Evolving into more than a perp DEX. - by 0xtiago

The team’s background comes from HFT market-making firms. Jeff Yan, Hyperliquid’s founder, stated that he and his team founded Hyperliquid with the ambition of bringing a CEX-like experience on-chain. Think of Binance, but decentralized and everything verifiable on-chain.

Enter Hyperliquid.

If you’re new to Hyperliquid, we did a report a year ago that you can check out here. A lot has changed since then so lets jump into it.

The protocol was then a recently launched small derivatives exchange with a slick UX and a promising future. Now, a little more than a year after they went live, the protocol is second only to dYdX in all-time volume for trading futures on-chain, with $270B (Hyperliquid) vs. $1.39T (dYdX). dYdX is 2 years older than Hyperliquid and reigned alone for a long time. If Hyperliquid’s team continues delivering like they are, I don’t see a future where it doesn’t surpass dYdX.

In April they announced some new interesting developments. They moved from their custom Tendermint consensus to a custom-built HyperBFT consensus algorithm, elevating its theoretical TPS from 20k to 200k.

They also introduced a spot market and added two Hyperliquid Improvement Proposals (HIP), which we will discuss later. Combined with its new spot market and exciting developments, these changes are very significant.

Let’s deep dive into the major changes they did to the protocol and what to expect from HL moving forward.

The evolution from a derivatives exchange with a great UX to an EVM compatible, parallelized L1, is a huge and ambitious shift from a self-funded team of 10 people.

On March 28th their official X account announced that the Hyperliquid L1 would begin supporting a spot market, with its HIP-1 and HIP-2 as the necessary steps towards general purpose infrastructure with performant native components. I will discuss both HIPs later in this report.

The $PURR airdrop on April 16th marked the start of HL’s spot market. Half of the PURR supply was airdropped to points holders while the other 50% was permanently locked as “hyperliquidity” on the PURR/USDC spot pair.

The protocol transitioned to its new consensus mechanism called HyperBFT, a custom-built consensus algorithm designed to optimally support Hyperliquid’s new set of products such as the parallelized EVM, spot markets, and future developments.

HIP-1 is Hyperliquid’s L1 native token standard, similar to an ERC-20 on Ethereum. All tokens deployed in the spot markets are HIP-1 tokens. For now, USDC is the only non-native asset supported on the Hyperliquid L1.

HIP-2 is an on-chain strategy called “hyperliquidity” designed to provide order book liquidity for HIP-1 tokens. It serves as an anti-rug measure because every new token must have permanently committed liquidity before it can be traded on Hyperliquid. This requirement ensures that teams launching tokens have skin in the game.

Can anybody deploy a HIP-1? Yes, it is permissionless, but it’s not as easy (and cheap) as creating a token on pump[dot]fun due to some unique features from Hyperliquid's design. Before a new token is released, the deployer must participate in a Dutch Auction with a 31-hour duration to secure their chosen ticker. The Dutch Auction winner will pay the gas fees in USDC as Hyperliquid’s native token is not yet live.

The community has already developed tools to monitor and interact with the auctions and new launches such as Hypurrscan, and Hypurrfun.

Hypurrscan is a block explorer for HL’s spot markets. You can use the website to track Dutch auctions, TWAP orders, addresses, and more.

Hypurrfan is a Telegram bot that offers a comprehensive set of tools including token information, new launches, auction details, and a trading bot similar to Banana Gun or Unibot. This trading bot allows users to trade any coin on the Hyperliquid L1 spot market.

Memecoins have been the hottest topic in this cycle, with pump.fun printing millions of dollars in revenue and helping launch countless coins since its inception. However, many ‘devs’ have been rugging after new coins start trading on the open market.

HIP-2’s ‘anti-rug’ feature from Hyperliquid addresses this issue by requiring every token deployer to commit liquidity permanently as soon as the coin is deployed. The spot market is growing rapidly, with 44 tokens deployed in just 2 months after permissionless spot deploys went live.

