How we rank LSTs

Our Approach

LST Rankings aims to help DeFi users make informed decisions about which Ethereum Liquid Staking Token to use. We believe that TVL alone is a poor proxy for quality — a token's oracle security, exit costs, peg stability, and fee structure matter just as much.

All data is fetched daily from primary sources (on-chain contracts, protocol APIs) rather than relying on secondary aggregators wherever possible. We cross-validate data between multiple sources and flag anomalies.

LST Score (0–100)

Each token receives a composite score based on five equally-weighted categories, 20 points each. Scores update daily as new on-chain data is fetched.

Category	Weight	What it measures	How it's scored
Yield	20 pts	30-day rolling APY	Linear normalization across all tokens. Highest APY = 20, lowest = 5.
Security	20 pts	Oracle mechanism	Trustless (EIP-4788) = 20, Committee = 12, Centralized = 5.
Peg	20 pts	Peg deviation from fair value	Inverse of absolute deviation, capped at 1%. Tighter peg = higher score.
Liquidity	20 pts	Slippage on 100 ETH exit	Inverse of slippage %, capped at 5%. Lower slippage = higher score.
Fees	20 pts	Performance fee on rewards	Inverse of fee %. 0% fee = 20, 25% fee = 5.

Missing data for any category defaults to the minimum score (5 points). All categories carry equal weight to avoid subjective bias in the ranking.

Metrics Explained

APY (Annual Percentage Yield)

The annualized yield from staking rewards. We show rolling averages over 30, 90, and 365 days to smooth out daily fluctuations. APY data is sourced from DeFi Llama's yield aggregator, which tracks actual pool performance.

Peg Deviation

Measures how closely the LST's market price tracks its fair value (the canonical exchange rate from the protocol's smart contract). Calculated as:

peg_deviation = (market_price - fair_value) / fair_value * 100%

A positive deviation means the token trades at a premium; negative means a discount. Tight pegs (close to 0%) indicate healthy market confidence and sufficient liquidity.

Exit Slippage

The actual cost of selling an LST for ETH on DEX markets. We simulate swaps at 4 tiers (1, 10, 100, and 1000 ETH) by querying Uniswap V3 and Curve pool contracts directly on-chain. For each token, we try all Uniswap V3 fee tiers and select the route that returns the most ETH. This reveals hidden exit costs that TVL alone doesn't show.

Oracle Mechanism

How the protocol gets off-chain data (validator balances, rewards) onto the blockchain. Rankings from most to least secure:

• Trustless (EIP-4788) — Beacon chain state root is exposed to the EVM, enabling cryptographic verification of validator balances directly on-chain. No oracle committee needed. (OETH)
• Committee (Oracle DAO) — Multiple independent operators reach consensus on validator balances. Robust, but requires trusting a majority. (stETH, rETH, ETHx)
• Centralized — A single entity controls the oracle that reports validator balances. (cbETH, WBETH, sfrxETH, swETH, mETH, ankrETH)

Token Type

Rebasing tokens (stETH, OETH) increase your balance daily to reflect rewards. Simple to understand but can break smart contract integrations. Value-accruing tokens (rETH, cbETH, etc.) maintain a fixed balance while the exchange rate increases.

Decentralization

• Decentralized — Permissionless node operators, no single entity controls validators. (Rocket Pool, Stader, Origin)
• Semi-Decentralized — Curated or permissioned node operator set. (Lido, Frax, Swell, Mantle, Ankr)
• Centralized — Single entity operates all validators. (Coinbase, Binance)

Data Sources

Source	Data	Frequency
DeFi Llama	TVL, APY, yield history	Daily
CoinGecko	Market prices, market cap, volume	Daily
Uniswap V3 / Curve (on-chain)	DEX swap quotes, slippage simulation	Daily
Ethereum RPC	Canonical exchange rates from smart contracts	Daily

Accuracy Safeguards

• Cross-validation: We compare data from multiple sources. TVL from DeFi Llama is checked against calculated values (supply × price).
• Validation rules: Hard bounds on every metric (e.g., APY must be between -2% and 20%, day-over-day TVL change must be <30%).
• Anomaly flagging: Data that fails validation is still recorded but flagged for manual review.
• Idempotent writes: Daily fetches are safe to re-run without creating duplicates.
• Audit logging: Every data fetch is logged with timestamps, record counts, and error details.

API Access

All data is available via our free JSON API:

• GET /api/rankings
  Current rankings
• GET /api/tokens
  Token attributes
• GET /api/token/{symbol}/history
  Historical data
• GET /api/health
  Health check