The Social Vault
A pooled Bitcoin trove that removes the minimum loan size, and the risk that comes with sharing it.
The Social Vault is Maren's differentiator, and its risk. It is one large, conservatively managed MUSD trove that many users share, each holding a proportional claim on it through a per-user ledger. It exists to do one thing the protocol cannot: let a small holder borrow.
The wedge: no minimum loan size
MUSD requires a minimum net debt of 1,800 MUSD, which becomes 2,000 MUSD of total debt once the 200 MUSD gas compensation is counted. That is a minimum loan size. At the 250% collateral ratio Maren recommends, 2,000 MUSD of total debt needs about $5,000 of Bitcoin — and a small holder cannot borrow at all.
Maren's edge is not "cheaper collateral" or an "80x" headline. It is that the protocol imposes a minimum loan size at all. You cannot borrow 50 MUSD against Bitcoin on Mezo, only 2,000 or more. The pooled vault removes that minimum entirely, which is a different and stronger claim than a discount on collateral.
Concretely, the vault lets someone borrow $25 against $77 of BTC, a position the protocol would reject outright. That is the whole product wedge in one sentence.
Who should not use it
The vault is for people the protocol prices out. Anyone with enough collateral for a trove of
their own is steered to a direct trove instead, which carries no pooling risk at all. The
SDK's assessBorrowRoute() makes this decision, and the interface explains it:
import { assessBorrowRoute } from '@maren/sdk';
// If the collateral supports a >= 1,800 MUSD direct trove, prefer it: no pooling risk.
const decision = assessBorrowRoute(collateralWei, priceWei);
// -> { eligible: true, route: 'direct' } enough for your own trove
// -> { eligible: true, route: 'pooled' } below the minimum, the vault is the only route
// -> { eligible: false, reason: 'Below the 0.001 BTC minimum deposit.', shortfallWei }
Maren loses pooled volume by routing capable users away from the pool. That is the correct trade.
How the shared position is kept safe
The vault keeps a wide margin above the protocol's liquidation cliff. These are the constants the deployed contract actually checks, transcribed into the SDK so the interface can evaluate the same conditions before you sign.
| Property | Value |
|---|---|
| Minimum deposit | 0.001 BTC |
| Protocol liquidation threshold (MCR) | 110% |
| Maren internal liquidation CR | 140% |
| Maren minimum healthy CR (borrow floor) | 180% |
| Maren target CR | 250% |
| Internal liquidation penalty | 5% |
The contract gives a KEEPER_ROLE holder the ability to top up collateral, repay debt, and
internally liquidate unhealthy users by seizing collateral at a 5% penalty and repaying their share
of pooled debt. A user must fall 110 percentage points below the 250% target — to 140% — before
Maren's own keeper may liquidate them, and 140 points, to 110%, before the MUSD protocol has any
claim on the trove.
On matsnet that role is held by a single team-operated key, every keeper action so far has been sent by hand, and no automated keeper is running yet. Automating it, then decentralising who may hold the role, is a prerequisite for the vault holding real money.
Borrowing pauses before withdrawal does. This is a hard requirement in the contract with a
dedicated test: withdraw, repay and closePosition carry no pause modifier, so no guardian
key can trap collateral. One state does freeze exits and it is not a pause — if the pooled trove
is liquidated, closed or fully redeemed it leaves Mezo's active set and every vault mutation
halts, including withdrawals, with no wind-down path built. See Risk R-10.
The risk, stated plainly
A pooled deposit is segregated in accounting, not in custody. If the shared trove were ever liquidated by the MUSD protocol, the loss would fall across pooled users in proportion to their share, including users whose own position was perfectly healthy. This is the one component in Maren that can lose user funds through a design failure.
The residual risk is a BTC drawdown faster than the keeper can act: a gap down through 140% to below 110% without an intervening block in which liquidation is possible. The 30-point buffer above the cliff makes this remote. It does not make it impossible, and the app says so where the decision is made rather than in a footnote.
There is also a binding commitment behind the vault: if the team cannot satisfy itself that the pooled vault is safe, Maren ships without it. Direct troves, gasless payments, and Earn are valuable on their own.
A sizing result worth knowing
A pooled vault seeded at the protocol minimum can never serve a borrow, because MUSD's 1,800 minimum net debt pins the seed below its own borrowing floor. Maren hit this in practice, and the pooled trove's move from a 153% ICR to a healthy one is on-chain in two transactions. The detail lives in How the pooled vault works.