At the time of writing this, HL’s spot tokens ecosystem has a total market cap of around $220 million, according to Liquid Start - a community developed platform to track Hyperliquid’s spot ecosystem

PURR still dominates with a market cap of $119 million, accounting for over half of the total spot coin market cap. The market is still dominated by memes, but this is about to change soon. Rage Trade, a derivatives exchange aggregator, announced that they will deploy their token on the Hyperliquid L1 through its founder’s official X account. Rage Trade’s team paid $14,000 for the ticker $RAGE to deploy on Hyperliquid.

Projects must participate in a Dutch auction for the right to buy a ticker, with only 280 auctions a year. This constraint on the number of tokens can bring more attention to existing assets on the platform, which can attract new launches to the chain.

According to Noodles, $RAGE will be airdropped to “Hyperliquid OGs” as well as Rage Trade’s users. They also highlighted that $HFUN holders will be airdropped $RAGE tokens. There was no clarification regarding a snapshot being taken, so it might be worth the risk.

There are two main tools to monitor HL’s spot markets, Liquid Start and Purrburn. I prefer the first for tracking general activity as it has more data, and the latter for checking PURR information.

When trading spot assets on Hyperliquid, there are two lists: a strict list composed of tokens considered safe by the team and an “all” list where every permissionless deployed token is shown. The strict list currently contains only four tokens: PURR, HFUN, POINTS, and JEFF.

In the last report from June 2023, HL had $3.77 million in TVL and had just experienced its best month to date in May with $3 billion in monthly volume. The average trading volume before the introduction of the points system was around $2.4 billion per month.

The points system began in November when the market started to heat up again, and points were not as controversial as they are today. This led to a surge in usage and TVL. During the first season of the points system, the protocol averaged $28.45 billion in monthly trading volume, representing over an 11x increase in the average monthly trading volume. We will dive deeper into the points season in the next section.

Now let’s compare the trading volume with its biggest competitor, dYdX, year-to-date to get a more realistic view of the competitive landscape.

dYdX amassed a staggering $260.77 billion in trading volume year-to-date, while Hyperliquid has achieved $237.2 billion in trading volume, which is 10% less than its main competitor. This gap seemed unattainable when looking at last year’s numbers.

*Both protocols are running incentive programs: tokens for dYdX and points for Hyperliquid so the comparison above is fair.

Regarding TVL, it has also grown significantly, increasing from $3.77 million to over $432 million on its derivatives exchange. This represents a staggering 114x growth over a period of 1 year and 1 month.

Hyperliquid started its first points program on Nov 1, 2023, and it lasted for 6 months, distributing 1 million points per week until May 1, 2024.

The team never clarified how the points were distributed and also mentioned that the criteria for earning points would be updated regularly.

Unlike most projects, users who started using Hyperliquid earlier received a real advantage. Since the total points amount was predefined, early adopters could farm cheaper points as the total amount remained the same, but was shared with an increasing number of participants as the campaign progressed.

The second points season is currently underway. It began on May 29 and will run for 4 months, distributing 700,000 points per week for 9 more weeks.

As highlighted in the Season 1 announcement, the team did not provide details on how points are distributed and regularly changes the criteria to prevent exploitation by farmers.

Trading spot and holding PURR have been the best way to farm points on season 2.

The new points season has become a cat-and-mouse game since the criteria for earning points are not clearly defined and seem to change every week. This encourages a more organic approach and reduces the impact of automated farming and wash trading systems.

Holding and trading spot tokens appear to be the best strategies for Season 2 so far. However, as mentioned earlier, the team can alter the points distribution criteria on a weekly basis. It makes sense for the new use cases to be incentivized, as highlighted in the official announcements.

Although the native token is confirmed, Hyperliquid’s first airdrop occurred on April 16th. The project’s official mascot, PURR, was airdropped to point holders as the first token deployed on its newly launched spot market. Fifty percent of the PURR supply was distributed linearly at a rate of approximately 22 tokens per point, while the remaining 50% was permanently committed as hyperliquidity for PURR.

PURR holders received over 20 different tokens as airdrops from new launches on Hyperliquid’s platform. Here is a list curated by 0xPasteke with all airdrops received by holding PURR.

Being self-funded with a standout product like this is rare nowadays as most of the industry relies on raising large amounts of capital from VCs, which leads to price discovery happening in private markets. By taking a different approach, Hyperliquid’s team holds significant power by owning all of the protocol’s equity (or token supply).

