How to Analyze Curve On-Chain Data: Liquidity Concentration and Pool Health Metrics
When analyzing Curve on-chain data, you only need to focus on three core priorities: whether the pool's liquidity is concentrated near the current market price, if transaction fees remain stable, and if any lending positions face risk of liquidation. Curve's mechanism differs significantly from Uniswap: it does not spread liquidity evenly across the entire price range, but concentrates it around an "internal anchor price" that adjusts slowly following market movements. Mastering these layers of data will let you accurately judge if a pool is "operating healthily" or showing early signs of failure.
1. Check Liquidity Concentration: Core Metric for Curve V2 Pools
What it does: Verify if a Curve V2 pool (such as 3pool, stETH/ETH pool) has liquidity concentrated near the current market price. The more concentrated the liquidity, the lower the slippage and the better the trading experience for users.
How to implement:
Curve V2 pools control liquidity distribution via two core parameters:
A (Amplification Coefficient): Determines how tightly liquidity is concentrated around the central zone. A higher A value means liquidity is more densely packed near the anchor price.
gamma: Controls the decay rate of liquidity as it moves away from the central zone. A higher gamma value means the liquidity distribution is wider.
Navigate to the official Curve interface or a Dune analytics dashboard to pull the following data for your target pool:
price_scale: The price point at the center of current liquidity, i.e. the fair price the pool calculates
price_oracle: The EMA price recorded by the internal oracle, used to trigger liquidity rebalancing events
Liquidity in the pool adjusts dynamically to follow the price_oracle. You need to compare the pool's liquidity concentration range against the current market price — if the deviation between the two exceeds 1-2%, the pool has not caught up with market movements yet, and trade slippage may be temporarily high.
Completion check: You can clearly state which price level the pool's liquidity is concentrated around, and confirm if the current market price falls within this concentration range.
Prerequisite: Access to the official Curve website or a Dune Curve analytics dashboard.
Common pitfalls: Curve V2 liquidity distribution is dynamic, shifting with trades and rebalancing events. A single snapshot of data at one point in time is not sufficient: you need to observe trend changes over a period of time.
2. Check Pool Balance: Watch for One-Sided Skew
What it does: Verify if the asset ratio inside the pool is severely imbalanced. If one asset is heavily withdrawn while another accumulates excessively, it signals the market is dumping one side of the pair, which is an early warning sign of pool trouble.
How to implement:
Open the Curve pool details page to view the asset proportion data:
Healthy status: The ratio of two or more assets stays close to equal distribution
Danger signal: The share of one asset keeps rising continuously while the share of another keeps falling
Take the USDC/USDT pool as an example: if USDC's share rises from 50% to over 80%, it means the market is selling large amounts of USDC for USDT, and the pool has effectively become a "dumping ground for USDC". In this scenario, the risk of holding LP tokens in the pool rises sharply, as the majority of your position is now made up of the asset the market is offloading.
On Curve's UI, asset proportions are usually displayed via pie charts or bar charts. If any single asset accounts for over 70% of the total pool value for multiple consecutive days, you should treat it as a high risk signal.
Completion check: You can clearly identify which asset makes up the largest share of the current pool, and confirm if this proportion is expanding or shrinking over the past several days.
Prerequisite: Access to the pool's asset proportion historical data.
Common pitfalls: Only checking static asset ratios without tracking trends. Short-term imbalance (lasting a few minutes) is normal and represents arbitrage opportunities, while sustained imbalance (lasting hours to days) is the actual risk factor.
3. Verify Transaction Fees and Real Usage: Distinguish Genuine Demand From Incentive Subsidies
What it does: Curve's core value comes from real stablecoin swap and trading demand. You can use transaction fee data to judge if a pool is being actively used by real users, or only kept afloat by CRV token emission rewards.
How to implement:
Curve's share of total DEX transaction fees is a key metric to measure real usage. Data from the end of 2025 shows that Curve's share of total Ethereum DEX fees has risen from roughly 1.6% one year prior to 44%, generating approximately 15.1 million USD in fees over the past 30 days.
