Temporary funding is not final repayment

An exit needs a first payment—and a final cash source

A temporary loan can solve the order of payments without solving the amount ultimately owed. A borrower may pay its lending protocol, recover collateral and still owe a new creditor who must wait for a sale or settlement. A shareholder may have a valuable claim, a documented exit interface and enough authorization, yet still lack sufficient shares, receiver admission or cash at the deadline. Neither case is answered by a near-par token price.

This analysis adds three connected tests to the 28 September financial continuation: the final cash required after temporary funding; the quantity, price and time needed to convert a principal-token exit into the creditor’s token; and the difference between a shareholder’s gross share-value debit and net receipt. The examples use the same frozen observations—not new balances or successful exits.

“Cash” throughout means the token required by the relevant obligation on the relevant network. Base USDC is not Ethereum USDC available locally, sUSDe is not USDC, and a USDe accounting equivalent is not an executed sale. Bank-dollar conversion is a further boundary, not implied by any token payment here.

The result is useful without calling every gap a failure. A smaller collateral release can need less initial funding; an independent repayment can replenish lenders; an admitted alternative withdrawal route can mobilize other markets. Each is a genuine mitigating possibility only to the extent that its own amount, timing and access conditions hold.

Source basis: Morpho’s 2023 funding description · the 28 August SDK design · frozen observations and calculation basis.

Temporary financing changes timing, not net funding

For the selected Base borrower, a one-shot release of all pledged collateral requires the original debt to be extinguished first under the retained financial convention. At the recorded 28 September state, that debt was 2,704,990.165885 USDC, before later accrual or costs. That is a prior-payment requirement for this sequence, not the minimum outside seed for every possible staged exit.

The earlier conditional 50,000-USDe slice lies within the simplified collateral-release headroom and therefore illustrates why some deleveraging need not wait for full-debt outside funding. It still depends on actual withdrawal access, a full fill, costs and no intervening adverse change. Repeating its quote does not establish that the whole position can unwind at the same rate.

Morpho’s 11 October 2023 design announcement describes public temporary liquidity repayable within the same transaction. This is positive evidence against the blanket assumption that only a negotiated private term loan could ever provide an initial payment. It is not a measurement of present Base capacity or an offer to this borrower. A same-transaction obligation also cannot simply remain unpaid while a later staking claim matures.

The financial distinction survives any financing label. A temporary lender advances principal and expects it back; repayment cancels that principal from the completed exit’s net external funding. If the replacement lender remains unpaid, the original debt has been refinanced or transferred—not eliminated for all creditors. Sufficient final sale proceeds or other independent repayment assets are still necessary.

Even a large nominal temporary-liquidity screen does not answer the last-payment question. The selected market’s 37,528,702.690816-USDC unborrowed screen in an earlier September 28 batch exceeded the later recorded borrower debt. But a market accounting difference is not reserved funding, a current singleton balance, an admitted loan or buyer demand for collateral. It cannot be credited simultaneously as temporary finance, final purchase money and a shareholder’s exclusive withdrawal resource.

Source basis: Morpho’s 2023 funding description · frozen observations and calculation basis.

The Base bound is a frozen four-pool scenario

The relevant borrower remains 0x5afe2414f865cbc5ff4e25979996c5b1252e0002 in the selected Base USDe-collateral/USDC-debt market. The debt and collateral observation is Base block 51,912,543, 28 September 2026 at 16:53:53 UTC: 2,704,990.165885 USDC debt and 3,039,240.113838519418177239 USDe collateral. The earlier account calculation supplies the upward-rounded debt convention and the limits of its release bounds.

Now take only the four previously examined Base outlets. The inputs are distinct pools’ recorded USDC token balances—not their total value, their USDe holdings or a sum of independent quote outputs. Each balance is frozen at its own observed time. Assign every recorded USDC unit to this borrower, even before accounting for price curves or competition. This deliberately generous treatment is an analytical assumption, not a claim that those units were simultaneously available or exclusive.

Assume no net replenishment, no added outlet and no independent borrower cash reaches this defined exit. With recorded debt D, assembled stock T, genuinely additional cash X and costs K, a necessary condition for finishing with no unpaid replacement creditor is:

Setting costs to zero makes the test least demanding. It gives X ≥ 1,918,946.446768 USDC. The assembled stock is approximately 29.0590% of the recorded debt. That fraction describes this denominator and scenario; it is not a reserve ratio, expected recovery rate or current Base-wide liquidity statistic.

