Risk
Liquidation & deleveraging
The part of a lending protocol that decides whether a bad week costs you a fee or costs you your position.
There are three stages, and you control the first two. Liquidation is the last one, not the first.
The three stages#
You act — any health factor, any time
Repay, deposit more collateral, or run self-liquidation. Cost: nothing, or a 0.30% swap fee if you have the protocol sell collateral for you. This window is open from the moment you borrow.
Smart Deleveraging acts — at health factor 1.10
The protocol sells the smallest slice of collateral that restores a health factor of 1.35 and repays that debt on your behalf. Cost: a 0.30% swap fee on the amount sold. You keep the rest of the position.
A liquidator acts — below health factor 1.00
Any keeper may close up to half your debt and take collateral at a 5–8% discount. Cost: that discount, paid out of your collateral.
Smart Deleveraging in detail#
This is the feature that most changes the experience of borrowing here. A conventional protocol has exactly one response to a falling position: wait until it is liquidatable, then let someone take a large penalty out of it. CredX intervenes earlier and smaller.
How the slice is calculated
The protocol solves for the minimum sale that restores the target health factor. Selling x of collateral repays x × (1 − fee) of debt, so:
(C − x) × t = target × (D − x × (1 − f)) ⇒ x = (target × D − C × t) ÷ (target × (1 − f) − t)
C is collateral value, D is debt, t is the weighted liquidation threshold, f is the 0.30% swap fee, and target is 1.35. It sells the least it can, never a fixed percentage.
$100,000 basket, $41,800 drawn, after a 48% market drop
Collateral value
was $100,000
Debt
unchanged — debt does not move with prices
Health factor
Collateral sold
31.5% of the position
Debt repaid
Swap fee at 0.30%
Health factor restored
A conventional liquidation at the same point would have closed $20,900 of debt and paid a liquidator roughly $1,344 in bonus out of your collateral. You keep that.
You can run it early, yourself
Self-liquidation is the same mechanism, triggered manually at any health factor. Same 0.30% fee, no liquidator, no penalty. It is the cheapest way to reduce a position you have decided is too large — and it is available on the borrow screen right now.
You can turn it off. You probably should not.
Smart Deleveraging is opt-out. Turning it off does not remove the liquidation risk — it removes the step that would have prevented one. The only coherent reason to disable it is that you are actively managing the position yourself and would rather choose which leg gets sold.
Liquidation in detail#
If health falls below 1, the position becomes liquidatable by anyone. This is not a special privilege — it is open, permissionless keeper work, and it is what keeps the pool solvent.
| Parameter | Value | Why |
|---|---|---|
| Trigger | Health factor < 1.00 | The point where collateral no longer covers debt at threshold prices |
| Close factor | 50% | One call may only close half the debt, so a dip cannot wipe the whole position |
| Liquidator bonus | 5–8% | Scales with the asset's volatility — riskier collateral pays more to clear |
| Who may call it | Anyone | Open competition keeps the bonus honest and execution fast |
A liquidation on the same position, with deleveraging disabled
Debt
Maximum closable in one call
50% close factor
Collateral seized
debt closed × 1.065 average bonus
Paid to the liquidator
Health factor after
Note that even a full liquidation call leaves you with a healthy position and most of your collateral. It is expensive, not fatal.
The 50% close factor matters more than it looks
Without it, a single momentary dip below 1 could close the entire position. With it, the worst a single call can do is halve your debt and push you back to health. Recovering prices then leave you with a smaller but intact position rather than nothing.
Where the liquidator bonus goes#
The bonus is the discount at which a liquidator receives your collateral. It exists because the work has real costs: gas, capital, and the price risk of holding the seized asset until it can be sold. Nobody would do it for free, and a pool with no liquidators is a pool with bad debt.
- 5.0% — broad-market ETF tokens, easiest to clear.
- 5.5–6.0% — mega-cap single names.
- 8.0% — volatile tech, where a liquidator carries real risk between seizing and selling.
What happens if liquidation is not enough#
Bad debt is possible
If collateral gaps below the debt it secures before any liquidator can act — an overnight gap in the underlying equity is the realistic scenario — the shortfall becomes bad debt. The protocol reserve absorbs it first. If the reserve is exhausted, lenders take the loss, spread across the pool.
This is the failure mode every other guardrail exists to prevent: market-hours haircuts, conservative LTVs, supply caps, and a collateral list restricted to assets that can actually be sold. It has not happened. It is not impossible.
How to never see any of this#
Draw half your limit instead of all of it. That single choice roughly doubles the drawdown you can absorb, and it costs you nothing but patience.
See it happen to your own position
The risk centre runs the full deleveraging plan and liquidation outcome for any shock you choose, with the real numbers.