Matching quantities is not the same as funding an exit

A consistent amount is stronger evidence—not a payment

The new evidence makes three different questions easier to separate. Does each conversion use the amount actually returned by the previous leg? Does the resulting asset cover the creditor’s whole claim? And has the asset actually been paid? The selected principal-token paths now satisfy the first measurement condition at one fixed Ethereum block. The full Aave comparison follows the actual USDT obligation. A separate Base transaction supplies a naturally observed payment to another vault holder. None makes the remaining funding conditions disappear.

For a 1,000,000-PT input, two named paths end at 329,793.046234 USDC and 309,302.853290 USDC. Both fail the retained hypothetical 900,000-USDC principal comparison before additional interest and external costs. These are alternative uses of the same token quantity, not proceeds that can be added. The two paths and their limits.

The already identified Aave borrower now has 5,873,310.573496 USDT of debt and positive accepted-value health factor. A conditional sale of the entire converted collateral plus free USDe, with idle USDT credited once, leaves a large named-route terminal gap. That is a statement about the measured outlet, not observed bad debt or proof that no outside resource exists. Whole-account comparison.

The natural vault record establishes something different: a 24,663.478510-USDC token payment to another holder’s receiver. A simultaneous fee-share mint creates a separate claim. It is not another cash outflow and does not reveal the selected shareholder’s gas, sponsorship cost or eventual net payout. Cash payment versus fee claim.

Two paths for one million PT now use matching quantities at one block

The instrument remains PT-sUSDE-26NOV2026, with maturity on 26 November 2026 at 00:00 UTC. Its previously identified Ethereum market again returns the same standardized-yield token (SY), principal token (PT) and yield token (YT). The price enquiry uses one million PT, not a small quote scaled upward. It returns 791,223.364535577181101772 sUSDe. A second SY-related output field is not a second asset to add to that amount.

The decisive measurement improvement is the common state. The initial quantities are observed at Ethereum block 26,117,487, 4 October 2026, 07:26:23 UTC. The subsequent financial batches explicitly request that block; their internal clocks agree even though the provider’s outer node height is later. Each next quote receives the exact integer output of the previous leg. This replaces a cross-block composition for the new comparison without rewriting the older evidence.

The first path is not a current receipt of USDe. The asset-conversion view supplies an accounting quantity. An eligible request must be initiated and completed, and the assets must actually be released before that USDe can be sold. A quote obtained now does not survive the waiting interval by assumption. The second path instead sells the native sUSDe output through crvUSD and then USDC. It avoids that particular unstaking step, not every admission, funding, market or cost condition.

The native route identifies where the local constraint sits. The first pool, 0x57064f49ad7123c92560882a45518374ad982e85, has observed coins crvUSD and sUSDe and 309,565.398356 crvUSD of token inventory. The next pool, 0x4dece678ceceb27446b35c672dc7d61f30bad69e, has USDC and crvUSD and 5,898,628.986674 USDC of inventory. A sizeable final USDC pool does not make the upstream native-sale output larger. These inventories are context for the quotations, not extra proceeds to add to them. Exact inputs and coin bindings.

This is also a bounded route screen. A registry result whose first two coins were crvUSD and DAI was not adopted as a verified two-coin sUSDe pool. Ten discovery results were not counted as ten measured cash sources. Pendle’s description of broader routing and off-chain limit-order liquidity is a reason not to call these static comparisons best execution. It does not supply a matched orderbook, an admitted buyer or a firm offer. Published routing scope.

The negative finding concerns the named paths, amount and state. Neither final quote can cover 900,000 USDC even with zero additional financing interest and external costs. The debt is illustrative, not a newly discovered PT borrower’s loan. A lower obligation, additional cash, a different split route or a firm counterparty could change the case. No realised loss against a known acquisition cost, issuer insolvency or market-wide depeg is established. An unlevered holder is not compelled to sell merely because this sized route is unattractive.

