>euler v2 · modular lending research terminal

Markets

Euler research terminal · Euler Labs

Euler V2 is a vault kit, so the book is whatever anyone chose to deploy. This page reads the two sides separately: what gets supplied, and what actually gets borrowed against it.

Assets in the book275distinct tokens supplied or borrowed

Stables borrowed71.0%of debt, major stables only

No borrow market41.1%of supply, assets under 2% utilization

Top 5 supplied55.7%share of supply in 5 assets

Vault utilization58.3%supply-weighted, vaults above $1M

Vaults tracked34EVK vaults above $1M

Two sides of the book

Supplied vs borrowed, by assetGross supplied (net TVL plus borrows) next to outstanding debt, per token, summed across all 16 chains. Assets are ranked by whichever side they are larger on, so a pure collateral asset and a pure debt asset both appear. An asset with a tall supplied bar and no borrowed bar is collateral nobody lends against.

USD

Top 12 assets by their larger side · all chains · Sep 2026

Source: DefiLlama@datumlabss

Major stablecoins are 71% of what gets borrowed but only 41% of what gets supplied, a gap of 30%. Both figures use a fixed list of widely-held dollars and exclude structured or yield-bearing dollar tokens, so they are floors rather than estimates.

Supplied assets with no borrow marketAssets with a real supplied balance and essentially no lending market against them. On a vault kit this is the signature of collateral posted purely to borrow something else: Pendle principal tokens, liquid staking wrappers and structured dollar tokens.

USD

Assets above $5M supplied with under 2% utilization · Sep 2026

Source: DefiLlama@datumlabss

No borrow market means no interest income for the depositor, so this collateral is held for the leverage it unlocks rather than for yield on Euler.

What gets borrowedDebt by token, from DefiLlama's per-chain borrowed token history. The chain filter is where the two deployment stories separate: Ethereum's debt is dollar stablecoins, Monad's is split between a dollar and ETH.

USD

Outstanding debt by token · filterable by chain and period

$377.0M total

USDT029.3%

WETH18.2%

AUSD17.0%

USDC14.5%

USDT5.9%

AVALANCHEUSDC3.7%

SCUSD3.3%

Other8.0%

Source: DefiLlama@datumlabss

How the book got here

Gross supplied by assetWeekly gross supplied per token, summed across chains. Assets outside the current top 10 are grouped into one band rather than dropped, so the stack always sums to the full book. A band that appears from nothing is a new collateral franchise being onboarded, which on a permissionless vault kit can happen without any governance step.

USD

Sep 2025 to Sep 2026 · top 10 assets, everything else grouped

USDT0WETHVUSDAUSDUSDCWSTETHPT-AUSD-8OCT2026GAUSDWEETHUSDTOther

Source: DefiLlama@datumlabss

Outstanding debt by assetWeekly outstanding debt per token. Compare the band count with the supplied chart above: collateral is spread across many long-tail assets while debt concentrates into a handful of dollars, which is the asymmetry a leverage venue produces.

USD

Sep 2025 to Sep 2026 · top 10 borrowed assets, everything else grouped

USDT0WETHAUSDUSDCUSDTAVALANCHEUSDCSCUSDUSD0USD1PYUSDOther

Source: DefiLlama@datumlabss

The five largest assets hold 56% of supply and 85% of debt. 41% of everything supplied sits in assets with no borrow market at all.

Vault level

Largest vaults by gross suppliedGross supplied per vault from the yields API's `totalSupplyUsd`, which is before debt. The pool-level `tvlUsd` DefiLlama shows elsewhere is net, and on Euler that understates the largest vaults by roughly ten times.

USD

EVK vaults above $1M · Sep 2026

Source: DefiLlama yields@datumlabss

34 vaults clear the $1M floor. Everything smaller is excluded rather than aggregated, because per-vault rates are not meaningful at that size. Curator attribution is not charted anywhere on this page: DefiLlama names a curator for only 28% of vault supply, of which 99% is K3 Capital, which is too thin to rank. Full coverage needs an on-chain governor read per vault.

Utilization by vaultOutstanding debt divided by gross supply, per vault. Read it in both directions: high utilization means the vault is earning, and it also means a depositor may not be able to withdraw in size without waiting for a repayment.

Vaults above $5M supplied with an active borrow market · Sep 2026

Source: DefiLlama yields@datumlabss

Euler's largest vaults run in the high eighties and above, well beyond where a monolithic money market would sit. That is the design working, and it is also the risk.

Borrow rates by vaultBase borrow APY excluding reward tokens, for vaults with a real drawn balance. This is the price of leverage on Euler and it maps directly onto utilization: the vaults at the top of this chart are the vaults at the top of the utilization chart.

% APY

Base borrow APY, vaults with over $1M drawn · Sep 2026

Source: DefiLlama yields@datumlabss

Rates above roughly 20% are not a market price so much as a liquidity signal. They sit on vaults that are 100% drawn, where Euler's rate model has run past its kink to push borrowers to repay and pull lenders in. Read the top of this chart as vaults with no free liquidity, and the middle as what borrowing actually costs.

Where the supply yield comes fromA lender's headline APY split into base, the interest borrowers actually pay, and rewards, tokens emitted on top. Only vaults above $5M are included, because a headline rate on a $1M vault is not one anyone can get filled at. Read against the borrow-rate chart: base yield is funded by that chart, reward yield is not.

% APY

Base interest vs token rewards · vaults above $5M · Sep 2026

Source: DefiLlama yields@datumlabss

Across the 17 vaults large enough to rate, 29% of the supply-weighted yield is token rewards rather than interest, and 8 of them carry any rewards at all. Emissions can be switched off; interest cannot, so the base bar is the part that survives an incentive programme ending.

Supplied assets with no borrow marketAssets with a real supplied balance and essentially no lending market against them. On a vault kit this is the signature of collateral posted purely to borrow something else: Pendle principal tokens, liquid staking wrappers and structured dollar tokens.

What gets borrowedDebt by token, from DefiLlama's per-chain borrowed token history. The chain filter is where the two deployment stories separate: Ethereum's debt is dollar stablecoins, Monad's is split between a dollar and ETH.

Outstanding debt by assetWeekly outstanding debt per token. Compare the band count with the supplied chart above: collateral is spread across many long-tail assets while debt concentrates into a handful of dollars, which is the asymmetry a leverage venue produces.

Largest vaults by gross suppliedGross supplied per vault from the yields API's totalSupplyUsd, which is before debt. The pool-level tvlUsd DefiLlama shows elsewhere is net, and on Euler that understates the largest vaults by roughly ten times.

Utilization by vaultOutstanding debt divided by gross supply, per vault. Read it in both directions: high utilization means the vault is earning, and it also means a depositor may not be able to withdraw in size without waiting for a repayment.

Borrow rates by vaultBase borrow APY excluding reward tokens, for vaults with a real drawn balance. This is the price of leverage on Euler and it maps directly onto utilization: the vaults at the top of this chart are the vaults at the top of the utilization chart.

Largest vaults by gross suppliedGross supplied per vault from the yields API's `totalSupplyUsd`, which is before debt. The pool-level `tvlUsd` DefiLlama shows elsewhere is net, and on Euler that understates the largest vaults by roughly ten times.