On-Chain Credit in 2026: Why Lending Protocols Are Moving Beyond Collateral

On-chain credit is reaching the limits of its founding design. DeFi lending began with a clever workaround: protocols couldn't know who the borrower was, so they demanded more value than the loan itself. Overcollateralization wasn't a feature. It was a substitute for trust.

That workaround built a multi-billion-dollar industry. Aave, Morpho, Maple, the CDP systems behind Sky and Liquity: these are battle-tested primitives, and nobody at Protocol Roundtable #3 disputed it. Leverage, loop strategies, and collateralized borrowing against native assets work. Builders now treat this infrastructure as a settlement layer rather than a frontier.

But the frontier has moved. Listening to a room of credit builders, two L1 chains, an invoice financing platform, an uncollateralized lending protocol, and a liquidity routing layer, the conversation kept circling back to a single structural shift: the industry is trying to replace collateral with knowledge of the borrower. Everything else, the privacy debates, the chain selection criteria, the institutional courtship, follows from that one move.

From Overcollateralized DeFi to Undercollateralized Lending

Undercollateralized lending is on-chain credit where the loan exceeds the value of posted collateral, or requires none at all, because the lender verifies the borrower's identity, cash flows, and creditworthiness instead of relying on locked assets.

Ale Losa, founder of Fuero, put the inversion plainly. The market already knows how to lend to an anonymous counterparty by trusting the collateral. His team is building the opposite: know the borrower well enough that you don't need the collateral at all. Fuero targets prime borrowers, the creditworthy counterparties that TradFi underwrites every day inside what Losa calls the credit black box, and brings that underwriting on-chain.

Muhammad Salman of DeFa by InvoiceMate is attacking a different segment with the same logic. Invoice financing runs on verified deal flow: receivables from payment networks and trade finance, vetted through due diligence, then connected to stablecoin lenders through permissioned vaults. The collateral is a real economic obligation rather than a token, and the protocol's value lives in verifying it.

Both models share a premise that first-generation DeFi rejected: identity and verification are not bugs to be engineered away. They are the product.

How Does Selective Disclosure Work in On-Chain Credit?

Selective disclosure is a privacy architecture in which transactions stay confidential by default, validity is proven with zero-knowledge proofs, and specific data is revealed only to authorized parties such as regulators or auditors.

Ask where institutional liquidity is, and the panel's answer was unanimous. It isn't waiting for higher yields. It's waiting for infrastructure that handles disclosure correctly.

Gareth from the Midnight Foundation framed the problem precisely: transparency isn't the solution, trust is. Institutions will not publish their lending criteria, counterparties, and positions to a public chain. Regulators, meanwhile, will not accept a black box. The naive answer to one requirement violates the other, and selective disclosure resolves the conflict.

The approach is already converting institutions. Monument Bank, an FCA-regulated UK bank, chose the chain for tokenized retail deposits because deposits could stay private while regulatory disclosure remained possible.

Losa, working from the borrower side, described his dream infrastructure in nearly identical terms: chain-agnostic compliance tooling that encrypts borrower data and decrypts it on a per-use basis for authorized parties. When the borrower side and the chain side converge on the same design unprompted, expect it to get built. Selective disclosure is on its way to becoming as foundational to on-chain credit as the price oracle was to collateralized lending.

Why Lending Protocols Are Specializing: RWA, Invoice Financing, and Consumer Credit

Salman was blunt: one size cannot fit everything. Every credit type carries its own operations, onboarding, risk rules, and investor base. Invoice financing needs receivables verification and 30-to-90-day duration management. Consumer credit needs underwriting at scale. Prime borrower lending needs TradFi-grade credit assessment. None of these fit cleanly into a shared liquidity pool with a utilization curve.

Ran Wei, Head of DeFi at Pharos, gave the sharpest technical example. RWA collateral often has no DEX liquidity and no oracle price feed, and fixed-income products carry lockups that make instant withdrawal impossible. A consumer credit product with a three-month lockup, priced on NAV, simply cannot serve as collateral on Morpho. Pharos onboarded TermMax specifically because fixed-term, fixed-rate markets are what these assets require.

The implication for builders is uncomfortable but clarifying: the era of launching a general-purpose lending fork and competing on incentives is closing. The next protocols will be credit-type-specific by design, and their moat will be operational, not contractual.