Specific query steps:
Open DefiLlama or the official Curve data dashboard to view the target pool's daily trading volume and daily fee revenue
Compare the pool's TVL against its daily trading volume: pools with high TVL but low trading volume indicate capital is sitting idle rather than being actively utilized
Check Curve's Gauge weight distribution: pools that receive more veCRV votes get a larger share of CRV emissions. If most of a pool's yield comes from token emissions rather than transaction fees, it is relying on subsidies to attract LPs, and will collapse once emission support is removed.
Completion check: You can clearly describe the yield composition of the target pool, confirming if it generates revenue primarily from transaction fees, or is sustained by CRV emission subsidies.
Prerequisite: Access to DefiLlama or the official Curve data portal.
Common pitfalls: Only checking total APY without reviewing yield sources. A pool with a 50% APY where 40% comes from CRV emissions and only 10% from fees will see its APY crash immediately once emission rates are reduced.
4. Check Lending Health: Risk Metrics for crvUSD and LlamaLend
What it does: If you are a crvUSD minter or LlamaLend lending user, you need to monitor your position's Health score. Curve's liquidation mechanism works differently from other protocols: there is no fixed liquidation price, and positions are only liquidated once the health score drops to 0.
How to implement:
Navigate to Curve's LlamaLend page to view your position data:
Health: A percentage value that represents the remaining buffer before liquidation. Higher values mean safer positions.
Liquidation Protection range: A designated price interval. If your collateral price falls into this interval, the system will automatically convert assets to protect your position from full liquidation.
The health score calculation formula is:
Health = (Estimated collateral value × (1 - liquidation discount) + value of portion of price above the interval) / debt - 1
Key rule: Once the health score drops to 0%, your position will be fully liquidated and cannot be recovered.
Reference safety thresholds:
Above 50%: Relatively safe, but still requires regular monitoring
30-50%: Medium risk, recommended to take proactive action
10-30%: High risk, strongly advised to repay part of the debt or add more collateral
Below 10%: Critical risk, immediate action required
0%: Liquidation has already been triggered
Note: Once your position enters the liquidation protection range, even if the collateral price rebounds, your health score may still continue to drop due to asset conversion losses.
Completion check: You can state the current health score of your lending position, and clearly know how far it is from the critical risk warning threshold.
Prerequisite: You have already opened a lending position on LlamaLend.
Common pitfalls: Only tracking collateral price movements without monitoring health score. Even when price stays unchanged, accrued interest will slowly reduce your health score over time, so long-term lending positions require continuous monitoring.
5. Use Third-Party Tools for Comprehensive Risk Assessment
What it does: Leverage market health scores from third-party platforms like LlamaRisk to quickly evaluate the overall risk profile of Curve's lending markets.
How to implement:
LlamaRisk provides quantitative health scores for Curve's crvUSD minting markets and LlamaLend lending markets. The scoring system covers multiple dimensions as shown below:
| Risk Category | Weight | Evaluation Criteria |
|---|---|---|
| Bad Debt Ratio | 10% | Percentage of bad debt out of total outstanding debt, score becomes 0 if the ratio exceeds 1% |
| Debt Ceiling Compliance | 30% | Check if the current debt ceiling exceeds the protocol's recommended value |
| Liquidation Efficiency | TBD | Verify if the liquidation mechanism is operating smoothly |
| Collateral Asset Characteristics | TBD | Evaluate the volatility and on-chain liquidity of the collateral asset |
Visit the LlamaRisk Risk Portal to view the detailed overall score and sub-indicators for your target market (such as the crvUSD/USDC market). A score of 0 represents maximum risk, while a score of 1 represents minimum risk.
Completion check: You can state the current 0-1 comprehensive health score of the target market, and identify which risk category has the lowest score (the largest risk factor).
Prerequisite: Access to the LlamaRisk Risk Portal page.
Common pitfalls: Relying solely on the aggregate score without reviewing underlying raw data. The final score is a high-level summary, and specific risk assessments should combine the on-chain data you collected in the previous steps for independent validation.
Next steps to take:
Spend 15 minutes today doing one simple exercise: open the official Curve website, pick a pool you frequently use (like 3pool or stETH/ETH pool), check its current asset ratio and daily trading volume. Compare these metrics to the values from one week ago: has the asset ratio developed a noticeable skew? Is trading volume rising or falling? The direction of change for these two metrics tells you far more about the real health status of the pool than any single static number.
