Fixed-Rate Protocols

Ten ways to lock a borrowing rate on-chain – and the honest accounting of how little money is actually behind them.

Almost every corner of DeFi lets the borrowing rate float. Fixed-Rate Protocols – published at fixedrateprotocol.com under the Rate.Lock motif – is a landscape brief on the small set of teams trying to nail the rate down for a fixed term: the niche living beneath variable-rate giants like Aave and Morpho Blue. It walks ten core protocols page by page, routes one worked example (borrow 10 ETH for three months) through nine different rate-discovery mechanisms, and then does the thing most landscape decks skip – nets the TVL out and admits how small the field really is.

The conclusion is deliberately unflattering to its own subject. By net TVL the DeFi-native field is Wildcat ($150M) and Liquity V2 ($74M) and not much else, while the institutional RWA tier next door – tokenized Treasuries, Maple, Centrifuge – is roughly 50× bigger. The mechanisms are inventive; the demand has not shown up yet.

Written as an R3 landscape brief; published in case it saves anyone else the fortnight of protocol-doc archaeology.

---

What's in it

A single scrolling deck of 13 numbered pages plus a closing panel:

• Cover – the field at a glance: ten mechanisms all now live, ~$300M combined, and the caveat that the RWA tier next door dwarfs it.
• 02 · Why the niche exists – what fixed-rate and fixed-term actually mean, why borrowers want them (cash-flow planning, institutional mandates, hedging, structured products), and the three structural reasons the slice stays small: liquidity fragmentation across maturities, rollover risk, and capital-efficiency drag.
• 03 · Worked example – borrow 10 ETH for three months, routed through nine native mechanisms side by side. The clearest single page for seeing how differently each protocol discovers a rate.
• 04 · Cross-cutting – the field sliced five ways: rate discovery, collateral model, capital efficiency and leverage, flexibility vs. predictability, and maturity/scale.
• 05 · The comparison matrix – all ten protocols against mechanism, fixed term, rate-setter, collateral/risk, status and TVL, and chains.
• 06–10 · Protocol deep-dives – Morpho Midnight; Term Finance; the tokenized-bond family (TermMax, Fira, Secured Finance); rate-as-a-token and per-maturity pools (Inverse FiRM, Exactly); and the overlay/borrower-led group (IPOR, Wildcat, Liquity V2).
• 11–12 · The institutional / RWA frontier – the ~$15B tokenized-Treasury market that functions as the on-chain risk-free benchmark, then the credit tier where real balance sheets sit (Maple, Centrifuge, Clearpool).
• 13 · Adjacent and wound down – the neighbours and the graveyard: Notional, Yield Protocol, Element/DELV, Goldfinch, TrueFi, plus an explicit note on what was cut from the list and why.
• Closing – the honest take, and the open question: what finally brings the volume – a better mechanism, clearer regulation, or institutional distribution?

The ten protocols

Protocol	Rate discovered by	Status · net TVL (Jun 2026)
Wildcat	Borrower decree, permissioned lenders	Live · ~$150M outstanding
Liquity V2	The borrower (rate sets redemption order)	Live · ~$74M
TermMax	AMM curve on tokenized bonds (FT/XT/GT)	Live · ~$33–49M
Inverse · FiRM	Market price of DBR borrowing rights	Live · ~$20M
Term Finance	Weekly sealed-bid auction, single clearing price	Live · ~$13M
Fira	Supply/demand on fixed-maturity bonds	Live · ~$5M net (large gross book)
Exactly	Per-maturity pool utilization	Live · ~$3.8M
Secured Finance	On-chain order book + Itayose auction	Live · ~$0.6M
IPOR	Interest-rate swap AMM (overlay on Aave/Compound)	Live · IRS now legacy
Morpho Midnight	Two-sided offers, no interest-rate model	Live 21 Jul 2026 · ~$1.9M (Aug 2026)

The headline findings

• The field is two names. Wildcat and Liquity V2 are most of it. Everything else is under ~$75M.
• Fira's headline was doing the marketing. A ~$425M gross loan book nets to roughly $5M of real on-chain value once the collateral it loops through itself is stripped out. Net TVL is the only honest cross-protocol ruler, so the brief ranks on it throughout.
• Scale was never the survival trait. Notional once topped ~$843M and still wound down after the November 2025 Balancer hack. Goldfinch – a16z-backed, once the flagship of undercollateralized RWA credit – passed GIP-87 unanimously in June 2026 and went into maintenance mode with GFI down ~99.8% from its 2022 high.
• The money already moved next door. $15B of tokenized Treasuries at a blended 3.3% yield, with Circle's USYC having overtaken BlackRock's BUIDL as the largest single product, plus Maple ($2.1B) and Centrifuge ($1.63B) in institutional credit.
• Rate discovery is the real taxonomy. Auction, bond price, order book, a two-sided offer, a token you hold, pool utilization, a swap overlay, or simple borrower decree – mechanism choice is what actually separates these protocols, not branding.
• The best-funded attempt launched into silence. Morpho shipped Midnight on 21 July 2026 and it drew ~$1.9M. The mechanism was built properly by the largest team in on-chain lending, and the volume still did not appear – which is the cleanest evidence the brief has for its own thesis.

