Dollar doubt, equity exuberance: two market signals pointing in opposite directions

At 00:58 UTC on 9 July 2026, a brief social-media note from a market-data account flagged a quiet but consequential inversion in global reserve management: for the first time since the GPI series began tracking central-bank reserve managers' long-term intentions in 2023, more institutions reported plans to decrease their dollar holdings over the next ten years than to increase them. The post described the result as a turning point in the slow choreography of how the world stores its savings — and, by implication, in the architecture of dollar-based finance that has underpinned US borrowing since the second world war.

Less than thirty hours earlier, the same feed's market desk had circulated a Bank of America note observing that the S&P 500 was "statistically expensive on 17 of 20 metrics," and trades rich against the late-1990s tech-bubble yardstick on eight of them. Taken together — the reserve-manager pivot on one side, the equity-multiple reading on the other — the two signals describe a market in which the marginal buyer of US financial assets is no longer obviously the world's central banks, while the price of US equity keeps climbing on the back of domestic flows that do not need a foreign reserve manager to underwrite them. Both can be true at once. The hard question is what they imply together.

Two data points, one tension

The GPI signal is the headline. The GPI — short for a longer internal title that has become standard shorthand among reserve managers — is a forward-looking survey, not a transaction tape. It asks institutions how their allocations will change over a horizon of roughly a decade. Until the latest reading, the cohort has consistently reported net intentions to add to dollar reserves. The shift documented this week — more banks pencilling in reductions than additions over a ten-year window — does not mean that any central bank has yet sold a meaningful position. It means that the marginal reserve manager, asked about the future, no longer treats the dollar as the default expanding share of the store of value.

The S&P signal is a mirror image. The BofA note circulated via the same channel on 8 July observed that the US benchmark trades rich versus the late-1990s technology-bubble yardstick on eight of twenty standard metrics, and "statistically expensive on 17 of 20." Equity multiples are not the same instrument as reserve composition. But they share a single underlying constraint: both reflect what the world is willing to pay, in real terms, for claims on the US economy. When reserve managers price the dollar lower in their intentions surveys while equity investors price US earnings higher in their tape, the simplest reading is that two different buyers — the official sector and the private sector — have diverged.

That divergence is not new. It has been latent since the 2008 financial crisis, when the Federal Reserve's balance-sheet expansion permanently altered the relationship between the official sector and US assets. But the divergence is, by the latest data, no longer latent. It is now self-reported in a survey instrument that the official sector itself uses.

The Fed minutes complicate the picture

The reserve-manager pivot does not exist in a vacuum. The same week, on 8 July, an account focused on monetary policy circulated a summary of the most recent Federal Reserve minutes indicating that support inside the rate-setting committee for further rate cuts was fading as inflation pressures proved more stubborn than expected. The two stories sit on the same page: a central-bank community that is simultaneously less willing to ease policy at home and more willing, at the margin, to diversify away from the issuer of the reserve currency.

This is the structural reading the wire has so far declined to print. The dominant framing — that the dollar's reserve role is inviolable because the US financial system is the deepest, the most liquid, and the most legally certain — has not been falsified by this week's data. It has, however, been weakened at its edges. The GPI signal does not say that central banks are abandoning the dollar. It says that, when asked about the next decade, more of them are no longer planning to increase their share. That is a different and more cautious statement — and it is the first time the survey has returned such a result.

The Fed minutes provide the mechanism. If US inflation remains sticky, the FOMC has less room to ease, which means US real yields stay higher for longer, which means the carry case for holding dollar reserves is unchanged in the near term but more politically and diplomatically costly over the medium term. Reserve managers do not have to dislike the dollar to diversify. They have to be uncomfortable enough with the concentration of their exposure to consider it.

What the equity market is telling us

The BofA note on the S&P 500 is, in its own register, telling a similar story about concentration. The bank counted twenty commonly used valuation metrics and found the index reading expensive on seventeen of them, and trading at or above late-1990s tech-bubble levels on eight. That does not mean the market is wrong to be there. It means the price investors are paying for US earnings is historically unusual, and that unusualness is concentrated in a small number of large-cap technology-adjacent names whose index weight has grown with each successive quarter.

