A creditor was paid. The debt did not disappear.

From an assumed funding step to a recorded partial refinance

Earlier investigations separated three questions: whether collateral has accepted value, whether it can be released, and whether the necessary debt token arrives in time. They also distinguished a temporary advance from the creditor’s ultimate repayment. The new event supports a stronger conclusion for one named account: a specific funding and repayment sequence actually occurred. Keeping all of that account’s funding wholly hypothetical would now understate the evidence.

The account is 0xea1776733cc969df4ff004a9e1a2d87f5b60add4, the direct-sUSDe Aave borrower already identified in the 2 October investigation. The new result does not select a different borrower or make the old hypothetical USDC example real. It follows the actual USDT account into a new state with both USDT and USDC debt.

A second advance examines an additional million-PT path through sDAI and DAI to Ethereum USDC. Exact successive quantities share one explicitly requested state block. This is useful route evidence, but it remains an indicative comparison: it neither identifies a funded PT owner nor shows an executed cash exit. The two findings have different evidentiary strength and must not lend each other unsupported certainty.

The event and later Ethereum state are 8 hours, 19 minutes, 36 seconds apart. The later quote is not the price executed by the earlier conversion. October 2 and 4 measurements, the 3 October analysis and the 5 October two-corridor limits keep their original inputs and assumptions. Nothing here refreshes every reserve, permission or legal entitlement in the wider assessment.

Basis: newly supplied October 7 financial records and the observation guide. Matched mined logs and historical states are not a separately obtained full receipt or a production security certification.

The actual payment sequence: trace each token and each creditor

The matched transaction is 0x65262caf428ff6865c63b76f6c094ee7c7a70a72a213e5df78f651e5127f012b. All nine retained records agree on Ethereum block 26,138,499, its block hash, the transaction hash and the timestamp. Their log indices establish the ordering. The public explorer link is an inspection locator; the investigation did not obtain a separate full transaction receipt.

USDC and USDT are both displayed in their own token units. Calling them simply “dollars” would hide the different reserve that provides replacement credit. The Aave interface supplies the reserve, debtor and repayer field meanings; actual transfers and adjacent debt states provide the account-specific evidence. The interface is not a complete deployment match.

The source and destination of the two Aave transfers matter. The USDT reserve and its suppliers receive cash. The USDC reserve supplies cash and receives a borrowing claim. Those are different financial effects even though both reserves belong to Aave. Morpho’s principal outflow and return settle its temporary advance, not the borrower’s new USDC obligation.

The internal conversion between the account’s outgoing USDC and incoming USDT was not reconstructed. The records establish the selected cash endpoints, not a complete inventory chain, a particular exchange fee schedule or a future conversion guarantee. No human or legal beneficial owner is inferred from the addresses.

Aave event-field source · Morpho temporary-loan description · Selected transaction and state data.

The account’s 4.688632 USDT contribution is not a fee

The received USDT does not by itself equal the payment to Aave. The recorded amounts differ by exactly the contribution shown below. Adjacent historical states support that interpretation: the account’s idle USDT falls from 9.689386 to 5.000754.

10,001.640594 USDT received + 4.688632 USDT own cash = 10,006.329226 USDT repaid.

This contribution pays debt. It is not gas, a withdrawal charge, a conversion fee or an unexplained loss. The cash record also cannot establish all-in profit: no usable separate receipt, gas payer, sponsorship record or complete service-charge account was obtained. The fact that a conversion delivered slightly more USDT units than the USDC units spent does not settle net economic return.

The USDC ledger balances separately. The account receives 10,000 from Morpho and spends 10,000 through the intermediaries. It later receives 10,000 from Aave and returns 10,000 to Morpho. The sampled pre/post USDC balance is zero in both states. 20,000 USDC of gross arrivals is not 20,000 of independent net funding. The second arrival creates a debt and pays back the first source.

Balance differences are not automatically payment amounts either. The pre/post USDT debt declines by 10,006.047631, whereas 10,006.329226 USDT was transferred as repayment. The 0.281595-USDT difference is retained as an interval accrual/rounding residual, not inferred to be a service fee. The post-block USDC debt is 10,000.000002, two base units above the borrowing event’s 10,000-USDC amount.

