67.5pp headroom to stress boundary (98%)
Blockworks Advisory illustrative proof of concept for the Curve Risk Assessment and Market Monitoring proposal. Figures and tiers support review discussion.
Utilization is 30.5%. It is below the 90% efficient band threshold.. Borrow crvUSD (WETH collateral).
0x5AE28c9197a4a6570216fC7e53E7e0221D7A0FEF
Updated Sep 25, 2026, 3:24 PM
67.5pp headroom to stress boundary (98%)
Soft liq share
0.0%
$45k LLAMMA inventory (mechanical conversion)
Available liquidity
$58k
$83k total supplied
Binding dimension
L2 80
LLAMMA soft liquidation · live data only
Risk dimensions
Aggregate tier is a triage summary. Each dimension has its own score and reading notes. Binding uses live status dimensions only.
Posture labels are monitoring thresholds for review. Partial enrichment cannot become the binding headline. See Methodology for the full register.
Live binding dimension L2
30.5% utilized · $58k available
Utilization is below the 90% efficient band threshold with $58k still available to borrow.
$45k in LLAMMA · 0.0% converted to crvUSD
LLAMMA holds $45k with 0.0% already converted to the borrowed asset. Soft liquidation depth is present but not dominant.
0.10% vs market · source DEX + CEX blend
Curve-reported WETH price tracks the external market within 0.10%. Oracle integrity looks healthy on this snapshot.
$546.6m DEX TVL · 21613.1× borrowed
External DEX depth for WETH is $546.6m across top pools (21613.1× outstanding borrows).
Semilog · borrow 0.74% (band 0.10% to 70.00%)
Semilog policy rates rise continuously with utilization (no hard kink). Current borrow APY is 0.74% within min 0.10% and max 70.00%.
30d fee stress 0.09× hist median · Ethereum
Ethereum fee regime is near historical norms (0.09× median). Liquidation execution costs look manageable on this snapshot.
Advanced risk analytics
Three views. Cross section shows where this market sits versus peers today. Stress shows what a chosen hypothetical shock does to util or soft liq inventory. Gas shows how liquidation execution fees have looked historically. Methodology.
Peer distributions, policy curve shape, and DEX exit venue concentration.
Where this market sits versus 23 listed LlamaLend markets (today's cross section). Soft liq and free liq sit as quartile text on the tiles. Only utilization keeps a full box plot.
Util percentile
P0
30.5% util · higher P = tighter
Soft liq percentile
P0
0.0% converted · peers Q1 0 · med 0 · Q3 0
Free liq percentile
P96
69.5% free · peers Q1 19 · med 27 · Q3 35
Utilization % (primary peer distribution)
How borrow APR rises with utilization under the on chain semilog IRM (rate = min × (max/min)^util). A steeper high util slope means rate pressure accelerates if utilization keeps climbing.
Current util
30.5%
Implied borrow APR
0.71%
Headroom to kink
59.5 pp
Daily loss distribution (positive equals a down day). Historical VaR and CVaR use the upper (right) tail of losses. Status is from price history. Sample size is n=3109 from 2018-02-13 to 2026-09-25. GARCH is a parametric regime estimate. Treat it cautiously.
CVaR₉₉ hist (1d)
23.57%
Upper tail historical simulation
CVaR₉₉ GARCH (1d)
1686.71%
Theoretically motivated · parametric
VaR₉₅ hist (1d)
7.07%
Ann. vol σ
106.2%
How concentrated exit liquidity is among sampled DEX pools (top 20 DEX pools (shares renormalized within sample)). Index 11 on house 0 to 100 scale (sum of squared shares × 100).
Top N concentration
11
0 to 100 house scale within sample
Top pool share (of sample)
20%
20 pools in sample · top pool 20% of sample TVL. Not full market HHI.
Structural PD (30d) · Theoretically motivated
2.09% to 13.59%
Asset value V is $45k. Default boundary D is $27k.
Base drift μ is 0%. Vol band is shown above.
Drift sensitivity: μ=0% → 7.22% · μ=3% → 7.11% · μ=5% → 7.04%
LlamaLend Merton uses on-chain borrower collateral mark-to-market (sum of user_state residual collateral + converted inventory). Unreliable under oracle staleness or thin liquidity. On-chain sum of user_state across 8 loans (residual collateral + converted inventory).
Single hypothetical events. Soft liq conversion under price drops is illustrative only. Shock sizes may use collateral loss quantiles when history exists.
If lenders withdraw a share of supply, how high does utilization go? Marks crossings of monitoring kink / stress bands. Accounting identity under a fixed borrow book.
Util after shock
38.2%
Headroom to stress
59.8 pp
Remaining available
$41k
Flags
Clear
X is percent of supply withdrawn. Y is utilization after the exit. Current util is 30.5%.
This view is illustrative only. The conversion path is a convex stress proxy on residual inventory. Shock sizes use collateral loss quantiles when history is available.
Converted share
43.9%
Baseline 0.0% → +43.9 pp
Converted USD
$20k
Residual collateral
$25k
Vs market borrows
77.6%
Illustrative conversion only. Inventory $45k · already converted $0 (0.0% today).
Realized Ethereum gas fees over the last 90 days (chain gas history for execution context).
Historical fee distribution over the last 90 days. Median and P75 are the primary typical range framing because fees are right skewed. Mean±1σ is shown only as a secondary reference.
Hist median
0.00100 ETH
Hist P75
0.00240 ETH
Stress ratio 30d
0.09×
90 daily observations · latest 2026-09-24T00:00:00+00:00. Mean±σ omitted as primary band because fee distributions are right skewed.
Market risk recap
| What is it? | Curve LlamaLend one way market WETH / crvUSD. Borrowers post WETH as collateral to borrow crvUSD. |
|---|---|
| Liquidity posture | Utilization is 30.5%. It is below the 90% efficient band threshold.. Available to borrow $58k of $83k supplied. |
| LLAMMA inventory | $45k is held in the LLAMMA AMM. 0.0% is already converted to crvUSD. 100.0% is still WETH. |
| Binding dimension | L2 LLAMMA soft liquidation (80/100). $45k in LLAMMA · 0.0% converted to crvUSD |
| Rates | Borrow APY 0.74% · Lend APY 0.23%. |
Main risks
Contracts
Vault, controller, and AMM (LLAMMA) addresses for this one way market.