Bitcoin rose about 24% between Monday and Friday, touching $79,400 on August 21 and trading above $77,000 — its strongest weekly advance since 2023, by CoinDesk's count. Smaller tokens moved further: ENA gained roughly 13% and ZEC 12.5% in Friday trading, with NEAR and LINK adding about 8%.

The cause was not crypto. On August 19 the Treasury doubled the ceiling on its buybacks of long-dated government debt, from $2 billion to at least $4 billion per operation across the 10-to-20-year and 20-to-30-year maturities, with operations running September 9 through November 4. Secretary Scott Bessent said the next day the program could go further still. Long yields fell, risk assets rallied, and the resulting short squeeze — forced buying by traders who had bet on lower prices — wiped out a record $2.7 billion in bearish positions on August 19 by Bloomberg's count, and roughly $3 billion by CoinDesk's a day later. Spot bitcoin exchange-traded funds took in $517 million on August 19 and $606 million on August 20; ether funds took $189 million and $221 million. The week before, spot bitcoin funds had shed about $390 million.

A bond buyback in the 20-to-30-year sector is not a statement about token supply, though the tape read it as one.

Where does a 24% week land in a rules-based assessment? Mostly nowhere, and the arithmetic is worth walking through.

Of the six scoring dimensions, two take direct input from price: Market Structure at 8 points of 100 (liquidity depth, spreads, realized volatility, drawdown, trend) and Relative Valuation at 12 (peer-relative ratios — MVRV, NVT, fees-to-market-cap). Twenty points between them. The other 80 sit with Tokenomics & Value Accrual at 28, On-chain Usage & Economics at 22, Security, Decentralization & Durability at 18, and Ecosystem & Moat at 12. Those read supply schedules, fee revenue, node counts, audit quality and developer depth. None of them moved because the Treasury extended a buyback.

The nine hard gates — applied before any weighted score is calculated — are untouched entirely. A rally does not make an anonymous team accountable, an unaudited contract audited, or an insider-dominated supply less insider- dominated. Gates are binary, and price is not an input to any of them.

Then the part that runs against intuition. Relative Valuation measures price against what a network actually earns and settles. When price climbs 24% in a week while fees, active users and non-incentivised volume stay flat, ratios like NVT and fees-to-market-cap get worse, not better: the asset now costs more per unit of the thing it produces. This is exactly the kind of input that dimension weighs, and the arithmetic points the opposite way from the mood. A green week is not automatically a scoring tailwind.

The uniformity is its own signal. When ENA, ZEC, NEAR and LINK all outrun bitcoin in the same session, the market is repricing risk appetite, not sorting assets by quality. Moves that correlated carry close to no information about any individual token. Of the 202 assets CryptoGrade covers, 2 grade A, 8 B, 31 C and 161 F, with an average composite score of 27.6 out of 100. That spread is built from supply, usage and security facts. A week when everything rises together does not compress it.

The framework takes no view on where the price goes next. It takes a view on what would have to show up in the data for an assessment to change — and this week, none of it did. See the track record for how the grades have actually moved.

CryptoGrade publishes educational research, not investment advice. Ratings and commentary are produced by a rules-based framework applied identically to every asset. Crypto assets are volatile and you can lose your entire investment. Do your own research.