OM repeats LUNA's script? Market makers control 90% of the market. Revealing the truth behind the price crash

Amid the rapid development of the digital economy, the crypto asset market is facing unprecedented risks and challenges - on one side, there is the cloak of compliance and regulation, while on the other side, there are serious manipulation and information asymmetry.

At 4 a.m. on April 14, 2025, the cryptocurrency market once again caused an uproar. The MANTRA (OM) token, once known as the "compliant RWA weathervane", was forced to close positions on multiple centralized exchanges (CEX) at the same time. The price plummeted from $6 to $0.5, a single-day drop of more than 90%, and the market value evaporated by $5.5 billion. Contract players lost $58 million in liquidation. On the surface, it looks like a liquidity storm, but in fact it is a premeditated highly controlled and cross-platform "harvesting game". We will deeply analyze the causes of this flash crash, reveal the truth behind it, and explore the future development direction of the Web3 industry and how to avoid similar incidents from happening again.

1｜Comparison between the OM flash crash and the LUNA crash

The OM flash crash is similar to the LUNA crash in the Terra ecosystem in 2022, but the causes are different:

LUNA crash: mainly caused by the de-anchoring of the stablecoin UST. The algorithmic stablecoin mechanism relies on the balance of LUNA supply. When UST deviates from the 1:1 US dollar anchor, the system enters a "death spiral" and LUNA falls from more than US$100 to nearly US$0. This is a systemic design flaw.

OM flash crash: The investigation showed that the incident was a market operation and liquidity problem, involving forced liquidation of CEX and high control behavior of the team, not token design defects.

Both caused market panic, but LUNA was an ecosystem collapse, while OM was more like an imbalance in market dynamics.

II｜Control structure - 90% of the team and the dealer secretly hold

》Super-high concentration of control structure

According to the on-chain monitoring, the MANTRA team and its associated addresses hold a total of 792 million OM, accounting for about 90% of the total supply, while the actual circulation of tokens is less than 88 million, accounting for only about 2%. Such an astonishing concentration of holdings has caused a serious imbalance in the trading volume and liquidity in the market, and large investors can easily influence price fluctuations during periods of low liquidity.

》Phase-based airdrop and lock-up strategy - creating false heat

The MANTRA project adopts a multi-round unlocking scheme, which continuously extends the cashing cycle to precipitate community traffic into a long-term lock-up tool.

• 20% will be released upon the first launch to quickly spread market awareness;

• Cliff-like unlocking in the first month, and linear release in the following 11 months to create the illusion of initial prosperity;

• The partial unlocking ratio is as low as 10%, and the remaining tokens will be gradually vested within three years to reduce the initial circulation.

This strategy seems to be a scientific allocation on the surface, but in fact it is to attract investors with high commitments. When user sentiment rebounds, the project party introduces a governance voting mechanism to transfer responsibility in the form of "community consensus", but in actual operation, the voting rights are concentrated in the hands of the project team or related parties, and the results are highly controllable, forming a false trading boom and price support.

》 OTC discount transactions and arbitrage takeover

50% discount shipment: Many revelations in the community pointed out that OM was sold at a 50% discount on a large scale in the OTC market, attracting private equity and large investors to take over.

Off-chain-on-chain linkage: After arbitrageurs purchase OM at a low price in the OTC market, they transfer OM to CEX, create on-chain trading heat and volume, and attract more retail investors to follow up. This double cycle of "cutting leeks off the chain and building momentum on the chain" further amplified price fluctuations.

Three｜MANTRA's historical problems

MANTRA's flash crash, its historical problems also laid hidden dangers for this incident:

The hype of the "compliant RWA" label: The MANTRA project has gained market trust with its "compliant RWA" endorsement. It has signed a $1 billion tokenization agreement with UAE real estate giant Damac and obtained the VARA VASP license, attracting a large number of institutions and retail investors. However, the compliance license did not bring real market liquidity and decentralized holdings. Instead, it became a cover for the team to control the market. With the help of the Middle East compliance license to attract money, regulatory endorsement has become a marketing tool.

OTC sales model: According to reports, MANTRA has raised more than $500 million through the OTC sales model in the past two years. The way it works is to absorb the selling pressure of the previous round of investors by continuously issuing new tokens, forming a cycle of "new for old, old for new". This model relies on continuous liquidity, and once the market cannot absorb the unlocked tokens, it may cause the system to collapse.

Legal disputes: In 2024, the Hong Kong High Court handled the MANTRA DAO case involving allegations of asset misappropriation. The court required six members to disclose financial information, and its governance and transparency itself had problems.

Four | A deeper analysis of the causes of the flash crash

1) Failure of liquidation mechanism and risk model

Split of risk parameters on multiple platforms:

The risk control parameters (leverage limit, maintenance margin rate, and automatic liquidation trigger point) of each CEX for OM are not unified, resulting in the same position facing completely different liquidation thresholds on different platforms. When a platform triggers auto-deleveraging (ADL) during a low liquidity period, the sell orders overflow to other platforms, causing "cascading liquidations".

