Partnered With:

Inventory of 60 perpetual contract protocols: Order book still dominates, while reliance on oracle intensifies.

Read this article in 44 Minutes

Perpetual contract market has achieved substantial product market adaptability, which may lead to market saturation to some extent.

Original Title: "60 perpetual contract agreements horizontal evaluation: order book still dominates, oracle dependence intensifies, market saturation begins"

Original Author: Pablo Bartol

Original Translation: Deep Tide TechFlow

Introduction

In the derivatives field, perpetual contracts are more popular than options, which have gained widespread recognition in both retail and institutional markets. The total trading volume of BTC perpetual contracts has increased from about $3 billion per day in early 2020 to $13 billion in 2023.

This series of articles aims to provide an overview of the current DeFi perpetual contract field, and compare different protocols, with a focus on their working principles, rather than their potential as investment tools.

The following 60+ protocols are involved in this analysis:

0x, Alex, Angle, Apollo, Avantis, BlueFin, bZx, Cap, Contango, ConvergenceRFQ, DDEX, DDX, Deri, Derivio, Digitex, Dolomite, Drift, dYdX, Futureswap, Gains Network, Gearbox, GMX, GMX Forks, Good Entry, Hubble, Hyperliquid, IDEX, Increment, InfinityPool, Injective, Kujira, Kwenta, Lendroid, Levana, Level, Lexer, Mango, MarketProtocol, MCDEX, Mux, NFTperp, Numoen, OpenBook, Opyn (Squeeth), Perennial, Perpetual Protocol, Perpy, Pika, Polynomial, Predy, RabbitX, RageTrade, Sentiment, Serum, STFX, Syndr, Synfutures Protocol, Synthetix, Tigris Trade, Tribe3, Unidex, Variabl, Vega, Vela, Vertex, Vyper, and Zeta Markets.

In this article, we will first briefly review perpetual futures, and then extensively classify the above agreements. Based on this classification, we will delve into the differences between these agreements and explore the trade-offs they make in providing liquidity and contract pricing methods.

Concept Review of Perpetual Contracts

In this section, you can find a brief introduction about perpetual contracts.

Futures contracts are a type of legal agreement that specifies the purchase or sale of a specific commodity, asset, or security at a predetermined price at a specific time in the future. Unlike options, futures contracts do not provide the option to choose and the agreement is binding.

Perpetual contract is a type of futures contract that has no expiration date, hence the name. The concept of perpetual contract was first proposed by Robert Shiller in 1992, and BitMEX was the first to implement perpetual contracts in 2016.

Perpetual contract is a type of Delta-1 product, which means that the contract price will change by 1 US dollar for every 1 US dollar change in the underlying asset. For example, assuming the trading price of Ethereum is 2000 US dollars. You open a perpetual long position with a nominal value of 1 Ethereum. After a period of time, if Ethereum reaches 2200 US dollars, the profit will be 1 Ethereum * (2200 US dollars - 2000 US dollars) = 200 US dollars, minus transaction fees and funding rates.

This allows traders to speculate on asset price fluctuations without worrying about time constraints. Traders can keep their long or short positions open for as long as they choose. It is important for traders to note that they will incur fees and, when using leverage, must closely monitor the health of their positions to avoid the risk of forced liquidation.

In most perpetual contract agreements, liquidation is a crucial component as the accumulation of bad debts may lead to the bankruptcy of the agreement and/or liquidity providers (LPs). Funding rates or borrowing rates (specific terms depend on the agreement) are implemented to promote market efficiency, provide arbitrage opportunities, and compensate participants who take on less desirable positions to balance open interest (OI).

Classification of perpetual contract agreements

You can classify extensively based on two main variables.

Does the protocol need to lock liquidity in it? This can be seen as: is there a liquidity pool, or is it completely peer-to-peer (P2P)?

