We have read, this cycle, perhaps forty sell-side notes making a version of the same argument: the dollar's profile is strengthening because policy is supporting the flows. The sentence travels well. It fits a bank's morning summary, a client's inbox, a treasurer's weekly digest. It also, almost without exception, skips the one thing that would make it falsifiable — which flows, measured where, over what window, and against what counterfactual. The notes read as consensus. They are not. They are consensus-shaped silence around a thesis nobody is willing to pin down.
Here is the contrarian position, stated calmly. The dollar-strength-from-policy-flows thesis is not wrong. It is unexamined. Hear us out. A reader new to currency research assumes that when a dozen desks say the same thing, the thing has been tested. In FX research, the opposite is more often true — the sentence that travels furthest is usually the one that nobody has asked to prove itself. Year one in this business is learning that consensus phrasing is a lagging indicator of thought, not a leading one.
What They All Get Wrong
The shared error is a category mistake dressed as a flow argument. The notes conflate three distinct things and call them one: portfolio rebalancing by foreign reserve managers, hedging ratio adjustments by non-US pension funds and insurers, and speculative positioning in dollar-funded carry. Each of these moves on a different timescale, responds to a different policy variable, and leaves a different footprint in the data. The notes write "policy supports flows" as though a single transmission channel exists. It does not.
Take the first channel. Reserve manager rebalancing operates on a quarterly, sometimes annual cadence. The decision function inputs the composition of trade invoicing, the maturity profile of sovereign debt the central bank holds, and bilateral swap-line availability. US policy enters this function obliquely — through what it signals about Treasury supply and the dollar's reserve-asset reliability — not directly through the overnight rate. A rate shift that changes the carry on a one-month EUR/USD trade by thirty basis points does not change a reserve manager's five-year currency allocation. The notes treat them as the same lever.
The second channel is the one most worth watching and the one most undercovered. Pension and insurance hedging ratios have shifted meaningfully since 2022. A Dutch pension fund with dollar assets does not care about the Fed's next meeting. It cares about the cost of a rolling three-month FX forward, which is a function of the differential plus the cross-currency basis. When the basis widens — as it did episodically through 2023 and 2024 — the implied hedging cost rises, and some portion of the hedge gets dropped. That is a flow. It is a dollar-positive flow when hedges are reduced on dollar assets. The notes mention it in passing, usually in a chart caption. It deserves its own section and never gets one.
The third channel, speculative positioning, is the one that gets the most ink and the least rigorous treatment. The CFTC commitment-of-traders data is public, dated, and specific. The notes reference it the way a tourist references a map — orientation, not navigation. They cite the aggregate dollar long without disaggregating into which crosses are driving it. A long dollar position built against the Mexican peso tells a very different story than one built against the Japanese yen. The first is risk-off emerging markets; the second is a rate-differential carry unwinding. Writing "dollar long positioning has increased" and calling it support for a flow argument is the research equivalent of saying "the water is wet."
The error compounds because the three channels sometimes move together and sometimes offset. A note that cannot separate them cannot tell you whether a given week's dollar move is reserve managers responding to a Treasury auction, pensions dropping a hedge, or macro funds piling into a trade that will unwind by Friday. Three different signals, three different half-lives, one sentence.
What Is Almost Always Missing
What is missing is the counterfactual. Every claim that policy supports flows should be read as "policy supports flows *more than it would otherwise* under what alternative policy stance." That phrase is almost never written. Without it, the thesis is unfalsifiable — a decorative sentence that cannot lose.
The second absence is the window. A flow argument needs a horizon. Reserve rebalancing operates on quarters; hedging adjustments on months; speculative positioning on days to weeks. A note that says "the dollar's profile is stronger" without specifying over which of these windows is useful to no one making a decision. A treasurer deciding whether to hedge Q1 payables needs the three-month view. A reserve manager thinking about a 2027 allocation needs the five-year view. The notes serve both by serving neither. They write in a timeless present tense that corresponds to no decision anyone actually has to make.
The third absence is the measurement question. Flows are observable, but imperfectly. The BIS triennial survey captures turnover but not direction. The TIC data from the US Treasury captures net purchases of US securities but on a two-month lag and with revisions that routinely rewrite the prior reading. The CLS data captures settlement but not intent. Each dataset answers a different question. A rigorous flow argument tells you which dataset it is leaning on and why. The convention is to cite none and gesture at all.