The defined frozen outlets alone cannot support complete debt-free closure. This is not a conclusion that all routes fail. A funded new buyer, a different same-chain outlet, an eligible issuer-redemption arrival or unencumbered borrower cash can change the resource set. Those arrivals need their own token, size, access and timing evidence. A continuing term loan may refinance the position, but does not on its own finish a debt-free exit.

The bound deliberately avoids guessing the consumed input of the old large quote. The million-unit input budget returned 785,605.527146 USDC at a price boundary; consumed input and leftover USDe were not returned. The full-position enquiry failed with no usable output. Neither is an observed zero recovery or a completed sale. Residual collateral is not written off or counted as cash, and the old Ethereum quote is not inserted as a Base resource.

Cash conservation explains why reordering does not defeat the stated bound. Without net cash entering the defined pools, their net outflow cannot exceed their frozen stock. A multihop route changes the conclusion only when it introduces a separately supported additional cash source. It does not turn repeated price enquiries into repeated inventory.

Source basis: frozen observations and calculation basis.

The PT exit must satisfy the final creditor’s token, size and date

The selected principal token is PT-sUSDE-26NOV2026, with recorded maturity on 26 November 2026 at 00:00 UTC. The September 28 work bound it to Ethereum AMM 0x47ad2cd1dd15739a7a035b9d3b7828d916fef77e and its SY/YT wrappers. No October market-activity or deployment reading refreshed that binding.

At Ethereum block 26,077,288, 28 September at 16:56:35 UTC, the million-PT quotation paid 790,975.183215710715129177 sUSDe. At the same batch’s recorded conversion, 1.250365749456529343 USDe per sUSDe, that corresponds to a conditional 989,008.2777630277 USDe. It is not received USDC, an executed staking withdrawal or a promise that the conversion ratio will be unchanged when a user actually initiates a claim.

A pledged-PT borrower faces two intervals: financing before the PT is released, and financing between native sUSDe receipt and the final owed-token receipt. The same lender could finance both, but that does not make their duration or risk identical. A direct sized sUSDe sale could avoid a queue; a staking exit could yield USDe later. Neither supplies USDC until a separately funded final conversion completes.

Let U be the conditional USDe quantity, N the initial USDC obligation, a a simple annual financing rate, t days to final cash, F extra USDC charges, and α the fraction that can actually be converted before the deadline. The financial analysis requires:

The price must apply to the sellable quantity at the relevant time. A marginal quote, NAV or accepted collateral value is not enough. At zero sellable quantity, no finite unit price repairs the missing capacity. Halving the timely sellable quantity doubles the necessary average price, with the other assumptions held fixed.

These are break-even conditions for an illustrative creditor, not a measured PT loan, financing offer, liquidation trigger or recommended discount. The apparent par-sale surplus is not verified profit: the quantity, route, access, price and finance are assumed. A separate illustrative 1,000,000-USDC obligation requires 1.01111388 USDC per USDe even at zero waiting cost, or 1.01344083 after seven days at the same assumed APR. A million-token label is not a million-unit cash payment.

The previously screened direct sUSDe/USDC pools held only 2.801682 USDC in total at Ethereum block 26,077,340, 28 September at 17:07:11 UTC. Those particular recorded outlets do not substantiate the modeled large final payment. This finite screen is not the complete liquidity universe, and it is not permission to substitute a USDe quote for an sUSDe quote.

Source basis: frozen observations and calculation basis.

A waiting claim is not an earning share—or a different debt token

LlamaRisk’s 16 March 2026 review describes burned staking shares, equivalent USDe placed in escrow, a later claim at the user’s recorded endpoint and no continuing sUSDe yield on those burned shares. Kairos’s 28 May 2026 retrospective reports that the dynamic framework had been activated in March with a one-day setting. These are documentary accounts, not newly tested October behavior or a commitment to an individual’s final cash time.

The waiting periods in the preceding table are therefore assumptions about total time to cash, not a list of current cooldown settings. PT release, inclusion, queue processing, final sale or eligible issuer settlement, and payment to the term lender all matter. A favorable protocol timer does not remove the steps before or after it. A free holder can choose to wait without the same debt deadline; a replacement USDC lender cannot treat that choice as repayment.