They have endless possibilities for designing a distribution model that can make a truly decentralized L1 out of the gate.

Jeff mentioned on the 0xResearch podcast that they don’t want insiders holding a large portion of the token supply, so we might see a unique token distribution model for their main token airdrop. The industry standard for airdrop distribution is around 10% nowadays, which I consider the minimum for Hyperliquid given the considerations mentioned above.

So I ran some simulations in a range between the industry standard, 10% and a whopping 30% token distribution. Here’s some hopium.

1. Season 1 > 26*1M = 26M points

3. Hiatus between s1 and s2 > 8M points

4. Season 2 > 12.6M points estimated

Total points estimated at the end of season 2 = 51,062,258

Let’s assume Hyperliquid “main competitors” as Sei and dYdX. I added Sei as a competitor because it launched as a trading-focused L1 and has now evolved into a parallelized EVM, following the same path as Hyperliquid.

I calculated a range from a $1B fully diluted valuation, which is dYdX’s current FDV, to $6B for an optimistic scenario. Sei’s valuation sits in the middle at $3.6B fully diluted.

Given the “L1 premium,” the product’s growth, and the hype around Hyperliquid, I would expect it to launch with a fully diluted valuation between $3 billion and $4 billion.

All the numbers are purely speculative and not financial advice. These estimations are based on an airdrop distribution at the end of season 2. If another points season is added, we will need to revise everything.

Despite the derivatives exchange running for over a year, the L1 is newly launched, so it hasn’t stood the test of time yet. This can bring new problems as more users begin to use the protocol. A non-battle-tested platform with huge success right out of the gate can be a victim of its own success, but who wouldn’t want to deal with problems like these?

Oracle issues are a big concern for all order book DEXes, not just Hyperliquid. This risk is highlighted in our first report and in the project’s documentation.

My main concern regarding Hyperliquid is that everything is still centralized, which brings two main problems: a central point of failure and performance.

Regarding the central points of failure, all validator nodes are run by the team, and there is a lack of transparency about the number of validators. The validators are responsible for bridge safety, oracle updates, and block production for the network. They recently published the specs to run a non validating node on their GitHub account.

On the performance side, when the project starts to migrate to a decentralized system, the chain’s performance may decline as nodes in different parts of the globe take more time to reach consensus, as happens with every decentralized network.

So, when the stack becomes fully decentralized, it will be a whole different scenario that could hurt the chain’s performance in the short to medium term. In the worst case scenario this could lead to users abandoning the product, especially after it has been running smoothly for over a year.

Despite being very bullish on Hyperliquid for more than a year already, the team’s lack of transparency regarding some data, such as validator addresses, multi-sig signers, and easy-to-find information on the core team, bothers me.

Discussing whether it will succeed as a derivatives exchange is no longer relevant, as it has been consistently alternating in the number one spot for daily trading volume in the entire space.

Hyperliquid pivoted from a single product to a complete system, moving in the right direction by owning the entire stack to really have a shot at becoming the decentralized Binance. The new consensus algorithm promising 200k theoretical TPS while maintaining low latency is very impressive.

Adding native support for the EVM can be a game-changer as an L1 since it is the most widely adopted virtual machine among smart contract platforms, which should attract many cool applications to the network. Hyperliquid’s EVM won’t be an isolated feature as it will be composable with the existing L1.

The spot market is still in its infancy, but it is the best beta to bet on HL’s mid to long-term success. PURR, their official memecoin, is the only asset doing more than $1 million in daily trading volume frequently. At the time of writing, mainly memecoins are deployed there ($HFUN, $VEGAS, and $FUN being the exceptions), but on August 7th, $RAGE will be deployed on the platform after their token sale..

The team is very lean, around 10 people, and is shipping as fast as one could wish, which is ideal for a startup in such a dynamic environment as crypto. Despite the worries mentioned above, the team seems to be on the right track to deliver on their vision.

If they can maintain high performance after decentralizing the blockchain, which their track record indicates they can deliver, the protocol should become the perfect L1 for all kinds of crypto-native trading and for a whole set of new financial products that could leverage a very liquid, fully on-chain order book.