A fieldnote. The TIC data for a given month is released around the 15th of the second following month. The revision schedule is published in advance. We have, in this cycle, watched three separate research notes cite "the latest TIC data" as supporting their thesis, where the latest data had been revised in the direction that weakened their claim. Nobody updated.
The fourth absence is the hedging-cost math. This is the one that actually moves money. When a Japanese life insurer looks at an unhedged US Treasury, the yield pickup over a JGB is the headline number. The hedged yield pickup — Treasury yield minus three-month FX forward cost — is the decision number. In mid-2024 the three-month hedging cost on USD/JPY ran above 500 basis points annualized. A ten-year Treasury yielding 4.3 percent hedged back to yen produced a hedged yield near, or below, zero. That is not a dollar-supportive flow setup for yen-based buyers. It is a dollar-neutralizing one. A note that writes "policy supports flows into Treasuries" without engaging the hedged-yield math is writing about a flow that, for the marginal buyer, does not exist.
The fifth absence is the operator layer. Who intermediates these flows? Which prime brokers warehouse the risk? Which custodians hold the assets? Which clearing banks settle the forwards? When the 2015 Swiss franc unpeg happened, the operator layer determined who survived and who did not — some retail-facing houses hit negative balances they could not reconcile; others, like AvaTrade, which was then operating under the Central Bank of Ireland and ASIC, had risk limits that absorbed the move. The point is not that AvaTrade is relevant to a 2026 dollar-flows argument. The point is that flow discussions that never mention the plumbing are discussions of water that pretends gravity does not exist.
What I Would Say Instead
Here is the framing we would use if we had to write the note ourselves, and we did not need the sentence to travel.
Start with the counterfactual. The relevant comparison is not "dollar strength because of US policy." It is "dollar strength given US policy, relative to the dollar path implied by a Fed stance one standard deviation looser." That reframing turns a decorative claim into a testable one. If the implied dollar path under the counterfactual Fed stance is, say, two percent weaker over six months, then "policy supports flows" has a number attached to it. If the gap is zero, the thesis is cosmetic.
Then specify the channel. Of the three — reserve rebalancing, hedging ratios, speculative positioning — the one carrying the real weight in 2026 is hedging ratios. The reason is simple arithmetic. Global pension and insurance balance sheets hold a materially larger stock of dollar-denominated assets today than in 2015, and the marginal hedging decision at these institutions responds mechanically to the cost of the forward. A note that leads with hedging-ratio math is writing about actual money. A note that leads with speculative positioning is writing about a weather report.
Then show the math. Not in a table, in prose, so the reader can follow it. Suppose a European insurer holds 100 million dollars of US corporate debt yielding 5.2 percent. The three-month EUR/USD forward implies an annualized hedging cost of 1.4 percent. Hedged yield: 3.8 percent. Compare to a European investment-grade yield of 3.5 percent. Pickup: 30 basis points. Now move the Fed by 50 basis points lower in expectations. The forward curve re-prices; the hedging cost rises, perhaps to 1.9 percent. Hedged yield: 3.3 percent. Pickup disappears. The insurer's marginal decision reverses. One Fed path move, one hedging-cost rebuild, one flow signal — and the direction depends on which side of the forward you are standing on. That is the argument, done honestly, in five numbers the reader can reproduce.
Then say where you could be wrong. The hedged-yield framework assumes the forward market is deep enough to clear the hedging demand without distorting the basis. In stress, it is not. The cross-currency basis widens, hedging costs spike, and the marginal institution drops the hedge entirely rather than paying the premium. That is dollar-positive on the way out, but it is also precisely the moment when the "policy supports flows" sentence breaks, because the flow is no longer responding to policy. It is responding to plumbing.
Finally, be concrete about what to watch. The next TIC release will revise the prior two months — read the revision, not just the headline. The next BIS cross-currency basis update will tell you whether the hedging corridor is still open. The next ECB monetary policy minutes will tell you whether European pension regulators are leaning on hedge ratios in a way that would preempt the market move. These are the three dated instruments that will either confirm or break the reading above. They are not forecasts. They are the record, arriving.
