fCash pool
Proportion = fCash balance ÷ (fCash balance + cash balance). A higher proportion means more fCash relative to cash in the pool, which prices a higher implied fixed rate.
Size a trade against the curve
Borrowing adds fCash to the pool and removes cash (pushes proportion and the rate up); lending adds cash and removes fCash (pushes proportion and the rate down). Execution price is approximated at the pre-trade marginal rate.
A curve, not a clearing price
fCash's implied rate comes from pool proportion on a continuous AMM curve, unlike the single uniform clearing price of the fixed-rate loan auction model or the oracle-free bucket walk of Ajna's LUP. Compare against a variable-rate utilization curve with the Morpho adaptive-curve IRM, a third fixed-yield mechanism with Pendle's PT/YT split, or scan rates across markets with the multi-protocol lending rate comparator.
Four ways DeFi discovers a fixed rate, and why the curve shape matters
Fixed-rate DeFi lending has converged on a handful of genuinely different price-discovery mechanisms, and Notional's fCash market is one of the oldest and most distinct. Term Finance and similar protocols run periodic auctions: bids and offers accumulate over a window, and the protocol finds one clearing price that matches supply and demand for that round — the rate simply does not exist between auctions. Ajna's bucket book is continuous but oracle-free and discrete: the Lowest Utilized Price is a mechanical readout of which lender-chosen price bucket is needed to cover total pool debt, moving only when deposits or debt at those buckets change. Pendle splits a yield-bearing asset into principal and yield tokens and lets their market price imply a fixed rate through discount-to-par. Notional does something closer to a standard token AMM: fCash (a claim on a fixed amount of the underlying at a future date) and cash sit in the same pool, and the ratio between them — the proportion — is fed through a logit function to produce an exchange rate, which is then annualized against time to maturity to give the implied fixed rate.
The logit shape is deliberate. Unlike a simple constant-product curve, it lets governance set both where the curve is centered (rateAnchor, the rate at a balanced 50/50 pool) and how steep it is (rateScalar, which controls how much a given trade moves the rate). A larger rateScalar flattens the curve, so the market can absorb bigger lend or borrow trades with less rate slippage; a smaller rateScalar makes the curve steep, so the same trade swings the rate much further. This is the same slippage phenomenon a token AMM has when a big swap moves further down a bonding curve — except here what is "slipping" is a fixed interest rate, not a token price. Because the rate is continuous, it also has to convert differently as time passes: the same pool proportion implies a higher annualized rate the closer the market sits to maturity, since a fixed compounding return over a shrinking window annualizes to a bigger number, which is why Notional (and protocols like it) widen rateScalar as a market nears its settlement date to keep the curve from becoming punishingly steep right before expiry.
None of this is better or worse than an auction or a bucket book in the abstract — each tradeoff shows up differently. An auction gives every participant in a round the exact same price but makes them wait for the round to clear. A bucket book removes the oracle but ties liquidation risk to lender behavior at specific price levels rather than a market price. An AMM curve like fCash's gives instant, always-on pricing and composability with other DeFi money legos, at the cost of slippage on size and a curve shape borrowers and lenders both need to understand before trading close to a pool's edges.