At the accepted debt prices returned at both endpoints, reported aggregate debt falls by 5.38082159 USD between the blocks. That is an accepted-value accounting change, not realized profit after costs. The endpoints are twelve seconds apart and can include accrual and other block effects; they are not transaction-local balance snapshots. The matching logs strengthen the attribution of the displayed payments without turning every book-value change into cash.

Exact quantities and the pre/post state are retained in the observation guide. No expense is filled in with a guessed or zero amount.

Follow the replacement creditor into the later state

At 14:04:11 UTC the same account still has 5,359,416.826188716790252832 sUSDe represented by its aSUSDe balance. Its outstanding obligations are 5,866,650.101604 USDT and 10,000.460197 USDC. The later record includes the USDC borrowing flag alongside the USDT borrowing and sUSDe collateral configuration. The current account description therefore needs two debt denominations; earlier USDT-only observations remain unchanged as historical facts.

The later accepted debt total is 5,875,672.29518107 USD and the category remains eMode 31. Health factors above one are favorable observations about accepted-value collateralization. They are not an owed-token payment or evidence that the collateral has been fully released. The small pre/post improvement does not establish a material reduction in total exposure.

The favorable funding conclusion is also narrow but meaningful: this borrower obtained a real repayment resource and returned the temporary principal. It would now be wrong to characterize all funding for this account as unobserved. It would be equally wrong to scale the 10,000-USDC draw into complete-position capacity. The draw was below the account’s available borrowing headroom; it does not empirically establish a refinance beyond that headroom or a standing facility at any other size.

The variable-rate borrowing event and its rate field do not commit terms through a future sale. No fixed tenor or final repayment date follows from the later positive balance. Without complete intervening history, the increase after the event is not attributed entirely to passive interest.

Finally, the zero/zero cooldown fields show no already-running request at the three sampled states. The observed 86,400-second setting does not start a funded wait merely because it was read. Adding one day to the event or research time would invent a request and cash-arrival schedule. A later direct sale could avoid unstaking, but still needs enough admitted final cash and repayment of every surviving funding obligation.

New event/state evidence, not a retroactive change to the 4 October account comparison. See exact observation scope.

A wider PT route reaches USDC through a different savings claim

The separate PT comparison still starts from 1,000,000 PT-sUSDE-26NOV2026 in the previously selected Ethereum market. The new route sells its native sUSDe output for sDAI, previews DAI redemption, then quotes DAI into Ethereum USDC. It avoids an Ethena unstaking step by selling the sUSDe claim. That removes one modeled waiting dependency, not the need to obtain, redeem and convert another claim under applicable conditions.

The selected first-hop pool is 0x167478921b907422f8e88b43c4af2b8bea278d3a. Its reported coins identify sDAI at index 0 and sUSDe at index 1. The sDAI asset getter identifies DAI; the final 3pool identifies DAI at index 0 and Ethereum USDC at index 1. The denomination chain is grounded in those specific reads rather than inferred from names.

The successive inputs retain full integer precision and the same explicit block. Rounded display quantities were not fed into later enquiries. That is stronger measurement consistency than combining unrelated amounts from drifting states. It remains a set of views, not an executed package, reserved future prices or an admitted financing commitment.

Pendle distinguishes net token output from the SY amount redeemed to produce it. These are not two balances to add together, and a fee already deducted from the net quote is not subtracted a second time. Other execution, financing and service costs remain unknown. Spark’s documentation identifies the observed sDAI address as a DAI savings receipt offered through Sky’s savings mechanism, not an Ethena payment or a corporate guarantee by Spark.

The selected pool holds 290,286.893683513943564680 sDAI at this state; the quoted output is approximately 99.9338% of that inventory. The final DAI/USDC pool holds 45,789,752.349654 USDC. A large final-pool cash balance therefore does not by itself enlarge the earlier receipt-token output. Neither inventory is owned repayment funding assigned to the PT holder.

