> ## 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. # Peg risk 101 > What it means for an asset to be 'pegged', why pegs break, and how peg risk shows up across the Solana asset classes Pegana tracks. A "peg" is a promise. An asset is **pegged** to a target — \$1, the SOL it represents, or the NAV of an underlying fund — because there is some mechanism that lets a holder redeem the asset for that target value, or arbitrage the price back when it drifts. **Peg risk** is the gap between the promise and the redemption path actually working. ## Why pegs break In every depeg event we've documented, the cause is one of three things: 1. **The redemption mechanism failed.** USDC during the SVB weekend (the bank holding the reserves was in receivership). UST in May 2022 (the mint/burn flywheel went reflexive). Any LST whose stake pool pauses redemptions. 2. **The market priced doubt faster than redemption could catch up.** mSOL in June 2022 — Marinade's on-chain redemption never moved, but a thin DEX pool printed a panic price for four minutes and an oracle ingested it. 3. **The mechanism was always asymmetric.** Aave's GHO at launch (mint cap, no PSM) floated below par for 8 months because arbitrage could only push the price down. The first kind is the one most monitors are designed for. The second and third hurt more, because they look like normal market noise until they don't. ## Seven asset classes, seven peg shapes Pegana tracks 63 active assets across all seven peg shapes below (`assets.toml` keys in parentheses) — FX-pegged is live too (EURC, monitoring\_only). Each shape has a different way of breaking. | Class | Peg target | Common break mode | | ------------------------------------- | ------------------------------------------- | -------------------------------------------- | | **SOL-pegged LSTs** (`lst`) | Stake-pool `sol_value` | Thin venue stress, validator slashing | | **Fiat stables** (`stable_fiat`) | \$1 redemption claim | Reserve composition shock (SVB-style) | | **Yield-bearing** (`stable_yield`) | Issuer-published NAV | NAV itself drops; negative funding for sUSDe | | **CDP-collateralized** (`stable_cdp`) | \$1 backed by collateral above CR floor | CR slides below 130%; SOL volatility | | **Delta-neutral** (`stable_dn`) | \$1 via funding-arb perp short | Funding regime flip; basis shock | | **FX-pegged** (`stable_fx`) | Issuer redemption claim in non-USD currency | FX volatility, issuer liquidity | | **Leveraged synthetic** (`synth_lev`) | Design-intended leverage ratio | Mechanism re-balance, oracle stress | See [LST peg drift](/docs/concepts/lst-peg-drift) for a deep read on the SOL-pegged LST case, and [yield-bearing NAV oracles](/docs/concepts/yield-bearing-nav-oracles) for the yield-bearing case. ## "Depeg" isn't binary A USDC trading at $0.998 has a small temporary spread driven by a Curve pool imbalance — not a depeg. USDC trading at $0.87 over a weekend, with intrinsic at \$1.00 and no redemption path until Monday — that's a depeg. The difference is **state** and **persistence**. Pegana resolves this with a 5-state FSM (PEGGED / DRIFT / DEPEG / CRITICAL / BLACK\_SWAN), hysteresis dwell timers (30s entry, 120s exit), and per-asset thresholds calibrated to each asset's noise profile. A 30 bps spread is DRIFT on jitoSOL and noise on INF. See [state machine](/docs/concepts/state-machine) for the FSM, and [hysteresis FSM](/docs/concepts/hysteresis-fsm) for why entry and exit thresholds differ. ## What Pegana is not * **Not a stop-loss oracle.** We publish state, not "sell now" signals. * **Not a single-source liquidation feed.** Cross-check with Pyth/Switchboard for liquidation pricing. See [honesty about limits](/docs/methodology/honesty). * **Not a price oracle.** For "fair price," use Pyth or Switchboard. We are a *state* oracle — the question we answer is "is this asset still doing what it promises." ## Where to read next The two-numbers framing every Pegana signal is built on. The five states, transitions, and what each state means. Why LSTs drift differently from stables, and what the signal looks like. Five real depegs analyzed end-to-end.