#termmax @TermMax Yesterday’s tweets reminded me of a previous promotional task I did: there were two projects at the same time. I posted a tweet for one project, and after I hit send, I realized I’d posted the wrong thread—the tweets meant for Project A ended up going to Project B. No choice but to rewrite and repost the tweets for Project A. Why am I bringing this up? Because yesterday, lots of people replied to my tweet as if it were about Project A, not realizing their replies were essentially going to Project B’s tweet. I looked at it for a while—turns out I hadn’t posted the wrong thing.
Over the past couple of days, I’ve been looking into the liquidation mechanism in lending agreements and noticed an easily overlooked point: many users choose an agreement based mostly on the interest rate, and very few think about whether the liquidation step is actually reliable.
I’ve seen this situation more than once—when the market is highly volatile, the liquidity of a certain collateral asset suddenly dries up. Liquidation can’t proceed, and ultimately lenders can’t get their principal back. The platform also has no good options; it can only wait or take the loss itself. These incidents usually aren’t a problem with the interest-rate design, but a design flaw in the liquidation mechanism. Yet when people study agreements, most of them basically don’t look at this part.
Following this line of thinking, I’ve been researching TermMax these past few days, and I found that its design approach here is quite different from most lending protocols. First, it isolates the market: each lending market is independent. If a specific collateral asset has a problem, the risk won’t propagate to other markets. This isn’t entirely new, but it’s a necessary baseline. What really made me pay extra attention is the liquidation design—if there isn’t enough liquidity in the market to complete normal liquidation, the lender will directly receive the corresponding proportion of the collateral asset in kind, rather than forcing the protocol to match orders in a bad way and create bad debt, or having liquidation fail altogether and leaving the lender to absorb the loss themselves.
At its core, this design means they’ve already thought through the worst-case scenario—“liquidation failure”—instead of assuming liquidation will always execute smoothly.
Of course, this doesn’t mean there’s no risk. Oracle dependency, and the collateral asset’s own liquidity depth—these old issues still exist. In-kind settlement only reduces losses to a relatively controllable range; it doesn’t eliminate risk. Also, the protocol’s current TVL isn’t that large yet. No matter how comprehensive the mechanism design is, it still ultimately needs to be validated by real lending demand and real trading depth.
My own view is that this kind of design may not attract attention in the short term as much as high-leverage products do.
Over the past couple of days, I’ve been looking into the liquidation mechanism in lending agreements and noticed an easily overlooked point: many users choose an agreement based mostly on the interest rate, and very few think about whether the liquidation step is actually reliable.
I’ve seen this situation more than once—when the market is highly volatile, the liquidity of a certain collateral asset suddenly dries up. Liquidation can’t proceed, and ultimately lenders can’t get their principal back. The platform also has no good options; it can only wait or take the loss itself. These incidents usually aren’t a problem with the interest-rate design, but a design flaw in the liquidation mechanism. Yet when people study agreements, most of them basically don’t look at this part.
Following this line of thinking, I’ve been researching TermMax these past few days, and I found that its design approach here is quite different from most lending protocols. First, it isolates the market: each lending market is independent. If a specific collateral asset has a problem, the risk won’t propagate to other markets. This isn’t entirely new, but it’s a necessary baseline. What really made me pay extra attention is the liquidation design—if there isn’t enough liquidity in the market to complete normal liquidation, the lender will directly receive the corresponding proportion of the collateral asset in kind, rather than forcing the protocol to match orders in a bad way and create bad debt, or having liquidation fail altogether and leaving the lender to absorb the loss themselves.
At its core, this design means they’ve already thought through the worst-case scenario—“liquidation failure”—instead of assuming liquidation will always execute smoothly.
Of course, this doesn’t mean there’s no risk. Oracle dependency, and the collateral asset’s own liquidity depth—these old issues still exist. In-kind settlement only reduces losses to a relatively controllable range; it doesn’t eliminate risk. Also, the protocol’s current TVL isn’t that large yet. No matter how comprehensive the mechanism design is, it still ultimately needs to be validated by real lending demand and real trading depth.
My own view is that this kind of design may not attract attention in the short term as much as high-leverage products do.