Original title: Measuring DeFi Risk
Jeremy Bertomeu, Xiumin Martin, Ibrahima Sall · September 2026
Decentralized finance lending, or DeFi, grew from nothing in 2017 to nearly 40 billion US Dollars deposited by May 2022. On these platforms, people lend out dollar-pegged cryptocurrencies called stable coins to earn interest. Borrowers use other cryptocurrencies as collateral to make risky trades. In exchange, lenders receive claims backed by this pool of collateral. The problem is knowing when this whole setup becomes dangerously unstable. Researchers built a tool to measure total risk using just overall deposit and borrowing numbers. Applying it to major protocols, they found that system fragility rose past safe levels around mid 2021. Extreme shifts in coin prices could even cause stable coins to lose their peg to the dollar.
Imagine a group of friends pooling money to cover each other's medical bills. Everyone contributes a set amount, and the pool promises to pay any member's bills. But the risk depends on how many friends take out large loans from the pool, and how much collateral they put up. If too many borrow too much with shaky security, the pool might not be able to pay out when someone gets sick. In this paper, the researchers measure the whole pool's risk by just looking at total contributions and total borrowings, giving an early warning before the system fails.
Anyone depositing stable coins into DeFi lending platforms faces the risk that the collateral backing their claims might fail. This paper shows that by simply tracking total deposits and borrowings across protocols, you can get an early warning of systemic fragility. Regulators and users alike could spot trouble months in advance, like the warning signs that emerged in mid-2021.
The abstract does not explicitly state limitations, but the framework relies solely on aggregate deposits and borrowings, potentially missing protocol-specific factors or off-chain risks. The evidence comes from major protocols, so applicability to smaller or newer platforms is not addressed. The model identifies fragility thresholds but does not prescribe specific policy responses.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Measuring DeFi Risk — Jeremy Bertomeu, Xiumin Martin, Ibrahima Sall
Abstract licence: CC0
Explanation generated by Synthesis from the abstract — it is not the original paper. This is a preprint, so it hasn't necessarily been peer reviewed.