This is a concept reference from the September 2025 design, archived as part of the cHYPE concept paper. There is no live cHYPE vault, no live APR, and no deposits open. Numbers below describe how the model would have worked at the time the paper was written — they are not a prediction or an offer. For where Compost is today, see the homepage.
Deposit HYPE. Vault routes it. cHYPE accrues value.
The receipt token in the design is cHYPE. As builder fees and staking rewards accrue to the vault, the cHYPE/HYPE exchange rate climbs — burning cHYPE returns the underlying HYPE plus accrued share of fees. Exchange-rate model, not rebasing. (See the vault demos for a working interaction of this mechanic on testnet.)
Two sources. One receipt token.
| Source | Target allocation | Indicative range |
|---|---|---|
| Validator staking | 20-30% | ~2.5% (baseline) |
| HIP-3 strategies | 50-80% | 0-15%+ (variable) |
Validator staking sits as the floor. HIP-3 routing is the variable upside. In the design, both would have compounded into a single cHYPE share price.
Five ways to touch the HIP-3 fee pool
HIP-3 markets require 500K HYPE to deploy. The deployer captures ~50% of trading fees. There are several ways a vault could route into that fee pool — the design considered the following five:
| Mechanism | What it is | Growth Mode | Standard Mode |
|---|---|---|---|
| Pooled wrapper | Stake into a pool with rights to market revenue (e.g. Kinetiq kmHYPE) | 0.5-2% | 4-8% |
| LP vault | Deposit into a market-making vault (e.g. Ventuals VLP) | -5% to +15% | -5% to +20% |
| Stake provider | Provide stake to a deployer, split fees (e.g. Felix arrangements) | 0.5-2% | 5-10% |
| Collateral yield | Use reward-bearing collateral (e.g. USDe on HyENA) | 5-15% | 5-15% |
| Direct deployment | Stake 500K HYPE, operate market, keep 50% of fees | 1-3% | 8-15%+ |
Default mix in the design: pooled wrappers (e.g. Kinetiq kmHYPE) and LP vaults (e.g. Ventuals VLP), with collateral yield where it makes sense and stake-provider deals only if terms were transparent.
At scale: direct deployment becomes possible past 500K+ HYPE.
LP vault caveat: market-making can lose money. VLP-style yield depends on trading P&L, not a clean fee share — the design sized this smaller for that reason.
$87.32B traded — and most of it ran cheap
HIP-3 markets have done $87.32B cumulative volume and are still growing weekly. The catch: most volume runs in Growth Mode, a 90% fee reduction. Figures as of April 14, 2026 (aggregate dashboards, e.g. Dune — for live ecosystem data, see the HIP-3 board).
Growth Mode (now): $87.32B × 0.45 bps = $3.93M total fees Builder share (50%) = $1.96M available to deployers Standard Mode: $87.32B × 4.5 bps = $39.3M total fees Builder share (50%) = $19.6M available to deployers
10x difference. Same volume. Same markets. Only variable: fee mode.
Now: $87.32B → $19.6M/year (if Standard Mode) 3x: $100B → $22.5M/year 30x: $1T → $225M/year Context: Nasdaq futures do $200B+ daily.
HIP-3 is early. Volume is ramping. As markets exit Growth Mode and Hyperliquid captures more global flow, the fee pool scales with it. The cHYPE design assumed this trajectory — whether a vault should chase it, and how, is a separate question, and one we stepped back from for now.
What the design's APR ranges looked like
HYPE holders who wanted directional exposure to HIP-3 builder fees without selling spot. 5-7% base would have been respectable, 10-15% upside was the bet, 2-4% floor was the honest worst case — wagering on HIP-3 growth, not promising a return. None of these are live numbers.
Plain HYPE validator staking pays ~2.5%. The product test in the design was simple: if cHYPE can't consistently beat that, there's no reason to wrap it.
A fee design built around the staking floor
The design pinned protocol take to the part of yield that beats validator staking — so a vault that didn't outperform vanilla staking would earn almost nothing.
Performance fee: 15% of returns above the HYPE validator-staking rate (~2.5%).
Management fee: 0.5%/year, charged on the HIP-3 portion only.
No deposit fees. No withdrawal fees.
Gross vault APR: 8% HYPE staking rate: 2.5% Excess return: 5.5% Performance fee: 15% × 5.5% = 0.825% Management fee: 0.5% × 50% = 0.25% Net to depositor: ~6.92%
What the protocol would have earned
| AUM | Vault APR | Protocol Revenue |
|---|---|---|
| $50M | 4% | ~$180K/year |
| $100M | 6% | ~$775K/year |
| $100M | 8% | ~$1.1M/year |
| $250M | 8% | ~$2.7M/year |
At low AUM and low APR the protocol barely covers costs. At scale with decent realised yield, it's self-sustaining. The financial logic behind the design — not a forecast.
What this concept ran into
The design was internally consistent. The reasons it isn't a live product are honest ones, worth naming:
• Fee routing complexity. The path from a HIP-3 trade to a depositor's cHYPE rate touches multiple counterparties — deployers, market-makers, custody. Each link adds operational and trust cost.
• Counterparty / default risk. Pooled wrappers and stake-provider deals carry credit risk that's hard to price up front and hard to communicate to a depositor.
• HIP-3 literacy bar. Outside HL-native audiences the model is hard to compress into a one-line pitch — and the people best placed to evaluate it are also the ones best placed to do it themselves.
None of that makes the underlying fee pool less real. It does mean a credible product needs HL-aligned operators, not a solo build. See the homepage for what Compost is doing now — live HIP-3/HIP-4 boards, vault-mechanic demos — and the cHYPE concept paper for the original full thesis.