The Refusal Ledger
Every ratings site publishes what it knows. We could not find one that publishes
what it does not. This page is that list, and it is generated straight from the
files the data pipelines actually read, so it cannot drift from what we really
believe.
As of 2026-08-26: 14 assets whose yield we will not publish,
16 whose size we will not publish, and 2 products we
carry as retired.
Where an asset appears below, the site shows no number for that field. Not a
stale one, not an estimate, not a dash with a footnote. Nothing.
Why this page exists
On 25 August 2026 we audited every rate and every size in our own corpus. The
results were bad. Sixteen of twenty rate matchers had silently died because
upstream project slugs were renamed, so four of fifty-nine assets were actually
updating and the site had advertised 14.5% on a vault retired after an exploit
for twelve months. Fifteen of sixty-four published sizes were wrong by more than
25%, four of them by over 10x.
Two of those errors were not staleness at all. They were the wrong metric
entirely: BlackRock BUIDL carried Securitize's platform-wide AUM rather than the
fund's, and Usual USD0 carried a collateral-TVL figure rather than circulating
supply. Both numbers were real, current, and correctly copied. They just
measured something else.
We published the audit rather than quietly fixing it, and we built the pipelines
to fail loudly rather than guess. This page is the visible half of that: the
assets where failing loudly means showing nothing at all.
Yields we will not publish
Currency One USD (C1USD)
No C1USD pool on DeFiLlama. Our 7.50% is unverifiable. The auto-resolver matched it to a BSC pool literally called 'U', which is exactly the failure mode this file exists to prevent.
ether.fi ETH (eETH)
No separate eETH pool exists on DeFiLlama. Our 2.31% appears to be a stale copy of the weETH reading. eETH and weETH are the same underlying yield; either merge the two assets or source eETH from ether.fi directly.
Ember rcUSD Vault (rcUSD)
Same Ember ambiguity.
Ember suiUSDe Vault (suiUSDe)
Four Ember USDC pools at 9.84% to 12.90%. One Sui pool does read exactly 12.000%, which may vindicate our figure, but 'may' is not good enough to publish. Needs a pinned pool.
GHO (GHO)
Four Aave GHO pools on Ethereum reading 0.00%, 1.17%, 2.07% and 5.96%. No way to know which we rate. Our published 5.53% matched Fluid's GHO pool, not any Aave market, so the figure was mislabelled from the start. Needs a pinned market address.
Idle Senior USDC (idleUSDC)
Idle has zero pools in the DeFiLlama dataset. Likely wound down. Verify and delist.
Kamino USDC Vault (kUSDC)
No Kamino 'Main Market' USDC pool exists any more. Named Kamino markets range 0.13% to 6.82%. Needs re-scoping to a specific market.
Steakhouse Prime USDC (spUSDC)
The contract address on record (0xbeef0e0834849aCC03f0089F01f4F1Eeb06873C9) does not appear in Morpho's top 1000 vaults at all. Morpho lists a 'Steakhouse Prime USDC' on Base at 0xBEEFE94c8aD530842bfE7d8B397938fFc1cb83b2 holding $21.6M at 4.20%, but our record says $106.79M. The address and the size both disagree, so this needs a human to identify which vault we actually rate.
Pendle PT-stETH (PT-stETH)
Multiple Pendle markets with different maturities. A PT rate is maturity-specific and meaningless without one. Needs a pinned market address and an expiry date shown in the UI.
Pendle PT-sUSDe (PT-sUSDe)
Same. Two Ethereum SUSDE markets at 4.72% and 4.91%, plus others on Monad.
Pendle PT-USDe (PT-USDe)
Same. No matching live market found for our 9.20%.
Solv BTC (SolvBTC)
The only SolvBTC pools are Venus and Tydro COLLATERAL markets paying 0%. Those are not the SolvBTC yield product. Either the yield went to zero or we were never reading the right thing. Treat as an incident and re-score.
Matrixdock XAUm (XAUm)
No XAUm pool on DeFiLlama. Matrixdock publishes its own figures; needs an issuer source.
Yearn USDC yVault (yvUSDC)
Four live Yearn USDC vaults on Ethereum reading 3.43%, 6.30%, 6.53% and more. No vault address recorded, so we cannot tell which one we rate.
