Crypto Lending Rebuilds: Two Models, One Unresolved Risk Question

Crypto lending is returning to scale. The shape of its recovery reflects an unresolved argument over how risk should be priced into loan structures. That disagreement determines where risk sits, who bears it, and how quickly it surfaces when markets turn.

The last cycle made one thing obvious: risk was often hidden behind yield. Centralized lenders like Celsius and BlockFi failed when rehypothecation and commingled assets could not withstand withdrawals. DeFi protocols cracked under stress when utilization spikes sent variable rates soaring and collateral values plunged. What replaces that system is still being contested.

The Two Rebuilding Philosophies: Simplify or Price Granularly

Two distinct rebuilding philosophies have emerged. Each makes a different bet about what went wrong and how to prevent a repeat.

The Simplification Bet

Lenders like Arch Lending are narrowing the product. Borrowers post crypto collateral, assets are held in qualified custody, and loan terms are defined upfront. The model reduces the number of moving parts.

The target user is not a trader chasing leverage-chasing traders. It is long-term holders who want liquidity without triggering a taxable event or losing upside exposure.

“The core advantage is preserving long-term exposure while accessing liquidity in the short term,” Sahay said.

The assumption: risk is best managed by reducing complexity. Fewer variables mean fewer failure points.

The Granular Pricing Bet

Fira takes the opposite starting point. “The right question isn’t whether the protocol is safe,” said CEO Pierre Person. “It’s whether a specific market is safe.”

Rather than one lending product, Fira breaks lending into components. Each market has its own collateral, loan-to-value ratios risk parameters. Users choose based on the returns and exposures they accept. Fira Money and Usual apply this logic.

Variable-rate lending is, in Person’s words, “a short-duration primitive dressed up as a lending market.” Fira locks rates at origination and holds them constant to maturity. Borrowers are insulated from the rate shocks that occur when utilization spikes.

“Once a position is opened, the rate is locked from origination to maturity. The borrower pays exactly that rate, regardless of what happens to market liquidity or spot rates,” Person said.

Where the Market Is Now: On-Chain Lending Scale and Stablecoin Supply

On-chain lending has recovered to scale. Aave alone has surpassed $40 billion in net deposits. Protocols like Morpho continue to grow across chains. Total lending activity across major DeFi protocols has rebounded meaningfully from post-2022 lows, according to recent industry research.

Stablecoin supply now exceeds $280 billion, reflecting continued on-chain dollar demand and the rise of yield-bearing collateral. That creates the foundation for lending systems where capital allocation across risk tiers becomes more central than simple leverage access.

The composition of activity has shifted. Less emphasis is on maximizing yield. More focus is on how yield is generated. The core question has moved from “What is the return?” to “What sits underneath it?”

Who Bears the Risk? Borrowers, Lenders, Platforms

Borrowers

Long-term holders using Arch Lending-style products face lower execution risk from rate volatility or collateral mishandling. They depend on the platform’s custody and regulatory resilience. Borrowers in Fira-style markets must evaluate each sub-market’s risk parameters themselves. That shifts responsibility to the user.

Lenders / Liquidity Providers

Lenders in simplified structures accept lower yields in exchange for more predictable risk. Lenders in granular markets can chase higher returns monitor individual market health. Both forms carry concentration risk if one collateral type or protocol dominates.

Platforms

Centralized lenders remain dependent on custody frameworks and regulatory environments that can change. On-chain protocols depend on users navigating increasingly complex market structures. Each introduces new forms of risk as they address old ones.

What Would Confirm or Weaken the Thesis

Factors That Reduce Risk

Clear regulatory frameworks for both centralized and on-chain lending that define custody, disclosure capital requirements.

Greater collateral diversity that reduces systemic dependence on BTC and ETH price moves.

Standardized risk disclosure across sub-markets, making it easier for users to compare loan terms and underlying exposures.

Stress testing of both models under historical volatility regimes (e.g., 2020 March crash, 2022 May UST depeg) – published by protocols, not just marketing materials.

Factors That Make It Worse

A repeat of rehypothecation in any new form – hidden leverage not disclosed to borrowers or lenders.

Regulatory crackdown on stablecoin lending that freezes capital flows before the market has adapted.

User confusion in granular systems leading to mispriced risk – for example, putting collateral into a high-risk sub-market without understanding the LTV floor.

Concentration of lending in one protocol that becomes a single point of failure, whether due to smart contract bugs or governance attacks.

The First Real Test: Market Stress Event

The next concrete catalyst will be the first meaningful market stress. If a sharp BTC or ETH drawdown hits, the simplified lenders will be tested on whether tight custody and defined terms prevent contagion. The granular lenders will be tested on whether users correctly priced their chosen sub-market’s risk.

The crypto lending rebuild is real. The data shows scale returning. The market has not converged on one model. It is being rebuilt around a disagreement over whether risk should be minimized through tighter structures or exposed and priced at the market level.

That disagreement is not academic. It determines where risk sits, who is responsible for it how quickly it surfaces when conditions change.

The last cycle showed what happens when that question is ignored. This time, the risk is not invisible. It is being pushed into the open – whether through structure or through design.

What remains uncertain is whether that makes the system more resilient, or simply makes the trade-offs harder to overlook. The first real drawdown will provide the answer.