Evolving into more than a perp DEX. - by 0xtiago

The team’s background comes from HFT market-making firms. Jeff Yan, Hyperliquid’s founder, stated that he and his team founded Hyperliquid with the ambition of bringing a CEX-like experience on-chain. Think of Binance, but decentralized and everything verifiable on-chain.

Enter Hyperliquid.

If you’re new to Hyperliquid, we did a report a year ago that you can check out here. A lot has changed since then so lets jump into it.

The protocol was then a recently launched small derivatives exchange with a slick UX and a promising future. Now, a little more than a year after they went live, the protocol is second only to dYdX in all-time volume for trading futures on-chain, with $270B (Hyperliquid) vs. $1.39T (dYdX). dYdX is 2 years older than Hyperliquid and reigned alone for a long time. If Hyperliquid’s team continues delivering like they are, I don’t see a future where it doesn’t surpass dYdX.

In April they announced some new interesting developments. They moved from their custom Tendermint consensus to a custom-built HyperBFT consensus algorithm, elevating its theoretical TPS from 20k to 200k.

They also introduced a spot market and added two Hyperliquid Improvement Proposals (HIP), which we will discuss later. Combined with its new spot market and exciting developments, these changes are very significant.

Let’s deep dive into the major changes they did to the protocol and what to expect from HL moving forward.

The evolution from a derivatives exchange with a great UX to an EVM compatible, parallelized L1, is a huge and ambitious shift from a self-funded team of 10 people.

On March 28th their official X account announced that the Hyperliquid L1 would begin supporting a spot market, with its HIP-1 and HIP-2 as the necessary steps towards general purpose infrastructure with performant native components. I will discuss both HIPs later in this report.

The $PURR airdrop on April 16th marked the start of HL’s spot market. Half of the PURR supply was airdropped to points holders while the other 50% was permanently locked as “hyperliquidity” on the PURR/USDC spot pair.

The protocol transitioned to its new consensus mechanism called HyperBFT, a custom-built consensus algorithm designed to optimally support Hyperliquid’s new set of products such as the parallelized EVM, spot markets, and future developments.

HIP-1 is Hyperliquid’s L1 native token standard, similar to an ERC-20 on Ethereum. All tokens deployed in the spot markets are HIP-1 tokens. For now, USDC is the only non-native asset supported on the Hyperliquid L1.

HIP-2 is an on-chain strategy called “hyperliquidity” designed to provide order book liquidity for HIP-1 tokens. It serves as an anti-rug measure because every new token must have permanently committed liquidity before it can be traded on Hyperliquid. This requirement ensures that teams launching tokens have skin in the game.

Can anybody deploy a HIP-1? Yes, it is permissionless, but it’s not as easy (and cheap) as creating a token on pump[dot]fun due to some unique features from Hyperliquid's design. Before a new token is released, the deployer must participate in a Dutch Auction with a 31-hour duration to secure their chosen ticker. The Dutch Auction winner will pay the gas fees in USDC as Hyperliquid’s native token is not yet live.

The community has already developed tools to monitor and interact with the auctions and new launches such as Hypurrscan, and Hypurrfun.

Hypurrscan is a block explorer for HL’s spot markets. You can use the website to track Dutch auctions, TWAP orders, addresses, and more.

Hypurrfan is a Telegram bot that offers a comprehensive set of tools including token information, new launches, auction details, and a trading bot similar to Banana Gun or Unibot. This trading bot allows users to trade any coin on the Hyperliquid L1 spot market.

Memecoins have been the hottest topic in this cycle, with pump.fun printing millions of dollars in revenue and helping launch countless coins since its inception. However, many ‘devs’ have been rugging after new coins start trading on the open market.

HIP-2’s ‘anti-rug’ feature from Hyperliquid addresses this issue by requiring every token deployer to commit liquidity permanently as soon as the coin is deployed. The spot market is growing rapidly, with 44 tokens deployed in just 2 months after permissionless spot deploys went live.

