SparkLend:
the Aave fork whose caution paid off when the real thing didn't.
We tore apart SparkLend, the lending market built on Aave V3's own code and operated under Sky (formerly MakerDAO) governance, across the same lending-adapted scorecard we've used throughout this series. From extensive audits (ChainSecurity, Cantina) inherited alongside Aave's own code lineage, one of the largest bug bounties in DeFi (up to $5 million via Immunefi), and a deliberately conservative, "blue-chip only" asset-listing philosophy more selective than Aave's own Prime Market, to a real, remarkable, well-documented decision worth centering directly: in January 2026, Spark proactively delisted low-usage assets including rsETH. Three months later, the April 2026 KelpDAO exploit devastated Aave's markets with unbacked rsETH collateral; SparkLend, having already removed that exposure, maintained healthy withdrawal liquidity while Aave's did not, and took in over $1.4 billion in "safe haven" deposits within days as capital fled toward it. We also found a real, disclosed governance-centralization concern, and a genuinely significant discrepancy in TVL reporting across Spark's several product lines; and landed on a score the marketing page won't show you.
Our take, up front: SparkLend is built on Aave V3's own contract code, deployed and configured under Sky (formerly MakerDAO) governance, with a licensing arrangement that pays Aave a share of profits in exchange for building on its codebase. Real, disclosed extensive audits (ChainSecurity for SparkLend and Savings, Cantina for the Liquidity Layer), alongside the code-level scrutiny inherited from Aave's own long audit history. Real, disclosed, substantial Immunefi bug bounty paying up to $5 million for critical vulnerabilities, one of the largest in DeFi. Real, disclosed, deliberately conservative "blue-chip only" asset-listing philosophy, described by at least one detailed source as more selective than Aave's own Prime Market. Real, disclosed mature, transparent governance inherited from MakerDAO's long track record, including monthly public risk reports and on-chain debt-ceiling votes. What we can't set aside, because it's the most remarkable finding in this review: in January 2026, Spark proactively delisted low-usage assets including rsETH and adjusted collateral rules, a decision that unsettled some leveraged traders at the time. Three months later, on April 18, 2026, the KelpDAO exploit let an attacker mint unbacked rsETH and deposit it into Aave as collateral, eventually leaving Aave with between $124-230 million in bad debt and a $10 billion TVL outflow within days. SparkLend, having already removed its rsETH exposure two months earlier, maintained healthy ETH withdrawal liquidity while Aave's markets faced shortages, and took in more than $1.4 billion in deposits within days as capital fled toward it as a perceived safe haven; SparkLend's own TVL surged from roughly $1.89 billion to $3.3 billion in the same window. Real, disclosed governance-centralization concern: SkyDAO holds significant power over Spark, including a disclosed, "controversial" ability to mint new SPK tokens in extreme scenarios. We also found a real, significant discrepancy in TVL reporting across Spark's several product lines (SparkLend, Savings, Liquidity Layer, and Morpho-based Isolated Markets), which different sources measure in different combinations. We weighted all of it below.
Real, disclosed extensive audits (ChainSecurity, Cantina) plus code-level scrutiny inherited from Aave V3's own audit lineage. Real, disclosed $5M Immunefi bug bounty, one of the largest in DeFi. Real, no confirmed direct hack or exploit of Spark/SparkLend found in our research. Real, genuinely remarkable, well-documented proactive decision: in January 2026, Spark delisted low-usage assets including rsETH; three months later, the April 2026 KelpDAO exploit devastated Aave's markets via unbacked rsETH collateral, while SparkLend, having already removed that exposure, maintained healthy liquidity and took in $1.4B+ in "safe haven" deposits within days. What tempers this: a real, disclosed governance-centralization concern, SkyDAO holds a "controversial" ability to mint new SPK tokens in extreme scenarios, per a detailed technical source.
