You see a protocol advertising "institutional-grade unsecured lending, 12% annual yield," and you get interested. But what you did not read closely is: if the borrowing institution defaults, how much can you, as a depositor, actually get back?
A leading global cryptocurrency platform,suitable for both beginners and experienced traders.
New user benefit: 20% off trading fees upon registration!！
On-chain unsecured lending is expanding fast. Just the first-quarter growth is surprising. But it has a long-standing unresolved problem: depositors bear the risk, while someone else controls risk management. Retail depositors are not the "beneficiaries" of this expansion. They are the "last line of defense"—when an institution defaults, you absorb the loss.
Tail Risks Retail Depositors Bear
Tail risk means the kind of risk that looks quiet in normal times, but hits you hard when something goes wrong. The structure of on-chain unsecured lending concentrates this risk on depositors.
Recovery after default is nearly zero
Clearpool's risk rating report says it bluntly: "When a borrower defaults, you get nothing back. Even if you go through legal channels, recovery takes years, and the recovery rate is close to zero." This is not theory—Goldfinch is the real-life version. One user posted on X that he started depositing in 2021, and after five years he only got back 30% of his principal, with an actual loss of about 70%.
How Risk Travels: The Goldfinch Case
Goldfinch's collapse clearly shows what "retail depositors bear tail risk" actually means.
Protocol background: Goldfinch was a decentralized credit protocol led by a16z in two funding rounds, raising $37.7 million in total. It lent USDC deposits to small businesses in Africa and Southeast Asia, where borrowers accepted high interest of 15%-25%.
Risk transmission chain:
Off-chain risk control failed: after the protocol converted USDC into local currency for lending, depositors could not track where the money went or how borrowers were doing. In 2022, local partner Tugende Kenya moved $1.9 million of a $5 million credit line to a related entity in Uganda without approval. Nearly 40% of the funds were misused, while depositors knew nothing and kept receiving 10%-12% book returns.
Loan portfolio deteriorated: within months of funding, problems surfaced one after another. Tugende Kenya defaulted; two underlying positions at U.S. credit fund Stratos fell to nearly zero; Singapore borrower Lend East repaid only 58% of principal.
Final result: the protocol officially began shutdown. Before closure, $56.15 million in lent capital was still outstanding, while TVL on Ethereum was only $1.63 million. Almost all deposit funds were locked in loans and could not be withdrawn. Governance token GFI fell from its peak of $32.94 to below $0.065, a drop of 99.8%.
Maple Finance: Intermediary Risk in the Delegate Model
Maple Finance's unsecured credit model introduced "delegates" to review loans, but this mechanism itself became a source of risk.
Orthogonal Trading's dual role: Orthogonal Trading acted both as a delegate reviewing loans and as a borrower that took about $36 million from the protocol. Later it was found to have misrepresented its financial condition and hidden the true scale of its losses on FTX. M11 Credit, a delegate, was also criticized—it allowed bad debt to pile up and even let some troubled borrowers extend their loans instead of declaring default.
Depositor losses: in this $36 million default, about $31 million USDC represented 80% of the relevant pool, while the pool's total insurance coverage was only about $1.85 million. Depositors faced severe losses.
The risk is not just "the borrower does not repay." It is that "you cannot know whether he can repay." The blockchain records that a loan was made, but it cannot record a factory's inventory in Kenya or a Singapore borrower's audited financial statements. When the only people who know that information choose to hide it, your risk becomes their room to profit.
Risk Multipliers: Asset Recycling and Leverage Transmission
Another feature of on-chain credit is that assets can be recycled as collateral. A credit asset can be tokenized, then repeatedly pledged, borrowed against, and re-pledged across DeFi protocols. That means if any underlying loan defaults, losses are amplified layer by layer along the leverage chain.
When money keeps flowing in, new deposits can cover redemption demand, and the risk is hidden. Once inflows slow down, the contradiction between promised token yield and the real repayment ability of the underlying loans is fully exposed, triggering a chain of runs.
A leading global cryptocurrency platform,suitable for both beginners and experienced traders.
New user benefit: 20% off trading fees upon registration!！
What Depositors Can Do
In the credit expansion on the institutional chain, depositors are in this situation: your returns are as limited as an institution's, but your losses may be unlimited.
Check the borrower list and concentration: if the top three borrowers took 80% of the pool, a single default can crash it.
Check the default handling mechanism: does the protocol have an insurance pool? What percentage of defaults does it cover? Who bears the uncovered part? In Goldfinch and Maple, insurance coverage was less than 5% of the default size.
Ask yourself one question: if this protocol announced a shutdown tomorrow, how long would it take to get all your money back? If you cannot answer, do not deposit.