Does the protocol directly affect the market, leading to or affecting the price discovery of the underlying asset or contract? Or does the protocol use an oracle to obtain a fair price for the asset?

Based on this, we can propose this basic (perhaps too simple) classification:

The classification matrix of perpetual agreements is based on two key criteria: 1) external price discovery and internal (oracle s and oracle less) and 2) liquidity locking and liquidity pool (P2P and P2Pool).

Order Book

In this section, we can find the following protocols:

Alex、Apollo、Blue Fin (formerly Firefly Exchange), DDX, Digitex, dYdX, Hubble, Hyperliquid, IDEX, Injective, Kujira, Lendroid, Mango, Market Protocol, RabbitX, Variabl, Vega, and Zeta.

AMM

Leveraged spot trading: bZx, DDEX, Dolomite, Futureswap, MCDEX.

Account

保证金账户：Margin Account

交割合约：Delivery Contract

永续合约：Perpetual Contract

ZKS：ZKS

STARK：STARK

SCROLL：SCROLL

永续期权：Predy v3。

Perpetual options: Predy v3.

VAMM

Drift, Increment, InfinityPools, NFTperp, Perpetual Protocol, Rage Trade, and Tribe3.

Although both Predy and InfinityPools use centralized liquidity to build "perpetual options" or perpetual contracts using AMM, please note that they are implemented in different ways and therefore belong to different categories. In our previous options series, we introduced some protocols built on the same concept, but focused more on providing options-like experiences, such as Gamma Swap, Panoptic, or Smilee.

Based on Oracle

In this section, we can find the following protocols:

Synthetic liquidity.

Avantis、Gains Network、Deri、Synthetix (Kwenta, Polynomial, etc.), Tigris Trade, and Vela.

A basket of assets.

Angle, Apollo V2, Cap, Deri, Derivio, GMX and its forks, Good Entry, Levana, Level, Lexer, Mux, Pika, and Synfutures Protocol.

Other

These protocols may not necessarily fit into the aforementioned categories, or may be better suited to be included in a separate category. Therefore, they are listed here for completeness, but the rest of this article will focus on the aforementioned categories.

Aggregator

In this section, we can find the following protocols:

Social Trading <> Copy Trading

In this section, we can find the following protocols:

Perpy and SFTX.

They allow users to perform copy trading on other platforms. For example, a trader trades on GMX and other users copy the same trade.

Infrastructure

Protocol that allows other protocols to be built on top of it.

0x, Serum, and OpenBook allow for the launch of an order book platform through their protocols.

Vyper, Derivio, Perennial, and Convergence RFQ have some more specific use cases for their solutions. Generally speaking, they can be seen as foundational layers upon which other protocols can build their own markets.

Brief introduction to RFQ. RFQ stands for Request for Quote, which is typically a request made by a trader to a market maker for a quote on a specific instrument/asset/quantity, etc.

Expired Assets

Contango provides expiring assets. This is neither permanent - as they obviously expire - nor a forward contract. Contango does not have an order book or liquidity pool. When a position is opened, Contango borrows in the fixed rate market, swaps in the spot market, and then lends in the fixed rate market. The trades are based on assets available in the fixed rate currency market, such as yield or nominal. At maturity, the contract can be settled through cash settlement or physical delivery.

Perpetual Agreement Market Analysis

Market Overview:

Please note that our research suffers from survival bias. Nevertheless, we focus on including the aforementioned protocols, despite their varying degrees of success.

Before we delve into exploring different types of perpetual agreements, there are two important considerations related to on-chain derivatives, namely regulatory uncertainty and the historical background prior to DeFi.

On-chain derivatives face regulatory uncertainty. They are facing regulatory challenges that affect the market. Here are two important cases:

Digitex, December 2017. The platform faced regulatory scrutiny when it was sued by the Commodity Futures Trading Commission (CFTC) for operating an illegal commodity futures trading platform.

bZx or bZeroX (Fulcrum Trade), June 2019. The history of bZx is more complex than Digitex. It has experienced four major losses (link). The protocol is managed by Ooki DAO. The CFTC accused the founder and Ooki DAO of illegally providing off-exchange digital asset trading.