Pros
- Jan 2026 proactive rsETH delisting directly anticipated and avoided the Apr 2026 KelpDAO crisis
- No confirmed hack or exploit of Spark/SparkLend found in our research
- $5M Immunefi bug bounty; extensive audits (ChainSecurity, Cantina) plus inherited Aave code scrutiny
Cons
- SkyDAO's disclosed ability to mint new SPK tokens in extreme scenarios is a real, flagged centralization point
Real, disclosed substantial TVL that surged from ~$1.89B to $3.3B within days during the April 2026 Aave crisis, per DefiLlama data cited directly by a detailed source. Real, disclosed second-largest DeFi lending venue status per at least one source. What we can't fully resolve: a real, significant discrepancy in TVL reporting across Spark's several product lines (SparkLend, Savings, Liquidity Layer, Isolated Markets), which different sources measure in different combinations, ranging from ~$3.6-4.65B for SparkLend alone up to $12.6B combined across all products.
Pros
- Well-documented $1.4B+ inflow and TVL surge during the April 2026 Aave crisis
- Second-largest DeFi lending venue by at least one detailed measure
Cons
- TVL figures range widely depending on which of Spark's several products a source includes
Real, disclosed mature, functioning on-chain governance inherited from MakerDAO's long track record, including transparent debt-ceiling polls and monthly public risk reports published via the governance forum. Real, disclosed SPK token with governance, staking, and farming utility. What tempers this: a real, disclosed, "controversial" centralization point, SkyDAO's ability to mint new SPK tokens in extreme scenarios, per a detailed technical source.
Pros
- Mature, transparent governance inherited from MakerDAO's long track record
- Monthly public risk reports published directly via the governance forum
Cons
- SkyDAO's disclosed emergency SPK-minting power is a real, flagged centralization point
Real, disclosed deliberately conservative "blue-chip only" asset-listing philosophy, described directly by a detailed source as more selective than Aave's own Prime Market on asset screening. Real, disclosed multi-chain deployment across Ethereum, Base, Arbitrum, Optimism, Unichain, and Gnosis. Real, disclosed proactive delisting of underperforming or risky assets (rsETH, Jan 2026), demonstrating disciplined curation in practice, not just policy.
Pros
- Deliberately conservative, "blue-chip only" listing philosophy, more selective than Aave's own Prime Market
- Proactive Jan 2026 delisting of rsETH, validated by subsequent events
Cons
- Conservative listing means fewer speculative or newer assets available than some competitors
Real, disclosed unified "Spark Borrow" experience combining SparkLend and Isolated Markets under one interface, with real-time metrics surfaced across both products. Real, disclosed, proactive frontend-risk disclosure directly in Spark's own documentation (phishing, DNS hijacking, frontend compromise), though no confirmed incident of this type found in our research.
Pros
- Unified Spark Borrow interface combining SparkLend and Isolated Markets
- Proactive, disclosed frontend-risk education directly in official documentation
Cons
- Understanding which of Spark's several products you're using requires some real effort
Real, disclosed, genuinely distinctive governance-set (not utilization-floating) rate for USDS/DAI specifically via the Sky Savings Rate, which moved from 5.5% (April 2026) to 3.75% (May 2026), a real, disclosed predictability feature relative to Aave, Compound, and Morpho's floating rates. Real, disclosed substantial annualized fee generation (~$147M per one detailed source).
Pros
- Governance-set, disclosed Sky Savings Rate offers genuine predictability for USDS/DAI
- Substantial, disclosed annualized fee generation
Cons
- Non-USDS/DAI assets still use variable, utilization-based rates like other protocols in this series
Real, genuinely extensive product suite: SparkLend, Spark Savings (sUSDS/SSR distribution), the Spark Liquidity Layer, and Isolated Markets built on Morpho infrastructure, reviewed earlier in this series. Real, disclosed institutional integration with Anchorage Digital, a federally chartered digital asset bank.
Pros
- Genuinely extensive, multi-product suite spanning lending, savings, and liquidity allocation
- Real institutional integration with Anchorage Digital
Cons
- The Morpho-based Isolated Markets inherit some of Morpho's own disclosed curator-layer risks
Access only through Spark's official app, and understand which specific product you're using.
Given that Spark spans SparkLend proper, Spark Savings, the Liquidity Layer, and Morpho-based Isolated Markets, confirm which specific product you're depositing into and what its specific risk profile is, since the Isolated Markets specifically inherit some of Morpho's own curator-layer risks covered in our separate review of that protocol.
Not just resilient, right: the highest score in this lending series so far, and for an unusually specific reason.
