Learn to read a crypto-lending market estimate without confusing loan balances, collateral values and new lending. Then follow the risks that a bigger loan book does not remove.
A larger loan book does not by itself mean safer lending
Borrowing can finance useful activity, leveraged speculation or both. A rising dollar loan balance does not reveal which use dominates. For example, if borrowers owe one bitcoin, its dollar value can rise even if nobody borrows more bitcoin. If borrowers instead owe a fixed dollar principal, an increase in the price of their collateral alone does not increase that principal. Changes in which lenders a survey covers can also move its total.
What the Q3 2025 estimate actually measured
Galaxy Research's report published 19 November 2025 estimated $73.59 billion of crypto-collateralized lending at Q3-end. Its categories included $40.99 billion on DeFi lending applications, $24.37 billion of tracked CeFi borrows and crypto-backed CDP stablecoins. It warned about possible double-counting and said private third-party lender figures had not been formally vetted.
How can DeFi be both 55.7% and 62.7% of lending?
Start with the same $40.99 billion outstanding on DeFi lending apps on September 30, 2025. Change only which loans count in the comparison.
| What counts? | Calculation |
|---|---|
| Apps and centralized lenders | 40.99 ÷ (40.99 + 24.37) ≈ 62.7% |
| Both, plus stablecoins issued against crypto collateral in debt positions (CDPs) | 40.99 ÷ 73.59 ≈ 55.7% |
The lower percentage describes the same day and the same app balances. It does not show a loss of market share over time. Galaxy also notes that a centralized lender could mint a CDP stablecoin to fund a customer loan: both financing steps may appear in the broader total. The report measures balances under its definitions, not the number of distinct borrowers.
New loans, unpaid principal and collateral: three different numbers
Tether's 18 November 2025 Ledn announcement reported more than $2.8 billion in bitcoin-backed loan originations since inception and more than $1 billion during 2025 to that point. Originations count lending activity over a period; they are not the amount still outstanding on that date. The announcement is an interested party's disclosure, not an independent loan-book audit.
Maple's institutional page describes permissioned secured lending. That establishes how the product is presented. It does not prove better underwriting, lower default risk or a market-wide return of institutional confidence. Those conclusions would require loan terms, realized performance and a comparable baseline.
Try it: where did the new loans go?
A hypothetical lender opens the quarter with $100 million in unpaid principal. It lends another $45 million, collects $35 million of principal and writes off $5 million from its books. At quarter-end it holds collateral worth $150 million. Assume dollar loans, no interest added to principal and no other balance changes. How much lending took place this quarter, and how much principal is still on the books?
Show the calculation and test a price drop
New lending: $45 million. Closing principal: $105 million. Start at $100m, add $45m, subtract $35m repaid and $5m written off. The book grew by $5m. The $150m is the value of assets pledged; adding it to $105m would count two different things.
Now cut the collateral value by 20%, to $120m. Before any repayment, liquidation or further write-off, booked principal is still $105m. What changed is the collateral cushion. These totals alone cannot tell you which individual loans can be liquidated.
Follow the collateral and the liquidation rule
For a centralized lender, relevant questions include custody, reuse of collateral, legal claims, withdrawal conditions and the lender's ability to meet obligations. For an onchain position, inspect accepted collateral, pricing, liquidation rules, upgrade powers and execution dependencies. Overcollateralization is a buffer valued at a moment in time; it is not a guarantee of recovery during a price gap or an illiquid sale.
Aave's liquidation explanation describes a health factor based on collateral, liquidation thresholds and debt. A position can become liquidatable when that factor falls below one. A reader should use the actual market rules rather than assume every DeFi loan follows the same parameters.
For an illustrative single-collateral case, $1,000 of collateral at an 80% liquidation threshold against $600 debt gives a health factor of 1.33. A fall to $700 gives 0.93. These invented inputs omit accrued interest and transaction costs. They demonstrate how a buffer can disappear; they are not recommended borrowing settings.
Rates are a signal to investigate, not a safety certificate
A low quoted borrow rate alone cannot establish that speculative leverage is low or that a market is healthy. Supply, utilization, incentives and rate-model settings all affect the interpretation. Nor does a stablecoin denomination eliminate redemption, issuer or collateral risk. Borrowing can be denominated in other assets, so a market estimate should say which loan currencies it covers.
A useful comparison records outstanding principal, originations, collateral valuation date, overdue debt, realized losses, liquidation performance and the methodology for counting borrowers. This creates a testable account of a credit product instead of inferring safety from market size or a familiar brand.
Revision history and scope
14 November 2026: Added a specific follow-up lesson to connect this article to the wider learning route. This navigation change does not extend the factual-review dates above.
Reviewed 19 October 2026. Separated dated research estimates from current market claims and removed the assertion that lending growth proves improved safety.
24 October 2026: Moved the earlier correction note below the lesson and added a reader-focused introduction. This is a presentation edit; the earlier factual review date is unchanged.
9 November 2026: Added a two-denominator check of Galaxy’s historical estimate and a hypothetical loan-book exercise. Clarified when a change in collateral value affects coverage rather than principal.
