Budget #69 — Voting Policy & Rationales

Treasury allocations to dApp-class proposals should produce on-chain economic activity proportional to capital deployed. Tx is directly causal for ecosystem activity. TVL alone is NOT — per Algorand Foundation research (and parallel observations across L1s), TVL does not correlate with token price or ecosystem prosperity. TVL must be evaluated by quality, not quantity. The criterion therefore distinguishes Tx (always good) from TVL (quality-dependent: ADA sell-pressure reduction or active Tx generation).

Lack of Cardano-native track record alone is not disqualifying — but for proposals where the proposer acts as judge / fund allocator (secondary-grant-dao) or claims to drive Cardano-specific outcomes (BD network, dApp), the absence of verifiable prior delivery on Cardano is a structural unsuitability signal. The criterion distinguishes 'we want to build for Cardano' (entrant-OK) from 'trust us to build/judge Cardano' (entrant-not-OK).

Process metrics (LOIs sourced, MoUs signed, attendance, masterclasses delivered) measure proposer activity, not Cardano ecosystem outcome. Tying milestone payments to outcome KPIs (Tx generated, TVL deposited, USD revenue collected) creates an accountability incentive. Without outcome KPIs, the proposer is paid for doing-the-work regardless of whether Cardano benefits.

When Treasury invests in a commercial entity that subsequently captures revenue, asymmetry occurs: public capital builds infrastructure whose commercial value is privately captured. C4 demands explicit symmetry — repayment, revenue share, fee return, or token allocation. The DeltaDeFi 'Treasury Loan with Perpetual Upside' structure is the constructive opposite of pure-grant-to-commercial-entity.

Treasury allocations to functionality already supplied by another funded source double-spend public capital. 3-tier comparison framework (refined to avoid circular logic):
• Tier A(valid duplication): comparison target has completed funding (past Catalyst delivery / past TW with milestones disbursed) OR is live in production with verifiable usage.
• Tier B(weaker; conditional): comparison target is in active TW currently in voting — citing it as 'duplication' is conditional because it's not yet ratified. Use as 'overlap concern pending TW outcome'.
• Tier C(cannot use): comparison target is another Budget #69 item in the same evaluation cycle — circular logic.

Cardano L1 has measurable economic constraints: per-tx fee ~0.17 ADA ≈ $0.04 (per ADAtool epoch fee data) and ~20-second block time (per Cardano Ouroboros Praos parameters). Use cases that economically require sub-cent fees or sub-second finality (continuous market-making with frequent order updates, micropayments at $0.001 unit cost, agent-to-agent commerce at high tx frequency) cannot operate viably on Cardano L1. C6 forces proposers to either (a) demonstrate L1 viability at empirical fee/block constraints, or (b) explicitly commit to L2 implementation (Hydra, Midgard, Partner Chain) — the latter then changes the L1-Tx-KPI denominator.

Conflict-of-interest is a structural risk when (a) proposer acts as fund allocator (deciding who else receives Treasury funds, e.g., secondary grant DAOs), (b) proposer is paid to amplify other commercial entities (marketing — selecting which projects get amplified), (c) proposer's tokenomics route Treasury investment back to their token holders (Libertum-pattern). C7 demands enforceable safeguards: exclusion thresholds (committee members not eligible for own funding), allocation caps (max % of grant pool to member-affiliated projects), independent review process. 'Will be published' is insufficient — concrete safeguards must be in proposal text before vote.

Repeat-funded proposers must demonstrate that prior cycles produced verifiable Cardano-specific outcomes. 'Completed' Catalyst status alone is insufficient — outcome must be Cardano builder adoption / production deployment / Tx contribution, not internal team activity. /tx-roi backend (https://adatool.net/tx-roi) provides the systematic prior-funding lookup.

Treasury investment ahead of demand validation inverts normal risk order. C9 demands proposers demonstrate that the central thesis (e.g., 'institutional self-custodial demand on Cardano', 'AI agents are new financial users', 'enterprise BD pipeline will convert') is empirically supported via pilot results, signed LOIs from named counterparties, or quantitative market-sizing analysis. Asserted thesis alone — without empirical anchor — fails C9. The criterion is most rigorous for speculative-finance class but applies broadly to demand-dependent proposals.

Net Change Limit (NCL) is the constitutional cap on Treasury withdrawals per cycle (currently 350M ADA for 2026-27 cycle per Cardano constitutional parameters). C10 prevents single proposals from monopolising scarce Treasury capacity at the expense of other ecosystem priorities. The 3-band threshold (<5% proceed / 5-15% scrutiny / >15% major concern) is calibrated against typical infrastructure proposal sizes (most well-scoped infrastructure asks land <5%; large initiatives 5-15%; multi-team strategic asks >15% should require strong justification).

Secondary grant DAOs split visibility into two layers, both practically unobservable to individual DReps. (1) Where funds actually flow to secondary recipients — even when secondary DAOs publish dashboards (CBDAO has one, Catalyst has long-running ones, this proposal will likely include one too), the structural constraint is NOT tooling availability. DReps are an unspecified mass of mostly-unpaid volunteers with day jobs; capacity for governance is bounded by personal time and resources. Core attention goes to reviewing on-chain GAs (constitutional minimum). Continuously tracking a secondary DAO's internal allocation across dozens of downstream proposals — voting cycles, milestone reviews, dispute resolutions — is outside what most DReps can fit in. Better dashboards don't reduce the cumulative attention required to exercise oversight. Empirically: among DReps I have spoken with, none can monitor Catalyst or CBDAO fund-allocation outcomes at the individual-recipient level. (2) Whether the secondary DAO's own internal governance operates without member capture / allocation bias / self-dealing — same DRep-capacity constraint applies, made worse by internal voting records / CoI disclosures / per-decision rationale not normally being produced at audit-grade granularity. Direct TW (one recipient, one set of deliverables) is straightforwardly handled within typical DRep attention budget by contrast. The Constitutional amendment lowering Direct TW barriers (17 active TWs requesting 74% of NCL, pending ratification) makes secondary mechanisms structurally less needed. C11 sets a high bar: default NO unless three exception conditions all met.

Filling a gap identified during by-category review of Policy v2: core-protocol / wallet-hardware / education / infrastructure proposals had previously only C5 + C8 + C10 (or fewer) applicable. This created 'too-loose' threshold — passing C8 (past funding outcome) alone could approve continuing-maintenance work without verifying that the proposed scope and team size are warranted by current user/adoption base. C12 demands maintenance-class proposals justify their scope+team via empirical user metrics + benchmarking against comparable maintainers.