Earn the Vault: An IPS, a One-Year Stage-Gate, and a Mandate That Expires Unless Renewed

My verdict up front: I support empowering the Foundation, and I propose amending how the power arrives. No enterprise on earth hands a new management team the entire balance sheet on day one. It runs a pilot: limited capital, pre-agreed metrics, a scheduled review, and a defined path back if results disappoint. This amendment restructures the temp check into exactly that — three instruments, one package: a ratified Investment Policy Statement, a Phase 1 mandate funded at operating run-rate rather than full transfer, and a stage-gate at month twelve where the mandate expires by default unless the DAO renews it on evidence. Start with the gap this thread has argued around but not closed. The responses so far propose caps, tranches, review rights, and scorecards — and I credit them: tranching, withdrawal caps, and measurement standards are all directionally right. But nearly all of them share one structural flaw: continuation is the default. The Foundation keeps its expanded powers unless the DAO organizes itself to take them back, and a body with sub-ten-percent turnout organizing a clawback against an incumbent operator is a fantasy. In finance we call this the reverse onus problem. My amendment flips the default: the expanded mandate sunsets automatically at month twelve. Reversion costs nothing and requires no vote. Renewal requires a vote, and the vote runs on numbers agreed before the pilot starts, not argued after it ends. Instrument one: the Investment Policy Statement. Before any fund-management authority activates, the DAO ratifies an IPS for the Endowment — the same boring, decades-proven document that governs every university endowment and Norway's sovereign wealth fund. Six sections: objective and horizon, with capital preservation in real terms ranked above return maximization; allocation bands, not points, giving the Foundation full execution freedom inside them and none outside; risk limits per protocol, counterparty, and custodian; a named benchmark so performance is a number, not a forum argument; a smoothed spending rule capping annual distributions at a fixed percentage of the trailing twelve-quarter average value — Alex Van de Sande's proposed 5% cap made operational, because a cap on spot value doubles the budget in a bull market and guts it in a bear; and an annual amendment process with comply-or-explain disclosure of any deviation within seven days. Instrument two: Phase 1 funding at run-rate, not transfer. The Foundation receives an operating budget sized to eighteen months of expenses, released quarterly. The Endowment and remaining treasury stay in DAO-controlled contracts, managed by the Foundation strictly under the IPS mandate. Custody of the base does not move in Phase 1. This delivers everything the temp check says it needs — legal capacity, staff, grants administration, operational speed — while deferring the one decision that is genuinely hard to reverse. Using Van de Sande's figures, roughly $50M left the treasury over five years, about 46% of it to Labs, and the Endowment holds around $130M; I cite whose numbers these are because this thread quotes treasury totals anywhere from $130M to $400M depending on definitions, and a debate that cannot agree on the balance sheet is itself evidence for standardized reporting. Instrument three: the pre-registered scorecard and the gate. At ratification — not at month eleven — the DAO and Foundation jointly publish the evaluation matrix with numeric thresholds. A word of measurement discipline here, because a badly built scorecard is worse than none: judge the Foundation on what it controls. Raw ENS revenue is mostly a function of registration demand and market cycles; punishing or crediting a board for macro is how enterprises teach their managers to game metrics. The scorecard should therefore weigh, roughly equally: budget discipline, actuals versus plan with variances explained; Endowment performance relative to the IPS benchmark, not in absolute terms; grants efficiency, cost per funded outcome with every grant published; ecosystem growth the Foundation can influence, including new builder integrations, developer activity, and named-contract or organizational adoption; revenue diversification, meaning progress on income streams beyond .eth registrations, judged on milestones rather than dollars in year one; and governance hygiene — reports delivered on time, conflicts disclosed, zero unexplained IPS deviations. Score well, and the DAO should renew without drama. Score badly, and nothing needs to be seized: the mandate lapses, budget authority reverts to the DAO's current structure, and active grantees are honored through a wind-down clause so third parties never bear the cost of our governance experiments. On responsibility, because accountability that belongs to everyone belongs to no one. The Executive Director is the accountable owner of scorecard results and presents them personally at months six and twelve. The board is responsible for oversight and certifies the reports. The DAO is the approver at the gate and is informed quarterly. One name answers for performance; that is the entire point of hiring an ED, and it is what Jeff Lau's honey-pot critique of the DAO era actually demands — his argument for accountability is my argument for a scorecard with a name on it. Phase 2, for completeness. Only after a successful month-twelve renewal does the DAO vote separately on whether custody itself should move — full transfer, permanent split, or continued mandate. By then the question is no longer theoretical: there is a year of audited, benchmark-relative, pre-registered evidence to vote on. Katherine Wu wants a DAO that votes rarely and matters every time it does. This structure is that DAO — two votes in eighteen months, both on evidence. The trade-offs, honestly. A one-year gate creates short-term incentives; the mitigations are multi-year metrics where possible and a renewal that extends three years, so the Foundation is not permanently campaigning. Run-rate funding means the Foundation cannot make decade-scale commitments in year one; I consider that a feature of a probation period, not a bug. The stage-gate adds one more DAO vote, which cuts against the vote-rarely objective — but one scheduled, evidence-based vote is cheaper than the emergency governance war this thread becomes if an irreversible transfer goes wrong. And this amendment does not settle board selection; it deliberately binds whichever board wins that argument. What