A common quote block does not create a funded waiting interval

The sampled normal cooldown-duration field remains 86,400 seconds. It is a setting, not a measured total time from pledged PT to USDC payment. The sequence can include debt repayment, collateral release, a qualifying staking request, the applicable wait, claiming and final conversion. No selected PT holder’s advance, request timestamp, agreed financing tenor or final cash-arrival deadline has been established.

For the sampled Aave address, the newly read cooldown fields are zero end timestamp and zero underlying amount. There is no recorded already-running request at that address from which to derive a cash-arrival schedule. This does not exclude another address, delegated arrangement or a later request. It prevents the narrower error of treating the research timestamp as the start of someone’s funded one-day wait.

The old issuer guide’s seven-day description remains a documentary statement; it does not override the measured one-day field. Neither the guide nor the getter certifies future configuration, exceptional operation or a maximum payment delay. Likewise, a successful empty event query covers only its stated 2,000-block window, while the unavailable wider query establishes nothing about events in that wider window. Neither proves that the PT market has no borrowers or financing arrangements. Duration and documentary limits.

The fixed-block views improve quantity consistency, not the feasibility of an unchanged-state transaction sequence. In particular, prior funding drawn from one of the quoted pools would change that pool’s later cash. The untouched quote cannot be credited again as though the first movement never happened. The temporary-versus-final funding analysis therefore remains relevant: paying the original lender with an advance leaves the advance’s own repayment outstanding.

The full Aave account owes USDT—not its accounting dollar value

This is the same activity-selected Ethereum account, 0xea1776733cc969df4ff004a9e1a2d87f5b60add4, established in the October 2 investigation. It is not a new hypothetical USDC-debt account or either of the smaller September USDe-debt examples. No beneficial owner, trading intention or representative-sample status is inferred.

The component token balances and accepted prices reconcile with the returned account totals. The positive health factor and available underlying quantity are useful favorable evidence. They do not establish release of all pledged collateral before debt payment, unrestricted withdrawal in every state, or conversion of the accounting value into the owed token at the necessary size. USDT’s dollar mark affects valuation; it does not turn a USDT obligation into USDC already available for payment.

For the conversion-through-USDe comparison, the whole recorded collateral quantity is supplied to the asset-conversion view. The resulting 6,706,342.569439 USDe-equivalent and the account’s 23,908.467042 free USDe form one combined conditional input of 6,730,251.036481 USDe. A single quote for that combined quantity returns 329,886.694793 USDT. Crediting the existing 9.689386 USDT once gives 329,896.384179 USDT.

Against the recorded debt, the supplied analysis calculates a 5,543,414.189317-USDT terminal gap, before subsequent interest or external costs. This requires successful release and conversion, realization of the reported conversion quantity, continued usability of the named route state and no additional resource. The selected USDe/USDT pool holds only 330,037.411541 USDT at the common block. The full Aave and PT enquiries do not yield proportionally scaled outputs; the investigation does not assign a cause to that nonlinear result or simulate repeated slices. Combined quantity and quoted output.

A second, collateral-only native-sale comparator reaches the actual owed token: 5,359,416.826189 sUSDe → 309,362.098311 crvUSD → 309,265.705257 USDC → 309,274.179330 USDT. The last quote uses the exact preceding USDC quantity and the observed USDC/USDT coin identities. Stopping at USDC would leave the loan’s denomination unresolved.

That native alternative leaves the account’s free USDe unconverted; it is neither written off nor credited as already paid USDT. The two alternatives cannot be added. Avoiding the modeled staking step changes the timing dependency but does not make the quoted collateral-only output sufficient for the whole debt. If an independent funder pays Aave first, the original suppliers receive cash while the funder becomes exposed to the remaining sale, timing and cost conditions.

A natural payment is stronger than a preview—and narrower than every holder’s recovery

The Base evidence begins with a bounded outflow query, not a withdrawal-rate estimate. It returned 35 USDC transfer logs from the selected vault over blocks 52,153,692–52,154,691. Twenty-nine went to the previously bound adapter. Those internal-link movements were not labelled shareholder payouts. All six other transfers were retained; the largest was selected for a transaction-level event match.

