Two readings from our research record, a week apart.

Avalanche (AVAX) read C and scored 52.9 on our record of October 24, 2026. On our record of October 30 it read C and scored 47.8. Five points and a bit, gone, and the letter did not move.

Beldex (BDX) read C and scored 40.5 on October 24. On October 25 it read F / Failed and scored 39.6. Nine tenths of one point, and the letter changed.

That is the whole week in one comparison. A letter change is not a measure of how far an asset travelled. It is a measure of where the asset was standing when it moved. Avalanche had room underneath it. Beldex was sitting on a line.

This is a review of the letter changes on our record between October 24 and October 30, 2026, with the arithmetic behind them and, where our record publishes one, the reason. It is a selection from our dated record rather than a complete weekly tally.

On our record of October 27, four assets crossed a letter line. In all four, our published change note names exactly one of the six scoring inputs, and that input moved exactly one step.

Four assets, four different shapes of business, and the same mechanic in each. No other input moved on any of them.

Every asset we rate is scored 0 to 5 on six weighted inputs. The weights are published, they are identical for every asset, and they are these: Tokenomics and Value Accrual 28%, On-chain Usage and Economics 22%, Security, Decentralization and Durability 18%, Ecosystem and Moat 12%, Relative Valuation 12%, Market Structure 8%.

Because each input runs 0 to 5, one full step is one fifth of that input's weight. A step on Tokenomics is worth 5.6 points of the composite score. A step on Relative Valuation is worth 2.4. A step on Market Structure is worth 1.6.

So Quant and Bitcoin Cash each took a single step on the heaviest input in the framework and collected 5.6 points. Decred and Kamino each took a single step on an input weighted less than half as much and collected 2.4, one up and one down. All four crossed a line, because all four were close enough to one that a single step carried them over it.

You can run this yourself on any asset. Bitcoin Cash on its October 30 record: Tokenomics 5 of 5 at 28% is 28.0 points, On-chain Usage 2.5 at 22% is 11.0, Security 3 at 18% is 10.8, Ecosystem 1 at 12% is 2.4, Relative Valuation 2 at 12% is 4.8, Market Structure 3 at 8% is 4.8. Add them: 61.8, which is the score we publish. We checked the same arithmetic on ten assets while writing this and it reproduced the published score on all ten.

Here is the note our model published against Decred's move, word for word:

That is the shape of all four. The note tells you which input moved and by how much. It does not tell you which piece of evidence moved it. That part is not in the published record, so we are not going to narrate it for you, and you should be wary of anyone who does.

What our record does carry, for these four, is a driver tag reading input-data: the move came from the evidence behind an input rather than from anything about the asset's price that day.

Earlier in the week, between our October 24 and October 25 records, four more assets crossed a line:

And at the end of the week, between our October 29 and October 30 records, one more:

For those five, our record for those dates attaches no cause. The scores moved, the letters moved with them, and there is no published change note behind any of them. We would rather tell you that than invent a story. The underlying input scores for every one of these assets are in the record itself, free with an account.

SafePal's is the largest single-day score fall in this piece, 11.2 points. We can tell you that it happened and when. We cannot tell you why, and our record does not either.

Our published bands are simple. A is 80 and above. B is 60 up to just under 80. C is 40 up to just under 60. F is below 40.

That is why Beldex's 0.9 points mattered and Avalanche's 5.1 did not. Beldex went from 40.5 to 39.6 and crossed out of C. Avalanche went from 52.9 to 47.8 and stayed in the middle of C the whole way.

One caveat on F that we repeat every time, because it matters: a low score is not the only road into an F. Our framework runs a set of screens alongside the score, and a failed screen places an asset in F at any score at all, including a score in the A band. Our record does not always say which road a given asset took, and where it does not, we do not guess. In Quant's case the change note names one input move and records no screen change, and the score crossed from 38.8 to 44.4. That is what the note says. It is not a statement that every screen on Quant is evidenced or clear, and we are not making one.

Letters read live on October 31, 2026. Each asset's own last scoring date is in brackets, which is what its page shows as "Data as of".

Two positions are missing on purpose. Raydium and Avalanche each sit on exactly the same score as another asset we rate, and a position between two tied scores is an ordering we are not willing to present as a judgement.

If you hold any of these, the finding is not "Decred went up". The finding is that a single input moving a single step was enough to change four letters in one day, so the letter on an asset you own can change on a week where nothing about that asset was in the news.

That is what CryptoGrade Alerts is for. You pick the assets you care about, and the alert is tied to our grade on one of them changing. It is free with a free CryptoGrade account, and the ask is deliberately small: one thing, about assets you choose, because the letter moving is the event most worth knowing about and the one you are least likely to notice on your own.

Create a free account here if that is useful to you. An account also opens the input scores behind every grade above, and our ranking past the top 10. One honest caveat on the ranking: assets we have taken in but not yet rated carry Not Rated rather than a letter, so it is not a list of letters all the way down, and we are not going to sell it as though it were.

If you want the longer read this week, our founder's guide to staking yields is published in full here: https://blog.cryptograde.ai/p/the-low-effort-guide-to-staking-in-2026-and-what-the-yield-really-is

Each one shows the same letter and score our record carries, with its own data as of date:

Quant (QNT): https://cryptograde.ai/tokens/quant

Bitcoin Cash (BCH): https://cryptograde.ai/tokens/bitcoin-cash

Decred (DCR): https://cryptograde.ai/tokens/decred

Kamino (KMNO): https://cryptograde.ai/tokens/kamino

Sky (SKY): https://cryptograde.ai/tokens/sky

Celestia (TIA): https://cryptograde.ai/tokens/celestia

Raydium (RAY): https://cryptograde.ai/tokens/raydium

Avalanche (AVAX): https://cryptograde.ai/tokens/avalanche

Beldex (BDX): https://cryptograde.ai/tokens/beldex

SafePal (SFP): https://cryptograde.ai/tokens/safepal

We wrote up the Bitcoin Cash move on its own when it happened: https://blog.cryptograde.ai/p/bitcoin-cash-s-crypto-rating-moved-bch-from-c-to-b-and-one-dimension-paid-for-all-of-it

All grades, tier names, scores, input scores, weights and bands above are CryptoGrade's own published readings, each carrying the date given beside it, and they change.

CryptoGrade publishes investment research, asset grades, scores and ratings. Nothing here is individualized investment advice, and no one at CryptoGrade is a registered investment adviser or broker-dealer. Invest wisely with AI-powered crypto scores and guidance.