Alpha Recap 37: LayerZero's ATLAS, yield on Liminal and AQAv2 on Hyperliquid

The Alpha Recap aims to bring you the most important Alphas from the crypto market this week. Every Friday, we offer a digest of the most valuable information from our Alpha Feed.

Reserved for OAK Premium members, the Alpha Feed brings together insights, yield and airdrop strategies, as well as key market information. In other words, what makes up OAK Research's DNA: providing you with filtered content that goes beyond the market's noise.

LayerZero announces ATLAS, should we see it as a direct competitor to Hyperliquid?

This week, LayerZero unveiled ATLAS, short for "Aggregated Trading Liquidity and Settlement," a trading infrastructure built on Zero, its blockchain announced last February. In concrete terms, the idea is to make ATLAS the reference infrastructure for tokenized assets and institutional players, with announced performance of 200,000 TPS and sub-millisecond latency.

On the operational side, ATLAS will be available in two configurations, both built on the same engine. Initially, Open ATLAS will allow any crypto-native application to build on the infrastructure permissionlessly, while Institutional ATLAS, as its name suggests, will target institutions looking to supervise their product in order to align with their regulatory requirements.

In practice, ATLAS's economic model relies on a single trading fee, of which 20 to 65% goes to the platform operating the front-end, depending in particular on the amount of ZRO tokens staked. The remainder is then split between the market creator, who receives 25%, while the remaining 75% is dedicated to the buyback & burn of ZRO tokens.

A turnkey infrastructure, simplified programmability allowing third-party platforms to plug into the protocol, market creators that echo how HIP-3 works and a token placed at the center of a circular economy... it is fairly difficult not to draw a parallel with what Hyperliquid has built.

That said, beyond what ATLAS offers on paper, let's remember that Hyperliquid built its "AWS of liquidity" thesis on the efficiency of its perps infrastructure, whereas ATLAS is starting from scratch and must now convince builders to come and build on its infrastructure. Nevertheless, several projects have already announced their arrival, among them GTE, Bullish, Defined and TrueNorth.

All this being said, it seems to us that the ground on which LayerZero could genuinely compete with Hyperliquid lies rather on the institutional side, with work already announced alongside the DTCC, ICE, Citadel Securities and Google Cloud. That is precisely the segment Hyperliquid is trying to win over with the Hyperliquid Policy Center. Ultimately, it could all come down to timing.

→ Is the comparison between ATLAS and Hyperliquid justified, and how should we really assess the significance of this announcement? The answers in our dedicated Alpha published this week.

Subscribe to Blocknote

Your weekly crypto digest delivered directly to your inbox.

---

Liminal capitalizes on the return of volatility on Hyperliquid

The market's recent rebound is enough to make anyone positioned short rather nervous. Bitcoin has gained more than 24%, ETH 31%, HYPE 40% and SOL 43%, a move that liquidated a record volume of short positions and sent funding rates soaring on perps markets, with Hyperliquid leading the way.

As a result of this pump, on August 22, annualized funding rates reached roughly 15% on BTC, 14% on ETH, 21% on SOL and 32% on HYPE, against levels usually closer to 10%. On some memecoins, the spike was even more violent, with 42% on XPL, 82% on PUMP and up to 89% on FARTCOIN.

It is precisely in this kind of environment that Liminal's strategies become profitable. As a reminder, the protocol aims to become Hyperliquid's "Native Yield Layer" by aggregating its various sources of return (funding, lending, BLP) to turn them into products that are simple to integrate, such as xTokens or limUSD.

xHYPE, for example, provides exposure to a delta-neutral strategy on HYPE/USDC. In this model, the yield comes from funding generated on the market, without directional exposure to the asset's price. It is currently showing around 30% APY over the last 7 days. In parallel, limUSD, which aggregates several strategies in order to diversify yield sources, has gone from around 8% before the rebound to more than 17% over the last seven days.

For the sake of transparency, let's remember that these yields are obviously not set in stone: a normalization of the market and of funding rates would mechanically bring them back down, and the strategies remain exposed to the risks specific to Liminal and Hyperliquid. But in the current environment, Liminal remains one of the most effective solutions for turning Hyperliquid's activity and volatility into yield, a conviction strong enough that we ourselves allocate around 10% of our treasury to it.

→ OAK Research Premium members can follow our Portfolio's movements in real time, where we detail the reasoning behind the rebalancing we carry out on it. Join us, and enjoy a 25% discount with the code "LILIAN25," on top of a 7-day trial period to test all of our platform's features.

Earn up to 10% yield

Access real yield generated on Hyperliquid through Liminal's tokenized strategies.

---

Hyperliquid launches manual borrowing on testnet: a first lever for AQAv2?

Since Wednesday, AQAv2 has been officially live on Hyperliquid. In concrete terms, the protocol now captures 90% of the revenue generated by its ecosystem's USDC reserves, with Circle keeping the remaining 10% and Coinbase acting as treasury deployer.

To give you an idea, on roughly $5 billion of USDC deposited, that represents nearly $178.5 million in additional annual revenue, entirely independent of trading volume, and therefore that much more buying pressure on HYPE once it is passed on to the Assistance Fund.

In parallel, on the same day, Hyperliquid announced that it was testing a manual borrowing feature on testnet, and this piece of news went relatively unnoticed.

As a reminder, borrowing on HyperCore currently only exists in automated form. More precisely, it is triggered as soon as an order exceeds the available balance of a user who has enabled Portfolio Margin. The rollout of manual borrowing changes things for depositors (manual borrowing being by definition far more flexible and practical than the automated kind), since a rise in borrowing demand mechanically pushes up the pool's utilization rate, which pushes up the yield paid out, which in turn attracts more USDC deposits, and so on.

This is precisely where the two mechanisms meet. Every additional dollar of USDC attracted by this yield mechanically widens the base on which AQAv2 takes its 90%, without the two setups cannibalizing each other, since they capture distinct flows. In short, the same dollar "works twice," for the user and for the protocol.

→ In our dedicated Alpha, we detail what opportunities this composability between HyperCore and HyperEVM could open up, particularly on the tokenization of lending positions. We also explained what the concrete impact of this new feature could be in light of AQAv2.

Become Premium

Unlock all our research and get the right insights, at the right time.