Size the flow that is in dispute. The window the fund describes opens in mid-2025 [1], and the government's Bitcoin Office was still posting the daily-purchase line on August 28, 2026 [4]; call the start July 1 and the span is 424 days [1], which at one coin a day is roughly 424 coins [2], or about 34.4 million dollars against the 81,249 dollar quote at the head of Decrypt's own price ticker [7][3]. Run it as a standing budget line instead and one coin a day annualises to about 29.7 million dollars [5]. That is the size of the appropriation that has been carried around as proof that sovereign treasuries will allocate to bitcoin. On the fund's account, that flow was private donations rather than a budgeted appropriation [2].
The phrase doing the work is "no public resources were used" [2], which answers which account paid rather than whether the wallet grew, and three descriptions of the same reserve are now on the record with only the issuing government's using the verb bought [4]. That gap matters more than the coin count, because a donation and a purchase leave identical on-chain balances and opposite fiscal signals. A sovereign bid is interesting to a market not for its volume but for its price insensitivity: a buyer with a mandate and a line item keeps bidding when the print falls, while a donor flow depends only on the donor's continued willingness to give.
The arithmetic runs the other way for El Salvador, which is the part the accumulation story never priced. A government that did not spend public money on coins [2] also did not forgo roughly 29.7 million dollars a year of anything else [5], so the fiscal-risk case against the reserve and the sovereign-demand case for it come apart at the same seam. Nothing in the fund's language describes a treasury choosing bitcoin over a road or a payroll.
The reading the evidence earns is narrow. There is no documented public purchase in the window, so the daily-buy line should not be priced as state balance-sheet demand, and the counter-position deserves its space: verification here runs through documentation the authorities supplied to the fund [2] rather than a published payment trail, and the fund's own account has already changed once [6]. A payment out of a public account would overturn the read; so would a fourth version of events, which on the record so far is not exotic. Until one of those arrives, what El Salvador has demonstrated is that a state can hold donated bitcoin - a narrower claim than a state choosing to buy it.
