The Journey of Hyperliquid (Part 1)

A specter haunts the blockchain: the specter of hyperliquidity. To carry out a holy siege against this specter, all CEX forces—offshore and onshore, Perp DEXs and FTX fragments, Solana's radicals and L2 upstarts—have united. The US dollar is the anchor of liquidity, Hyperliquid is the equivalent of crypto liquidity, and trading is increasingly concentrated in mainstream currencies like BTC and ETH. People, willfully or deliberately, have forgotten that Hyperliquid seized the opportunity to grow and develop, inadvertently blossoming into a towering giant. People are talking about the possibility of Hyperliquid flipping Binance, and it seems that Aster's attack is just the beginning of a long offensive. However, the 20x leverage limit in 2021 did not end the competitive situation between FTX and Binance, and today's Aster 1001x will not hinder Hyperliquid's pace. Mystery is the source of charm. The Hyperliquid ecosystem is too complex. This article only focuses on how Hyperliquid survived the siege of CEX and turned the tables. As for the pedigree of HyperEVM and its DeFi ecosystem, as well as the capital operation stories such as $HYPE's flywheel, $USDH's closed loop, ETFs and DATs, they will be described in subsequent articles.

The Price of Time

Don't hate your enemies, it will make you lose your judgment.

The history of finance is as long as the history of arbitrage. From November 2, 2022, when Coindesk exposed FTX's secrets, to November 11, 2022, when FTX filed for bankruptcy, the attacks were aimed at the security of $FTT as FTX's reserve and the misappropriation of FTX user assets by market maker Alameda. CZ then initially offered to acquire FTX but then refused, disrupting SBF's fundraising and self-rescue efforts. During its fundraising efforts in the Middle East, SBF clearly expressed its resentment towards CZ. Emotions are the enemy of business, and ultimately, SBF failed to save FTX.

Looking at it today, whether it is the prediction market, AI investment, the Solana chain itself, the gradual progress of FTX asset liquidation, or even the collaboration between perpetual contract exchanges and market makers, we will find that SBF’s ideas are correct, but it’s a pity that he is unstable.

Image description: The early development history of Hyperliquid

Image source: @zuoyeweb3

After the collapse of FTX, CZ began to negotiate with the US judicial system for ticket repayment. The price of getting on the bus first and then paying the ticket was US$4.2 billion, but unfortunately, the high price did not restore Binance's solid position. The seeds of Hyperliquid had already been planted. The US defines a challenger as one whose GDP reaches approximately 60% of its own, a similar figure for the Soviet Union and Japan. Binance sets a 10% threshold for challengers. As long as the threshold falls below this safety line, both on-chain dYdX and centralized FTX are negotiable, with Binance's main site and the BNB ecosystem's ApolloX collaborating on defense. Hyperliquid is practically mirroring Binance's development, launching in June 2023 and launching its token in November 2024. Who remembers CZ speaking publicly about Bio Protocol and Giggle Education at this very moment? As for Hyperliquid, it's just another Perp project focused on token issuance.

Image Caption: Hyperliquid OI Market Share

Image Source: @0xhypeflows

It's a pity that Hyperliquid's growth flywheel did not stagnate after the epic 31% airdrop of $HYPE. Instead, trading volume really began to rise. Counterintuitively, GMX also experienced a surge in trading volume, exchanging trading volume for tokens, only to see both the price and trading volume return to zero after the token issuance. It's foreseeable that Aster, Avantis, Lighter, Backpack, edgeX, StandX, Drift, and BULK will all follow this path. This doesn't mean they aren't good projects, nor does it mean tokens shouldn't be considered. We're simply emphasizing that Hyperliquid's ability to survive after issuing its token is inherently anomalous, and this is why we're examining its ability to navigate cycles and become a major player in the crypto industry. The remaining similar projects will all lose their appeal after launching their tokens. Sushiswap can absorb Uniswap's trading volume, GMX can absorb dYdX's trading volume, and Blur can absorb OpenSea's trading volume. Binance's offshore platform is profiting from Coinbase's onshore operations, SBF is using FTX to profit from Binance's regulation, and Hyperliquid is profiting from CEX's operations. The time window was merely a passive window; proactively keeping up with the situation is what led to our current situation. The centralization of BTC/ETH trading mirrors the meme craze. It's precisely because large funds and ETFs dominate the price trends of mainstream cryptocurrencies that the on-chain market initiated by PumpFun has emerged. Desperate retail investors who traded GameStop would face sanctions like disconnections, but memes offer no such restrictions. In mid-2024, Hyperliquid launched its Dutch auction model for listing coins. Beyond innovating the opaque CEX listing model, more importantly, Hyperliquid is gradually transitioning to a dual-track model for both spot and futures trading.

