Staking, Lending & Yield
Best Crypto Lending Platforms
Borrow against your holdings, or lend them out. The category with the worst historical record on this site — the weighting reflects what actually went wrong, repeatedly.
All 4 compared
Sorted by computed score. Column headers carry this category's weighting; the highest score in each pillar is marked. Tap a name for the full assessment.
| # | Service | Score | Custody35% | Cost20% | Regulation10% | Performance25% | Access10% |
|---|---|---|---|---|---|---|---|
| 01 | AaveOver-collateralised on-chain lending with a public risk framework | 8.0Strong | 8.4 — best in table | 8.2 | 6.4 | 8.0 — best in table | 7.6 |
| 02 | CompoundPlain single-asset lending markets, if you check the collateral list | 7.5Solid | 7.8 | 8.0 | 6.2 | 7.4 | 7.4 |
| 03 | NexoCustodial borrowing against crypto, for users who accept counterparty risk | 6.7Adequate | 6.0 | 6.4 | 6.6 — best in table | 7.4 | 8.0 — best in table |
| 04 | MorphoBetter rates through isolated, permissionless lending markets−1 penalty | 5.5Mixed | 7.4 | 8.4 — best in table | 6.2 | 7.6 | 7.2 |
The assessments
what each score rests on- 018.0/10
Aave
Best for over-collateralised on-chain lending with a public risk framework
Company file: Aave, owners, incidents →Strong◆ Sources checked · 27 Sept 2026The benchmark for on-chain lending, with about $19.2B locked (27 Sep 2026). Every position is over-collateralised and every parameter is visible on-chain. Its code has held, but its collateral choices have not always: in April 2026 an attacker borrowed about $193M against unbacked rsETH, with the gap covered by liquidations and the DeFi United coalition.
Pillar scores
- Custody & Security35%
- 8.4
- Cost & Fee Transparency20%
- 8.2
- Regulation & Legal Standing10%
- 6.4
- Performance & Reliability25%
- 8.0
- Access & Support10%
- 7.6
Strengths
- Over-collateralisation enforced by contract — rehypothecation is not possible
- Liquidation machinery has cleared crashes that broke smaller lenders
- All parameters visible on-chain with a published risk framework
Against it
- Liquidation risk is entirely yours to manage
- A bad collateral listing can put nine figures at risk, as rsETH did in Apr 2026
- Interface assumes real familiarity with collateralised borrowing
Full assessment of Aave →strong · 8.0/10 - 027.5/10
Compound
Best for plain single-asset lending markets, if you check the collateral list
Company file: Compound, owners, incidents →Solid◆ Sources checked · 27 Sept 2026One of DeFi's oldest lenders, with one borrowable asset per Compound III market. Its collateral choices are not as conservative as its reputation: deUSD was listed at an 88% collateral factor, and its November 2025 collapse left $15.57M of exposure. Elixir's $11.65M settlement fell through when it entered liquidation on 16 April 2026.
Pillar scores
- Custody & Security35%
- 7.8
- Cost & Fee Transparency20%
- 8.0
- Regulation & Legal Standing10%
- 6.2
- Performance & Reliability25%
- 7.4
- Access & Support10%
- 7.4
Strengths
- Simple single-asset markets with a long live record
- Public risk parameters per market
- Transparent on-chain rates with no operator spread
Against it
- deUSD listed at an 88% collateral factor; $15.57M exposure after its Nov 2025 collapse
- Elixir defaulted on the $11.65M settlement (Apr 2026)
- Same liquidation risk as any collateralised borrowing
Full assessment of Compound →solid · 7.5/10 - 036.7/10
Nexo
Best for custodial borrowing against crypto, for users who accept counterparty risk
Company file: Nexo, owners, incidents →Adequate◆ Sources checked · 27 Sept 2026The most substantial survivor of the custodial lending cohort: it kept paying withdrawals through 2022 and has no loss of client funds on record. It now lists security credentials (SOC 2 Type 2, SOC 3, ISO 27001, CCSS Level 3) rather than a live liabilities report, and it is still a custodial lender. Back in the US since 16 February 2026 through Bakkt.
Pillar scores
- Custody & Security35%
- 6.0
- Cost & Fee Transparency20%
- 6.4
- Regulation & Legal Standing10%
- 6.6
- Performance & Reliability25%
- 7.4
- Access & Support10%
- 8.0
Strengths
- Kept processing withdrawals through 2022 when its entire peer group failed
- SOC 2 Type 2, ISO 27001 and CCSS Level 3; named custodians
- Simple onboarding and clear interface
Against it
- Custodial — your deposit is an unsecured claim, and the terms permit it to be lent out
- Best rates are gated behind holding the platform token
- $45M US settlement (2023) and a $500K California fine (Jan 2026)
Full assessment of Nexo →adequate · 6.7/10 - 045.5/10
Morpho
Best for better rates through isolated, permissionless lending markets
Company file: Morpho, owners, incidents →Mixed◆ Sources checked · 27 Sept 2026Isolates each market so a failure in one collateral type cannot contaminate the rest, which is a genuine structural improvement on pooled designs. The permissionless market creation that makes it flexible also means some markets are configured by people with no reason to be careful.
Pillar scores
- Custody & Security35%
- 7.4
- Cost & Fee Transparency20%
- 8.4
- Regulation & Legal Standing10%
- 6.2
- Performance & Reliability25%
- 7.6
- Access & Support10%
- 7.2
Deductions applied
- −2.0Unrecovered user-fund loss. Users lost funds in the last 24 months and were not made whole. Applies once, regardless of cause.
Strengths
- Market isolation contains a collateral failure instead of socialising it
- Narrower borrow-lend spread than pooled designs
- Immutable, audited core contracts
Against it
- Permissionless market creation means parameter quality varies a lot
- Requires the user to assess individual markets
- Depositors lost money in curator vaults in Nov 2025 and Mar 2026
Full assessment of Morpho →mixed · 5.5/10
The category with the worst historical record on this site. Celsius, BlockFi, Voyager and Genesis all failed inside eighteen months, and in every case the mechanism was the same: undisclosed rehypothecation and a yield that was never coming from where users thought. We weight custody at thirty-five per cent and read the terms of service line by line, because that is where the last collapse was written down in advance.
Frequently asked questions
Why do the on-chain protocols score so much higher than the custodial platforms?+
Because over-collateralisation enforced in code is a structurally different promise to a company undertaking not to misuse your deposit. Aave cannot lend your collateral out to a hedge fund; a custodial platform can, and several did, and their users are still waiting. That gap is what the custody weighting exists to express.
Is a high advertised yield a warning sign?+
It is a question, and the question is where the yield comes from. Over-collateralised on-chain lending pays what borrowers pay, which is knowable. A double-digit rate on a stablecoin from a custodial platform is being paid from somewhere else, and every collapse in this category started with users not asking where.
What did the 2022 failures actually have in common?+
Undisclosed rehypothecation, and terms of service that permitted it in language nobody read. In each case the platform was legally entitled to do what it did. That is why the compliance pillar here includes reading the contract rather than just checking for a licence.
Assessed by
Blockchain Security Researcher · September 27, 2026