We've now reviewed several protocols in this series that survived a crisis they didn't cause. Spark's story is different in a way we think deserves real, distinct credit: it wasn't passively spared, it made a specific, disclosed, proactive decision two months in advance that turned out to be exactly right. Delisting rsETH in January 2026 wasn't a response to anything that had happened yet; it was a judgment call about risk that a majority of the market, including Aave, didn't make at the time. When the KelpDAO exploit hit in April, that earlier decision is the specific, documented reason SparkLend had liquidity to spare while Aave didn't, and it's the reason capital moved toward Spark rather than away from it. That's a genuinely rare form of evidence: not "nothing bad happened to us," but "we made the harder, more conservative call before we had to, and it worked." We still think the disclosed SkyDAO minting power and the real confusion across sources about which of Spark's several products a given TVL figure refers to are worth real, honest weight, which is why this isn't a perfect score. But on the question that matters most in this category, whether the people managing your collateral make good decisions before they're forced to, Spark has the single best piece of evidence we've found in this entire series.
The scorecard above is deliberately general. Whether SparkLend is right for you depends heavily on which of these you already are.
The user who wants deliberately conservative, blue-chip-only asset curation and a proven, tested risk-management track record
This is exactly where Spark's genuinely disciplined, validated decision-making delivers real, demonstrated value.
The USDS/DAI holder who wants a predictable, governance-set rate rather than a floating, utilization-based one
The Sky Savings Rate mechanism is a genuinely distinctive, well-disclosed feature most competitors don't offer.
The user who confirms exactly which Spark product, SparkLend, Savings, Liquidity Layer, or Isolated Markets, they're using
Given the real, disclosed reporting confusion across these product lines, this specific habit genuinely matters here.
Anyone who wants the widest possible asset selection or a single, simple product rather than a multi-line suite
Aave and Kamino, both reviewed earlier in this series, offer broader asset listings for users who specifically want that breadth.
The scorecard covers the headline judgment calls. These four tables cover the specifics we didn't want to bury in prose; the January-to-April 2026 timeline precisely, the Aave-fork relationship explained, TVL broken out by product, and our eighth lending-protocol comparison entry.
January to April 2026, timelined
| Date | Event |
|---|---|
| Jan 2026 | Spark proactively delists low-usage assets including rsETH and adjusts collateral rules; some leveraged ETH traders are unsettled at the time |
| Apr 18, 2026 | KelpDAO's LayerZero bridge is exploited; an attacker mints unbacked rsETH and deposits it into Aave as collateral to borrow ~$190M in WETH |
| Apr 18-22, 2026 | Aave's TVL falls from >$26B to ~$16B; Aave is left with an estimated $124-230M in bad debt depending on loss allocation |
| Apr 18-22, 2026 | SparkLend receives over $1.4B in deposits; its TVL rises from ~$1.89B to $3.3B; active loans grow by ~$500M |
We want to be precise that Spark's January decision predates and was independent of the April exploit; this was a proactive risk call, not a reaction to it, which is precisely why we're weighting it as meaningfully stronger evidence than after-the-fact resilience.
SparkLend's relationship to Aave, explained
| Detail | |
|---|---|
| Code base | Built directly on Aave V3's own contract code |
| Licensing | A share of profits is paid to Aave in exchange for using its codebase, per an original protocol engineer's own account |
| Governance | Operated under Sky (formerly MakerDAO) governance, not Aave's own DAO |
| Asset listings | Diverge sharply from Aave's; Spark applies its own, more selective blue-chip criteria |
Same underlying contract code, different governance, different asset-listing philosophy, and, as of April 2026, a genuinely different outcome during the same industry-wide stress event.
TVL, broken out by product
| Product | Approximate figure | Source date |
|---|---|---|
| SparkLend (alone) | $3.6-4.65 billion | May-Aug 2026, varies by source |
| Spark Savings (sUSDS) | $6.4 billion | May 2026 |
| Spark Liquidity Layer | $2.6 billion | May 2026 |
| Isolated Markets (Morpho-based) | $157.1 million | Recent, per a detailed source |
Combined figures cited elsewhere ($7.96B to $12.6B+) depend entirely on which of these product lines a given source includes; we're breaking them out separately rather than citing a single combined number as if it were self-evident.