Before the arrival of DeFi Summer in 2020, not much happened on-chain, and some projects that had been working on before did not see the light of day.

Market Protocol. The final update was made on Twitter in December 2019.

The goal of the order book is to fully realize on-chain peer-to-peer transactions. However, the main issue is that most popular blockchains are not suitable for a fully on-chain order book.

Therefore, different solutions have been adopted.

1) Place some of the system's content off-chain. This mainly refers to the matching engine, as it is a resource-intensive (high computational cost) component in the order book.

2) Fully on-chain. Transferring to the Alt layer, such as Solana, has lower security for the underlying blockchain, but higher throughput. Updates to the order book can occur every block or every half second.

3) Establish your own L1. There are two ways to achieve this: one is to build an application chain on the OP stack or Cosmos, and the other is to build an independent L1 from scratch.

Each method has its advantages and disadvantages. For example, placing part of the application off-chain introduces a trust assumption. Trading off lower security for faster throughput and less decentralization on the underlying chain. Application-specific chains can provide protocols to capture MEV, but at the cost of requiring more permission settings and increasing friction layers.

Automated Market Maker (AMM)

Unlike order books, orders are settled in P2P in which AMMs need to lock liquidity in liquidity pools. The key difference between AMMs and vAMMs is that AMMs require real liquidity, while vAMMs' liquidity comes directly from the insurance pool outside of vAMM, so LP's presence is not necessarily required.

Spot AMM

In this case, although AMMs are not perpetual futures, they are included because they allow traders to leverage the underlying assets with a Delta=1 exposure by borrowing funds.

However, this type of protocol also has some advantages.

AMMs like Uniswap provide flexibility to trade various assets as long as there is a liquid market available. This allows traders to access assets that may not be eligible on other perpetual futures protocols with leverage. Additionally, using liquidity already included in spot AMMs ensures continuous liquidity to match buy and sell orders, thereby aiding in price discovery of assets.

However, there are also some trade-offs. In order to provide traders with leverage, it is necessary to incentivize borrowers with attractive interest rates and/or rewards for deposits. Additionally, while we mentioned that all assets can potentially be traded through the use of spot markets, there is also a need for some degree of regulation to prevent users from losing borrower funds through honey pots, hacks, and other means.

Due to the fact that LP is not a direct counterparty to traders, they avoid bankruptcy risk. However, this comes at the cost of becoming borrowers and assuming credit risk as traders.

About Power perpetual futures, they were introduced by Paradigm in 2021. Opyn (SQUEETH) was the first team to implement this protocol. In short, they have returns of n^x, such as ETH².

VAMM

vAMMs provides a decoupled market structure that facilitates independent price discovery from underlying spot prices. This may result in price discrepancies between futures prices and asset spot prices, creating arbitrage opportunities. On the other hand, what is the true price of perpetual contracts?

vAMMs faces another challenge in terms of liquidity. The lack of deep liquidity may cause unexpected or unwanted price fluctuations when opening or closing positions.

The vAMM model relies on the clearing position and balanced holding to make the marked price close to the spot price. However, this can create inherent imbalances and deviations, especially over time and during market crashes or extreme market volatility.

Oracle-based Protocol

Broadly speaking, oracle-based protocols can be defined as protocols with liquidity pools that act as counterparties for transactions. Unlike AMMs, the protocol itself does not have price discovery, but instead uses oracles to price underlying assets from other venues.

We can find two prototypes of perpetual futures based on oracles.

合成流动性。Single assets are used as the basis for all trading markets. Through the use of oracles, LPs can become counterparties to all of these markets.