However, apart from $GOD reaching $100,000 at the end of 2024, the remaining bidding prices have been hovering at low levels, and the currencies launched throughout 2024 are basically "non-mainstream" Meme coins, which is incomparable to the current BTC spot trading volume approaching Binance.

Image Caption: Hyperliquid Dutch Auction Data

Image Source: @asxn_r

To be more precise, Hyperliquid's development started from Perp DEX, focusing on conventional processes such as low latency, token incentives, and permissionless, but following the Meme wave, it launched an auction mechanism, and since then it has truly penetrated the spot market and the minds of "retail investors", and then cultivated a complete trading ecosystem.

The same is true for $USDH and the start of HyperEVM. System Thinker is the nature of Hyperliquid founder Jeff. You will reuse this set of methodologies at every stage of Hyperliquid's development.

Image Description: Hyperliquid user growth

Image source: @Hyperliquidx

From the intuitive point of view of user growth, the first 10,000 users were accumulated in the internal test in October 2023, but in the S1 season from November 2023 to May 2024, the number of Hyperliquid users increased to 120,000. Of these, the spot market saw a second small peak in April and May, with auctions in the Netherlands. At the height of the meme craze at the end of 2024, $GOD was quickly auctioned and listed, and $Solv subsequently entered the Hyperliquid spot market, becoming the first mainstream BTCFi project token. Of course, I can't quantify the causal effect of the spot market on Hyperliquid's user and trading volume growth, but the two are highly correlated over time. Hyperliquid didn't simply rely on high leverage and no KYC to achieve its market dominance. We must change the stereotype of Hyperliquid. From the very beginning, Hyperliquid has been an all-around exchange, simply entering the market with its Perp product. The Price of Liquidity To date, Hyperliquid's repurchase amount has reached US$1.4 billion.

In November 2024, after the $HYPE airdrop is completed, the main spot trading will be concentrated within itself, but a fork is needed at this time because Hyperliquid is no longer just a spot and contract exchange. HyperEVM is still in pre-research and will not be launched until early 2025.

To classify it, Hyperliquid is a bit like Ethereum with a consensus layer + two execution layers. HyperBFT is its consensus layer, and the so-called nodes maintain the consensus of HyperBFT, while HyperCore is the contract and spot exchange L1, and HyperEVM is an open L1 with no permission access.

Through the CoreWriter system, HyperEVM can call and allocate HyperCore liquidity. For example, the LST protocol does not need an intermediary to package assets, but can directly reuse the native staking standard. In addition, Unit Protocol can bridge external ecosystem assets to Hyperliquid, while Builder Codes allow any front-end to access HyperCore liquidity and participate in fee sharing. Rabby Wallet and Based App are both examples of this type of front-end access. The above article describes how Hyperliquid relied on passive time windows to proactively build two supermarkets: spot and futures. However, it did not address how to avoid the liquidity split before and after $HYPE's launch. Here is a complete description.

Let's first piece together the appearance of Hyperliquid's growth flywheel. The following is the most rational path:

1. S1 and points system, token airdrop in November 2024, 31% (31M) tokens distributed to S1 and Closed Alpha users, the standard at this time is mainly to examine the user's contract trading volume;

2. However, it should be noted that after the airdrop, the hidden S1.5 and S2.5 have been ongoing. This public and semi-hidden activity gives the Hyperliquid team great flexibility;

3. Hyperliquid claims to have 10 people, no VC, and community interests first. Its fees are in HLP (liquidity treasury, responsible for liquidation) and repurchase of $HYPE The airdrop is roughly evenly distributed, which is also an important source of support for the token price.

That is, it relies on points before the airdrop and on buybacks after the airdrop.