Tail risk blind spot of risk model:

Most CEX adopts VAR (Value at Risk) model based on historical volatility, which underestimates extreme market conditions (tail events) and fails to simulate "gap" or "liquidity exhaustion" scenarios. Once the market depth drops sharply, the VAR model fails, and the triggered risk control instructions increase liquidity pressure.

2) On-chain fund flow and market maker behavior

Large hot wallet transfer and market maker withdrawal:

FalconX hot wallet transferred 33 million OM (≈20.73 million US dollars) to multiple CEXs within 6 hours, which is suspected to be caused by market makers or hedge funds liquidating positions. Market makers usually hold a net neutral position in high-frequency strategies, but in the face of extreme volatility expectations, in order to avoid market risks, they often choose to withdraw the two-way liquidity they provide, resulting in a rapid expansion of the bid-ask spread.

The amplification effect of algorithmic trading:

When the automatic strategy of a quantitative market maker detected that the OM price fell below the key support (5% below the 10-day moving average), it activated the "flash selling" module to arbitrage across products between index contracts and spot, further exacerbating the spot selling pressure and the soaring funding rate of perpetual contracts, forming a vicious cycle of "funding rate-spread-liquidation".

3) Information asymmetry and lack of early warning mechanism

Long-delayed early warning and community response on the chain:

Although there are mature on-chain monitoring tools (Arkham, Nansen) that can warn of large transfers in real time, the project party and major CEX have not established a closed loop of "early warning-risk control-community", resulting in the failure of on-chain capital flow signals to be converted into risk control actions or community announcements.

The herd effect from the perspective of investor behavior:

In the absence of authoritative information sources, retail investors and small and medium-sized institutions rely on social media and market push notifications. When prices fall rapidly, panic liquidation and "bottom-fishing" are intertwined, which amplifies the trading volume (the trading volume increased by 312% month-on-month within 24 hours) and volatility (the historical volatility of 30 minutes once exceeded 200%) in the short term.

V. Industry Reflection and Systematic Countermeasures and Suggestions

In order to deal with such incidents and prevent the recurrence of similar risks in the future, we put forward the following countermeasures and suggestions for reference only:

1. Unified and dynamic risk control framework

• Industry standardization: For example, the formulation of a cross-exchange liquidation protocol (CELP), including: interoperability of liquidation thresholds, real-time sharing of key parameters (maintenance margin rate, ADL trigger line) and large-scale position snapshots by various platforms; dynamic risk control buffer, starting the "buffer period" (liquidation grace period, T+δ) after the liquidation is triggered, allowing other platforms to provide limit buy orders or algorithmic market makers to participate in the buffer to avoid instantaneous large-scale selling pressure.

• Tail risk model enhancement: Introduce stress testing and extreme scenario simulation (scenario analysis), implant "liquidity shock" and "cross-product squeeze" simulation modules in the risk control system, and conduct regular systematic drills.

2. Decentralization and insurance mechanism innovation

• Decentralized Liquidation Chain

  Based on the smart contract-based clearing system, the clearing logic and risk control parameters are put on the chain, and all clearing transactions are open and auditable. Use the cross-chain bridge (Cross-chain Bridge) and the oracle (Chainlink) to synchronize the prices of multiple platforms. Once the price falls below the threshold, the community nodes (liquidators) will bid to complete the clearing, and the income and fines will be automatically allocated to the insurance pool.

• Flash Crash Insurance

  Introduce an option-based flash crash insurance product: When the OM price falls by more than a set threshold (such as 50%) within a specified time window, the insurance contract automatically compensates the holder for part of the loss. The insurance premium rate is dynamically adjusted based on historical fluctuations and the concentration of funds on the chain.

3. On-chain transparency and early warning ecosystem construction

• Large-holder behavior prediction engine

  The project party should cooperate with data analysis platforms such as Nansen and Dune to develop an "Address Risk Score" (ARS) model to score potential large-amount transfer addresses. Once a large-amount transfer occurs to an address with a high ARS, it will automatically trigger a platform and community early warning.

• Community Risk Committee

  Composed of project owners, core advisors, major market makers and representative users, it is responsible for reviewing major on-chain events and platform risk control decisions, and issuing risk notices or suggesting risk control adjustments when necessary.

4. Investor education and market resilience enhancement

• Extreme market simulation platform

  Develop a simulated trading environment to allow users to practice stop-loss, lightening, hedging and other strategies in simulated extreme market conditions, and enhance risk awareness and response capabilities.

• Leveled leverage products

  For different risk preferences, we launched leveled leverage products: low-risk level (leverage ≤ 2×) uses the traditional liquidation model; high-risk level (leverage ≥ 5×) needs to pay an additional "tail risk margin" and participate in the flash crash insurance pool.

VI. Conclusion

The flash crash of MANTRA (OM) is not only a major shock in the field of cryptocurrency, but also a severe test for the overall risk management and mechanism design of the industry. As we have discussed in detail in the article, the extreme concentration of holdings, the false prosperity of market operations, and the lack of cross-platform risk control linkage have jointly created this "harvest game".

Only through cross-platform standardized risk control, decentralized clearing and insurance innovation, on-chain transparent early warning ecosystem construction, and extreme market education for investors can we fundamentally enhance the impact resistance of the Web3 market, prevent similar "flash crashes" from happening again in the future, and build a more stable and credible ecosystem.