A local quote is not yet a funded exit

Follow the earlier payment, not just the later asset

A useful local quote is not yet a financed exit. Before treating one as repayment capacity, identify the debt token that must arrive first, the token the quoted route actually pays, and the creditor who is still waiting afterward. The same position can have positive accounting equity, a useful small sale, and no established funding for its complete exit.

“Cash” here means the token needed for the next payment at the relevant account and network, not automatically bank dollars. USDe is distinct from its staking-vault shares, sUSDe, and from USDC. A principal token (PT) is a maturity-specific claim—not the same asset as any of those payment tokens.

The 28 September evidence makes four previously open questions more concrete. A Base borrower now has a same-chain, sized USDe/USDC quote alongside its debt calculation. Two real Aave accounts replace a generic account discussion with measured collateral and USDe debt. The active November principal token has a matched pre-maturity market, but its quotes deliver sUSDe. The selected Base vault’s eight allocated claims reconcile, while only part of their value enters a conditional cash screen. None of these results demonstrates that a customer completed the entire payment chain.

The positive findings matter. A smaller release can avoid a need to prefinance the entire Base debt, and liquidity outside the dominant market can materially reduce an assumed supplier gap. The limitations matter just as much: a small quote cannot be repeated indefinitely by assumption, and a portfolio total cannot be presented as the amount one shareholder can withdraw now.

The 27 September chapter remains the foundation for accepted collateral, releasable collateral and repayment cash. The new result below narrows its missing local-sale, account, PT and supplier-resource questions rather than erasing that earlier evidence.

Base: how much USDC must arrive before collateral can leave?

The same activity-selected borrower remains in the Base Morpho market lending USDC against USDe. It is not a largest-borrower ranking, a representative sample or an attribution of the wallet’s beneficial owner. At Base block 51,912,543, the position, market accounting, normal oracle output and local quotes were returned together. The accepted price was one USDC per USDe and the liquidation threshold was 91.5%; neither proves how the oracle will behave during an exceptional event.

A health factor compares threshold-adjusted accepted collateral value with debt. A value above one passes that financial adequacy comparison; it does not make every unit of collateral free or fund the next payment.

Debt is calculated from the account’s borrow shares and same-batch market totals, using the published share-conversion convention with one virtual asset, one million virtual shares and upward rounding. A borrow-share integer is not itself a human-unit USDC balance. The accounting convention is inherited; the new financial reading does not independently establish build equivalence or every repayment-path rule. Earlier convention and account-selection basis.

For collateral q, debt d, accepted collateral price p in USDC per USDe, liquidation threshold l, and a release budget x, the supplied financial comparison uses:

These are necessary financial bounds, not executable withdrawal limits. Exact equality leaves no margin for later interest, rounding or price movement. Permissions, actual account rules, transaction inclusion and fees remain additional conditions. A positive bound is evidence of financial headroom, not a guarantee that a withdrawal succeeds or a sensible target for action.

Earlier on 28 September, at Base block 51,912,061, this address’s free USDC and USDe balances both returned zero. Those observations do not establish the owner’s total assets, other wallets, credit lines, pledged resources or incoming transfers, and they are not silently refreshed to the later quote block. No committed third-party financing for this position was identified.

Saved position/market/price comparison: Q2 records Q210–Q212. Earlier idle balances: F07–F08. All observations are provider-mediated; included block times are not measured retrieval times. Observation scope and source trail.

A quoted input budget is not necessarily an amount sold

The new local route is the Base Uniswap V3 USDe/USDC pool 0xedaf6ca46fb852d4ab0a2e9449d267cf03213f05. The official Base factory and QuoterV2 locators, factory lookup, and pool token/factory/fee readings agree. Its fee value of 500 represents 0.05%, not 500 basis points. This is a same-chain route for Base USDC debt, not an Ethereum quote imported as local liquidity. P01

A preceding pool-balance observation at Base block 51,912,502 found 785,959.347619 USDC and approximately 214,446.95059945 USDe. Inventory is not evenly distributed execution depth. The same-batch quotes and borrower comparison at block 51,912,543 provide the more relevant size-specific evidence:

The million-unit result returns square-root price 4,295,128,740, exactly the published minimum plus one used as the default extreme boundary when the caller’s price-limit field is zero. QuoterV2 returns output and post-quote price, but not consumed input. Consequently, one million USDe is the offered budget—not an established amount sold in full. No leftover quantity was returned. P02 · P03

That missing quantity is not written off, treated as cash, or silently counted as newly pledged collateral. It prevents a realized-price or loss conclusion for a completed million-unit sale. The useful supported statement is narrower: this enquiry does not establish enough USDC to repay the earlier financing needed to release the stated budget. There is no demonstrated 21.4% realized loss, full-position insolvency, issuer impairment or global USDe depeg in these observations.