Sizes we will not publish
Aave V3 USDC (aUSDC)
We publish $4.80bn. The pinned rate pool holds $174.3M, and the ENTIRE Aave v3 USDC complex on Ethereum in the yields feed is about $235M across four markets. The published figure is roughly 20x anything findable and has no source. This also calls the RATE binding into question: the note on that pool claims it is 'the largest by TVL', which it is not obviously true of. Re-scope both.
Currency One USD (C1USD)
No Currency One USD anywhere in the stablecoin feed or the pools feed. Our $2.87bn is unsourceable, and this is the same asset whose rate the auto-resolver bound to a BSC pool called 'U'.
ether.fi ETH (eETH)
No separate eETH pool; the figure appears to be copied from weETH. Merge the two assets or source from ether.fi.
Ember rcUSD Vault (rcUSD)
Four Ember USDC pools exist at different sizes and we have no vault address on record, so we cannot tell which one we rate.
Ember suiUSDe Vault (suiUSDe)
Four Ember USDC pools exist at different sizes and we have no vault address on record, so we cannot tell which one we rate.
Idle Senior USDC (idleUSDC)
Idle has zero pools in the feed. Likely wound down. Verify and delist.
Kamino USDC Vault (kUSDC)
No Kamino Main Market USDC pool exists any more.
Steakhouse Prime USDC (spUSDC)
The recorded address is not in Morpho's vault list at all. Same block as the rate binding.
Pendle PT-stETH (PT-stETH)
A PT position is maturity-specific, so its size is meaningless without a pinned Pendle market address and an expiry date. We record neither.
Pendle PT-sUSDe (PT-sUSDe)
A PT position is maturity-specific, so its size is meaningless without a pinned Pendle market address and an expiry date. We record neither.
Pendle PT-USDe (PT-USDe)
A PT position is maturity-specific, so its size is meaningless without a pinned Pendle market address and an expiry date. We record neither.
Solv BTC (SolvBTC)
No SolvBTC supply source; the only pools are Venus and Tydro collateral markets.
Staked Sky USDS (sUSDS)
SUSDS shares rate pool d8c4eff5 with USDS. That pool measures the Sky Savings Rate, not sUSDS supply. Our $6.90bn is suspiciously close to USDS's own $6.62bn circulating, which suggests the wrapper's size was copied from the underlying. Needs a separate supply source.
Wrapped Staked ETH (wstETH)
THE STRUCTURAL CASE. wstETH shares rate pool 747c1d2a with stETH, but that pool measures stETH supply ($23.76bn), not wstETH's. Every wstETH entry in the yields feed is a LENDING market (Aave $3.01bn, Spark $2.96bn, Morpho $334M, Fluid $220M) which measures wstETH posted as collateral, and summing those would double-count the same tokens across venues. Our $7.81bn may well be right, but nothing in any feed can confirm it. Needs the token's own circulating supply from CoinGecko or an on-chain totalSupply read.
Matrixdock XAUm (XAUm)
No XAUm pool. Matrixdock publishes its own figures; needs an issuer source.
Yearn USDC yVault (yvUSDC)
No vault address recorded, so we cannot tell which of four Yearn USDC vaults we rate.
Issuer-sourced, and what that means
For 16 assets the only source is the
issuer's own disclosure. Off-chain funds, exchange programmes and allocated
bullion have no on-chain reading to check against. We publish those figures and
label them, because an issuer-published number carries real information. It is
just not independently verified, and we would rather say so than imply a
verification that did not happen.
Retired products
A retired product keeps its page and its history, and shows no live rate or
size. Delisting it entirely would erase the record of what happened, which is
usually the most useful thing about it.
How to check us
Both source files are public in the repository. lib/rate-sources.json pins a
yield source per asset. lib/tvl-sources.json pins a size source and, more
importantly, declares a measure per asset, because three defensible
definitions of TVL can differ by 12x for the same product. Every entry carries
the reasoning, including the entries recording our own errors.
If a number on this site is not in those files, it does not get published.
Generated from source on 2026-08-26. Corrections to research@realworldtokenspace.com.
