Angle's Wind-Down Shows How Curve LPs Become the Exit Queue

Stablecoin shutdowns are usually framed as solvency stories. That is not wrong. It is also usually too late. The better question is what happens to liquidity before solvency becomes the issue.

Angle is a good example. On February 20, 2026, an Angle governance proposal introduced an orderly wind-down for EURA and USDA, arguing that activity had declined enough that keeping the stablecoins alive no longer made sense. The proposal said the protocol still held about $2.41 million in assets backing USDA and about EUR5.3 million backing EURA, and that holders would have a one-year redemption period to exit at 1:1 on Ethereum before the protocol stops active operations (Angle governance).

On paper, that sounds clean. No haircut. No panic. No insolvency.

But by March 17, 2026, Curve governance had already moved to kill gauges on pools containing EURA, explicitly because the assets were being deprecated. The proposal was blunt: stop CRV emissions, stop incentivizing new users into a dying asset, and reduce systemic exposure, while leaving the pools themselves technically live (Curve governance).

That is the more interesting story, because once a stablecoin enters managed decline, the first thing that usually breaks is not redemption. It is the economics of being the person still warehousing the exit flow.

As of April 1, 2026, Angle's own site still says the protocol is in its "final chapter," remains fully collateralized, and that every USDA and EURA is redeemable 1:1 before March 1, 2027 (Angle site). So this is not a fraud story and it is not a sudden depeg story.

It is a market-structure story about how liquidity surfaces decay.

Redemption Can Stay Intact While Market Liquidity Gets Worse

People hear "fully collateralized" and assume the market should remain orderly. But collateral sufficiency and secondary-market quality are different things.

A holder who can bridge back to Ethereum and redeem through the Transmuter may still be fine. A Curve LP sitting in a deprecated EURA pool is solving a different problem. That LP is not relying on the abstract existence of collateral. That LP is relying on:

• enough two-way flow to keep fees attractive,
• enough incentives to justify idle inventory,
• enough confidence that routers and arbitrageurs still care,
• and enough market depth that exits do not turn the pool into a one-way drain.

Once a gauge gets killed, that equation changes immediately.

Curve's March 17 proposal makes the mechanics clear. The pools are not shut down. They are simply no longer paid to matter, removing one of the main reasons third-party liquidity would keep volunteering to sit there.

For a deprecated stablecoin, that matters more than people think.

A Dead Stablecoin Usually Dies in Stages

The popular mental model is binary:

• either the stablecoin is alive and redeemable,
• or it has broken and everyone runs.

Real DeFi shutdowns are uglier than that. They usually die in stages.

Stage one is narrative death. Governance or the core team tells the market the product has no future.

Stage two is incentive death. Gauges, emissions, or side rewards get removed because continuing to subsidize a sunset asset no longer makes sense.

Stage three is routing death. Aggregators, arbitrageurs, and active traders give the pool less attention because it is no longer a strategic place to keep inventory.

Only after that do you get the slow grind where the asset may still be redeemable, but market liquidity is thin enough that selling first and asking questions later becomes expensive.

That is what the Curve gauge-kill proposal is really signaling. It is not merely a housekeeping vote. It is an admission that deprecated stablecoins should stop occupying valuable incentive real estate inside the wider Curve ecosystem.

That is rational from Curve's perspective. It is also the moment LPs should realize they may be turning from fee earners into exit facilitators.

Why LPs Should Care Even If They Trust Angle's Backing

I do not think the core risk here is that Angle secretly lacks reserves. The governance proposal and website messaging both argue the opposite. My point is different.

LPs do not get paid by being technically correct about backing. LPs get paid when other market participants still need the pool.

If the market learns:

• the protocol is winding down,
• the long-term integration path is over,
• gauges are being killed,
• and the official user journey is "redeem and move on,"

then the pool's role changes. It stops being a strategic liquidity venue and starts becoming a transitional off-ramp.

A less attractive business.

A transitional off-ramp tends to produce the flow profile LPs dislike:

• more one-sided exits,
• less balanced organic volume,
• lower marginal routing interest,
• and weaker reasons for fresh capital to replace departing liquidity.

You may still collect some fees during the unwind. But the pool is no longer competing to become more useful. It is managing a shrinking user base on its way out.

The Undercovered Risk Is Cross-Chain and Operational Friction

Angle's February 20 proposal is reasonably generous in one sense: it gives holders a year to bridge assets back to Ethereum and redeem them 1:1. But that detail is exactly why market liquidity can degrade faster than "no haircut" headlines imply.

The clean redemption path is not the same as instant universal exit liquidity across every venue and chain where the token circulates.

That means some users will still prefer to sell in the market rather than:

• bridge,
• wait,
• redeem through the canonical system,
• and operationally unwind from there.

When that happens, secondary pools become the convenience layer. And convenience layers get abused hardest when the official path is slower, more manual, or chain-specific.

Curve's move matters beyond EURA itself. It highlights a recurring pattern in DeFi: redemption infrastructure can remain intact while tradable liquidity becomes increasingly thin and adverse for the marginal LP.

That is not a contradiction. It is normal.

Curve Is Protecting Itself, Not Protecting LP Returns

The Curve proposal gives a straightforward rationale for killing the gauges:

• do not attract new users into deprecated assets,
• reduce ecosystem risk,
• and align with prior risk-management actions.

That is sensible governance.

But LPs should read the action correctly. Curve is optimizing the ecosystem's incentive budget, not guaranteeing that existing liquidity providers get a graceful economic exit.

In fact, removing emissions is effectively Curve saying the social contract has changed. If you keep capital in those pools after the warning signs are obvious, that is increasingly your choice, not the DAO's job to subsidize.

Again, that may be the right call. My point is that it reveals how DeFi really handles product shutdowns.

In practice, the moment an asset is strategically dead, incentive systems start steering attention elsewhere even if the contracts remain operational. The chain does not close. The economics close first.

My Take

Angle's wind-down looks orderly. Curve's response looks rational.

The underexplained part is what this teaches LPs.

When a stablecoin enters its final chapter, the key question is not only whether reserves are there. The key question is whether you are now sitting inside the market's preferred exit queue.

If you are an LP in a deprecated stablecoin pool, you should assume the business model is changing under you:

• incentives will fade,
• routing attention will weaken,
• the most informed users will likely migrate early,
• and the pool can remain live long after it stops being strategically healthy.

That does not mean immediate disaster. It means the highest-quality capital usually leaves before the formal shutdown date.

So the lesson from Angle and Curve on April 1, 2026 is not that redemptions are broken.

It is that in DeFi, liquidity usually dies before solvency does. If you are still providing that liquidity after the protocol itself has announced the end state, you are no longer being paid to support growth. You are being paid, if at all, to absorb the inconvenience of everyone else's exit.