The preview implies approximately 1.1814448464 DAI per sDAI for this quantity, rather than one. The route uses that specific preview instead of a new par assumption. But a favorable share-conversion ratio cannot compensate for not obtaining enough shares. sDAI, previewed DAI and final USDC are successive forms of the same route resource, not three reserves.

Registry discovery returned four candidates, not four established cash paths. One other full-budget enquiry returned only 32.557419537236092841 sDAI; another requested quote reverted and remained nonnumeric; another was not carried through a full final-token route. Neither the small quote nor a failed call is added to the selected path. A rejected enquiry is not a zero-liquidity finding.

This is an additional route, not an additive correction to the earlier two-corridor envelope. Adding a new whole-input quote to the old whole-input amounts would spend the same PT budget again and mix state dates. The old restricted assumptions remain intact. No optimized split, best-execution guarantee or all-market upper bound is claimed.

For the retained hypothetical 900,000-USDC principal, the selected path is insufficient even before extra costs. That bounded negative says nothing about every possible route, a lower actual debt or an independent contribution. The observed 10,000-USDC Aave refinance belongs to a different account/obligation case; it cannot supply financing to this hypothetical PT holder by association. An actual owner, credit interval, applicable redemption conditions and final receipt remain unestablished.

Pendle net-output definitions · sDAI claim and preview · Indexed quote interpretation · Exact amounts and pool identities.

Base: a remeasured obligation still needs an identified payer

The selected Base borrower remains 0x5afe2414f865cbc5ff4e25979996c5b1252e0002 in the same USDe-collateral/USDC-loan market. At Base block 52,296,246, 14:03:59 UTC, its collateral is 3,042,881.283763481768838477 Base USDe. Upward-rounded stored-share debt is 2,707,724.528483 Base USDC, using the retained virtual-share convention.

The market’s recorded accrual is 34 seconds older than the batch clock. The calculated amount is not an executable future payoff invoice; later settlement can require a different amount. No guessed rate or unobserved repayment is added to make the stored figure current at another time.

The inspected direct Base-USDC balance, free Base-USDe balance and supplied shares in this same market are all zero. Those are scoped observations, not failed requests converted into zeros. They exclude those particular objects as positive owned resources at that state, but do not establish complete wealth or rule out another wallet, other markets, a sponsor or an admitted facility. Matching borrow-share counts at the October 4 and 7 endpoints do not prove an absence of repayment and redrawing between them.

The market’s nominal supply-minus-borrow amount is 38,816,025.713668 USDC. It does not identify a willing lender, the account’s admitted draw, net proceeds after fees or the replacement obligation. Similarly, October 4’s protocol-held inventory was not assigned borrower cash. The actual Ethereum refinancing cannot pay this Base loan simply because both networks use a token named USDC.

The funding branch therefore remains open: a sufficient, entitled same-network resource or a supported final sale must still be identified for complete repayment. That is narrower than saying no Base refinancing or alternative route exists. Earlier unsuccessful route enquiries retain their original dates and do not become new zero-proceeds findings.

New Base borrower records and the retained location-specific payment boundary. See the clocked resource screen.

A larger preview and an empty event window are not a net receipt

The selected shareholder remains 0xe5a8476199a4e213fe594d1a2f5702f06f26a776. Its measured raw share balance is 2,963,919,352,231,113,055,472,892; the new preview is 3,001,245.554234 USDC. The increase of 778.879531 USDC from the October 4 preview is a claim comparison, not realized interest, cash received, an expense or a reconstruction of all intervening accounting changes.

A successful owner-indexed standard Withdraw query returned no matching records over Base blocks 52,294,247–52,296,246 inclusive. This is a genuine bounded negative: one event type, one indexed owner and 2,000 blocks. It does not establish that a withdrawal was attempted and failed. Nor does it provide lifetime history, exclude every transfer or in-kind route, or identify receipts outside the chosen signature and period.