All quote rows are independent enquiries, not sequential fills that consume the pool and can then be added together. A zero supplied price-limit field is not a protective minimum-received order. Not hitting the extreme boundary on a small quote also does not establish future execution, admission, inclusion cost or unchanged liquidity. No recommendation to accept any quoted terms follows.

The larger full-position enquiry failed with success=false and 0x; no reason was supplied. It is preserved as unavailable output, not interpreted as a token worth zero, a transport failure or proof that every alternative route fails. No better-route or market-wide capacity claim is established either.

A finite supplementary screen found zero actual USDe/USDC balances at an identified six-coin Curve pool, about 12.042525 USDC in a classic Aerodrome stable pool, and about 72.328973 USDC in the sampled Slipstream pool. Those particular resources do not substantiate another large buffer. The screen does not exhaust fee tiers, multihop routes, aggregators, private counterparties or all deployments. P06 · P07 · P08

Quote inputs and returned fields: Q205–Q209. Same-call pool binding: Q203–Q204. Pool inventory is a preceding observation, not the quote batch. The failed quote and unknown consumption remain explicit. Observation scope and source trail.

A smaller sale can help; a prior lender can still be exposed

The 50,000-USDe budget is below the sampled no-paydown bound. The supplied analysis therefore describes a conditional smaller deleveraging sequence without assuming the borrower first borrows the entire 2.705 million USDC debt. It is not an executed route or an instruction to repeat slices.

The cash is counted once: pool USDC decreases, borrower cash passes through, and lending-market USDC increases by the same amount. Repayment converts part of an existing receivable into cash. It is not a new reserve on top of that receivable, an automatic PSM contribution, or a second liquidity resource labelled “arbitrage.” Deducting actual costs from the sale cash would reduce debt repayment.

This is meaningful counterevidence to an overly strong claim that no deleveraging is possible without all of the debt funded externally. It is not evidence that every later slice gets the same price, that a buyer finances all collateral, or that the full position carries the small-slice average value. The research neither prescribed nor simulated repeated execution.

A larger release changes the funding order. At the sampled state, releasing a 100,000-USDe budget requires at least about 15,585.46 USDC first. Releasing a million-unit budget requires about 839,085.46 USDC first. The cash quoted afterward does not fund that earlier obligation by travelling backward in time. A funder, other available asset, or independently evidenced alternative sequence is needed before those later proceeds can be relied on.

This is the same creditor-ordering principle as the 27 September assumed term-funded Aave illustration, now constrained by a measured Base route. It does not make that earlier term loan an observed facility, and it does not supply an external commitment for this Base borrower. A funded alternative must identify who provides the correct token, under what terms, and before which deadline.

The 50,000-unit full-fill/no-intervening-change comparison is the supplied conditional case using Q206. No actual repayment, sale, withdrawal or funding commitment was recorded. Observation scope and source trail.

Aave: two measured accounts owe USDe, not USDC

The account question has advanced, but in a different debt denomination from the earlier example. A bounded natural sUSDe-supply window identified two Ethereum accounts. Later financial reads found sUSDe collateral and USDe debt in both. They are activity-selected examples, not a ranking of the largest or most fragile borrowers, and no identity or intention is assigned to the addresses.

User configuration identifies reserve 30 as borrowed and reserve 32 as collateral. Fresh reserve-address observations identify them as USDe and sUSDe. The observed USDe variable-debt amounts reconcile to reported account debt values; sampled USDC and USDT debt-token balances were zero. USDC must not be substituted for the token these accounts owe. P09

Both use Aave’s efficiency-mode category eMode 32, with effective 92% LTV and 94% liquidation threshold; the returned bonus factor is 102%. The category name includes a PT maturity, but its current bitmaps—not the name alone—include sUSDe collateral and USDe borrowing. These are not a refresh of category 2’s settings for every account, and a bonus does not establish that a liquidation is funded.