The two measured Aave examples remain sUSDe collateral with USDe debt, at their September 28 observations and eMode 32. They are not renamed USDC or USDT borrowers. A common dollar-mark movement can partly cancel in their relative account-health calculation while changing the dollar value of USDe-denominated wealth. A cross-token creditor needs a different token and a supported conversion basis.

The intended actual direct-sUSDe/USDC-or-USDT comparison still lacks a compatible-time account record tying together collateral, collateral enablement, all liabilities, category, accepted values and independently available repayment assets. A supplied sUSDe balance does not reveal debt denomination; an aggregate market’s USDC debt cannot be assigned to a holder. The earlier hypothetical USDC account remains hypothetical, and the bounded zero-aSUSDe candidate screen does not prove such accounts are absent.

Any primary issuer alternative also retains the Mint User eligibility and supported settlement-asset conditions. A PT, a staking share and issuer backing are successive claims and resources—not three independently spendable amounts.

Source basis: the 16 March queue review · the 28 May activation account · frozen observations and calculation basis.

A published shareholder route still needs an admitted holder and receiver

The new documentary evidence makes the vault-mobilization question more precise than “other markets may help.” Morpho’s 28 August 2026 SDK design record describes VaultExitBundlesV1 for an updated force-withdrawal interface. The pinned registry lists a Base bundle at 0xe52e169c342c096c4949abb944dc9f30e3f5ea84. These are a specific public design and locator—not verified runtime, current cash, a personal approval or an instruction to use the address.

The design uses available ordinary liquidity first, then additional market resources subject to cash availability and the adapter’s own claims. Its amount convention includes a share-value penalty. It is distinct from the older caller-planned path and a full-share redemption interface. The word “force” does not remove ownership, authorization or receiver restrictions, nor does it turn lent-out value into cash.

The selected vault’s dated one-adapter/eight-market structure resembles the described route’s intended shape, but does not establish current compatibility. Most importantly, the route makes the bundle the asset receiver. Earlier zero sendAssetsGate and sendSharesGate readings are not a reading of receiveAssetsGate or proof that the bundle is admitted. This is a missing financial-access fact, not evidence that the gate is closed or a vulnerability exists.

A positive resource screen cannot skip these conditions. Equally, a missing holder record is not proof that the route can never work. Public account and configuration evidence could advance several of the missing rows; this analysis does not make private documentation a universal starting requirement.

Source basis: the 28 August SDK design · the pinned Base registry · frozen observations and calculation basis.

Gross share-value debit is not net USDC

Use the inherited selected-market cash screen L = 37,528,702.690816 USDC and other-market claim-capped screen O = 11,166,614.381566 USDC from Base block 51,912,330, 28 September at 16:46:47 UTC. Separately retain the earlier F-batch penalty fraction p = 0.00001, or 0.001%, read at block 51,912,061, 16:37:49 UTC. These are mixed-time frozen scenario inputs—not current route terms.

For a hypothetical gross share-value debit budget E, the continuous-money model first uses ordinary cash A = min(E, L). Additional cash is B = min(O, max(0, E − A)/(1 + p)); the share-value penalty is p × B. Cash before external charges is A + B. A separate fee fraction f paid outside the holder reduces the net receipt to (A + B) × (1 − f).

The share penalty represents value charged through shares while assets stay in the vault. It is not an outward USDC payment and must not be deducted a second time. An external referral charge is different: it reduces the cash reaching the holder. Share-price drift, performance/management fees and exact share rounding are excluded from this approximation, not assumed absent from production.

The 50-million row does not predict a partial fill. It describes the maximum resource-supported sub-budget in the model. The published route validates coverage; an under-covered request need not pay partially at all. The uncovered gross budget is not a measured remaining share balance.

The cash ceiling before external fees stays 48,695,317.072382 USDC. The corresponding modeled gross debit support is approximately 48,695,428.738526 USDC-equivalent; the difference is share cost, not added liquidity. Neither figure is the SDK’s exact integer-boundary field or an executable shareholder withdrawal limit.

Keep the two 50-million questions separate. A 50-million cash demand still has the old 1,304,682.927618-USDC cash gap. A 50-million gross share-value budget has approximately 1,304,571.26 USDC-equivalent unsupported by the cash/debit envelope. The smaller second number does not correct the older dataset or show that the cash gap improved.