How it's built

One file. index.html, ~68KB, 1,264 lines, 14 <section> blocks. No build step, no bundler, no dependencies to install, no JavaScript.

• Type – Fraunces for display and Inter for body, loaded from Google Fonts. These are the only external network requests the page makes.
• Palette – defined as CSS custom properties in :root: paper #F2EBDD, ink #1A1816, accent #C44A36, plus muted/rule/highlight/olive supporting tones. Change a value there and it propagates through the whole deck.
• Favicon – an inline data: URI SVG padlock (the Rate.Lock motif), so there is no icon file to keep in sync.
• Responsive – a single @media (max-width: 900px) breakpoint collapses the multi-column layouts and lets the wide matrix scroll horizontally inside its own container.

The same design system backs the sister brief at trancheprotocol.com.

Run it locally

Nothing to install:

git clone https://github.com/mishablank/fixedrateprotocol.git
cd fixedrateprotocol
open index.html          # macOS – or just drag the file into any browser

If you want it served over HTTP (closer to production, and required if you add relative assets such as the social image):

python3 -m http.server 8000
# then visit http://localhost:8000

Deploy

The site is hosted on Cloudflare Workers with the static-assets integration, wired to this repository. Push to main and the live site rebuilds within seconds – there is no manual deploy step and no CI workflow in the repo.

.assetsignore keeps .git, .gitignore, .wrangler, .claude, wrangler.jsonc and itself out of the uploaded asset bundle. Anything added to the repo that should not be publicly fetchable needs a line there. Note that the Worker configuration itself lives in the Cloudflare dashboard rather than a committed wrangler.jsonc.

Editing conventions

• En dashes, never em dashes. Use – (or –) throughout. Three separate commits in this repo exist purely to undo em-dash regressions, so it is worth getting right the first time.
• Every number carries an as-of date. Figures are stamped in the page footer and in the source note under each table rather than floated as timeless facts.
• Net, not gross. When a protocol's headline TVL includes collateral it recycles through itself, the brief quotes the net figure and says so inline. Fira is the worked case.
• Scope discipline is explicit. Page 13 lists what was deliberately excluded and why – Pendle fixes yield rather than borrowing, Contango pivoted to variable-rate looping, Royco is an incentivized action market. Keep that list current when the scope is challenged.
• Page numbering is hardcoded in each section's data-slide attribute and footer. Inserting a page means renumbering both, in every section.

Data & methodology

• Sources – DefiLlama for TVL, rwa.xyz for the tokenized-Treasury tier, and individual protocol documentation for mechanism detail.
• As-of – protocol figures verified 21 June 2026; the page carries a revision note of August 2026.
• Scope – fixed-rate, fixed-term borrowing. Yield-fixing protocols, tranching chassis, and variable-rate loopers are covered only as adjacent context.

Because the underlying figures move continuously, treat every number here as a dated snapshot rather than a live feed.

Known gaps

One thing a next revision should address:

1. Social previews are broken. og:image and twitter:image point at the relative filename Fixed-rate-v4-og.png, which is not in the repo and 404s in production. Link unfurls on X, LinkedIn, Slack and iMessage need an absolute https://fixedrateprotocol.com/… URL and the actual image committed. A leftover https://your-domain.com/… placeholder also sits in an HTML comment. There is no og:url or <link rel="canonical"> either.

Closed since this README was first written:

• Morpho Midnight is no longer pre-mainnet. Fixed 1 Aug 2026. Midnight launched publicly on 21 July 2026 – offer-based fixed-rate, fixed-term credit, cbBTC/USDC on Base across a limited set of maturities, ~$1.9M TVL. The cover headline, subtitle, matrix row and takeaway, worked example, both cross-cutting bullets, the deep-dive on page 06, the closing stat block and the OG/Twitter meta were all reframed, and the mechanism was corrected from intent-based (the pre-launch whitepaper design) to offer-based (what actually shipped).
• A stale cross-reference. Fixed 1 Aug 2026. Page 13 pointed at "pages 10–11" for the institutional/RWA tier; it now reads pages 11–12.

Credits

Written by Mike Blank. Companion brief: trancheprotocol.com – yield tranching protocols, eight chassis compared.

License

No license file is currently present, so default copyright applies and all rights are reserved. If you want this to be reusable – quotable charts, forkable layout – add a LICENSE (MIT for the markup, or CC BY 4.0 if the intent is to license the research and prose).