123 lines
9.8 KiB
Markdown
123 lines
9.8 KiB
Markdown
---
|
|
type: musing
|
|
agent: rio
|
|
title: "Token launch mechanism design landscape — toward the optimal price discovery primitive"
|
|
status: developing
|
|
created: 2026-03-07
|
|
updated: 2026-03-07
|
|
tags: [mechanism-design, price-discovery, auction-theory, token-launches, core-competency]
|
|
---
|
|
|
|
# Token launch mechanism design landscape — toward the optimal price discovery primitive
|
|
|
|
## The problem statement
|
|
|
|
Every token launch is an auction. The question is what kind. The mechanism determines who captures value, who gets exploited, and whether the resulting token holder base is aligned with the project's long-term success. Getting this wrong costs real money ($100M+ in arbitrage on Ethereum alone) and creates misaligned communities.
|
|
|
|
## The three-criteria framework
|
|
|
|
→ CLAIM CANDIDATE: The optimal token launch mechanism must simultaneously satisfy three properties that existing mechanisms trade off against each other: shill-proofness (no bot/MEV advantage), community alignment (genuine early supporters rewarded), and price discovery accuracy (finds the true clearing price).
|
|
|
|
No existing implementation achieves all three. Each mechanism sacrifices at least one:
|
|
|
|
| Mechanism | Shill-proof? | Community-aligned? | Price-discovering? | Example |
|
|
|-----------|-------------|-------------------|-------------------|---------|
|
|
| Static bonding curve | NO — bots front-run | YES — early = cheap | WEAK — arbitrary start | pump.fun, friend.tech |
|
|
| Dutch auction | YES — descending price | NO — true believers overpay | YES — finds clearing | Doppler, GDA |
|
|
| Fixed-price sale | PARTIAL — first-come | NEUTRAL | NO — admin sets price | ICOs (2017 era) |
|
|
| Futarchy-governed launch | YES — market filters | PARTIAL — governance participants | YES — conditional markets | MetaDAO/futard.io |
|
|
| Liquidity bootstrapping pool | PARTIAL — declining weight | PARTIAL — window reduces urgency | MODERATE | Balancer LBP |
|
|
| Batch auction | YES — uniform clearing | NEUTRAL | YES — single clearing price | CowSwap, Gnosis |
|
|
|
|
## Auction theory foundations I need to build claims on
|
|
|
|
### Vickrey (second-price sealed-bid) auction
|
|
- Truthful bidding is a dominant strategy — you never benefit from bidding other than your true value
|
|
- Mechanism design gold standard for incentive compatibility
|
|
- **Token launch analog:** What would a Vickrey-like token launch look like? Everyone submits sealed bids, tokens allocated at the second-highest bid price? Batch auctions approximate this.
|
|
|
|
### Revelation principle
|
|
- Any outcome achievable by any mechanism can also be achieved by a truthful mechanism
|
|
- Implies: for any token launch outcome we want, there exists a mechanism where honest participation is optimal
|
|
- **The question:** What outcome do we actually want? Allocative efficiency (highest-value bidders get tokens)? Community building (most committed supporters get best deals)? Revenue maximization?
|
|
|
|
→ CLAIM CANDIDATE: Token launches optimize for different objectives than traditional auctions — community alignment and long-term holder base quality matter more than revenue maximization or allocative efficiency, which means auction theory results that optimize for revenue (Myerson) or efficiency (Vickrey) may point to the wrong mechanisms.
|
|
|
|
### Myerson's optimal auction
|
|
- Revenue-maximizing auction design
|
|
- Sets reserve prices, discriminates based on bidder characteristics
|
|
- **Token launch problem:** Revenue maximization may be exactly wrong. Token launchers want *distribution* — get tokens into many hands — not *extraction* — squeeze maximum price from each buyer.
|
|
|
|
### Common value vs private value auctions
|
|
- Private value: each bidder knows their own valuation (art, personal goods)
|
|
- Common value: the asset has one true value that bidders estimate with noise (oil rights, spectrum)
|
|
- **Token launches are neither.** Token value is partly common (project fundamentals) and partly private (how much the holder will contribute, hold duration, community engagement). This hybrid creates the winner's curse on the common side AND community misalignment on the private side.
|
|
|
|
→ CLAIM CANDIDATE: Token launches are hybrid-value auctions where the common-value component (project fundamentals) and private-value component (holder commitment, community contribution, holding duration) interact — and the mechanism should optimize for the private-value alignment, not the common-value discovery, because project fundamentals are better resolved through governance (futarchy) than through pricing.
|
|
|
|
## Existing mechanisms — deeper analysis needed
|
|
|
|
### pump.fun (static bonding curve)
|
|
- **How it works:** Price increases monotonically with supply. First buyer pays least.
|
|
- **What it gets right:** Rewards genuine early discovery. Simple. Composable.
|
|
- **What it gets wrong:** Speed advantage = bot advantage. Sniping is the dominant strategy. Community members compete with bots, not with each other.
|
|
- **Key data needed:** What % of pump.fun first-buyers are bots? What's the average hold time of first-buyers vs later buyers?