A basket of assets. Use a basket of assets as a counterparty in trading. That is to say, in the assumed BTC/USDC, BTC serves as collateral for long positions and USDC serves as collateral for short positions. More assets can be added, but this is the general basic principle.

Composite assets. Their advantage is that, since a single asset is a counterparty, there is no actual restriction on which markets can be added. As long as there are reliable enough oracles. This makes the protocol more flexible in adding markets. However, this is more risky for liquidity providers, as they bear the risk of certain assets that cannot be hedged on-chain. For example, as a counterparty to gold or USD/JPY. The good side is that if the asset basket maintains a 50% USD balance in the pool, there will be no unexpected temporary losses, similar to the temporary loss situation of Uniswap V2.

The protocol that uses a basket of assets provides more strict settings for liquidity providers, which is beneficial for risk management (volatile assets support long positions, stable assets support short positions, so even if the price fluctuates sharply, the system still has collateral). This is at the expense of traders who do not have as many markets to trade. For example, in GMX v1, you can only trade five assets. However, a very reasonable rebuttal or question is whether most traders really want to access markets beyond Bitcoin and Ethereum. Trading volume data broken down by asset can be found here. A quick glance at the data shows that for Gains Network, 50-70% of the trading volume is generated on Bitcoin and Ethereum, while for Kwenta, it is about 40% to 80%.

Protocol Segmentation

In the previous section of the market overview, we turned to explore the different parts that the protocol can be decomposed into. This article will focus on liquidity provision and pricing.

Liquidity Provision

From the perspective of protocols, we can broadly divide them into two groups.

P2P trading. Therefore, there is no need to lock liquidity on the platform (TVL=0).

Liquidity pool. Trading through AMM, vAMM or oracle.

This section covers the following topics:

Order Book

In short, market makers and market takers trade with each other. The agreement may incentivize liquidity for the market makers/takers.

For protocols, the complex task is how to match market makers and market takers' orders in a timely manner. For example, according to a report from Deutsche Brse Group in May 2023, participants need to react in less than 2770 nanoseconds. Can this be handled by a general-purpose chain like Ethereum? The answer is no.

This forces the protocol to either move off-chain or to dApps, forcing validators to also run order books. It can be seen that liquidity providers are active agents interacting with other participants. Apart from market makers who may not provide liquidity for meme coins such as HarryPotterObamaSonic10Inu, there are no real barriers preventing order books from listing these currencies.

Spot AMMs

Here, the liquidity used for settlement transactions comes from AMMs such as Uniswap. However, the protocol itself needs to incentivize borrowers to come forward as they are the ones providing leverage for traders. Sentiment and Gearbox, while allowing users to do more, also fall into this category. Liquidity providers play a relatively passive role, simply providing leverage.

Power Perpetuals

Power Perpetuals, such as SQUEETH, consist of two participants. One party is bullish on ETH², while the other is bearish on ETH². Due to each protocol following a different design, it is impossible to summarize them. For example:

In Opyn, bears use ETH as collateral to mint Power Perpetual. Bulls can only buy SQUEETH on open markets such as Uniswap AMM. Therefore, in order to profit, they also need to be able to sell.

In Numoen, multi-party minting tokens, while liquidity providers provide LP tokens as collateral, which they lend to traders.

Deri uses a single liquidity pool as the counterparty for all derivatives (including Power Perpetuals). Therefore, although Deri is included here, its liquidity model is essentially the same as that of oracle-based protocols.

vAMMs

When it comes to providing liquidity, vAMMs have adopted various different methods.

Perpetual Protocol and its protocols that follow its model (such as NftPerp). In theory, liquidity providers are not actually needed here because vAMM does not require counterparties when minting virtual tokens. What happens if there are no counterparties? That is why in Perp v1, the team was one of the main liquidity providers for a period of time because they acted as market makers to facilitate trades. Later, liquidity providers were introduced, although the goal is still to match longs and shorts, the introduction of liquidity providers ensures continuous liquidity. Therefore, in this model, liquidity providers can be passive (provide liquidity, traders solve it themselves) or more active (arbitrage prices and charge funding fees).

