Sovereign Funds Are Quietly Fleeing US Treasuries — Will the Next Neutral-Reserve Bid Land in Bitcoin?

America's longest-maturity bond just delivered a message the Treasury would rather not hear. The September 20-year auction cleared at a record 5.420% — the highest since the security was reintroduced in 2020 — and the share taken down by foreign and other indirect bidders fell to a record-low 52.5%.

A single auction is noise, though a record low inside a trend that has been running for two decades is less so. What makes it worth a few minutes of your time is who is behind it. Two of the world's biggest official pools of reserve-style money are now openly, if quietly, re-allocating away from US Treasuries. Japan's GPIF could shed as much as $62 billion within its current allocation bands, in the estimate of Santander's analysts. Norway's roughly $2.4 trillion sovereign wealth fund has proposed cutting its US Treasury holdings by around $80 billion.

And here is the twist that should sharpen your attention: so far, that rotating capital has chosen gold, not Bitcoin. Central banks keep buying the metal — global official reserves are now the better part of $4 trillion, at record prices well above $4,500 an ounce. Bitcoin, by contrast, is negative for the year, its market cap stuck around $1.6 trillion, and its old habit of trading in step with gold has broken down.

The question the headline wants to answer — will the next neutral-reserve bid land in Bitcoin? — is the wrong frame until you understand why none of it has arrived yet, and why the answer is not a foregone conclusion either way. Let me walk the mechanism.

Why official rotation lands in gold first

A "neutral-reserve bid" is the polite name for a bid that is not yield-sensitive. It is the money that wants deep, liquid, ownable stores of value that carry no single sovereign's credit — the dollars that once defaulted quietly into US Treasuries as the world's de facto safe asset. When a pension fund or central bank stops trusting that default, it does not rotate into a riskier thing. It moves one step sideways, into the incumbent.

That incumbent has a 5,000-year head start and billions already sitting in official vaults. It has no custody question, no accounting headache, and a volatility profile that does not threaten a balance sheet. Bitcoin has none of those comfort features. So the first, natural, dominant stop for official rotation is gold — and the data says that is exactly where it has gone.

There is a second, nastier reason it has not reached Bitcoin, and it is the one that matters for your own judgment. When official demand steps back from Treasuries, long yields go up — the 5.42% 20-year is that friction showing up in price. Higher long yields mean tighter financial conditions, and tighter conditions are precisely the headwind that hits long-duration, high-beta risk assets such as Bitcoin hardest. Pull the thread and the same structural story that lifts gold works against crypto twice: once by excluding it from the new reserve bid, once by tightening the liquidity that prices it.

What it would actually take to move the cap

Now the numbers, because they tell you this is a narrative trade more than a size trade. Bitcoin's market cap is on the order of $1.6 trillion. The entire official rotation in the headlines — GPIF's $62 billion plus Norway's $80 billion — is a little under $150 billion, and that is the total Treasury reduction before pensions recycle most of it into other bonds and equities. Even a generous "marginal slice" reaching crypto would be single-digit billions.

Here is the thing: that is small enough to matter enormously and small enough to be a rounding error, at the same time. On-exchange bitcoin liquidity is thin — effectively a few billion a day. So a sovereign committing even a few billion to a reserve-lite allocation would show up instantly as a marginal buyer and could move price. Measured against a $1.6 trillion market cap, though, it is a rounding error — which is exactly why a real reserve bid would be a signal about durability and expectations, not about the size of any single purchase. It would change what investors believe about who is holding at the bottom — price-insensitive official buyers — rather than about the float. That is the mechanism that compounds: a neutral bid that does not sell.

Confirm it, or kill it

So how do you tell a genuine reserve bid from high-beta noise? Three observable signals hang together.

Watch the Treasury auction indirect share. If it keeps sliding, or stays pinned near record lows, the official retreat is real and the search for neutral stores is on. If it recovers, the thesis loses its evidence base.

Watch the dollar. Reserve rotation into neutral assets is, at heart, a story about the dollar leaving the system. A falling dollar while gold absorbs the flow is the tell that dollars are genuinely exiting toward reserve alternatives. A firm dollar swallowing the same flows means gold is just a crowded hedge — and Bitcoin has no more claim on "neutral" than any other risk asset.

Watch the BTC-gold relationship. For years the two traded as fraternal twins; 2026 broke that, with gold up roughly a fifth while Bitcoin fell. Only when Bitcoin starts sharing gold's driver — moving with it on debt and dollar impulses instead of as a pure 3x-beta tech proxy — would you have evidence that a neutral bid is actually reaching it.

And the falsification, cleanly: kill the thesis if indirect participation recovers, the dollar stays firm while gold absorbs the flow, AND no sovereign discloses a crypto allocation. All three. Any one alone is ambiguous; all three together, and the next reserve bid goes nowhere near crypto.

Where I land

The retreat from Treasuries is real and structural — the auction data and the pension proposals point the same direction, and whatever you think of it, the question of the dollar's reserve status is a debt-and-trust story that does not un-ring itself. But a neutral-reserve bid is a conservative, sticky animal: it goes to the incumbent first, and it always has. Bitcoin will only capture a marginal slice once the data starts treating it like a store of value rather than a leveraged growth trade. That is the price of admission. The door is open, conditionally — and nothing in the last six months says the key has turned yet.