For the illustrative 45-million gross budget, a separately assumed 0.1% external charge reduces modeled receipt from 44,999,925.29 to 44,954,925.36 USDC. Its approximately 44,999.93-USDC outward cost is distinct from the approximately 74.71 share-value penalty. The design’s share-price floor precedes referral deduction, so a price bound alone does not specify final after-fee cash. A fee returned to the same economic holder would not be the external loss modeled here.

The rounding check is narrower still. Rounding penalty amounts upward on seven positive other-market legs gives 111.666148 USDC-equivalent, against 111.666144 when rounding the aggregate once: a difference of 0.000004, below the stated maximum of 0.000006. This does not determine actual shares burned. The decisive unknowns remain ownership, admission, cash coverage and final charges rather than that tiny arithmetic difference.

Source basis: the 28 August SDK design · frozen observations and calculation basis.

Competition consumes cash; recycling does not replenish it

The eight-market total does not reserve its cash for one claimant. In the illustrative sensitivity, an independent earlier 5-million-USDC withdrawal consumes selected-market cash with no replenishment. The later claimant’s assembled screen becomes 43,695,317.072382 USDC; the gap against an assumed 50-million cash demand becomes 6,304,682.927618 USDC, before external charges. No such withdrawal or pending demand was observed.

For a smaller requested budget, reduced ordinary liquidity can move more of the exit into the penalized additional-market leg even before total cash is exhausted. Competition can therefore change both availability and who bears mobilization cost. A portfolio reconciliation alone answers neither priority nor net receipt.

Both scenarios can put a repayment transaction in view, but only the second adds net cash to the defined resource set. Counting the first as new support would credit the same principal twice. This does not claim the vault has authority to finance the borrower or that it did so; it tests the accounting of an alleged source of support.

In-kind exit is another distinct result. A shareholder might leave a wrapper while receiving a direct lending claim. Without a funded buyer or repayment, that transfers the remaining collection and liquidity exposure rather than paying a USDC obligation. Similarly, repaying an original lender with temporary credit protects that lender only to the extent actually received; the replacement creditor’s claim persists until its own settlement.

Genuine outside repayments, funded buyers and admitted mobilization remain meaningful stabilizers. Their value lies in cash arriving at the correct entity before the deadline, not in giving the same unit a new label. The calculations here do not quantify a whole-system rescue capacity or assume an issuer reserve, PSM or bridge supplies the missing asset.

Source basis: frozen observations and calculation basis · the 28 August SDK design.

What changed—and what remains partial

The advance is a more precise financing test, not fresher state. The analysis can reject the claim that temporary principal automatically solves final cash insufficiency, or that a positive gross budget promises the same amount of net USDC. It cannot establish the real financing commitment or admission that was not observed.

Financed-recovery coverage remains partial for financial reasons, not only because security questions remain open. Completion need not require executing a trade; adequate passive account, route and funding evidence could support a stronger conclusion. But an unavailable record cannot be turned into evidence that a route fails, and a public address cannot become a personal entitlement.

The consequences remain position-specific. An unlevered holder may wait through a recoverable discount. A borrower faces a debt and collateral deadline. A term lender can inherit realization risk after the original protocol is repaid. A vault shareholder can face delay, mobilization share cost or receipt of a noncash claim. A principal loss requires its own supported recovery and priority analysis; no realized loss, global depeg, current reserve ratio or universal exit guarantee follows from these examples.

The whole-system assessment retains the issuer, backing, control and integration context. This continuation does not replace it with a liquidity score or expand into a new whole-system joint stress investigation.

Evidence dates, calculation basis and source trail

What was new on 1 October: public financial-route documentation and offline analysis. What was not new: pool balances, borrower positions, quotes, gate readings, financing commitments or successful payouts. Separate September batches and stored accounting times remain separate. A retrieval date does not turn a prior observation into a live one.

All new documentary acquisitions above were made by the 1 October investigation on 1 October 2026. Access does not establish an effective date. Public data/API descriptions and unsuccessful retrievals supplied no new usable borrower, shareholder or final-sale record. The readable pinned design—not an unavailable tutorial body—supports the load-bearing gross/net and admission analysis.