The positive ordinary sender and receiver findings from October 2 remain supported at that original date. They were not remeasured by the new event query, and an empty later window does not reverse them. The newly observed native-ETH balance is zero, but no conclusion about absent sponsorship or the ultimate gas payer follows from that direct balance alone.

A separate historical request for the denominator needed to attribute fee dilution at Base block 52,154,674 failed with HTTP 403. It produced no numeric fee observation. The October 4 other-holder payment and performance-fee share creation remain their own event, and the October 5 fee interpretation remains conditional. Neither can be charged wholesale to this shareholder or relabelled as its gas.

A receipt-level net payout would require the selected holder’s cash receipt, share/asset change, attributable expenses and competing cash uses over the relevant interval. An earlier payment already precedes a later stock; subtracting it again would double-count the outflow. This continuation does not provide the missing full-period cash/competition ledger. Missing expense evidence is not zero expense, and it is not evidence of a failed withdrawal.

Positive October 2 ordinary access · Other-holder payment and fee shares · New shareholder query and claim.

Who benefits from the payment—and who still bears the risk?

The original USDT repayment is a stabilizing counterflow for that reserve. Suppliers there receive underlying cash for the repaid slice rather than merely a promise of eventual collateral recovery. At the same time, another reserve supplies USDC and holds a continuing receivable. Calling the whole sequence an undifferentiated improvement in “Aave liquidity” would conceal both the denomination and the creditor.

An unleveraged holder may have time to wait or choose another route; a creditor with an earlier obligation may not. A fixed-block quote can show that a named amount misses an illustrative debt without proving the underlying asset is worthless or requiring every holder to sell. Conversely, a positive health factor cannot substitute for a funded final-token arrival.

The whole-system boundaries remain essential. Issuer backing claims are not personal assets of these borrowers. PT, SY, sUSDe and USDe are not separate reserves to add together; neither are sDAI, its DAI preview and final USDC. The foundation and backing-recovery investigation preserve the ownership, location and timing distinctions behind those conclusions.

Exceptional oracle response, authority changes, bridge continuity and PSM funding remain conditional. A natural refinancing event, an ordinary quote or a normal setting does not establish those protections. The new evidence strengthens the financial account of one partial repayment; it does not complete deployment assurance or a system-wide solvency assessment.

What is now established—and what would overstate it

The full funding and recovery investigation remains partial. The progress is not merely another quote: an actual named-account funding sequence is observed. The remaining gap is also not merely a missing timestamp: most original debt and replacement debt persist, and complete final settlement has not been traced.

The evidence is provider-mediated. Matching transaction/block hashes, log order, reserve fields and adjacent states substantiate the described cash sequence, but no separate complete receipt or independent finality certification was obtained. C0/C1 are end-of-block observations rather than internal transaction traces. The broader intervening history is incomplete, and the internal conversion intermediary chain is not reconstructed.

The historical access error, rejected candidate quote and initially unsuccessful metadata call remain unavailable or nonnumeric where appropriate. A subsequent documented token-binding enquiry supplied the route identity; none of these acquisition outcomes is interpreted as a contract-security defect. Research’s offline consistency checks support decoding and bookkeeping at the saved inputs, not successful future execution.

No transaction instructions, financing recommendation or promise of par exit follows. The decision-changing next evidence would link a whole obligation to final settlement with attributable costs, rather than scale a partial draw or relabel an indicative amount as committed liquidity.

Source trail and dated observation data

The reader dataset below provides selected token movements, state clocks, route quantities and scope limits. It is a companion for inspection—not an additive balance sheet, live feed or replacement for the full historical investigations. Numeric token amounts and large integers are strings so readers can preserve precision. An explorer’s default current view may differ from the stated historical block.

Download October 7 observations (JSON)

The newly inspected public documents have an October 7 access date. That date does not refresh their entire contents, older balances, maturity records, normal admission or exceptional safeguards. Source descriptions, saved observations and conditional financial deductions remain different evidence.

Earlier exact reader data remain available as the 20 September backing-recovery dataset and 27 September integration dataset. They are not October 7 measurements. Read the October 4 chapter and October 5 analysis at their original scopes, or return to the complete research catalogue.