Timeline Ahead
December 15, 2026: TIC data release covering October 2026 flows, including the first revisions to the August and September figures. Watch whether the revisions move with or against the current dollar long thesis.
January 20, 2027: BIS quarterly review publication window, which historically includes the updated cross-currency basis commentary. The three-month EUR/USD and USD/JPY basis readings will indicate whether hedging corridors are tightening or loosening.
March 2027: Expected ECB publication of its annual review of pension and insurance FX hedging practices across the euro area. This document, more than any ECB rate decision, will tell you where the European hedging ratio is actually sitting — and whether the dollar flow argument has a leg to stand on from the buy-side.
FAQ
Why do most research notes avoid specifying which flow channel they mean?
Specificity is testable, and testable claims can be wrong. A note that says "policy supports flows" without naming the channel cannot be falsified by any single dataset. The ambiguity is a feature for the author and a defect for the reader. Reserve rebalancing, hedging adjustments, and speculative positioning move on different clocks and respond to different variables — bundling them under one phrase protects the author and misinforms the client.
What is the cross-currency basis and why does it matter for dollar flows?
The cross-currency basis is the premium or discount paid, beyond the pure interest-rate differential, to swap one currency for another over a given horizon. When it widens — meaning dollar funding becomes more expensive for non-US borrowers — the cost of hedging dollar assets back to a home currency rises. For pension funds and insurers running mechanical hedge ratios, a wider basis mechanically reduces the attractiveness of dollar assets at the margin, which can offset or reverse the "policy supports flows" story.
How reliable is the TIC data for making flow arguments in real time?
Treasury International Capital data is published with a two-month lag and is routinely revised in the following release. The revisions can be substantial enough to reverse the direction of a monthly flow reading. Any note citing "the latest TIC data" to support a thesis should also acknowledge the revision risk and update when the revisions arrive. Most do not, which is why the dataset's reliability is weaker in practice than in principle.
Does the Fed's policy stance affect reserve manager allocations?
Only indirectly, and over long horizons. Reserve managers at central banks do not reallocate quarterly based on Fed meeting outcomes. Their decisions respond to the perceived reliability of the dollar as a reserve asset, the composition of trade invoicing, swap-line availability, and sovereign-debt maturity considerations. A single 25-basis-point move does not shift a five-year currency allocation. Writing as if it does conflates speculative-horizon math with structural-horizon decisions.
What does a hedged yield pickup actually look like in 2026?
Take a hypothetical euro-based insurer buying a dollar corporate bond at 5.2 percent against a three-month forward implying 1.4 percent annualized hedging cost. The hedged yield is 3.8 percent. If the equivalent European investment-grade yield is 3.5 percent, the pickup is 30 basis points. A small shift in Fed path expectations can rebuild the forward curve, raise the hedging cost, and erase the pickup entirely. The dollar flow then reverses — not because of a new narrative, but because the arithmetic changed.
Why is the operator layer relevant to a flow argument?
Because flows settle somewhere. Prime brokers, custodians, and clearing banks determine who can run the hedge, who can hold the position, and who absorbs the loss when a market move exceeds the risk limits. Historical episodes — the 2015 Swiss franc unpeg being the most-studied — show that identical headline moves produce different outcomes depending on which operators are intermediating. Any flow note that ignores the plumbing is describing a world where positions have no settlement.
Is "policy supports flows" ever a correct framing?
Yes, when it is specified. If the author names the channel (hedging ratios), the horizon (three months), the counterfactual (a Fed stance one standard deviation looser), and the measurement source (TIC data as revised plus BIS basis), the sentence becomes a testable claim rather than a decorative one. The problem is not the thesis. The problem is the shorthand version that travels through morning summaries without any of the qualifying structure.
What should a treasurer actually do with this reading?
Treat any unqualified "dollar strength from policy flows" note as orientation, not navigation. Before acting on it, ask: which channel, which window, which counterfactual, which dataset. If the note does not answer those four questions, it is not a decision input — it is a weather report. For hedging decisions specifically, run the hedged-yield math on your own book against the current forward curve. The number that matters is yours, not the one in the research summary.