Increment adopts a model similar to Perpetual Protocol, but instead of using the Uniswap v3 model, it concentrates liquidity by utilizing Curve's V2 mathematics. Each liquidity pool in the protocol may use different parameterizations based on the volatility of the assets in the pool.

Rage Trade has introduced the concept of "full-chain circular liquidity". What is full-chain circular liquidity? Full-chain circular liquidity is a treasury that divides liquidity into 80-20 portions, where 80% of the liquidity is redeployed to provide liquidity on Curve, while the remaining 20% of the liquidity provides centralized liquidity on Rage Trade (supported by Uni v3). The virtual liquidity of vAMM is supported by the 80-20 treasury.

InfinityPools achieves its liquidity by building on the centralized liquidity concept of Uniswap V3. By rebalancing liquidity positions similar to Uniswap, it can achieve very high leverage without the risk of liquidation. For example, a liquidity provider deposits assets worth $1000 in the ETH/USDC pool, with a price range of 900-1000. Then, a trader can borrow the liquidity provider's position and redeem all liquidity with 1 ETH (1ETH = $1000).

Drift V1. Drift uses a VAMM called Dynamic VAMM (DAMM), where liquidity comes from liquidity providers on Drift. The AMM implements a set of parameters to fine-tune market depth, such as price multiplier, fee pool, or segmentation. Drift V1 AMM is now part of Drift V2.

Hybrid Mode: AMM + Order Book

The problem that the order book may face is: what if there is no market maker to post buy and sell orders? Therefore, some protocols have introduced a hybrid model between the order book and AMM to ensure continuous liquidity.

Drift v2 uses three different liquidity mechanisms. First, an immediate Dutch auction provided by market makers. This process lasts approximately 5 seconds. Second, there is a limit order book that only executes limit orders and is run by an administrator bot network. Third, if there is no market maker participation in the first two steps, a constant product AMM - Drift V1 AMM - will be used to ensure continuous liquidity for traders.

Vertex uses a price/time priority algorithm, so orders will be executed based on the best price, whether provided by an AMM or a market maker. AMMs quote discrete price levels on the order book, approximately xy=k.

Syndr combines an off-chain order book with a liquidity pool deployed on Arbitrum. The AMM will integrate liquidity provided by retail investors, while the order book is intended to provide quotes from market makers.

Based on oracle, synthetic assets

Liquidity provision is relatively simple. An insurance pool that serves as a counterparty for all trades. While the main liquidity providers are passive liquidity providers, we can also identify active participants who seek to obtain funding rates. By doing so, they balance the open interests of the platform, thereby encouraging more trades to occur.

The liquidity of the insurance pool can be provided in the following ways:

Stablecoins priced in US dollars, such as Avantis, Gains Network, Deri on zkSync, and Tigris Trade, are supported by funds in DAI, USDC, or USDT for all transactions, depending on the protocol. Another difference in the protocol is whether they have implemented a layered mechanism like Avantis.

Volatility assets. For example, Kwenta and other protocols are built on top of Synthetix, where SNX stakers are counterparties to trades.

Based on oracle, a basket of assets

The main idea is the same as above, but the foundation is not a single asset, but a basket or multiple baskets of assets.

Similar to the previous category, protocols can also have a hierarchical mechanism - Level - or none - GMX, Pika, and MUX.

Although Angle is here, the feature has been disabled due to the Euler hack. In brief, Angle is the issuer of agEUR (Euro stablecoin). The collateral assets used to mint agEUR are used as counterparties for traders. Therefore, someone can provide 1 ETH to receive 1000 agEUR, allowing them to go long on ETH and short on Euro, thus keeping the protocol neutral. This also allows Angle to have perpetual contracts in the forex market, such as the agEUR/USDT trading pair.