would change my mind. Show me a comparable institution that granted a new executive team full, irreversible control of a nine-figure endowment in year one, with no ratified investment policy and no probationary review, and a clean ten-year record afterward — and I will withdraw this amendment and support the transfer as drafted. I do not believe that case exists. Enterprises pilot before they scale, endowments ratify policy before they delegate, and boards that skip both steps become case studies. ENS should be the case study taught for getting it right: empower the Foundation, fund the pilot, publish the scorecard, and let the Foundation earn the vault. SUMMARY OF AMENDMENTS PROPOSED Amendment 1 — Ratify an Investment Policy Statement before any fund-management authority activates. Why: it converts board discretion into a bounded mandate and makes performance a benchmark-relative number instead of a forum argument. Binds any board, under any custody outcome. Amendment 2 — Replace the full treasury transfer with a Phase 1 operating budget sized to eighteen months of run-rate, released quarterly; Endowment and remaining treasury stay in DAO-controlled contracts, managed under the IPS. Why: delivers every operational benefit the temp check asks for while deferring the only decision that is hard to reverse. Amendment 3 — Adopt the smoothed spending rule: annual distributions capped at a fixed percentage of trailing twelve-quarter average value. Why: makes Van de Sande's 5% cap operational and stops market cycles from silently rewriting the budget. Amendment 4 — Pre-register the twelve-month scorecard at ratification, with numeric thresholds: budget vs actuals, benchmark-relative Endowment performance, grants cost per outcome, builder integrations and developer growth, revenue diversification milestones, governance hygiene. Why: metrics agreed before the pilot cannot be gamed after it; measuring only what the Foundation controls makes the scorecard one the Foundation itself can accept. Amendment 5 — Sunset by default: the expanded mandate expires at month twelve unless the DAO votes to renew on scorecard evidence; renewal extends three years. Why: flips the burden of proof — continuation must be earned, reversion costs nothing and needs no vote from a low-turnout DAO. Amendment 6 — Named accountability: the Executive Director personally owns and presents scorecard results at months six and twelve; the board certifies; the DAO approves at the gate. Why: accountability that belongs to everyone belongs to no one. Amendment 7 — Grantee wind-down clause: if the mandate lapses, active grant commitments are honored through transition. Why: third parties should never bear the cost of our governance experiments. Amendment 8 — Phase 2 custody vote only after successful renewal. Why: by then the DAO votes on a year of audited, pre-registered evidence instead of projections and trust. PROS OF THIS AMENDED PATH 1. Both camps can vote for it: the Foundation gets funding, staff, and operational authority on day one; the DAO keeps custody and a costless reversion path. Nobody concedes the custody war in Phase 1. 2. Accountability becomes cheap and objective: quarterly benchmark-relative reports replace ad hoc forum litigation of performance. 3. The default protects the principal: a sub-ten-percent-turnout DAO never has to organize a clawback against an incumbent operator. 4. It answers the strongest pro-Foundation argument on its own terms: the honey-pot critique demanded accountability, and a pre-registered scorecard with a named owner is more accountability than either the current DAO or the unamended temp check provides. 5. It is precedent-backed: pilot-then-scale is how enterprises deploy capital, and IPS-governed delegation is how every comparable endowment operates. ENS would be adopting the base rate, not experimenting against it. CONS, STATED HONESTLY 1. One year of probation creates short-horizon incentives; mitigated by multi-year metrics and a three-year renewal, but not eliminated. 2. Run-rate funding limits the Foundation's ability to make decade-scale commitments in year one. 3. Adds one scheduled DAO vote, cutting slightly against the vote-rarely objective; I judge one evidence-based vote cheaper than the emergency governance war an irreversible failure would trigger. 4. A badly drafted IPS can force selling into market stress; mitigated by allocation bands, annual review, and comply-or-explain deviations rather than hard reverts. 5. Does not resolve board selection; it deliberately binds whichever board wins that separate argument. A closing word on legitimacy, because structure is not the whole story. I respect Nick Johnson, and I take seriously where he is coming from: years of watching governance dysfunction up close, and a genuine desire to protect the treasury and the protocol's long-term mission from it. That concern is valid and this amendment shares it. But ENS did not have one founder. Alex Van de Sande has publicly proposed a different path — DAO-approved talent and a 5% withdrawal cap rather than a full handover — and Brantly Millegan has come out clearly against the proposal as drafted, arguing the board is selected entirely by the proposing party and that months of good-faith talks on independence were bypassed. When the founding team itself is split on a change this irreversible, that is not noise to be voted past. That is precisely the signal that says: pilot first, prove it, then scale. And we should be honest about where the treasury actually comes from. Every dollar in the Endowment exists because a community chose to register names, build integrations, and make ENS the default identity layer of Ethereum. The community is not a stakeholder to be managed around; it is the revenue engine. If builders and users lose confidence and stop using the protocol, there is no treasury left to steward — no Foundation structure, however well designed, survives the loss of the thing that funds it. That is exactly why this amendment insists on a scored pilot rather than a permanent transfer: the twelve-month scorecard measures whether an empowered Foundation actually grows what the community built — revenue diversification, builder integrations, developer activity — or erodes the trust that generates it. If the experiment grows the engine, renew it with a bigger mandate and my vote. If it does not, the mandate lapses and the community's asset comes home by default. Hold the people accountable, score the experiment, and let the results — not the loudest voices on either side — decide the next era of ENS.