This is the second half of the divergence. Reserve managers are quietly preparing for a world in which the dollar is a smaller share of the global savings pool. Equity investors are paying the highest multiples in a generation for claims on the same US economy those reserve managers are preparing to underweight. The two groups can both be right for years — the official sector moves slowly, and the equity market can stay expensive long past the point at which rational observers call it expensive. But the gap between what central banks say they will do and what equity investors are doing is now wide enough to register.

The structural frame

Stripped of academic scaffolding, the structural argument is straightforward. The post-1945 financial order rested on a quiet bargain: the United States would provide the world's reserve asset, and the world would recycle its savings into US Treasuries at rates the US Treasury could live with. That bargain held because there was no serious alternative store of value at scale, and because the US was willing to run the current-account and fiscal deficits required to absorb the world's excess savings.

What we are watching, in plain language, is the slow renegotiation of that bargain. The renegotiation is not being driven by any single event. It is being driven by the cumulative weight of three developments: the increased financial integration of economies — including China — into global capital markets; the maturation of alternative safe-asset providers; and the willingness of US policy itself to use sanctions and financial infrastructure as instruments of statecraft, which has given non-US reserve managers an additional reason to think about concentration risk that did not exist twenty years ago. The GPI signal is the first time the official sector has put a number on its intentions. It will not be the last.

The China angle, briefly

The same week, on 8 July, a Japan-based English-language outlet reported that Beijing was strengthening protections for older workers in order to expand the effective labour force — a domestic policy that, taken alongside the slow diversification of the official reserve base, points to a Chinese state apparatus that is preparing for a longer, slower-growth horizon rather than the high-velocity convergence of the 2000s. The policy framing is unremarkable: ageing populations across East Asia are now a structural variable in capital flows, not a footnote. What is notable is the timing. Beijing is, in effect, leaning into its labour-supply ceiling at the same moment that the world's reserve managers are leaning into their diversification ceiling. The two adjustments are not formally coordinated. They rhyme.

Stakes and time horizons

If the GPI signal hardens — if a second or third reading shows the same inversion, and if it is corroborated by actual reserve-flow data rather than intentions alone — the implications arrive in three layers.

First, US borrowing costs. Treasury issuance priced off a base of foreign reserve managers carries a different risk premium than Treasury issuance priced off domestic flows alone. The difference is not catastrophic; the US has domestic savings at scale. But it is real, and it would show up first in the long end of the curve, where reserve managers have historically been most active.

Second, the equity market. If the BofA note is right that the S&P 500 trades rich on seventeen of twenty standard metrics, then the cushion against any disappointment is thinner than the headline indices suggest. Multiple compression is not the same as a crash. It is a slow grinding lower in the multiple investors pay for the same earnings, and it is the historical pattern of equity markets after extended episodes of expensive valuations. The eight metrics on which the index trades at or above late-1990s levels are the ones to watch.

Third, the political economy. A world in which the US economy is funded increasingly by its own households and institutions is a world in which the distributional consequences of US fiscal and monetary policy land harder on those same households and institutions. Reserve diversification, in other words, is also a quiet transfer of who pays for the US safety net.

What remains uncertain

The GPI signal is a survey, not a transaction. It captures what reserve managers say they will do over a ten-year horizon, and surveys of intent have a long history of overshooting actual behaviour. The 8 July Fed minutes are also a snapshot, not a trend; one inflation print and a softer labour-market release would re-open the door to cuts and reset the carry calculus. The equity valuation reading is a comparison to history, and history rhymes rather than repeats. None of the three signals is, on its own, decisive. The novelty this week is that they are pointing in the same direction — away from the assumption that the post-1945 financial architecture will look in 2036 much as it looked in 2016.

What the sources do not yet show is whether the reserve-manager shift is broad-based or concentrated in a handful of large holders; whether the Fed's hesitation on cuts reflects a one-quarter inflation surprise or a more durable re-anchoring; and whether the equity-multiple reading reflects genuine earnings power or a thinner mix of index concentration. These are the questions the next quarter's data will answer.

Desk note: The wire this week framed the Fed minutes as a stand-alone inflation story and the S&P valuation note as a stand-alone equity story. Monexus reads them together with the GPI reserve-manager signal as one connected narrative about who, in 2026 and beyond, is willing to pay what price for claims on the US economy.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://x.com/unusual_whales/status/1800000000000000001
• https://t.me/CryptoBriefing
• https://t.me/NikkeiAsia
• https://t.me/TSN_ua
• https://t.me/TSN_ua