Follow us for more threads and research like this. For professional research services, contact us at

7/ Final Takeaway 2025 proved this simple rule: ➡️ Market making is NOT a commodity. ➡️ Vault architecture determines your destiny. ➡️ Most users are blindly farming without knowing their actual risk exposure. We analyzed the NAV, drawdowns, correlations, and failure paths of

6/ The Real Categorization Forget APY. Here’s the true classification: -EdgeX → Savings Account with volatility farming -Hyperliquid → Macro crash hedge -Lighter → High-frequency latency bet -gTrade/Aster → You vs. Trader PnL -Ostium → Oracle latency lottery These are not

5/ The Hidden Risks No One Talks About Every vault is “selling” a specific risk: -Execution Risk → Lighter (sequencer lag) -Inventory/Delta Risk → Hyperliquid (unhedged exposure) -Counterparty Risk → gTrade/Aster (trader PnL) -Oracle/RWA Latency Risk → Ostium (lag = free

4/ Sharpe Ratios Tell the Truth Risk-adjusted performance matters more than APY. EdgeX Sharpe: ~8.4 (elite) • Almost no drawdowns • Delta-neutral, consistent spread farming Hyperliquid: 1.8 (solid) • “Lumpy” but lucrative during liquidations Lighter: 0.8 •

3/ The Crash That Exposed Everything — Oct 10 Tariff Meltdown A $19B liquidation wave hit crypto. This single event cleanly separated scalable models from fragile ones. Lighter: –5.35% in 24h • ZK sequencer lag → stale orders → arb bots drained it • Gave back ~45

2/ Wildly Different Returns NAV, June 11 → Nov 28, 2025: EdgeX: +24.8% Lighter: +18.6% Hyperliquid: +10.5% Aster: +10.9% gTrade: +4.9% Ostium: –0.6% ➡️ Mechanism dictates performance. The biggest winner wasn’t the highest APY. It was the vault with the most

1/ LP Vaults Are NOT Homogenous Yield Products Most users see “20% APY” and assume identical risk. Reality: Perp LP vaults fall into 3 completely different architectures, each with distinct failure modes: (A) CLOB Market Making • Hyperliquid (HLP) — unhedged inventory •

Not All Perp DEX Vaults Are Created Equal EdgeX | Hyperliquid | Lighter | gTrade | Aster | Ostium Most LPs think these yields are the same. They’re not. They are algorithmic hedge funds with totally different risk profiles. Here’s what we found. 👇

Full Report & Professional Research Services For the complete report: 👉

Conclusion USDai stands out for its thoughtful engineering and high-quality backing today. But it is not a fixed-risk instrument — it is a dynamic collateral system with a roadmap that introduces additional complexity and counterparty exposure. For allocators, USDai should be

Key Risks 1. Collateral Drift Risk The most significant risk factor. As AI infrastructure loans increase, USDai’s credit profile may diverge from today’s T-Bill–driven safety. 2. Governance Centralization All risk parameters depend on core-team discretion. No guardrails exist

Operational Risk Security foundations are strong for an early-stage protocol: two audits (Cantina, KTL), active bug bounty, clean findings. Weaknesses include: Short operational history No post-launch re-audit No live attestation partner Heavy reliance on off-chain enforcement

Governance USDai governance is fully centralized: Core-team multisig controls parameters, upgrades, collateral onboarding No DAO, timelock, or community veto Limited transparency on internal decision-making This is a key non-technical risk. For a protocol with evolving

Secondary Liquidity & Peg Stability Liquidity conditions have improved markedly since launch. USDai now trades close to par, supported by growing depth on Fluid/Plasma and other venues. However: Liquidity is venue-concentrated Depth outside ±2% remains thin Large redemptions

Minting & Redemption USDai operates a dual-token model: -USDai for liquidity -sUSDai for yield accrual Minting is straightforward and backed by the protocol’s reserve ratio. Redemptions for T-Bill–backed supply are near-instant, while redemptions tied to loan-backed collateral

Collateral & Backing USDai’s current collateral mix is heavily weighted toward short-duration U.S. Treasuries (~99%) via the M0 $M token, providing strong credit quality. A small portion (~1%) consists of AI infrastructure loans, with the protocol intending to increase this

Issuer Profile USDai is developed by Permian Labs, a well-capitalized, venture-backed team with institutional recognition. The project benefits from reputable investors and a coherent technical roadmap. However, the issuer retains full operational control, and no legal

Assessment: “Prime, but Unseasoned.” Today, USDai’s credit strength is anchored by short-duration U.S. Treasuries (~99% exposure). This gives it a strong starting point relative to most new stablecoins. However, the protocol is young, governance is centralized, and several

🧵 USDai — Risk Report USDai is one of the fastest-growing entrants in the synthetic-dollar and RWA-backed stablecoin space. Its design is ambitious: blend the safety of T-Bills with the yield of AI infrastructure. That combination deserves attention — and careful monitoring.