IRIS brings fixed-rate borrowing to SparkLend lending

By Daniel Okafor

IRIS Protocol said on November 9, 2026, that it plans to launch fixed-rate borrowing on SparkLend in mid-November, with loan terms available from one day to 90 days.

IRIS describes itself as an origination layer sitting atop variable-rate lending venues, “starting with SparkLend.” Its first integration lets borrowers fix their borrowing cost while solvers absorb and manage the floating-rate position underneath.

How does IRIS turn variable lending into fixed terms?

A borrower submits a loan intent specifying the asset pair, amount, duration and maximum fixed annual percentage yield, or APY. The request enters an off-chain request-for-quote, or RFQ, system, where market makers send direct prices instead of the borrower using a public order book.

Competing solvers return signed fixed-rate quotes, each backed by a bond. That bond gives the solver something to lose if it fails to honor the commitment. Once the borrower accepts, the fixed terms apply and the winning solver manages the corresponding floating-rate position on SparkLend.

The separation fixes the borrower’s rate but does not eliminate rate risk from the system. If SparkLend’s variable rate rises during the loan, the solver bears the difference. After accepting, the borrower knows the cost for the selected term even as the solver continues managing the underlying loan.

IRIS also gives solvers room to refinance positions during the term. That flexibility covers cases involving overdue rates and grace periods after a loan reaches maturity.

Does IRIS create a separate liquidity pool?

No. IRIS does not create prebuilt fixed-rate liquidity pools or a new liquidity silo. Capital comes directly from the underlying venue and from solvers, so the fixed-rate market depends on active solver participation rather than a dedicated pool.

That structure leaves quote quality and delivery risk with the solvers. Their bonds stand behind the signed terms, while their rate forecasts determine whether the spread between the fixed quote and SparkLend’s variable borrowing cost is profitable. If forecasts are wrong, their bonds are at risk.

SparkLend held approximately $4.66 billion in total value locked as of November 9, 2026. Outstanding borrows stood at about $2.64 billion, a 47% increase over the preceding three months, according to figures cited in the IRIS announcement.

What does the launch change for borrowers?

For borrowers, the immediate benefit is a known cost of capital over a selected 1-to-90-day window. A user locking 30 or 90 days can plan positions with less guesswork than someone watching a variable APY drift. The maximum-APY parameter provides a hard limit: if no solver can meet it, no deal proceeds.

For solvers, IRIS creates a new spread business. A solver earns the difference between the fixed rate it quotes and the variable rate it pays on SparkLend, provided its forecast holds. When the forecast misses, the same bond backing the quote is exposed to the loss.

The RFQ process is not fully onchain. This structure can support speed and pricing, as is common in RFQ systems, but borrowers still rely on quoting infrastructure to remain available and function properly.

Is expansion beyond SparkLend confirmed?

No. IRIS’s description leaves room for the layer to expand to other variable-rate markets, but the November 9 announcement includes no additional integration or timetable for one. SparkLend is the only named venue at launch.

Three measures will show whether the model gains traction: the number of participating solvers, the competitiveness of their quotes relative to SparkLend’s floating rates, and the share of roughly $2.64 billion in outstanding borrows that eventually routes through IRIS.

The next checkpoint is mid-November, when the integration is scheduled to go live and demonstrate competitive quoting and measurable routing through IRIS; until then, fixed-rate borrowing remains an unproven addition to SparkLend’s market structure.