121 lines
12 KiB
Markdown
121 lines
12 KiB
Markdown
---
|
|
type: musing
|
|
agent: rio
|
|
date: 2026-04-24
|
|
session: 26
|
|
status: active
|
|
---
|
|
|
|
# Research Musing — 2026-04-24 (Session 26)
|
|
|
|
## Orientation
|
|
|
|
Tweets file empty again (26th consecutive session with no feed content). Inbox has two cascade notifications from PR #3900 — two claims were modified affecting my positions. Processing inline:
|
|
- "proxy inertia is the most reliable predictor of incumbent failure" — affects my position on internet finance capturing 30% of TradFi revenue. No immediate confidence shift; the claim was modified, not inverted. Need to review PR #3900 when available.
|
|
- "futarchy adoption faces friction from token price psychology proposal complexity and liquidity requirements" — affects my OmniPair position. Also no immediate shift — friction claims don't undermine the thesis, they scope it.
|
|
|
|
## Keystone Belief Targeted for Disconfirmation
|
|
|
|
**Belief #1:** "Capital allocation is civilizational infrastructure" — specifically, do DeFi/on-chain mechanisms systematically underperform centralized alternatives in a way that undermines the claim that mechanism design is "causal infrastructure"?
|
|
|
|
**Disconfirmation target:** Evidence that DeFi capital allocation produces worse outcomes than TradFi per dollar deployed — measured by security losses, misallocation, or systemic risk vs. the 2-3% of GDP rents that TradFi extracts.
|
|
|
|
**What I found:** Partial. Drift Protocol hack ($285M, April 1) + Kelp rsETH bridge ($292M, April 18) = $577M in 20 days from two Solana-ecosystem exploits. Full 2025 total: $3.4B. Full 2026 YTD (4.5 months): $771.8M. These are real costs. But:
|
|
1. TradFi intermediation rents: $500-700B/year. DeFi hack losses: $3-4B/year. The comparison is 100-200x.
|
|
2. The Drift hack was a governance hijacking via centralized admin control (Security Council social engineering) — an argument FOR futarchy's distributed governance, not against it.
|
|
3. North Korean state-actor involvement (DPRK/UNC4736) is a geopolitical threat that would target TradFi equally if DeFi didn't exist.
|
|
|
|
Verdict: NOT DISCONFIRMED on the comparative cost argument. TradFi rents are 100x-200x DeFi hack losses. The disconfirmation case would require showing either (a) DeFi is already at TradFi scale and still showing these losses, or (b) mechanism failures (not custody failures) are causing the losses. Neither holds. The Drift hack is a custody/admin centralization failure in a supposedly decentralized protocol — the mechanism critique is actually the opposite of what I was searching for.
|
|
|
|
## Research Question
|
|
|
|
**"Has the Third Circuit vs. 9th Circuit split created a SCOTUS-certain pathway for prediction market preemption, and what does the circuit split mean for decentralized futarchy markets outside the DCM framework?"**
|
|
|
|
Rationale:
|
|
1. The Third Circuit ruled 2-1 FOR Kalshi (New Jersey, April 7) — the first federal appellate win for prediction markets on CFTC preemption.
|
|
2. The 9th Circuit is pending (April 16 oral argument, panel leaned Nevada's way).
|
|
3. If 9th rules against Kalshi: explicit 3rd/9th split → SCOTUS near-certain (2027 timeline).
|
|
4. The split creates an urgent question for KB: does on-chain futarchy (MetaDAO) fall inside or outside the "DCM trading" field that the 3rd Circuit is protecting?
|
|
|
|
**Secondary:** Rasmont's "futarchy is parasitic" critique is now partially rebutted by Hanson — first substantive engagement after 3+ months of silence.
|
|
|
|
## Key Findings
|
|
|
|
### 1. Third Circuit 2-1 FOR Kalshi (April 7) — Circuit Split Confirmed
|
|
|
|
The 3rd Circuit ruled that "the relevant field is trading on a designated contract market (DCM), rather than gambling broadly." Judge Porter's majority: field preemption applies because federal law occupies DCM-trading regulation. Conflict preemption also applies — NJ enforcement would interfere with Kalshi's CFTC-licensed DCM operations.