Chains Now Compete as Ecosystems, Not Infrastructure

When the builders on the panel were asked how they choose where to deploy, nobody led with TPS. Luis Medeiros of Arenas.fi said the chain conversation used to be about technology and now it's about ecosystems: which chain aggregates real borrowers and originators. Losa named distribution as his second criterion after baseline infrastructure. Salman looks for what he calls enterprise-focused chains, plus foundations that open doors to institutional LPs rather than handing out grants.

Medeiros added a sharper test: does the chain have skin in the game? A foundation that pushes RWA products while keeping its own treasury parked in safe assets is sending a signal, and LPs read it.

The chains are behaving accordingly. Pharos is replacing token subsidies with yield engineering, and the numbers are explicit: its consumer credit vault yields 14%, borrowing against it costs 9%, and reinvesting the proceeds in a 14% real estate vault takes the looped LP yield to 16%. Real cash flows fund the premium instead of emissions. Midnight is building privacy-preserving cross-chain liquidity transfer so capital arrives without bridge friction and structuring products to keep it circulating rather than mercenary.

The chain that attracts borrowers may end up more valuable than the chain with the fastest technology, because borrowers, not blockspace, are the scarce asset in credit.

Who Gets to Launch a Lending Protocol in 2026?

The honest answer from the roundtable: anyone can still deploy, but deployment is no longer the bar. The question that matters is who can attract real borrowers, satisfy compliance, and convince increasingly demanding capital.

Liquidity providers have grown up. Medeiros, who sits closest to them, listed what they now ask for: risk-adjusted yield, withdrawal clarity after watching utilization spikes lock lenders in major markets, fixed rates and fixed terms, borrower transparency, and capacity. That last one is underrated. Allocating ten million is easy; the question institutions ask is whether a protocol can absorb two hundred million without breaking its own model.

The panel's most revealing moment came from an audience provocation: in three years, on-chain credit belongs to BlackRock and the banks. Nobody took the bet against it. On-chain RWA volume is already 60 to 70 percent concentrated among institutional giants. But the consensus wasn't surrender. As Medeiros noted, large institutions want large retail markets; niche credit strategies are too expensive for them to operate, in DeFi exactly as in TradFi. The likely outcome is a hybrid market: institutional rails dominating headline volume, with specialized, permissioned-but-open protocols owning the segments institutions can't economically serve.

What Comes Next for On-Chain Credit

The forward trajectory follows directly from the trust thesis. RWA collateral is forcing fixed-term and fixed-rate lending markets into the mainstream, because assets with lockups and NAV pricing cannot live on variable-rate utilization curves. Selective disclosure is moving from research topic to deployed infrastructure, with regulated banks already onboarding. Credit delegation and cross-chain routing are connecting idle mainnet liquidity to the chains that actually hold borrowers.

Even the AI conversation reduced to a trust question. Know Your Agent (KYA) is the emerging verification standard for autonomous agents that borrow, lend, and manage positions on behalf of users, an extension of KYC logic to non-human actors. The panel agreed agents will first take over operations: onboarding, due diligence, monitoring, refinancing. Ran's observation cut deepest: curators have slept through liquidation windows while multisigs waited for signatures. The technology to automate that exists today. What's missing is governance, the policies defining how much authority a protocol grants an agent. Verifying and bounding a new class of counterparty is the same problem this entire market is built on.

That's the pattern across the whole conversation. The hard problems in on-chain credit are no longer cryptographic. They are about trust, disclosure, and distribution, and the builders solving those problems are the ones who will own the next cycle.

FAQ

What is undercollateralized lending in DeFi? Undercollateralized lending is on-chain credit where the loan exceeds the value of posted collateral, or requires none, because the protocol verifies the borrower's identity, cash flows, and creditworthiness instead of relying on locked crypto assets.

Why do institutions need privacy in DeFi lending? Institutions cannot publish lending criteria, counterparties, and positions on a public chain, but regulators require visibility. Selective disclosure resolves this: transactions stay private by default, and specific data is revealed only to authorized parties such as financial regulators.

Can real-world assets be used as DeFi collateral? Yes, but they need specialized infrastructure. RWA collateral often lacks DEX liquidity and oracle price feeds and may carry fixed lockups, so it requires fixed-term, fixed-rate lending markets rather than standard variable-rate pools.

Will banks and asset managers dominate on-chain credit? Institutional players already account for an estimated 60 to 70 percent of on-chain RWA volume and will likely dominate headline volume. Specialized protocols are expected to keep the niche credit segments that are too expensive for large institutions to serve.