At the time of writing this, HL’s spot tokens ecosystem has a total market cap of around $220 million, according to Liquid Start - a community developed platform to track Hyperliquid’s spot ecosystem

PURR still dominates with a market cap of $119 million, accounting for over half of the total spot coin market cap. The market is still dominated by memes, but this is about to change soon. Rage Trade, a derivatives exchange aggregator, announced that they will deploy their token on the Hyperliquid L1 through its founder’s official X account. Rage Trade’s team paid $14,000 for the ticker $RAGE to deploy on Hyperliquid.

Projects must participate in a Dutch auction for the right to buy a ticker, with only 280 auctions a year. This constraint on the number of tokens can bring more attention to existing assets on the platform, which can attract new launches to the chain.

According to Noodles, $RAGE will be airdropped to “Hyperliquid OGs” as well as Rage Trade’s users. They also highlighted that $HFUN holders will be airdropped $RAGE tokens. There was no clarification regarding a snapshot being taken, so it might be worth the risk.

There are two main tools to monitor HL’s spot markets, Liquid Start and Purrburn. I prefer the first for tracking general activity as it has more data, and the latter for checking PURR information.

When trading spot assets on Hyperliquid, there are two lists: a strict list composed of tokens considered safe by the team and an “all” list where every permissionless deployed token is shown. The strict list currently contains only four tokens: PURR, HFUN, POINTS, and JEFF.

In the last report from June 2023, HL had $3.77 million in TVL and had just experienced its best month to date in May with $3 billion in monthly volume. The average trading volume before the introduction of the points system was around $2.4 billion per month.

The points system began in November when the market started to heat up again, and points were not as controversial as they are today. This led to a surge in usage and TVL. During the first season of the points system, the protocol averaged $28.45 billion in monthly trading volume, representing over an 11x increase in the average monthly trading volume. We will dive deeper into the points season in the next section.

Now let’s compare the trading volume with its biggest competitor, dYdX, year-to-date to get a more realistic view of the competitive landscape.

dYdX amassed a staggering $260.77 billion in trading volume year-to-date, while Hyperliquid has achieved $237.2 billion in trading volume, which is 10% less than its main competitor. This gap seemed unattainable when looking at last year’s numbers.

*Both protocols are running incentive programs: tokens for dYdX and points for Hyperliquid so the comparison above is fair.

Regarding TVL, it has also grown significantly, increasing from $3.77 million to over $432 million on its derivatives exchange. This represents a staggering 114x growth over a period of 1 year and 1 month.

Hyperliquid started its first points program on Nov 1, 2023, and it lasted for 6 months, distributing 1 million points per week until May 1, 2024.

The team never clarified how the points were distributed and also mentioned that the criteria for earning points would be updated regularly.

Unlike most projects, users who started using Hyperliquid earlier received a real advantage. Since the total points amount was predefined, early adopters could farm cheaper points as the total amount remained the same, but was shared with an increasing number of participants as the campaign progressed.

The second points season is currently underway. It began on May 29 and will run for 4 months, distributing 700,000 points per week for 9 more weeks.

As highlighted in the Season 1 announcement, the team did not provide details on how points are distributed and regularly changes the criteria to prevent exploitation by farmers.

Trading spot and holding PURR have been the best way to farm points on season 2.

The new points season has become a cat-and-mouse game since the criteria for earning points are not clearly defined and seem to change every week. This encourages a more organic approach and reduces the impact of automated farming and wash trading systems.

Holding and trading spot tokens appear to be the best strategies for Season 2 so far. However, as mentioned earlier, the team can alter the points distribution criteria on a weekly basis. It makes sense for the new use cases to be incentivized, as highlighted in the official announcements.

Although the native token is confirmed, Hyperliquid’s first airdrop occurred on April 16th. The project’s official mascot, PURR, was airdropped to point holders as the first token deployed on its newly launched spot market. Fifty percent of the PURR supply was distributed linearly at a rate of approximately 22 tokens per point, while the remaining 50% was permanently committed as hyperliquidity for PURR.

PURR holders received over 20 different tokens as airdrops from new launches on Hyperliquid’s platform. Here is a list curated by 0xPasteke with all airdrops received by holding PURR.

Being self-funded with a standout product like this is rare nowadays as most of the industry relies on raising large amounts of capital from VCs, which leads to price discovery happening in private markets. By taking a different approach, Hyperliquid’s team holds significant power by owning all of the protocol’s equity (or token supply).