Aave’s accepted prices in the final comparison were $1.24997886 per sUSDe and $0.99969009 per USDe, with a same-batch staking conversion of 1.250366364562567020 USDe per sUSDe. Those are accepted prices and accounting conversion, not market-sale receipts. Using the actual owed asset and effective threshold gives:

For AV2, the approximately 45.5182 USDe needed before a 50-share release cannot be financed by merely pointing to the later approximately 62.5183-USDe accounting claim. The shares must first become free and then complete the applicable staking path. A term lender could bridge that timing if actually willing and funded; a conventional same-transaction flash loan cannot simply remain unpaid through a later cooldown.

For full repay-first release, the starting obligations are about 311.5015 USDe and 413.3412 USDe. The released shares’ sampled accounting equivalents are about 481.8339 USDe and 453.8194 USDe. Positive accounting residuals do not establish an instant conversion, a term facility or issuer admission. Nearby idle-token observations found no free USDe, sUSDe, USDC or USDT in those addresses, but no complete external wealth or financing inventory was obtained.

Relative account health and the lender’s dollar exposure are different

In the supplied conditional comparison, lowering the accepted dollar marks of both sUSDe collateral and USDe debt by 5% leaves their relative health factors unchanged. A USDe supplier can nevertheless own a token worth fewer dollars. A dollar-funded intermediary may face that value or currency mismatch even when the protocol ratio remains stable. This is not a prediction that the installed feeds move together.

Reducing only collateral’s accepted value relative to debt by 5% produces approximately 1.38130218 for AV1 and 0.98045083 for AV2. These are hypothetical valuation branches, not a realized discount, observed liquidation or a USDe spot-trigger estimate. Likewise, holding the separately sampled 6.32203% annualized USDe debt rate and conversion ratio fixed for one-day arithmetic consumes headroom; it is not a future rate or staking-yield promise.

Final account, accepted-price and ratio evidence: S2, Ethereum block 26,077,313 at 17:01:47 UTC. Category, idle balances and the 86,400-second cooldown were checked separately in R/S1. Natural logs provide selection, not independently reconciled funding histories. Observation scope and source trail.

The active PT now has a sized sale—but the output is sUSDe

The selected principal token remains PT-sUSDE-26NOV2026 on Ethereum, maturing on 26 November 2026 at 00:00 UTC. A bounded transfer window led to AMM 0x47ad2cd1dd15739a7a035b9d3b7828d916fef77e; its own token tuple and expiry matched the specific PT, SY and YT. This closes the earlier missing AMM identity for this instrument without substituting another maturity or implying an Aave PT listing.

The official Ethereum deployment manifest supplies the static pricing router, whose published interface identifies net output and fee fields. At Ethereum block 26,077,288, approximately 58.294 days remained to maturity. Independent quotes for free PT returned the following net sUSDe amounts. P10 · P11

The last column multiplies quoted sUSDe by the same-batch SY exchange rate of 1.250365749456529343. It is not USDC, dollars, a second sale quote or a promised future unstaking output. Because the interface’s output is already net, the separately returned SY fee is not deducted a second time.

At the same observation, PT supply was approximately 15.849 million, with approximately 1.064 million PT and 2.376 million SY at the AMM. These are inventory/accounting quantities, not proof of usable depth. The difference between a PT’s pre-maturity accounting equivalent and one USDe can include time value, market pricing, fees and size effects; it is not all issuer loss or all slippage. The four independent quotes cannot be summed as sequential fills.

The PT’s lending market still has a separate USDC loan. A borrower must first release encumbered PT; a free holder avoids that particular debt-release step. After the quoted AMM sale, sUSDe still needs a funded sale or a financed staking withdrawal and later USDe-to-USDC conversion when USDC is the obligation. The earlier accounting-unit and maturity analysis remains valid: maturity changes a claim path, not the currency or existence of an independent loan.

The remaining cash leg was screened, not assumed

Four directly identified Ethereum Uniswap pools were checked after the PT quote: two sUSDe/USDe pools and two sUSDe/USDC pools at sampled 0.05% and 0.01% tiers. The two USDC inventories were 2.761749 and 0.039933 USDC. The USDe-pair inventories were about 0.0044631954 USDe and zero. A separate query in one Curve factory returned no matching pool at the queried index. Those particular balances do not substantiate a large final cash leg.