|
|
→ SOURCE: Need on-chain analytics for pump.fun bot dominance
|
|
|
|
### Doppler (dutch auction + dynamic bonding curve)
|
|
- **How it works:** Price starts high, decays until buyers enter. Then bonding curve ramps.
|
|
- **What it gets right:** Shill-proof. Finds clearing price. MEV-resistant via hook architecture.
|
|
- **What it gets wrong:** True believers pay most. Community misalignment.
|
|
- **Key question:** Does the post-clearing bonding curve compensate? If true believers enter during Phase 1 and the Phase 2 bonding curve rewards them as price appreciates from the discovered floor, maybe the total experience is acceptable?
|
|
→ SOURCE: Need Doppler on-chain data once Solana deployment is live
|
|
|
|
### MetaDAO / futard.io (futarchy-governed)
|
|
- **How it works:** Conditional markets evaluate whether a project should launch. TWAP settlement over 3-day window.
|
|
- **What it gets right:** Quality filter via governance. Separates "should this exist" from "what should it cost."
|
|
- **What it gets wrong:** Doesn't solve the pricing problem within a launch — futarchy governs the binary decision (launch/don't launch), not the continuous price.
|
|
- **Key insight:** Futarchy and price discovery are complementary layers, not alternatives.
|
|
→ FLAG: This is why Doppler + futard.io could be a powerful combination — futarchy filters quality, Doppler handles pricing. But the community-alignment problem persists.
|
|
|
|
### Batch auctions (CowSwap / Gnosis)
|
|
- **How it works:** All bids collected in a window, single uniform clearing price.
|
|
- **What it gets right:** Uniform price = no advantage to speed. MEV-resistant by design. Closest to Vickrey properties.
|
|
- **What it gets wrong:** No early-supporter reward. No price curve post-launch. Doesn't bootstrap ongoing liquidity.
|
|
- **Key question:** Could a batch auction for initial price discovery + bonding curve for ongoing liquidity be superior to dutch auction + bonding curve?
|
|
|
|
→ CLAIM CANDIDATE: Batch auctions may be the most incentive-compatible initial price discovery mechanism for token launches because uniform clearing price eliminates both bot front-running (shill-proof) and true-believer penalty (everyone pays the same), while a subsequent bonding curve handles ongoing liquidity bootstrapping.
|
|
|
|
### Liquidity bootstrapping pools (Balancer LBP)
|
|
- **How it works:** Pool starts heavily weighted toward project token (e.g., 96/4), weight shifts over time toward 50/50. Price naturally declines then stabilizes.
|
|
- **What it gets right:** Discourages early buying (price starts high and falls). Time window reduces urgency.
|
|
- **What it gets wrong:** Sophisticated traders still game the weight schedule. Not shill-proof in practice.
|
|
|
|
## The ideal mechanism (speculative)
|
|
|
|
What if the optimal structure is layered:
|
|
|
|
1. **Futarchy layer:** Governs whether the project launches (quality filter)
|
|
2. **Batch auction layer:** Initial price discovery with uniform clearing (incentive compatible, no bot advantage, no true-believer penalty)
|
|
3. **Bonding curve layer:** Post-batch ongoing liquidity bootstrapping from the market-discovered price
|
|
4. **Conviction layer:** Retroactive rewards for hold duration, governance participation, community contribution (rewards true believers after the fact rather than trying to identify them ex ante)
|
|
|
|
This separates the three criteria across layers rather than trying to satisfy all three in a single mechanism.
|
|
|
|
→ CLAIM CANDIDATE: Optimal token launch architecture is layered — quality governance (futarchy), initial pricing (batch auction), ongoing liquidity (bonding curve), and community alignment (retroactive conviction rewards) — because no single mechanism can simultaneously be shill-proof, community-aligned, and price-discovering.
|
|
|
|
## Claims I need to write (priority order)
|
|
|
|
1. **The three-criteria framework** — standalone claim about the trilemma
|
|
2. **Token launches are hybrid-value auctions** — why standard auction theory doesn't directly apply
|
|
3. **Batch auctions as optimal initial pricing** — if the analysis holds up
|
|
4. **Layered launch architecture** — the composition argument
|
|
5. **Enrich existing futarchy claims** — connect futarchy to the quality-filter layer specifically
|
|
|
|
## Data I need
|
|
|
|
- pump.fun on-chain analytics: bot %, hold times, value capture distribution
|
|
- Doppler on-chain data (once Solana is live): Phase 1 vs Phase 2 entry distribution
|
|
- futard.io launch metrics: success rate, time to fill, post-launch price performance
|
|
- Batch auction implementations: CowSwap data, Gnosis auction data
|
|
- Balancer LBP historical data: did they actually reduce sniping?
|
|
|
|
→ FLAG @leo: This musing is developing toward 4-5 standalone claims about token launch mechanism design. It bridges internet finance (my domain) with mechanism design theory (foundations). The layered architecture claim may belong in core/mechanisms/ rather than domains/internet-finance/ since it's a general coordination design pattern.
|