> ## Documentation Index
> Fetch the complete documentation index at: https://docs.monolith.market/llms.txt
> Use this file to discover all available pages before exploring further.

# Liquidations

> How unhealthy positions are force-repaid

# Liquidations

## Overview

Liquidations close unsafe positions when a borrower’s debt exceeds their borrowing power at the current oracle price. Anyone can repay part of an unsafe borrower’s debt in `Coin` and seize collateral with a variable incentive. This keeps solvency intact and prevents bad debt.

## Triggers

Liquidations are enabled only when a valid, sufficiently fresh oracle price is available. A position becomes liquidatable when:

* Borrowing power \< debt at the current price, where borrowing power = price × collateral × collateralFactor.
* When liquidations are enabled. Liquidations are disabled when no valid price is provided by the feed.

## Process

High‑level flow (per `liquidate(borrower, repayAmount, minCollateralOut)`):

1. Accrue interest and update borrower’s internal accounting.
2. Read oracle price; require liquidations to be allowed.
3. Compute the maximum chunk that can be liquidated now:
   * If the position is healthy, liquidation is not allowed.
   * If unhealthy, the protocol allows liquidating up to 25% of the borrower's debt, but at least a minimum chunk size (10,000 `Coin`) when applicable.
4. Reduce the borrower’s debt by `repayAmount`.
5. Compute the liquidator’s collateral reward at the current price with an incentive (see Incentives below).
6. Cap the reward by the borrower’s remaining collateral and enforce `minCollateralOut` (slippage protection for the liquidator).
7. Transfer collateral to the liquidator; pull `repayAmount` of `Coin` from the liquidator and burn it.
8. If the position is undercollateralized, the protocol attempts a write‑off: delete remaining debt, redistribute it across the system’s borrowers, and hand the borrower’s remaining collateral to the liquidator.

## Incentives

The liquidation incentive scales with how far the position is beyond the collateral factor:

* 0% incentive when loan‑to‑value (LTV) is at or below the collateral factor.
* Up to 10% when LTV is 5 percentage points above the collateral factor.
* Linear between those points.

Collateral reward calculation (conceptually):

* collateralReward = repayAmount × (1 + incentiveBps/10,000) ÷ price
* Reward is capped so it cannot exceed the borrower’s available collateral.

This design pays more to liquidators as positions drift further into unsafe territory, incentivizing liquidators while minimizing loss to borrowers.

## Limits and safety valves

* Chunked liquidations: At most \~25% of a borrower’s debt can be liquidated per call (subject to the 10,000 `Coin` minimum). This ensures that the liquidation incentive is sufficient to attract liquidators.
* Oracle‑gated: If the oracle is stale/invalid, liquidations are temporarily disabled to avoid mispricing.
* Write‑off path: If a position is effectively irrecoverable (debt > 100× collateral value), remaining debt can be redistributed among borrowers, and the liquidator receives the borrower’s remaining collateral (negligible).

## Example

* Suppose a borrower has debt of 200,000 `Coin`, collateral value of 230,000 at the oracle price, and a collateral factor of 80% (borrowing power = 184,000). The position is undercollateralized (200,000 > 184,000).
* Maximum liquidatable chunk is 25% of debt = 50,000 `Coin` (greater than the 10,000 minimum).
* If the incentive computes to 8%, and the price implies 1 `Coin` buys \$1 of collateral value, then the liquidator who repays 50,000 `Coin` receives collateral worth 54,000 in value (capped by what’s available), transferred at the oracle price.