This is a finite screen, not the complete sUSDe liquidity universe. It excludes neither other fee tiers or venues, multihop routes, institutional counterparties nor a properly financed wait. No funding commitment for that wait was identified. The new evidence therefore establishes the sized pre-maturity native-output leg, not a same-size PT-to-USDC repayment.

Matched AMM and four quotes: T2 records T203–T213, Ethereum block 26,077,288. Direct exit-pool screen: W/X, with X at block 26,077,340, 17:07:11 UTC. These are different states and different stages of the exit. Observation scope and source trail.

Eight claims reconcile; a shareholder’s cash access still needs proof

The selected Base vault has one observed adapter whose enumeration returns eight market IDs. Each market’s loan asset is Base USDC. At Base block 51,912,330, the eight expected claims sum exactly to 442,162,661.601103 USDC of returned vault total assets. Direct USDC balances at both vault and adapter were zero. This is a complete reconciliation of the enumerated accounting boundary, not a global issuer-reserve, beneficial-ownership or withdrawal-entitlement reconciliation.

The earlier “other assets” residual can now be examined market by market. For each market, the supplied analysis caps the screen by both the vault’s expected claim and nonnegative nominal unborrowed accounting:

That avoids assigning the vault another supplier’s entire market balance. It still does not reserve the cash, establish access or measure a shareholder’s executable withdrawal. Stored market accrual times can precede the common batch time; no uniform accrual adjustment is invented here.

Outside the selected USDe market, 57,267,018.473168 USDC of claims yields 11,166,614.381566 USDC under the cash screen, about 19.4992% of those other claims. The remaining lent-out amounts are not valued at zero. They are simply not opening unborrowed cash in this comparison. Collateral symbols are returned labels, not independent reviews of those tokens.

The selected USDe market remains approximately 87.04843% of vault accounting assets. That is a new, dated vault-specific denominator—not an issuer backing percentage and not a redating of the September 27 value of 87.753850%. A changed ratio alone does not prove the vault became safer.

Why the default route does not make every screened dollar immediately available

The vault’s observed liquidity-adapter pointer and decoded liquidity data identify the selected USDe market. The published liquidity-curation description distinguishes that chosen route from other allocated markets; ordinary withdrawals are not described as automatically exhausting every market in turn. The other 11.17 million USDC cannot therefore be added without qualification to a default withdrawal promise. P04

Other resources may become usable through permitted mobilization, deallocation, another supported pathway or later repayments. Access, timing and competing demand still have to be satisfied. An in-kind loan claim is not USDC paid to a shareholder, and a conservative zero maximum getter does not establish absence of valuable or usable assets. The published V2 reference explicitly treats its maximum functions as conservative zero returns. P05

The sampled force-deallocation penalty is 0.001%. Under the documented share-burn treatment, applying that fraction to the entire outside cash screen would be about 111.6661 USDC-equivalent of share value. It is not USDC sent out to an external fee recipient, so it is not subtracted as cash leaving the vault. Actual shareholder share conversion, approvals, rounding and successful withdrawal were not measured. A mobilizing shareholder may bear a share-value cost while remaining shareholders retain the associated value; loan impairment remains a separate possible loss, not a finding from this screen.

Enumeration and market parameters: G/H; common-state expected claims, market accounting and vault/adapter direct balances: N03–N21. Selected liquidity data and penalty: earlier F14–F16. No shareholder execution or complete operational authority review accompanies the screen. Observation scope and source trail.

Credit the extra resources once—and only when they can arrive

The supplied comparison retains 50,000,000 USDC of assumed withdrawal demand. It is not an observed queue, a particular shareholder’s entitlement or a total-vault maximum. It assumes no new borrowing or competing use, then varies which identified resources can become usable before the same deadline.

The final row is a stabilizing countercase, not credited current capacity. The borrower's debt and quote belong to a later batch than the vault screen; treating them as future arrivals is explicitly conditional, not a claim of a common-state measurement. A late repayment does not cover an earlier obligation. Nor can the same cash be counted as pool output, borrower balance, lender repayment and issuer reserve simultaneously.

The result improves on both extremes. Ignoring the other markets understates the potential resource set by roughly 11.17 million USDC under the stated screen. Crediting their entire 57.27 million claim value as immediately available overstates it. The unresolved question is which screened assets can be mobilized for the relevant shareholder, through the applicable permissions, after competing use, by that shareholder’s deadline.