Lending protocols, side by side (series continues)
| Aave | Compound | CoinRabbit | Kamino | Binance Loans | Morpho | Nexo | SparkLend | |
|---|---|---|---|---|---|---|---|---|
| Model | Non-custodial | Non-custodial | Custodial CeFi | Non-custodial | Custodial CeFi | Non-custodial | Custodial CeFi | Non-custodial (Aave V3 fork) |
| Most severe disclosed history | $292M bridge exploit | $161.7M governance bug | Transparency gaps | None found | $4.3B DOJ settlement | $18M vault loss | $500K CA penalty | None found; benefited from Aave's crisis |
| Distinctive model | Multi-network V3/V4 | Isolated Comet markets | Fixed-rate, no-KYC | Curator-managed markets | Tiered CeFi products | Immutable core + vaults | Loyalty-tier CeFi | Aave fork, Sky-governed, blue-chip only |
SparkLend is the only protocol in this series whose most notable security story is a proactive decision that paid off during a crisis that hit a different, related protocol.
We don't just want to hand you our number; we want to show you how it sits next to what other review desks and comparison sites have published. We've excluded Spark's own research and documentation pages from this comparison, treating them as primary sources we weighted directly in the scorecard.
Our score lands closely aligned with the aggregated industry average, one of the smaller gaps we've found across this lending series; the genuinely well-documented January 2026 risk decision we weighted heavily is the same event most independent sources single out favorably.
| Source | Score | Type |
|---|
Scores compiled by our editorial team from publicly available reviews as of August 2026. "Editorial estimate" means the outlet didn't publish a single numeric score, so we converted their published verdict and sentiment into a comparable 100-point figure. Verify current figures directly with each source before citing them elsewhere.
We found no confirmed hack or exploit of Spark or SparkLend in our research. Notably, Spark actually benefited during the April 2026 KelpDAO exploit that hit Aave, having proactively delisted the affected rsETH collateral three months earlier.
No, though it's built on Aave V3's own contract code, with a share of profits paid to Aave for the license. SparkLend is operated under Sky (formerly MakerDAO) governance, applies its own, more conservative asset-listing criteria, and diverged sharply from Aave during the April 2026 crisis as a direct result.
Spark proactively delisted low-usage assets including rsETH and adjusted collateral rules, unsettling some leveraged ETH traders at the time. Three months later, this decision proved prescient when the April 2026 KelpDAO exploit specifically targeted rsETH collateral on Aave.
After the KelpDAO exploit left Aave with an estimated $124-230 million in bad debt and constrained liquidity, capital fled toward SparkLend as a perceived safe haven, since Spark's earlier rsETH delisting meant it had no exposure to the same crisis. SparkLend received over $1.4 billion in deposits within days.
It depends which product you mean. SparkLend alone holds roughly $3.6-4.65 billion depending on the source and date; combined across SparkLend, Spark Savings, the Liquidity Layer, and Isolated Markets, sources cite figures from $7.96 billion up to $12.6 billion or more.
A governance-set (not utilization-floating) rate that applies specifically to USDS/DAI borrowing and lending on SparkLend, offering more predictability than the variable rates other assets use. It moved from 5.5% in April 2026 to 3.75% by May 2026.
Independent-risk borrowing venues built on Morpho infrastructure, reviewed separately in this series. They inherit some of Morpho's own disclosed curator-layer risks, distinct from SparkLend's own Aave-derived contract code.
Yes, SPK, used for governance voting, staking, and farming. SkyDAO retains a disclosed, "controversial" ability to mint new SPK tokens in extreme scenarios, a real centralization point worth understanding.
More Reviews
Ledn – Crypto Lending & Borrowing Review
Score: 75/100. Highest CeFi score in this series: 10 straight PoR attestations, explicit no-re-lending pledge, never lost customer funds.
Read More
Aave – Crypto Lending & Borrowing Review
Score: 69.5/100. Category leader hit hardest by the Apr 2026 KelpDAO crisis: $124-230M bad debt, its largest liquidity event yet.
Read More
Compound – Crypto Lending & Borrowing Review
Score: 63.25/100. Pioneer whose 2021 self-inflicted $161.7M COMP bug remains its scar, though its caution spared it in Apr 2026.
Read More
CoinRabbit – Crypto Lending & Borrowing Review
Score: 50.5/100. Fast, no-KYC lending with a great product, but no named founder, no PoR, and vague licensing behind it.
Read More