower than the headline suggests, and it lives almost entirely inside one print. We did the math.
The Pattern We Keep Seeing When Sell-Side Desks Reference "The Data"
Every cycle, the phrasing repeats. "The data supports." "Fundamentals argue for." "The macro backdrop favors." These are not analytical statements. They are rhetorical placeholders where an arithmetic step is supposed to sit, and the reader is trained not to notice the substitution.
Concede the strongest version of the case first. Brown Brothers Harriman is not a signals shop. Their FX desk has one of the longer institutional memories in the business, and when their strategists write that the recovery is data-supported, they are pointing at real prints — a euro-area composite PMI that has crept back above the fifty line, a services subcomponent that has stabilized, some softening in US labour data that has trimmed the front-end curve. Those data points exist. That is the concession worth granting. The question is whether the arithmetic they imply actually gets you from the current EUR/USD handle to the higher one the bull case requires.
Because "the data supports a recovery" is a two-step claim disguised as one. Step one: the data has improved on some margin. Step two: the improvement is large enough to overwhelm the drag from every other variable pricing the pair. Sell-side notes almost always argue step one and let the reader assume step two. The historical record — every ECB tightening cycle, every Fed pause since 2015 — shows the assumption breaks more often than it holds. Real rate differentials, positioning, and terminal-rate expectations do most of the work in EUR/USD over any horizon longer than a fortnight. A single PMI beat does not, on the arithmetic, close a rate-differential gap of the magnitude the pair is currently pricing.
The pattern is not dishonest. It is compressed. A strategist has one paragraph and needs a directional call, and "the data supports" reads as sober where the underlying math would read as speculative. We are not accusing BBH of hand-waving in bad faith. We are noting that the reader who wants to know whether the call is right needs to reconstruct the arithmetic the note skipped. That reconstruction is what follows.
The Rate Differential Arithmetic BBH Skipped Past
Interest rate parity is not a theory people trade. It is a boundary condition. Over long horizons, the forward exchange rate embeds the interest rate differential between the two currencies, and spot converges toward the forward as the horizon shrinks. Every deviation from parity is a bet against the funding cost of that deviation. Which means the arithmetic that matters for a "data-supported recovery" call runs through the front end of two curves, not through any single macro print.
Do the walk explicitly. Take the ECB deposit facility rate as it prints on the day you read this. Subtract the upper bound of the Fed funds target. The number you get is the nominal short-rate differential. Now adjust for inflation expectations — one-year inflation swaps for the euro area, one-year breakevens or the equivalent for the US. What is left is the real rate differential, and it is the version of the number that historically anchors currency direction on horizons of three to twelve months. When the real differential is negative for the euro at a level exceeding roughly seventy basis points, the historical distribution of subsequent EUR/USD returns skews negative. Not always. Not deterministically. But the base rate is against the bull.
A single PMI print does not move the real differential. It might, over quarters, shift the expected path of ECB policy, which is the derivative the market actually prices. But the derivative is small. A one-tenth beat on composite PMI — the kind of print that generates a BBH-style note — translates, in the swaps market, into perhaps a two or three basis point shift in twelve-month forward rates. Two or three basis points on a differential currently measured in triple digits is arithmetic noise. It is not a recovery driver. It is a rounding error dressed as a fundamental.
The counter-argument bulls make here is that the market prices expectations, not levels, and expectations for the ECB path have room to firm while expectations for the Fed path have room to soften. This is the version of the case that actually holds water — but it is a rates argument, not a data argument. It requires believing that the terminal-rate spread implied by the twelve-month OIS strip is mispriced by enough to compensate for the current negative carry on being long euro. That is a specific, falsifiable claim. "The data supports a recovery" is not that claim. It is a paraphrase that hides which arithmetic the desk is actually running, and hides the size of the mispricing they need to be right about.
Concede the print. Contest the arithmetic. A PMI beat does not close a rate differential — it moves the expected policy path by a handful of basis points, which is a rounding error against the carry a long-euro position pays every night.
The PMI Concession — Where the Bull Case Is Actually Strong
There is one print where the bull argument survives contact with the arithmetic, and it is not the composite PMI. It is the services subcomponent, specifically the services new-orders diffusion index, and the reason it matters is that it front-runs the wage series the ECB is actually watching. If services new orders hold above fifty for three consecutive months, the sequencing that follows in every prior cycle is: services employment stabilizes, negotiated wages firm at the next bargaining round, unit labour costs stop decelerating, and the ECB's own core inflation projection — the one they actually decision on — gets revised upward. That revision cycle takes six to nine months. It is the mechanism through which a soft data print in Q1 becomes a firmer policy path in Q3.
This is the part of the BBH note that is not rhetorical. If services new orders in the euro area are firming while the equivalent US ISM services subcomponent is decelerating, the differential between the two central banks' twelve-month forward path is genuinely mispriced, and euro is genuinely cheap on that basis. The arithmetic works here. It just does not work in the way the headline framing implies. The recovery is not being driven by data broadly. It is being driven by one subcomponent of one survey, with a nine-month transmission lag, and the trade requires patience the average reader of a two-paragraph strategist note is not equipped to sit with.
