Credit risk has moved onchain. Active on-chain private credit rose from roughly $0.40 billion at the start of 2025 to $2.29 billion by March 31, 2026, and that surge is overwhelmingly concentrated in a single venue: Maple Finance, which expanded its active loan book from $0.21 billion to $2.13 billion over the same period, or about 93.1% market share. The numbers come from CoinGecko’s sector reporting based on DeFiLlama data, which shows the step change in scale and concentration across protocols through the end of Q1 2026 (CoinGecko 2026 RWA Report).
That shift is timely because on-chain private credit relies on borrower underwriting and real counterparty performance, not only on-chain collateral. CoinGecko also highlights that Maple’s recent growth is driven by loans to crypto-native market makers and trading firms, a borrower mix whose balance sheets can be tightly coupled to crypto market volatility. Correlation risk is therefore embedded in the leading venue for this activity (CoinGecko).
At the same time, the yields on offer have been compelling. Maple reported its High Yield Secured product delivered a 16.83% net APY during 2024, with secured pools remaining overcollateralized through the year, a partial liquidation executed in August 2024, and 61 margin calls cured on average within roughly three hours (Maple Yield Performance 2024). These datapoints, alongside raw on-chain dashboards that catalog private-credit assets and TVL by product line, such as DeFiLlama’s RWA listings where Syrup USDC shows an active market cap above $1.3 billion, indicate both scale and investor appetite (DeFiLlama RWA dashboards).
From overcollateralized DeFi to concentrated on-chain credit
What changed is not just growth, but composition. Early DeFi lending skewed to overcollateralized, algorithmic markets. The current expansion routes more capital through credit underwriting and borrower performance, often with limited or offchain collateral. CoinGecko’s analysis shows this activity has not diversified evenly across protocols; it has concentrated in Maple, and within Maple, in crypto-native credit exposures (CoinGecko).
Maple’s own disclosures show a platform that scaled meaningfully before the broader sector inflected: the protocol said TVL reached as high as $600 million during Q4 2024 and loans outstanding grew 23% quarter over quarter that quarter (Maple Q4 2024 Treasury Report). As liquidity migrated into private credit strategies, on-chain datasets tracked by DeFiLlama captured an expanding roster of RWA and private-credit products, making the growth legible to the market in near real time (DeFiLlama).
The data: $2.29B in loans, with Maple at ≈93%
The clearest evidence is the loan book arithmetic through March 31, 2026. Sector totals and Maple’s share, as compiled by CoinGecko from DeFiLlama’s on-chain metrics, point to a market that grew and centralized at the same time.
Metric
Start 2025
Mar 31, 2026
Total active loan value (on-chain private credit)
~$0.40B
$2.29B
Maple Finance active loan value
$0.21B
$2.13B
Maple share of total
n/a
≈93.1%
CoinGecko also underscores a qualitative point behind these figures: Maple’s growth has been powered by loans to crypto-native firms, which can experience correlated stress during market drawdowns (CoinGecko). That concentration heightens tail risk for lenders, even when headline default rates remain muted.
Implications for LPs and borrowers: yield versus correlation
For liquidity providers, the trade-off is straightforward. High net yields like Maple’s reported 16.83% APY in 2024 are attractive, but they sit atop borrower performance and, per CoinGecko, a borrower base exposed to crypto market cycles. Inference: if volatility spikes and liquidity thins across exchanges and OTC venues, multiple borrowers could face simultaneous pressure, raising the probability of clustered credit events. This is a structural difference from overcollateralized lending where liquidation mechanics, not borrower solvency, are the primary defense.
Maple’s 2024 disclosures provide some comfort. The project says its secured pools remained overcollateralized, it executed a partial liquidation in August 2024, and it issued 61 margin calls that were cured on average in roughly three hours (Maple Performance 2024). Verified fact: operational controls handled episodic stress in that period. Reasonable inference: such responsiveness may reduce loss severity, but cannot eliminate losses if borrower stress becomes systemic.
Protocol design adds a second loss channel
Credit underwriting is not the only risk. Protocols that depend on price oracles to trigger margin calls or liquidations face an on-chain attack surface. Academic work on Protocols for Loanable Funds formalizes how oracle manipulation, often enabled by flash loans, can impose losses on lending pools independent of any borrower default (arXiv: SecPLF). This is a distinct channel from credit risk and raises the bar for security engineering in undercollateralized or hybrid collateral models.
DeFiLlama’s product-level dashboards make the aggregate exposure visible, including private-credit assets with billion-dollar scale like Syrup USDC (> $1.3B active market cap), underscoring why oracle robustness and pool-level risk controls are not theoretical concerns (DeFiLlama).
Why the bear case isn’t foregone
There is a cogent counterargument: execution to date has been strong. Maple’s secured pools remained overcollateralized through 2024; when stress emerged, a partial liquidation was carried out, and margin calls were generally resolved in hours, not days. Loans outstanding also expanded 23% quarter over quarter in Q4 2024 while protocol TVL reached as high as $600 million, suggesting lenders continued to supply capital even as risk management was tested (Maple Q4 2024 Treasury Report; Maple Performance 2024).
Opinion: strong historical performance and fast operational responses deserve weight. But they are not dispositive when concentration and correlated borrower exposures are rising. The system can be robust in normal times and still fragile at the tails.
What would confirm or weaken this thesis
Investors and builders can track concrete indicators that would sharpen, confirm, or challenge the case that credit risk is building onchain:
- Borrower mix and concentration: Disclosures showing diversification away from crypto-native market makers and trading firms would weaken the correlation risk argument; deeper concentration would confirm it (CoinGecko).
- Active loan growth and share: Continued sector growth with Maple’s share near ≈93% keeps systemic concentration high; new entrants gaining share would diffuse it (DeFiLlama).
- Performance metrics: Future Maple reports on net APY, margin-call counts and cure times, and any liquidations or realized losses will be decisive datapoints (Maple).
- Stress events: Evidence of clustered borrower stress during sharp market drawdowns would validate the correlation thesis; resilient performance would argue the opposite (CoinGecko).
- Protocol risk posture: Transparent changes to oracle designs, circuit breakers, and pool-level protections would mitigate the attack surface identified by SecPLF; notable oracle incidents would elevate concern (arXiv: SecPLF).
Editorial conclusion: On-chain private credit has arrived at scale, but with concentration and borrower correlation that import traditional credit risk into DeFi. Strong recent performance shows the model can work, yet the system’s true resilience will be tested not by averages, but by outliers. Watching the mix of borrowers, the dispersion of platforms, and the quality of protocol safeguards will determine whether this credit cycle gets safely securitized onchain or finds its limits.
Disclaimer: This article is provided for informational purposes only. It is not offered or intended to be used as legal, tax, investment, financial, or other advice.
