Needed by the Chain, Optional for the Business: What MANTRA Teaches Token Designers

MANTRA's coin is mandatory for every transaction on its network. It's optional for everything the network was built to do.

That gap is the core finding of our tokenomics audit of MANTRA (formerly OM), which closed at 22 out of 100, rating CC, on data as of Sep 9, 2026. The coin trades at a $29,1m market cap. The network under it holds $540k in TVL and processes 726 transactions a day. Market cap sits 54 times above network TVL and 78 times above the $370,9k of real-world assets actually issued on-chain.

This isn't a recap. The full report is here. This is about the design decisions behind those numbers, and what a team preparing an RWA chain or any infrastructure token can take from them before TGE.

A token can be required and still capture nothing

The heaviest block in our methodology, Token Product Linkage, carries 40% of the score. MANTRA scored 1,32 out of 5 there. The problem isn't missing utility. It's utility wired to the wrong layer.

For UAE-based teams this is the closest case study available. MANTRA holds a VARA licence through Mantra Finance FZE in Dubai and announced its tokenization deals with Dubai real-estate names, yet none of that reaches the coin. A VARA or ADGM permission attaches to a legal entity; it never attaches itself to a token, and no licence obliges anyone to settle in one.

On the infrastructure side, the link is strong. Gas is paid only in MANTRA. You can't become a validator without it. Governance runs on it.

On the business side, the link doesn't exist. The VARA licence belongs to a separate legal entity, Mantra Finance FZE. Tokenized assets on the network are issued by Ondo and Circle, and settlements on them don't run in the coin. Validation by Google Cloud and Binance is paid through emission, not through demand for the token.

So ask the question nobody on the team wants to hear: if tokenization in the UAE grows exactly as the 2024 and 2025 press releases promised, what does a MANTRA holder get? The extra gas. Nothing else in the model routes business growth to the coin.

If your token only pays for the rails, it earns only what the rails earn. At 726 transactions a day, that's close to nothing.

Staking paid from emission is dilution on a delay

The network's mint module is fixed at 16% a year. That's 1,30b new MANTRA annually, roughly $16,1k a day at the audit price. Stakers receive it, and aggregators quote 16% to 29% a year depending on the date.

It looks like yield. It isn't. Here's how the money layers break down: user fees are negligible at 726 transactions a day, there's no measurable protocol revenue (DefiLlama has no fee adapter for the network), and nothing from usage reaches holders. The payout is new supply. A staker keeps their share. A non-staker loses 16% of it every year.

About 29% of issued supply is staked. Everyone outside that 29% pays for the rest.

Now the other side of the ledger. There's no fee burn, no buyback with destruction, no collateral lock against asset issuance. Emission of 1,30b a year meets nothing. At this rate the 10b cap is reached in roughly 15 months, by the end of 2027. Then emission stops, and the project hasn't published how it will pay for network security after that.

Staking rewards are a security budget. They need a sink on the other side and a funding source for the day emission ends. Without both, you haven't solved a problem. You've scheduled one.

A burn is only as final as the address that holds it

After the April 2025 crash, the project burned 300m OM. It was the only absorption measure in the token's history.

The coins sat at a burn address the project didn't monitor. On Aug 20, 2026, an exploit of a known Cosmos EVM vulnerability drained that address together with a dormant multisig: 720,9m MANTRA, about $3,6m, went back into circulation. The theft went unnoticed for about four hours. Nothing has been recovered.

The buyback announced in August 2025 doesn't close the gap. At least $45m was promised, but the bought coins aren't burned. They're staked by the project's validator set, where they keep earning emission and voting power for the issuer. Supply doesn't shrink. It changes owner. No consolidated report on actual volumes has been published.

Two rules follow. Irreversibility has to be engineered, not assumed: a module-level burn or a provably unspendable address. And a buyback that doesn't destroy coins is a treasury operation, not a supply mechanism. Call it that in your docs, or the market will call it something worse.

Your documentation is part of your monetary policy

MANTRA's docs state 3% inflation. The live network parameter is 16%. That's a 5,3x gap in the single most important monetary number, and no project source explains it.

It isn't the only one. The Token Transparency Filing sends 40% of emission to the association. The distribution module's community tax is zero, so through the standard Cosmos path the entire emission goes to validators and delegators. The supply cap has changed three times in two years: 888 888 888 on the ERC-20, no cap at genesis, then 2,5b, then 10b after the redenomination. Treasury, team and vesting addresses aren't labeled, so nobody outside can check the published unlock schedule against the chain.

For a project that positions itself on regulatory compliance, this is the costliest gap of all. When one published parameter is off by a factor of five, every other number the team publishes starts at a discount.

The fix is cheap: one source of truth generated from chain parameters, plus labeled addresses. It costs less than any campaign and does more for trust.

Announced deals aren't on-chain assets

MANTRA announced a deal with MAG for $500m in July 2024 and one with DAMAC for at least $1b in January 2025. On-chain, as of the snapshot, $370,9k of assets have been issued, both by third-party issuers. The gap between announced and executed is more than four thousand times. In May 2025, the MAG partner signed with a competing network for the same properties.

Network data tells the same story: 11 537 addresses over two years, network load of 0,093%, TVL down 88,1% from its peak, and 99,1% of what's left sitting in one external app, Quickswap V4.

A market can price a narrative for a while. MANTRA's 2024 rally ran roughly seventyfold on one. But the metric that decides the long run is the one a founder can't announce into existence: assets actually issued, settled and paying fees on your chain.

What the audit doesn't claim

A fair reading needs the limits.

Two assets passed verification. VARA licence VL/25/02/001 is active in the regulator's public register. Inveniam Capital Partners announced the acquisition of the project in June 2026, a year after investing $20m in it, though formal closing hasn't been publicly confirmed. If the rating ever moves up, it'll move on these two.

The cause of the April 2025 crash hasn't been established. Coins moved to exchanges beforehand, but who owned them is unknown, and both existing investigations were run by interested parties and never published. The audit doesn't endorse claims of an organized team dump.

The score itself sits on a boundary. Under conservative assumptions it's 19, under optimistic ones 26, which would put it in CCC. Confidence is 74 out of 100. Unlabeled addresses and a 6,9x gap in daily volume between CoinGecko and CoinMarketCap keep it there.

Five questions to ask your own token before TGE

1. If the business grows tenfold, through which exact flow does that growth reach the token?
2. What pays stakers, revenue or new supply, and what happens when emission ends?
3. Which mechanism takes coins out of circulation, and can anyone reverse it?
4. Do your docs, your filing and your chain parameters show the same numbers today?
5. Which addresses would an outside analyst need to verify your unlock schedule, and are they labeled?

MANTRA answers all five badly. Most projects answer two or three well. The ones that answer all five are the ones worth funding.

What would change the verdict

The audit names three moves, each verifiable on-chain. A mechanism that removes coins from circulation independently of emission would lift tokenomics sustainability. A first material issuance of tokenized assets on the project's own network, settled in the coin, would lift Token Product Linkage. Labeled treasury addresses and remediated findings would lift governance and security: 14 of the 16 findings from the Hacken audit still sit in accepted risk.

We run these model reviews out of Dubai, so if your token is licensed here we can go through the absorption side, the disclosure set and the unlock schedule in one pass.

None of these needs a press release. All of them show up on-chain.

Read the full MANTRA ($OM) Tokenomics Audit. If you're designing a token and want these five questions answered before the market asks them, that's what a Tokenomics Audit is for.