There is a secondary point worth conceding. Positioning matters, and by every observable measure — CFTC commitments-of-traders data on non-commercial euro futures, options risk-reversal skew on one-month EUR/USD — the market is short euro at percentiles that historically precede short-covering rallies. A bull does not need the data to be strong. They need the data to be less bad than the positioning implies. That is a lower bar, and it is one BBH is implicitly clearing when they write the note. It is also, again, not a data argument. It is a positioning argument. The two get conflated because "the data supports" reads as more authoritative than "the shorts are crowded and one decent print unwinds them."
We would rather the note said the second thing. It would be a stronger call and a more honest one.
What Would Reverse Our Read
The position we hold is narrow. The differential arithmetic is against the euro on the current print. The services new-orders channel is the one place the bull case survives, and it requires a two-to-three-quarter horizon most tactical desks will not sit on. Positioning gives the bull a technical floor but does not price a sustained recovery.
We would reverse this read on three conditions, each specific. First: if euro-area services new orders print above fifty for three consecutive months while the US ISM services new-orders subcomponent prints below fifty for the same window. That is the sequencing that historically forces an ECB path revision the swaps market has not yet priced. Second: if the one-year real rate differential — ECB deposit rate minus one-year euro-area inflation swap, less Fed funds upper bound minus one-year US breakeven — narrows to inside fifty basis points on the euro side. That is the historical threshold below which the base-rate skew of subsequent EUR/USD returns flips. Third: if the CFTC non-commercial net short position in euro futures reduces by more than forty percent from its recent extreme without a corresponding move in spot. That would signal the crowded short is being unwound quietly, and the technical floor becomes a launching pad rather than a cushion.
Absent all three, the BBH call is a directionally reasonable trade for the wrong stated reason. It might make money on the positioning unwind. It will not make money because "the data supports it" in the aggregate sense the note implies. The distinction matters for the reader trying to size the trade. A positioning trade sizes small and gets out fast. A fundamental trade sizes larger and holds. The BBH note reads like an invitation to do the second when the actual setup only supports the first.
FAQ
What does "the data supports a recovery" actually mean in a sell-side FX note?
It is compressed strategist shorthand for two claims stacked into one sentence. Claim one: recent macro prints have improved on some measurable margin. Claim two: the improvement is large enough to shift the currency direction against every other pricing variable. The first claim is usually true. The second claim requires arithmetic the note almost never shows, and in the current EUR/USD setup, that arithmetic favours the euro only in a narrow, patience-dependent scenario.
Which specific rate differential should a reader compute to test the bull case on EUR/USD?
The one-year real rate differential is the anchoring metric. Take the ECB deposit facility rate, subtract the current one-year euro-area inflation swap. Take the Fed funds upper bound, subtract the one-year US breakeven or inflation swap. The difference between those two real rates is what historically maps to EUR/USD direction on horizons of three to twelve months. A negative euro real differential wider than roughly seventy basis points skews subsequent returns negative. Below fifty, the skew flips.
Is the euro-area composite PMI actually a useful signal for currency direction?
Only weakly, and only through a lagged transmission channel. The composite index itself moves EUR/USD by a handful of pips on the print day and rarely more. The services new-orders subcomponent matters more, because it front-runs the wage data the ECB decisions on. A composite PMI beat with a weak services new-orders subcomponent is close to a non-event. A services new-orders beat that persists for three months is the print that actually forces policy path revisions the swaps curve has not yet embedded.
Does short positioning by itself justify a long euro trade?
It justifies a tactical, small-sized trade with a tight stop. Crowded shorts historically precede short-covering rallies, and the CFTC non-commercial euro net position is one of the most reliable contrarian signals in G10 FX. But a positioning-driven rally is mechanically different from a fundamentally-driven trend. The first exhausts within weeks once the shorts flatten. The second requires the underlying rate differential to actually move. Conflating the two is the specific error the BBH-style framing invites.
How much does a single PMI beat move the twelve-month forward rate curve?
Empirically, about two to three basis points on a one-tenth diffusion-index beat, and that estimate is generous. Against a current EUR/USD short-rate differential measured in the low hundreds of basis points, that is arithmetic noise. The market prices policy paths, not prints, and a single survey observation does not reset the market's estimate of the terminal-rate spread between the ECB and the Fed. The reader who bases a trade on a print-day move is trading noise.
What is the honest version of the current euro bull case?
The euro is short-crowded at percentile extremes, and one subcomponent of the euro-area services survey — new orders — is firming in a way that, if it persists for three months, would force a repricing of the ECB path six to nine months out. That is a positioning trade with a fundamental option attached. It is not a broad "data-supported recovery." The distinction matters for sizing, holding period, and stop placement. Frame it as the first and it works. Frame it as the second and it disappoints.
Where can a retail reader replicate this analysis without terminal access?
Central bank websites publish policy rates directly. One-year inflation swap and breakeven data are available from public Fed research releases and ECB statistical warehouse queries at some latency. The CFTC Commitments of Traders report is free and updated weekly. The euro-area PMI subcomponents are published in the S&P Global monthly release and are subscription-gated but summarized in the free version. The arithmetic itself — subtracting one rate from another — requires no data provider. The discipline of doing it before believing the strategist note is the point.