Two amendments are sitting in front of XRP Ledger validators right now, and between them they would do the thing Ripple has spent years telling banks the ledger can do. XLS-65 adds Single Asset Vaults — pooled deposits of one asset. XLS-66 adds a lending protocol, the plumbing for institutional on-chain credit. Ripple’s own validator has voted yes on both.
Neither is close. SingleAssetVault sits at roughly 37% support, or 13 of the 35 validators on the default trusted list. LendingProtocol is at about 34% — 12 of 35. An amendment needs more than 80% support, held continuously for two consecutive weeks, before it activates on mainnet. On a 35-validator list that is 28 yes votes. Both proposals are less than halfway.
The instructive part is what is passing on the same ledger in the same window. fixCleanup3_3_0 — a bundle that hardens existing code in vaults, the lending protocol, automated market makers and pseudo-accounts, and adds no new capability whatsoever — reached 82.86% validator consensus on 28 August and entered its 14-day activation window. Earliest possible mainnet activation is 11 September. The patch is sailing through. The features being patched are stuck in committee.
What the vote actually measures
This is the part that outsiders consistently get wrong about the XRP Ledger, and it cuts both ways for Ripple. The company writes a great deal of the code. It does not ship it. Amendments are ratified by validators, and Ripple operates one validator among the 35 on the default Unique Node List. Writing the feature and voting for the feature buys you one vote out of thirty-five, the same as anyone else running infrastructure on the network.
The 80% threshold is also not a one-off hurdle to clear and forget. Support has to hold above 80% for the full two weeks. If it slips to 80% or below at any point in that window, the amendment is rejected and the countdown restarts from zero. That is why an amendment can look imminent for a month and never activate — and it is why the 11 September date on fixCleanup3_3_0 is a floor, not a promise.
Our take: A validator vote is the cheapest honesty test in crypto infrastructure. Marketing can claim a capability exists; a governance tally either shows 28 of 35 or it does not. Right now the XRP Ledger’s operators are enthusiastically willing to make existing features safer and roughly one-third willing to add credit primitives. That is not a rejection of the roadmap. It is a statement about sequencing — and about who actually controls it.
The gap between the pitch and the ledger
Ripple’s institutional story — tokenised funds, custody, cross-border settlement, on-chain credit — increasingly depends on primitives that are proposed rather than live. We have watched this pattern for a month. Version 3.3.0 put six amendments in front of validators, including encrypted institutional balances, atomic batched payments and delegated signing authority; two of them had already been through the process once and had not survived it. Confidential Transfers, the headline feature for institutions that do not want balances readable by the network, is still waiting on the same 80% bar.
None of that means the features die. Validators are conservative by design, and the caution looks a lot more reasonable when the fix bundle in front of them exists specifically because vaults and lending had bugs worth fixing. But there is a real distance between “the ledger supports institutional credit” and “twelve of thirty-five operators have voted to turn it on,” and anyone modelling adoption timelines off press releases should be pricing that distance.
What to watch
- 11 September. The earliest fixCleanup3_3_0 can activate — and a live test of whether consensus above 80% actually holds for the full fourteen days.
- Whether XLS-65 and XLS-66 move together. Vaults are the dependency; lending is the product. If SingleAssetVault breaks away and LendingProtocol does not, validators are drawing a line at credit specifically.
- The gap between 34% and 80%. Sixteen more validators have to flip. Watch which operators publish reasoning — on a 35-node list, a handful of custodians and exchanges moving together decides it.
- Whether the fix bundle unblocks the features. The strongest argument against activating lending was that the code had known defects. That argument expires in September.