Liquidity delay and economic loss remain different. Recoverable loans that have not repaid can make shareholders wait without proving principal impairment. A deteriorated loan can instead reduce eventual value. A borrower repaying from independent cash may improve supplier access, while a lender funding that borrower acquires the remaining timing and realization risk. These mechanisms can occur without a change in a token’s headline market price.

The Maple and JAAA recovery chains do not supply cash to this vault merely because they belong to the broader USDe discussion. Their rights, owners, external releases and locations remain separate. Neither a proposed reserve use nor a documentary PSM deployment is entered as an available arrival. Recovery timing and non-overlapping resources and local conversion versus replenishment retain their original boundaries.

What advanced—and what is still incomplete

The comparative financed-recovery study remains partial for financial reasons. That is not simply because the separate control study is unfinished, and it is not because Research was required to execute a trade. Adequate passive evidence of an admitted, sized and funded route could support a stronger conclusion; that evidence is still missing for the complete position-level comparison.

For a plain holder, the relevant question is the accessible payout route at the needed size and date. A leveraged borrower adds a payment that may have to occur before collateral is free. A USDe lender bears the value of the owed token even if a relative collateral ratio stays stable. A USDC supplier needs USDC, not a PT accounting value or a healthy collateral mark. An outside funder can inherit risk after the original protocol has been paid.

Useful mitigants remain conditional rather than assumed absent: free collateral can shorten a route; independent repayment funds can replenish suppliers; a financed staking wait can avoid a poor immediate sale; and genuinely mobilizable assets outside one market can reduce a cash gap. None creates value or cash twice. None establishes universal issuer eligibility or an unconditional bank-dollar exit.

No realized loss, global depeg, funding commitment, complete security audit, current issuer-wide solvency ratio or universal exit assurance is delivered. The source, amount, token, network and timing qualifications belong beside each result—not only in this closing limitation.

Evidence, clocks and source trail

This presentation uses the supplied financial investigation dated 28 September 2026 UTC. Its quotations and observations are passive provider-mediated returns, not transactions. Sixteen retained batches contain 213 subcalls, including 212 successful returns and one failed quote; block/time getters are included in that count. Discovery reads and bounded natural-event queries are separate. These are not counts of independent security checks.

The record retains requested targets, signatures, arguments and returned values, with distinct chain, block and time. Successful quotations do not establish consent, customer access, available credit or completion of an exit. Source text used to interpret financial output is not independently rebuilt deployment code. The observation table below identifies the principal comparisons relied on here.

These times are included block timestamps, not measured HTTP retrieval times. An earlier enumeration batch G returned inner block 51,912,087 while the outer reader reported 51,912,088; the inner value identifies that batch. Later observations must not be used to recreate an earlier state. Some observations occurred on September 29 in Bangkok, but this chapter deliberately retains the investigation’s September 28 UTC date.

Exact selected objects

What calculation checks mean

Research reports 37 retained and 51 new offline financial/data checks, including debt conversion, eight-claim reconciliation, quote-boundary handling, debt-token/account reconciliation, native PT output and non-duplicated cash. They validate arithmetic and data handling, not contract security, future rates, current permissions or an executed trade. The checks do not establish a funded route or a new setting beyond the observations identified here.

The relevant Aave account flags are reserve 30 borrowing and reserve 32 collateral. Unrelated reserve lookups do not establish account exposure. The large Base quote is likewise intentionally interpreted as a boundary-limited input budget, not a completed full-fill price. Unavailable public discovery responses were not treated as zero balances or complete market lists.

There is no independently reconciled complete repayment/withdrawal/sale chain, external credit commitment, all-route depth survey or bank-dollar receipt in this evidence. The absence of those observations is not an allegation of misconduct or insolvency. The quoted source and amount remain useful without implying those stronger conclusions.

Public interpretation and discovery sources

Research accessed the sources below on 28 September 2026 UTC. Access dates are not publication, implementation or economic cutoffs. The scope descriptions distinguish source reading from the separate financial observations. Earlier methods and sources remain in their dated chapters.

Retained research and exact reader downloads

The 18 September whole-system foundation, 20 September recovery chapter, 22 September documentary chapter, 26 September control continuation, and 27 September financial chapter remain separately dated. The earlier 20 September recovery data and 27 September selected financial data retain their original bytes; neither download contains refreshed September 28 observations. All research also preserves the historical reports.