> ## Documentation Index > Fetch the complete documentation index at: https://pegana.xyz/docs/llms.txt > Use this file to discover all available pages before exploring further. # Hylo CDP CR — the under-collateralization signal > Hylo's hyUSD is a CDP-style stablecoin on Solana. Its peg holds as long as the collateral ratio (CR) stays above the protocol's floor. When CR drops below 130%, Stability Mode triggers — and that's the alert you actually want. Hylo runs a CDP-style stablecoin (**hyUSD**) collateralized by a basket of Solana LSTs (hyloSOL, hyloSOL+). Unlike USDC, hyUSD doesn't claim to be backed 1:1 by a dollar in a bank. It claims to be backed by SOL collateral worth more than the hyUSD outstanding. The exact ratio — how much collateral per dollar of hyUSD — is the **collateral ratio**, or CR. The peg holds because if hyUSD trades below $1, an arbitrageur can buy hyUSD on the market, redeem it for $1 of LST collateral, and pocket the difference. If hyUSD trades above \$1, they can mint by depositing collateral and sell. Either side closes the gap. But only when CR is healthy. ## What CR actually measures ``` CR = total_collateral_value_usd / hyUSD_supply ``` If Hylo holds 500,000 SOL worth $80M and 50M hyUSD are outstanding, CR = 160%. Every dollar of hyUSD is backed by $1.60 of collateral. That cushion is what lets the system absorb a SOL price drop without becoming undercollateralized. Hylo's protocol parameters define the floor: * **CR \< 130%** → **Stability Mode / DRIFT**: minting is paused, redemptions are throttled, and the Stability Pool (sHYUSD) starts absorbing liquidations of the riskiest positions * **CR \< 105%** → **CRITICAL** * **CR \< 100%** → **BLACK\_SWAN**: the system is genuinely undercollateralized and recovery depends on collateral recovery, not arbitrage ## Why market price is the wrong signal here For USDC, watching market price is sufficient — a sub-$1 print is the alert. For hyUSD, **market price can stay at $1.000 for hours while CR quietly grinds down from 160% to 135%\*\*. By the time market price reacts, you're already inside Stability Mode and the protocol is making forced unwind decisions. CR is the **leading indicator**. Market price is the **lagging indicator**. If you're watching the wrong one, you'll get to the alert too late. ## How Pegana reads CR There's no API for this — you have to decode it from on-chain state. Hylo publishes an IDL for its Exchange program. Pegana's `indexer-rs` reads Hylo's root Exchange PDA every 30s and decodes the protocol's own canonical `total_sol_cache.total_sol` — the same field Hylo's own app uses for pricing. Multiply by Pyth's SOL/USD live, divide by hyUSD supply, and you get CR with no source-of-truth lag. Per-asset thresholds for hyUSD use **CR** (not spread %). The live bands (`assets.toml:228` — `drift=130, depeg=115, critical=105, black_swan=100`) are: | State | CR band | | ------------ | ----------------- | | `PEGGED` | CR ≥ 130% | | `DRIFT` | 115% ≤ CR \< 130% | | `DEPEG` | 105% ≤ CR \< 115% | | `CRITICAL` | 100% ≤ CR \< 105% | | `BLACK_SWAN` | CR \< 100% | The `/explain hyUSD` command on `@PeganaWatchBot` walks through each transition's meaning in plain English. ## The IDL drift risk Reading on-chain state via a hardcoded layout has a known failure mode: the protocol upgrades, the account layout changes, and your decoder silently produces garbage. We guard against it by validating the account's 8-byte Anchor discriminator and layout on every read — if the discriminator doesn't match (an IDL upgrade or the wrong PDA), the read fails fast instead of decoding garbage. On that failure, Pegana does **not** flip hyUSD to a separate state: it skips the recompute and holds the last good CR rather than publishing a confidently-wrong number. That path emits a warn log and increments a Prometheus counter. Better to hold a stale-but-true read than report a confidently-wrong CR. ## Adjacent assets in the Hylo system * **hyloSOL** — the LST that backs hyUSD. Behaves like any other LST (drift in SOL terms). * **hyloSOL+** — yield-amplified Hylo LST that captures CDP-mint fees. Grows faster than hyloSOL when the system is healthy. * **sHYUSD** — staked hyUSD in the Stability Pool. NAV grows when liquidations earn yield; can drop sharply during black-swan absorption events. * **xSOL** — the leveraged-SOL synthetic that absorbs the volatility hyUSD sheds. Intentionally volatile; "peg" here means the design-intended leverage ratio. Pegana tracks all five (hyUSD, hyloSOL, hyloSOL+, sHYUSD, xSOL) and emits class-appropriate alerts for each. The Hylo design is interlocking: stress on hyUSD shows up first in CR, then in xSOL repricing, then in sHYUSD's absorption mechanic. ## What this means for builders If your protocol accepts hyUSD as collateral, you want to know the moment CR breaches 130% — when Hylo enters Mode 2 and redemptions get re-prioritised. Pegana's `/v1/assets/hyUSD/extra` endpoint exposes CR directly with timestamp; the WebSocket `/v1/ws` pushes a hyUSD update every 30s and a transition alert the moment CR crosses a threshold. ## Read next Current CR, history, threshold visualization. What can happen when CDP collateral itself depegs. ### Sources * [Hylo docs](https://docs.hylo.so) * [Hylo Exchange program (Solscan)](https://solscan.io) * [Sanctum docs — `sol_value`](https://docs.sanctum.so)