LRT rehypothecation risk: looping, depegs and collateral vs TVL

Source note. The AutoLayer-specific content here comes from the LRT scoring section of the team’s Arbitrum LTIPP application, dated 26 February 2024. LRT scoring was a planned V2 feature at that time, and we describe it as planned, not live. The mechanics of looping and liquidation are general DeFi knowledge, explained in our own words. Not financial advice.

Start with the definitions

A liquid staking token (LST) is a receipt for staked ETH that you can still trade or use elsewhere. A liquid restaking token (LRT) goes one step further: it is a receipt for ETH or LSTs that have been restaked, so the position can earn restaking rewards and also stay liquid. The upside is flexibility. The cost is another layer of risk, because the token’s value depends on the restaking system behaving as expected.

Rehypothecation means that the same underlying asset is reused as collateral more than once. In traditional finance the term describes a broker reusing client collateral. In DeFi it appears more informally: an LRT is a claim on staked assets, and that claim can itself be deposited into a lending market as collateral.

The mechanics of looping

The application notes that LRTs can function as collateral in markets, which lets users build larger positions through looping. Here is how that works in practice.

1. A user holds an LRT and deposits it as collateral in a lending market.
2. The user borrows ETH (or another asset) against it.
3. The borrowed ETH is swapped into more of the LRT, which is deposited again.
4. Steps 2 and 3 repeat, often several times.

The result is a position much larger than the starting deposit, all of it exposed to the LRT’s price and to the lending market’s rules. If the LRT earns more than the borrowing costs, the loop magnifies the yield. If the LRT falls relative to what was borrowed, the loop magnifies the loss.

From depeg to cascade

The post names the danger in one sentence: “an extreme event leading to a sudden depeg could initiate a cascade of automatic liquidations, potentially resulting in the creation of bad debt in money markets.”

Walk through the chain:

Step	What happens
1. Trigger	An LRT trades below its expected value, because of a slashing event, a contract issue or panic selling.
2. Collateral value drops	Looped positions that borrowed close to the maximum fall below their safe threshold.
3. Liquidations	The lending market sells collateral automatically to repay loans.
4. Price pressure	Those sales add supply of the LRT into thin liquidity, pushing the price down further.
5. More liquidations	The lower price pushes more positions past their threshold.
6. Bad debt	If collateral is sold for less than the loan, the market is left with a shortfall.

The key feature is the feedback loop in steps 4 and 5. Each forced sale makes the next one more likely. The more of an LRT’s supply sits in borrowed positions, the stronger that loop can be.

The metric: collateral relative to TVL

The post points out that, despite the risks, “transparency is a hallmark of DeFi markets”. Lending-market balances are public. That makes it possible to measure how much of an LRT is deposited as collateral and compare it with the token’s total value locked (TVL).

The application says this comparison offers insight into “the rehypothecation level of an asset”, and that it “contributes to a risk/reward profile”. As a reader, you can think of it as a ratio:

• Low ratio: most of the LRT is held directly or sits in ordinary pools. A depeg still hurts holders, but there is less forced selling waiting to happen.
• High ratio: a large share of the supply is pledged in lending markets. A depeg is more likely to trigger liquidations.

The post does not publish a formula, thresholds or a score scale. We therefore cannot say what counts as high or low, and any number we attached would be invented. The idea is the direction of the signal, not a cut-off.

A worked example, with made-up numbers

Suppose an LRT has 100,000 ETH of TVL, and lending markets hold 10,000 ETH of it as collateral. That is a 10% ratio. Now suppose a second LRT with the same TVL has 60,000 ETH in lending markets, a 60% ratio. Both numbers are invented for illustration; they are not data about any real token.

If both tokens lose 5% of their value in an hour, the first sees limited forced selling, because most holders are not borrowing against it. The second may see a large share of collateral sold at once, because many positions sit near their liquidation threshold. The price impact of the second sale is larger, which feeds the next round of liquidations. The ratio does not tell you the outcome, but it tells you which token has more fuel for a cascade.

What this signal misses

Even a clean collateral-to-TVL ratio has blind spots:

1. Borrow health is hidden. Two markets with the same collateral can have very different loan-to-value levels. A ratio does not show how close to the edge positions are.
2. Liquidity matters. A deep market can absorb liquidations that a thin one cannot.
3. Several markets. An LRT can be collateral on multiple platforms, each with its own parameters.
4. It says nothing about the cause. The score does not predict the trigger, only the amplification.

This is why the application frames it as a risk/reward profile for decisions “about purchases”, not as a safety rating.

Why it matters on Arbitrum

AutoLayer’s application is written for Arbitrum, where many lending markets and DEXs sit close together. That proximity is what makes looping cheap: all the steps can be bundled in one transaction. The intent-based routing article explains how AutoLayer describes that bundling. The same convenience that makes strategies easy can make borrowing easy too, so a visible rehypothecation indicator is a sensible companion.

What to do with this as a user

You do not need a dashboard to start thinking this way. Before buying or looping an LRT, consider:

• Where the token is accepted as collateral and in what quantity.
• How the token has traded relative to its underlying during stress.
• Whether the lending market has a record of handling liquidations in size.
• Whether you are comfortable losing the whole looped position, not just the yield.

If any of those answers is unclear, the position is probably larger than your understanding of it. The risk disclosure covers the general hazards, and how AutoLayer works describes the product context.

Planned as of February 2024

LRT scoring appears in the application under “AutoLayer V2 Features”. It was a plan, with no stated release date, and it sits alongside the AVS scoring idea. Together, the two aim to describe risk at the service level and at the token level. For the origin of the project and its stated limits at the time, see From Tortle Ninja to AutoLayer.

Takeaway

Looping turns a liquid restaking token into a debt-funded position, and a depeg can turn that debt into forced selling. Comparing the collateral in money markets with an LRT’s TVL is a simple, public-data way to see how exposed a token is to that chain reaction. It is a useful first look, not a forecast, and the source leaves the precise method open.