#termmax @TermMax $BTC I tested TermMax’s Range Order AMM over the past two days. I originally just wanted a lending pool with a fixed interest rate, but I ended up staring at three symbols—FT, XT, and GT—for twenty minutes without daring to hit confirm.

At first glance, I thought it was like Pendle, where yield is split into components. But when I actually break down the trade, the underlying logic is completely different. 1 FT + 1 XT = 1 unit of debt token. This equation isn’t a division of yield rights—it’s making the “fixed interest rate” itself into a tradable mathematical object.

In traditional AMMs, you set the price; in TermMax, you set an APR range. The curator isn’t setting buy/sell prices—instead, it sets “within what interest-rate range am I willing to lend/borrow.” This design made me pause. The hardest part of fixed-rate protocols isn’t the matching—it’s making “interest-rate preferences” expressible and executable on-chain.

FT anchors the principal-repayment right, XT anchors the interest right, and GT is the governance credential. Together they equal the complete debt, but once split, each one becomes liquid inside the AMM. Before maturity, you can sell FT to get your principal back as expected, or sell XT to get your interest back as expected—you don’t have to wait until maturity to exit your fixed-rate position.

But the part I couldn’t fully understand is physical delivery settlement. Most protocols settle via auctions, with lenders receiving stablecoins back. TermMax’s documentation says “the lender receives the collateral asset directly.” What you lend out is USDC, but what you get back could be a bunch of RWA you may need to handle manually, or long-tail assets. Whether the curator’s pricing curve can handle extreme conditions is one question; a more practical one is: do ordinary users have the ability to deal with these “physical” collateral assets?

TermMax doesn’t share bad debt, doesn’t blur risk exposure—each range order is independent and profit/loss is on your own. This kind of design is unique in the fixed-rate space, but the complexity is shifted onto users.

BABY is “no custody, no bridging,” and TMX is “no pooling, no averaging, no acting on your behalf.” But how much power does the curator actually have? The width/narrowness of the range, collateral eligibility, and liquidation thresholds—these parameters are concentrated in whose hands. That ultimately determines whether this is truly decentralized fixed-rate lending, or just a CeFi product in a different shell.

What do you think? Does physical delivery make DeFi more real, or does it raise the bar too high? Let’s discuss in the comments. #tmx