At Base block 52,154,675, 4 October 2026, 07:24:57 UTC, a Withdraw event, a burn of 24,363.064586372408152658 vault shares and a 24,663.478510-USDC transfer agree on the payment receiver and share owner. The selected transaction hash and block hash agree across the two provider-mediated log queries. This is scoped positive payment evidence, rather than another preview.

The share owner is 0x2f3a43fd3c0c79fd8be9fd527eb4dd9db3e9ceab; its receiver is 0x2a60fbd700b6f49b827a2dfb8dcf9b6cdc19273d. Neither makes this a payout to the selected approximately three-million-USDC shareholder. No independently retrieved receipt, gas/L1-fee record, sponsor reimbursement or full net-expense record was obtained. A recorded underlying transfer is not automatically net of costs paid in another token. Matched transaction and fields.

Fee-share creation is a different financial object

An AccrueInterest event in the same transaction reports an accounting-asset increase of 13.399412 USDC, a performance-fee mint of 3.309050057993409206 shares and zero management-fee shares in that event. A transfer from the zero address to fee recipient 0x255c7705e8bb334dfcae438197f7c4297988085a matches the performance-fee amount. The published event interface supports that field/topic/value match, not certification of the entire deployed system.

The fee shares can become a later claim on assets. That is why they matter economically even though they are not cash paid out now. The data needed to attribute precise dilution to the selected holder—particularly the pre-event denominator and applicable fee state—were not obtained. Its unchanged share balance does not prove zero fee burden; the available event also does not justify charging the entire mint to that holder. Published fee-share distinction and scope.

The event is already before the new snapshot

The payment precedes the new Base state by 16 blocks and 32 seconds. Subtracting it again from a later observed stock would manufacture an extra withdrawal. Nor does gross outflow alone identify a net liquidity change without inflows, allocations, borrowing and repayments. The matched deallocation event does not establish which underlying market supplied this payment; no selected USDe-market, issuer-recovery or other-market funding attribution is made.

The six transfers outside the adapter link total 32,165.478837 USDC, but only the chosen transaction received the full withdrawal/burn/payment match. That total is not a verified redemption queue or projected withdrawal rate. The useful conclusion is narrower: one other holder received a matched payment, while the conditions for a later claimant remain separate.

The selected shareholder keeps its positive admission evidence—not an invented receipt

The selected address remains 0xe5a8476199a4e213fe594d1a2f5702f06f26a776 in Base vault 0xbeeff2490feffa212fac2f6553682c219e6a8845. At block 52,154,691, 07:25:29 UTC, its 2,963,919.352231113055472892 shares preview 3,000,466.674703 USDC. The share quantity is unchanged from October 2; the preview is a new accounting view, not a cash receipt.

The positive October 2 ordinary sender and receiver checks remain established at their own dates. They are not reversed into rejection merely because this continuation does not repeat them. The earlier allowance and normal-cash comparisons also retain their own scope. A new preview cannot transfer another holder’s transaction permission, priority, actual costs or liquidity reservation to this address.

The directly measured native ETH balance is zero. That does not rule out another legitimate initiator or sponsor, nor establish who would bear or later recover the expense. A complete net-payout assessment still needs the relevant costs and timing, not a zero-fee assumption.

The adapter reports seven markets in the new state. A complete seven-market claim/cash reconciliation was not performed. The five-market October 2 and eight-market September 28 tables must therefore stay dated; neither is silently relabelled as the current portfolio. Similarly, the in-kind exit analysis still separates receipt of a direct lending claim from USDC paid. An observed payment and an in-kind alternative answer different questions.

Base: a large protocol balance is not this borrower’s credit line

At Base block 52,154,691, 4 October 2026, 07:25:29 UTC, the selected borrower has 3,041,694.618425122434917758 USDe collateral and calculated stored-share debt of 2,706,734.095152 Base USDC. The sampled ordinary price remains one USDC per USDe, with 91.5% liquidation threshold and accounting health factor 1.0282319866. These are ordinary financial observations, not verification of an exceptional oracle transition.

