The one Monad LST that doesn't expose its own exchange rate on-chain — a small but real gap that means its TVL comes from DeFiLlama rather than NullTerminal's own index.
Kintsu issues sMON, a liquid-staking token representing staked MON. Mechanically it works the same way as Monad's other LSTs from a user's perspective — deposit MON, receive sMON, keep earning staking rewards while sMON itself stays usable elsewhere — but its contract architecture differs under the hood.
Three of Monad's four liquid-staking tokens are plain ERC-4626 vaults, which means their exchange rate is a simple, standard, on-chain-readable function call — NullTerminal reads it directly via `convertToAssets()`. Kintsu keeps its rate in a separate staking contract that sMON itself doesn't expose the same way, so only `totalSupply()` is read on-chain for sMON directly; NullTerminal can't read Kintsu's true exchange rate the way it does for shMON, gMON or aprMON. Instead, Kintsu's TVL is sourced from DeFiLlama's protocol feed (matched by name, cached 15 minutes upstream), converted to a MON-denominated figure, and divided by the on-chain-read supply to imply a rate — a derived number rather than a directly measured one. If DeFiLlama's data is ever unavailable entirely, there's a documented last-resort fallback that assumes a flat 1:1 rate rather than showing a blank or zero value, so a temporary DeFiLlama outage degrades to "probably close, not exactly right" instead of "no data."
NullTerminal's own indexer polls every few minutes for all four LSTs, but Kintsu's figure additionally depends on DeFiLlama's own 15-minute cache sitting upstream of that — the two caching layers stack, so Kintsu's numbers can lag real-time by meaningfully more than shMON, gMON or aprMON's near-live, directly-read figures. This is a genuine data-quality distinction, not a reason to distrust Kintsu's underlying protocol — it's specifically about how confidently NullTerminal can timestamp the number it shows you.
The same underlying-validator risk applies as with any Monad LST — a slashing event or validator incident would ultimately show up in a rate that stops growing. The added wrinkle for Kintsu specifically is that its externally-sourced TVL is a step removed from direct on-chain verification, so it's marginally more exposed to reporting lag or discrepancy than an LST whose rate can be checked directly against its own contract at any moment.
Because its rate is implied from DeFiLlama's TVL feed rather than read directly from an on-chain function — and DeFiLlama's own 15-minute cache sits upstream of NullTerminal's own indexing cycle, so two caching layers stack instead of one.
There's a documented last-resort fallback that assumes a flat 1:1 MON-per-sMON exchange rate rather than showing no data at all — a deliberate degrade-gracefully choice, though it means the figure shown during an outage is an approximation, not a measurement.
Sources: GET /v1/defi/staking — NullTerminal
Last reviewed 2026-07-08. More Monad research.
Monad DeFi board · Markets · Swap on NullTerminal