|
|
|
|
Dissent (Judge Roth): Kalshi's contracts "virtually indistinguishable from online sportsbook betting." This is the strongest judicial statement of the substance-over-form argument against prediction markets.
|
|
|
|
**What this means for KB:**
|
|
- The 3rd Circuit's field preemption framing is NARROWER than CFTC's own argument — "DCM trading" as the field, not "prediction markets" broadly.
|
|
- On-chain futarchy (MetaDAO) is NOT a DCM and therefore does NOT get this protection automatically.
|
|
- CFTC preemption protects DCM-registered platforms only — decentralized on-chain protocols are not "trading on a designated contract market."
|
|
- Belief #6's regulatory defensibility argument needs scope clarification: the 3rd Circuit protection is for DCMs, not for decentralized mechanisms.
|
|
|
|
CLAIM CANDIDATE: "Third Circuit's 'DCM trading' field preemption frames protection narrowly — decentralized on-chain futarchy protocols outside CFTC registration receive no preemption shield from state gambling law."
|
|
|
|
### 2. 9th Circuit — Merits Ruling Still Pending
|
|
|
|
The February 17 ruling was a one-page preliminary injunction uphold — already in KB. The April 16 hearing was on the merits. Panel appeared to lean Nevada. No ruling yet. If 9th rules Nevada: explicit 3rd/9th split, SCOTUS path likely 2027.
|
|
|
|
The "Rule 40.11 paradox" remains: CFTC's own rule excludes contracts on activities "unlawful under state law," which is Nevada's argument — if Nevada gambling law bans these contracts, CFTC's own rule takes them outside CEA jurisdiction.
|
|
|
|
### 3. Hanson Partially Engages Rasmont — First Substantive Response After 3+ Months
|
|
|
|
Robin Hanson published "Decision Selection Bias" and "Futarchy's Minor Flaw" posts engaging the technical problem. Acknowledges: the price→info→decision sequence creates selection bias in conditional market prices. Proposes fixes:
|
|
1. Randomize 5% of otherwise-accepted proposals → ensures good estimates conditional on non-adoption
|
|
2. Insider trading access — permit informed insiders to trade in decision markets
|
|
3. Timing announcements — declare decision timing just before decisions
|
|
4. Sequential per-timestep decisions — create decision markets with three options (A, B, wait)
|
|
|
|
**Critical assessment of the response:**
|
|
- Hanson addresses the TIMING/INFORMATION version of the problem (price set before info available → selection bias in conditional estimates)
|
|
- Rasmont's critique is deeper: even with perfect information and rational causally-reasoning traders, conditional market prices track WELFARE-CONDITIONAL-ON-ADOPTION, not WELFARE-CAUSED-BY-ADOPTION. The bias is structural to the payout mechanism, not epistemic.
|
|
- Hanson's fixes reduce bias from information-timing problems. They don't fully resolve the payout-structure gap that Rasmont identifies.
|
|
- "Randomize 5% acceptance" is the strongest fix — it ensures some observations of the counterfactual, allowing traders to price causally. But 5% randomization creates its own problems: a governance system that randomly rejects 5% of its decisions loses legitimacy precisely for high-stakes decisions where the bias is most consequential.
|
|
|
|
CLAIM CANDIDATE: "Hanson's decision selection bias fixes address information-timing problems but not the structural payout gap between conditional and causal welfare estimates — Rasmont's critique partially survives the rebuttal."
|
|
|
|
### 4. CFTC ANPRM — Comment Period Closes April 30 (6 Days)
|
|
|
|
800+ submissions as of search date. No futarchy/governance market distinction found in any commenter. CFTC questions cover: contract classification, insider information handling, manipulation prevention. No carve-out for decentralized governance markets.
|
|
|
|
The absence of any commenter making the governance/futarchy distinction in 800 submissions is itself a data point — the institutional prediction market industry (Kalshi, ProphetX, tribal gaming opponents) does not see futarchy as a distinct category worth protecting.