Image Description: Hyperliquid Points System

Image Source: @zuoyeweb3

Hyperliquid and Solana founder Anatoly are very similar. They both attach great importance to the market price of tokens and regard it as an important indicator to continuously stimulate the activity of the ecosystem. This is not common. Vitalik cares more about technology and the value of "doing good". Public chains and large projects with VC participation are always under game theory pressure from selling off from all parties. In reality, it's hard to believe Hyperliquid lacks external funding. Initial market making required proprietary or third-party investment to attract both retail and institutional investors. Just look at how Aster's trading volume instantly surpassed Hyperliquid's. Currently, Hyperliquid nodes include multiple market makers, such as Infinite Field, Alphaticks, CMI, Flowdex, and FalconX. Galaxy has also become a node, and Paradigm has confirmed its purchase of $HYPE in November 2024. The most plausible hypothesis is that market makers participated in Hyperliquid trading early on. However, unlike previous VC investment models of equity and token distribution, market makers gradually purchased tokens from the foundation. This doesn't contradict the company's VC-free designation, and MM is certainly not a VC. For MM, Hyperliquid's strong buyback mechanism also ensures their long-term interests. Hyperliquid also obtains long-term liquidity support, thus breaking the trilemma of declining trading volume after issuance, a subsequent drop in token price, and ultimately no subscriptions for the node. According to @Mint_Ventures's estimates, approximately $50 million flowed into the buyback fund, the Assistance Fund, before and after the airdrop. Adding to the $40 million allocated to HLP, the latter can be considered market-making fees, while the former represents overall marketing expenses. As for VCs and market makers, it's difficult to accurately distinguish them. In addition to market making, HLP and the buyback mechanism also contribute to price support. However, the ultimate use of HLP rests with the Hyperliquid team. In the $JELLYJELLY incident, the team ultimately decided to use HLP reserves to cover $20 million in bad debt. However, in the $XPL hedging incident, they chose to let users bear the losses. However, there is good news: Arthur Hayes sold $5 million, and DragonFly followed closely behind, buying $3 million. Institutions haven't abandoned Hyperliquid; the only debate remains: what is $HYPE actually worth? The price of leverage: Extending the selling curve slows the sell-off. Interest rates are the speed of money flow, and prices are the difference in valuations between the two parties. If we simply compare Hyperliquid's revenue and expenses, revenue comes from spot listing fees (Dutch auctions), spot transaction fees, contract transaction fees, liquidation fees, and commissions from Builder Codes. Expenses include buybacks and burns. However, it's not that simple. Otherwise, the price of $HYPE would become another proxy for trading volume, assuming it's around 10% of $BNB, or $100. Unfortunately, it's currently only $40-50.

Image description: $HYPE repurchase price

Image source: @asxn_r

Continuing from the previous article, $HYPE repurchase and the Ethereum Foundation selling at the current high point are reflected in each other, which neither affects the normal development of the ecology nor relies entirely on repurchase to raise prices and create false prosperity. On the other hand, if EF makes a move, you'd better run away.

There are three factors that disrupt the $HYPE valuation system:

1. High control. HYPE's main spot and contract trading volumes occur in HyperCore, and the foundation controls most of the $HYPE pledge and circulation;

2. Price-to-sales ratio. P/S must be effective in a fully traded market. Coinbase and Circle represent the prices of CEX and stablecoins in the US stock market respectively, but $HYPE can be "manipulated" by high control;

3. Institutional price. The main participation of ETH and DAT, as well as the mutual mixing of the staking system and the HyperEVM ecosystem are still in the growth stage and have not experienced the market consensus of $BNB crossing the cycle.

However, P/S can give us an illusion, and false illusions can also appear objective through numbers. Let's simply calculate the P/S to see how far HYPE is from $1,000:

1. As of now, Hyperliquid's revenue in 2025 is $730 million. Assuming that the annual revenue is $1 billion, it is more reasonable, with a market value of $15 billion and a P/S of about 15

2. Coinbase P/S is 11.8, Binance's annual revenue is about $10 billion, BNB's market value is $136 billion, and its P/S is about 13.6. Robinhood's current value is 30, which is obviously too high. It was also 11.4 before the Cannes press conference in June. A less accurate estimate suggests that the P/S ratio for US stocks on a typical CEX/crypto brokerage is around 11, meaning the token price includes 10x leverage, a discount to the 10x potential expected by many. However, Hyperliquid's price is manipulated in various ways. The only problem is that Arthur Hayes believes $HYPE buybacks won't outstrip selling pressure. He predicts that 237.8 million $HYPE will need to be unlocked on November 29th, creating massive selling pressure and potentially crippling HYPE's future. However, he maintains his long-term forecast of 126x. Facing the end of S2, the Hyperliquid team opted to empower NFTs rather than continue direct token airdrops. This can be seen as a diversion of liquidity to the HyperEVM rather than a direct cause of selling pressure on $HYPE. $BNB embodies Binance's unique position in the trading landscape. It's not enough for Hyperliquid to dominate the Perp DEX space; it needs to defeat Binance to maintain $HYPE's highs. Once the market fluctuates drastically or turns bearish, the current massive buying pressure will reverse into selling pressure. UST is not Bitcoin, and $HYPE may not be $BNB. It will either be the next Binance or the next FTX. Conclusion: Sow the seeds of a gentle breeze, and leave the storm for future generations. Hyperliquid didn't leap ahead of its time; instead, it seized a rare window of opportunity, creating the strongest synergy from market combinations. This was also true of Midjourney in its early days. Hyperliquid truly brought Perp to the retail market, enabling everyday use on DeFi's OG chains, serving as a testing ground for institutions and a hunting ground for whales. It unleashed the meme, seizing the opportunity to grow. After achieving this growth, Hyperliquid chose not to enter a sell-off mode, but instead maintained the $HYPE price within a moderate range. It's important to note that it took Binance eight years, Hyperliquid three, and $HYPE two to catch up.