did you hear what is happening with @alturax i spent some time going through the founder response, accountable dashboard, merkl campaign route, morpho vault data, hyperevm contract trail and dium’s withdrawal queue scan, and my current read is that this is not presenting like a normal smart contract exploit. the more accurate failure mode looks like a redemption-liquidity crisis inside a product where users treated the liability side as short-duration redeemable dollars, while the asset side was spread across rwa exposure, exchange balances, custody, trading venues and a fully utilized morpho lending market... the founder said altura processed more than 8.5 million usdt in instant redemptions over 24 hours, and then decided to begin an orderly wind-down of the vault. that sequence is important because it does not point to a single contract-level break. it points to the vault being able to process some liquidity at the start, but then reaching the point where continuing instant withdrawals would require unwinding assets that are not all sitting in the same liquidity layer or on the same settlement timeline... accountable currently shows around $32.41m total supply against $33.99m total vault reserves, which gives a collateral ratio of roughly 104.9%. on the surface, that does not show an immediate reserve shortfall. the problem is that the reserve composition is not equivalent to instant redemption capacity. the dashboard shows $21.81m in inessa rwa, $6.86m on okx, $3.08m on hyperliquid, $1.21m with cobo, $524k on hyperevm, $401k on ethereum and $101k in tauri vault. so the majority of reserves were sitting in offchain or counterparty-dependent buckets, while the immediately accessible onchain liquidity was small relative to the total claims users were trying to redeem... this distinction matters because proof of reserves only tells you that reported assets are above reported liabilities at a point in time. it does not tell you whether those assets can be converted into usdt fast enough during a withdrawal run. a vault can be overcollateralized and still become illiquid if the asset side is mostly venue balances, rwa claims, private-credit style exposure, custodial balances or deployed lending positions. that looks like the central issue here... the morpho data makes the onchain bottleneck more obvious. the Alpha USDT Prime vault on hyperevm, address 0x242572d6f1AF7111bcA807ECDd0f74108cEAeD5d, had around $5.9m in assets, but the important fields were idle assets at 0, liquidity at 0 and force deallocatable liquidity at 0. the vault was not holding redeemable idle cash. it had routed into one morpho market adapter, 0x4651f49F4AFB050e9B1cdB212d2655fF647C6f80, which supplied USD₮0 into a market backed by AVLT collateral... the AVLT contract, 0xd0Ee0CF300DFB598270cd7F4D0c6E0D8F6e13f29, is also the address that connects the withdrawal queue trail and the morpho collateral trail. that market showed roughly $5.90m supplied, roughly $5.90m borrowed, 0 liquidity, 100% utilization and 91.5% lltv. at that point, depositors cannot exit from normal market liquidity unless borrowers repay, new lenders enter, the curator reallocates, or the team injects external capital. that is not a frontend issue. that is the actual lending-market state... then the slow-redemption queue started giving the market a visible stress signal. dium’s scan over roughly 72 hours showed 2.91m AVLT queued, 1.31m AVLT claimed, 150k AVLT cancelled, 1.45m AVLT net queue growth, 223 open withdrawal requests and around 1.55m AVLT still outstanding. that lines up with users reporting that instant withdrawals were unavailable and the frontend was pushing them toward slow redemption. once the queue becomes visible, the psychology changes because users no longer underwrite the product as a yield vault. they underwrite it as an exit line... the timeline therefore looks fairly straightforward. the mainstreet/msy depeg created the initial fear. altura said it had no exposure to mainstreet, but that was no longer the only question users cared about. users started underwriting altura’s own liquidity stack. the accountable dashboard showed reserve coverage, but it also showed that most reserves were not sitting in instantly accessible onchain liquidity. the morpho vault had no idle liquidity. the AVLT-backed lending market was fully utilized. the slow-redemption queue expanded. altura processed a large amount of instant redemptions, and then the founder moved to an orderly wind-down once continuing the product required unwinding positions across venues, counterparties and settlement paths... the largest caveat is the inessa rwa bucket. accountable notes that the statutory audit confirms a related-party balance payable to altura, but also says the audit does not independently verify asset-level existence, custody, segregation, valuation or backing of the specific assets referenced in the reporting data. that does not mean the exposure is fake. it means that the largest reserve bucket does not have the same liquidity and verification quality as onchain usdt sitting in a redemption contract. in a normal environment, that caveat is easy to ignore. during a run, it becomes the main thing users care about... what likely happens now is not a single clean redemption event, but a staged wind-down. exchange and venue balances should be faster to move. morpho liquidity depends on repayment or new liquidity entering the market. cobo and other custody balances depend on operational movement. rwa or private-credit style exposure depends on settlement and counterparty timelines. that is why “orderly wind-down” is the correct phrase. it means the vault is moving from instant redemption UX into portfolio liquidation mode... my conclusion is that the altura failure mode is asset-liability maturity mismatch. the liabilities behaved like short-duration redeemable dollars, but the assets were a mixed portfolio of offchain, venue-based, custodial, rwa and fully utilized lending exposure. that structure can work when inflows are steady and withdrawals are normal. it breaks when users all ask for instant liquidity at the same time... this is why proof of reserves is not enough for yield vaults. the real question is proof of liquidity. how much is idle? how much is instantly redeemable? how much is deployed into lending markets? how much depends on borrower repayment? how much sits on exchanges? how much sits with custodians? how much is rwa/private credit? how much has asset-level verification rather than just balance-level reporting? because reserve coverage answers whether assets exist above liabilities... it does not answer whether users can exit before the line gets longer...
· 22K Views