They have endless possibilities for designing a distribution model that can make a truly decentralized L1 out of the gate.

Jeff mentioned on the 0xResearch podcast that they don’t want insiders holding a large portion of the token supply, so we might see a unique token distribution model for their main token airdrop. The industry standard for airdrop distribution is around 10% nowadays, which I consider the minimum for Hyperliquid given the considerations mentioned above.

So I ran some simulations in a range between the industry standard, 10% and a whopping 30% token distribution. Here’s some hopium.

1. Season 1 > 26*1M = 26M points

3. Hiatus between s1 and s2 > 8M points

4. Season 2 > 12.6M points estimated

Total points estimated at the end of season 2 = 51,062,258

Let’s assume Hyperliquid “main competitors” as Sei and dYdX. I added Sei as a competitor because it launched as a trading-focused L1 and has now evolved into a parallelized EVM, following the same path as Hyperliquid.

I calculated a range from a $1B fully diluted valuation, which is dYdX’s current FDV, to $6B for an optimistic scenario. Sei’s valuation sits in the middle at $3.6B fully diluted.

Given the “L1 premium,” the product’s growth, and the hype around Hyperliquid, I would expect it to launch with a fully diluted valuation between $3 billion and $4 billion.

All the numbers are purely speculative and not financial advice. These estimations are based on an airdrop distribution at the end of season 2. If another points season is added, we will need to revise everything.

Despite the derivatives exchange running for over a year, the L1 is newly launched, so it hasn’t stood the test of time yet. This can bring new problems as more users begin to use the protocol. A non-battle-tested platform with huge success right out of the gate can be a victim of its own success, but who wouldn’t want to deal with problems like these?

Oracle issues are a big concern for all order book DEXes, not just Hyperliquid. This risk is highlighted in our first report and in the project’s documentation.

My main concern regarding Hyperliquid is that everything is still centralized, which brings two main problems: a central point of failure and performance.

Regarding the central points of failure, all validator nodes are run by the team, and there is a lack of transparency about the number of validators. The validators are responsible for bridge safety, oracle updates, and block production for the network. They recently published the specs to run a non validating node on their GitHub account.

On the performance side, when the project starts to migrate to a decentralized system, the chain’s performance may decline as nodes in different parts of the globe take more time to reach consensus, as happens with every decentralized network.

So, when the stack becomes fully decentralized, it will be a whole different scenario that could hurt the chain’s performance in the short to medium term. In the worst case scenario this could lead to users abandoning the product, especially after it has been running smoothly for over a year.

Despite being very bullish on Hyperliquid for more than a year already, the team’s lack of transparency regarding some data, such as validator addresses, multi-sig signers, and easy-to-find information on the core team, bothers me.

Discussing whether it will succeed as a derivatives exchange is no longer relevant, as it has been consistently alternating in the number one spot for daily trading volume in the entire space.

Hyperliquid pivoted from a single product to a complete system, moving in the right direction by owning the entire stack to really have a shot at becoming the decentralized Binance. The new consensus algorithm promising 200k theoretical TPS while maintaining low latency is very impressive.

Adding native support for the EVM can be a game-changer as an L1 since it is the most widely adopted virtual machine among smart contract platforms, which should attract many cool applications to the network. Hyperliquid’s EVM won’t be an isolated feature as it will be composable with the existing L1.

The spot market is still in its infancy, but it is the best beta to bet on HL’s mid to long-term success. PURR, their official memecoin, is the only asset doing more than $1 million in daily trading volume frequently. At the time of writing, mainly memecoins are deployed there ($HFUN, $VEGAS, and $FUN being the exceptions), but on August 7th, $RAGE will be deployed on the platform after their token sale..

The team is very lean, around 10 people, and is shipping as fast as one could wish, which is ideal for a startup in such a dynamic environment as crypto. Despite the worries mentioned above, the team seems to be on the right track to deliver on their vision.

If they can maintain high performance after decentralizing the blockchain, which their track record indicates they can deliver, the protocol should become the perfect L1 for all kinds of crypto-native trading and for a whole set of new financial products that could leverage a very liquid, fully on-chain order book.