The same batch finds 234,176,758.575636 USDC at the Morpho core, a 38,897,920.872781-USDC stored supply-minus-borrow difference in the selected market, and 459,215,884.166077 USDC of vault accounting assets. These are different objects, not three independent buffers. No evidence assigns any of them to this borrower as an unencumbered, admitted source for the whole repayment.

The narrower owned-resource screen is negative for the inspected objects: zero direct USDC, zero free USDe and zero supplied shares in this same market at the selected address. That does not establish absence of other wallets, other-market positions, custody resources, sponsors or facilities. The missing link is from the actual resource through entitlement and permitted use to a timely Base-USDC payment—not a lack of large numbers elsewhere in the system.

The debt uses matching stored market totals and upward share conversion. The market’s accrual timestamp is 32 seconds before the batch clock. It is not a payoff invoice for future inclusion; no unobserved interest or fee is inserted. The public request for a route using the full recorded collateral amount returned no usable data. That is an unresolved broader route, not a zero-output quote. Stored-state calculation and exact bindings.

The October 2 three-pool bound remains its original dated result. It is not refreshed into an October 4 whole-Base liquidity estimate. A documented same-transaction temporary loan can alter payment order, but principal returning to its lender cannot also count as permanent repayment proceeds. Ethereum quotes, issuer-owned recovery assets and a possible PSM are not imported into this borrower’s cash column. Temporary-funding scope.

Name the creditor and the event before describing recovery

A recoverable temporary discount can favor a funded holder who avoids a forced sale. A native-collateral sale can avoid one waiting step. The positive account headroom, ample underlying collateral quantity, observed final-pool inventory and real vault payment are useful evidence for those narrower propositions. None commits the resources needed by every creditor together.

If value is genuinely impaired, postponing sale or changing the wrapper cannot create the missing value. An accepted mark can preserve borrower optionality while suppliers or a replacement lender remain dependent on later realization. Conversely, a thin named outlet is not proof that all collateral is worthless, and an unavailable fee record is not proof that a withdrawal failed. The supplied analysis keeps those cases separate rather than selecting an unverified exceptional oracle, bridge or authority response as a guaranteed safeguard.

The whole-system foundation and recovery investigation retain their original boundaries. A lender’s receivable, a receipt token, its underlying and a cash inventory are not additive backing dollars. No ownership or financing agreement linking the sampled accounts, no reconciled issuer coverage ratio and no funded emergency backstop is inferred.

What the October 4 evidence closes—and what it leaves open

Quantity consistency has advanced; financing remains partial. The new named PT paths have exact successive quantities and a common observation block. The actual Aave account is measured at full size in its owed token. The natural vault payment is matched to a withdrawal and burn. Those are substantive results, not just a more recent date.

A named-route negative can be conclusive within its stated amount and assumptions without proving that all alternatives are absent. Failed access to a wider route, historical fee denominator or receipt does not establish a zero balance, no facility, no sponsor or a failed transaction. The new snapshots also do not independently verify source/build correspondence or make old observations current.

Evidence, dates and public source trail

The observations below are the recorded provider-mediated results used in this chapter. Explorer links identify the contracts, blocks and transaction; opening them does not guarantee reproduction of the saved historical view. Public documentation supports only its stated financial definitions. A source describing execution is not evidence that the described operation occurred here.

Public documentation inspected for the 4 October investigation

These descriptions identify the inspected source versions and their limits. Retrieval on 4 October does not refresh the underlying date of older explanatory material. No new build, receipt or security assurance follows from the documentary links.

Earlier material and exact reader downloads

The October 2 account/cash chapter, October 3 repayment/maturity chapter and October 1 frozen-input financing chapter retain their original evidence dates. The 20 September recovery data and 27 September integration data remain unchanged downloads; neither contains refreshed October 4 observations.

Saved-input checks distinguish new payloads from retained historical records and verify quantity, denomination and event matches. They do not establish endpoint consensus, future execution or loan availability. A wider event request, pre-event fee-denominator request and receipt retrieval remained unavailable. Those gaps are not represented as zero results, failed withdrawals or absent facilities. No security test or financial transaction was performed by the investigation.