Quantifying Contagion Risk in Real-Time

Onchain finance is deeply interconnected. Composability is one of DeFi’s differentiators, but it also makes risk much harder to understand and quantify. When an asset fails, knowing who holds it directly is not enough. Risk teams need to understand every path through which that failure can propagate and, ultimately, calculate how much each participant could lose.

Mapping how a failure propagates

A recent analysis of USDe exposure in Aave v3 on Ethereum shows why. The report models a severe scenario in which USDe becomes worthless. At the September 1 snapshot, the estimated loss to Aave v3 would be approximately $940.20M, or 3.82% of assets. In a rush scenario, where borrowers extract the remaining liquidity available to them, losses could reach $1.26 billion, or  5.10%.

The interesting part is not only the size of the loss, but how it gets there. Some exposure is direct through Aave’s USDe reserve, while other paths run through assets such as sUSDe and Pendle principal tokens that ultimately redeem into USDe. And the exposure is not static. If an asset starts failing while its oracle still reflects the old price, holders may have an incentive to post it as collateral, borrow good assets and abandon the position. In the scenario analyzed, another $315.33 million could leave the market before liquidity itself becomes the binding constraint, increasing the modeled loss from $940 million to $1.26 billion

The loss doesn’t stay where the risk originated

The most important result appears when we stop looking at Aave as one large pool and ask who actually loses the money.

Consider a plain USDT depositor. USDT is the deepest stablecoin market in the protocol, with $2.97 billion supplied at the time of the analysis. This user has is no leverage and no USDe anywhere in their position.

Yet if USDe was compromised, USDT suppliers would face an estimated loss of  16.68% of their supplied assets. And this could rise to 22.42%, with the reserve’s remaining liquidity drained entirely. The loss reaches them because other accounts borrowed USDT against USDe-linked collateral. When that collateral becomes worthless and those borrowers abandon their positions, the bad debt is left with USDT suppliers. Their exposure is determined not only by what they hold, but by positions elsewhere in the protocol that they had no visibility into and no say over.

‍

This is why protocol-level risk tells only part of the story. The same failure can affect each depositor, vault or institution very differently depending on how their capital is connected to the rest of the system.

For a vault curator, the relevant question is how much a particular vault could lose. For an asset manager, it is which positions are economically exposed to a failure even when the failing asset does not appear anywhere in the portfolio. For a protocol, it is how much bad debt can there be and where it will ultimately land.

A real-time system requires a real-time view of risk

This USDe scenario is a stress test analysis, not a prediction or an expected-loss calculation. The broader lesson is that understanding onchain risk requires more than knowing what assets you own. It requires a live view of the dependencies between assets and protocols, the paths through which failures can propagate, and the ability to quantify what those failures mean for each participant.

That is what Forta Risk is designed to make visible: not simply what you hold, but a real-time view of what you depend on, how a failure can reach you, and how much you could lose under different scenarios.

Your portfolio does not define the boundaries of your risk. Request a demo to explore how Forta Risk can provide a real-time view of your onchain exposure.

‍