> For the complete documentation index, see [llms.txt](https://docs.dano.finance/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.dano.finance/news/lending-v2-will-price-your-collateral-from-many-sources-not-one.md).

# Lending V2 Will Price Your Collateral From Many Sources, Not One

Every lending protocol is, underneath, a machine that acts on a single number. That number decides how much you can borrow, whether your position is healthy, and — at three in the morning when nobody is watching — whether your collateral gets sold.

Today on Cardano that number comes from one place. We have decided that is the wrong architecture for a protocol holding your collateral.

**Dano Lending Smart Contract V2 will price all collateral through Chorus, a multi-source oracle aggregation layer, from launch.**

### Why one source is not enough

A single feed goes quiet, and the protocol either freezes you out or runs on a stale number. It gets pushed, when someone moves a thin source far enough to make healthy positions liquidatable. And it disagrees with other sources, because every protocol prices independently.

This is not a criticism of any provider — Cardano's oracle teams do careful work. The problem is structural: depending on exactly one source means inheriting all of its failure modes and none of its redundancy.

### How it works

We rejected the obvious design, where a service collects prices off chain and publishes an aggregate. That trades a dependency on one oracle for a dependency on one aggregator.

Chorus aggregates with no publisher at all. Your transaction carries the source observations and executes the validator directly. Before your action proceeds, it checks on chain that a **minimum quorum** of independent sources is present, that each sits inside its **freshness window**, that observations **deviating beyond the configured band** from the median are excluded, and that the price used is the median of what survives. Any failure aborts the transaction.

So a source that goes offline stops counting. A manipulated source falls outside the band and is ignored — the attacker has spent capital moving a price we discard. And if too many sources fail at once, actions abort rather than proceed on thin data, because refusing to act beats liquidating someone on a number nobody can stand behind.

The result is computed from data your own transaction carries. We are as unable to influence it as anyone else is.

### What changes for you

You will see a price panel before signing, showing which sources contributed and which were excluded. There should be fewer unfair liquidations. And transactions will be slightly heavier, since validating several sources costs more than reading one number — a fair trade for not being liquidated on a manipulated price.

### The commitment

* Chorus prices **all** collateral valuation and liquidation checks — not an optional mode, not a secondary feed
* It ships **with** V2, not after. 
* Initial adapters cover \[\[Liqwid]], \[\[Minswap]], \[\[SundaeSwap]] and \[\[Splash]], with dedicated oracle providers to follow
* The full methodology is published before mainnet, so it can be reviewed rather than announced

Source configuration lives in an on-chain registry, timelocked — adding a source needs no redeploy, and the parameters your loan depends on cannot change without notice.