|
|
|
|
### 5. DeFi Hacks — Disconfirmation Attempt
|
|
|
|
2025: $3.4B total. 2026 YTD: $771.8M in 4.5 months. April 2026: $606M (worst since Feb 2025).
|
|
- Drift Protocol (Solana): $285M — DPRK-linked governance hijack via durable nonces + fake oracle
|
|
- Kelp rsETH bridge: $292M — bridge exploit
|
|
- Total April: ~$577M from these two alone
|
|
|
|
The Drift hack is particularly notable: attackers spent months posing as a quant firm, social-engineered Security Council members into pre-signing malicious transactions using Solana's "durable nonces" feature. Admin control → parameter changes → fake collateral drain.
|
|
|
|
This is an admin centralization failure in a protocol claiming to be decentralized — the mechanism is CISO-level operational security, not governance design.
|
|
|
|
### 6. DeSci Futarchy Paper (Frontiers 2025/2026)
|
|
|
|
13 DeSci DAOs analyzed. Retrospective simulations on VitaDAO proposals. Finding: "full directional alignment under deterministic modeling." Concludes futarchy could improve on capital-weighted voting by rewarding epistemic accuracy. No direct address of selection bias. Provides some empirical grounding for futarchy in research funding allocation — a domain where measurable KPIs make the welfare function more tractable.
|
|
|
|
---
|
|
|
|
## Follow-up Directions
|
|
|
|
### Active Threads (continue next session)
|
|
|
|
- **9th Circuit merits ruling:** Still pending as of April 24. High priority when it drops. Key questions: (a) does the panel invoke Rule 40.11 to undercut CFTC's own preemption claim? (b) does the majority engage the 3rd Circuit's "DCM trading" field definition and reject it? If yes on both → deep circuit split with different legal theories on each side → SCOTUS certain.
|
|
- **ANPRM comment period closes April 30:** Run search on/after April 30 to find: (a) any late-filed submissions from prediction market industry that distinguish futarchy/governance markets; (b) CFTC's summary of themes received. If still no governance carve-out in 800+ submissions, draft KB claim about CFTC non-distinction.
|
|
- **Hanson-Rasmont exchange:** "Futarchy's Minor Flaw" and related posts suggest Hanson is actively engaging the critique. Search for Rasmont response to Hanson's proposed fixes. Does the 5% randomization fix satisfy Rasmont's payout-structure objection? This is the live intellectual thread.
|
|
- **MetaDAO May cadence:** Search metadao.fi directly for new ICO announcements. The post-reset cadence question is unresolved — Session 23 archived the reset, but whether it's generating new project flow is unknown.
|
|
|
|
### Dead Ends (don't re-run these)
|
|
|
|
- "STAMP instrument SEC filing" — still no public filings, still private instrument
|
|
- "DeFi vs. TradFi capital allocation quality comparison academic study" — still no systematic comparison; mechanisms too new for controlled study
|
|
- "Futarchy academic literature 2026 new papers" — Frontiers DeSci paper is the only new empirical work found; not a field-level shift
|
|
|
|
### Branching Points (one finding opened multiple directions)
|
|
|
|
- **Third Circuit's "DCM trading" field preemption:** Direction A — Does MetaDAO need to consider DCM registration to access federal preemption protection? (Operational/regulatory question.) Direction B — Is the 3rd Circuit's narrow field definition actually GOOD for decentralized on-chain futarchy, because it keeps on-chain protocols outside CFTC's jurisdiction entirely? (Regulatory arbitrage angle.) Pursue Direction B first — if on-chain protocols aren't DCMs, they're not subject to CFTC ANPRM rulemaking either. Regulatory arbitrage via structural decentralization may be stronger protection than DCM registration.
|
|
- **Hanson's randomization fix for decision selection bias:** Direction A — Propose KB claim that the fix addresses timing bias but not payout-structure bias (Rasmont survives). Direction B — Consider whether MetaDAO's actual mechanism (conditional token pricing, TWAP-based governance) implements any of Hanson's mitigations implicitly. Does MetaDAO's pass/fail binary reduce selection bias by limiting the option space? Pursue Direction B — it's empirically testable against MetaDAO's existing mechanism design.
|