Here it is worth mentioning Lexer Markets, who have utilized asset baskets as well as synthetic liquidity.

Price Discovery

We can determine whether the price discovery is taking place within or outside the protocol, and then roughly categorize the protocol through the broadcast of the oracle.

This section covers the following topics:

Price Discovery Methods - Price Slippage - Settlement.

Order Book

Due to the order book facilitating P2P transactions between buyers and sellers, it creates a market for price discovery and transaction execution. In this case, the protocol allows traders to execute transactions at any price without the need for an oracle.

The reason for slippage is that the order may be too large to be executed at a specific price, or there may not be enough liquidity at that price level, so the order pushes the price towards an unfavorable execution price.

The settlement process of order book trading involves two participants, namely market makers and order takers. Market makers place limit orders on the order book, while order takers are traders who accept and execute existing orders on the order book. When an order taker's market order matches a market maker's limit order on the order book, a trade occurs.

The existence of funding rates is to balance long and short positions, while keeping the price of derivative contracts close to the price of underlying assets.

Spot AMM

Price discovery occurs outside of the platform. As these protocols are integrated into spot AMMs or DEX aggregators, trading takes place on platforms such as Uniswap and Curve. Therefore, price discovery is limited by the number of integrated trading venues and the liquidity within these venues.

For spot AMM, oracle is very important, mainly to ensure the best price execution for buyers and sellers. Oracles are also used to track the health of their positions; detailed information on this will be described in the risk section.

Settlement is conducted for the liquidity that exists (or is conducted) on the AMM executing the transaction. However, it is important to note that two transactions actually need to be conducted: purchasing/selling assets on margin when opening a position, and selling/purchasing assets on the market when closing a position.

vAMM

Generally speaking, we can classify all vAMMs into the same category, excluding InfinityPools. vAMMs will face slippage issues in their virtual liquidity reserves. Transactions are settled against assets used to mint virtual liquidity.

Infinity Pools is built on the concept of Uniswap V3, and allows for very high leverage without the risk of liquidation by rebalancing its liquidity positions in its vAMM.

A liquidity provider deposits assets worth $1000 in the ETH/USDC pool, with a price range between 900-1000.

The trader borrowed and redeemed all liquidity, receiving 1 ETH (1 ETH = $1000) and using $100 as collateral. This will result in one of the following three scenarios based on the price of ETH.

If 1 ETH > $1000, the trader will profit. The profit is the difference between the ETH spot price and $1000.

If 1 ETH < $900, the trader's entire capital will be converted to ETH. For liquidity providers, there is no "risk" as their entire liquidity provision will be in ETH if on Uniswap. In this particular example, the trader will have 1.11 ETH as collateral on their 1 ETH position.

Drift V2 has a very clear pricing method. Trades are first priced through auction, then through order book pricing, and finally through AMM pricing.

Vertex will provide quotes to users using an order book or AMM. Currently, market makers mostly provide tighter quotes, but if the scale of AMM significantly expands, AMM may also provide tighter market quotes.

On Syndr, the AMM portion works similarly to the GMX/GLP pool, but includes buy/sell spreads.

Based on oracle, comprehensive assets and a basket of assets.

These protocols use oracles (hence the name) to determine the price of contract transactions. In turn, oracles determine the price of assets through various methods, such as using only the TWAP of Uniswap (which is not very effective) or aggregating prices on the most liquid CEXs (primarily Binance). To further illustrate this point:

Gains Network prices are relayed through a set of semi-permissioned robots that require NFT ownership to operate. These robots use Chainlink asset prices when executing limit, stop-limit, take-profit, stop-loss, and liquidation orders.

GMX. GMX v1 uses custom price feeds and a combination of CEXs like Binance and Chainlink. If the fast price is earlier than 5 minutes, only the Chainlink price is considered. For GMX v2, a new low-latency Chainlink oracle will be implemented. The oracle maintainers are operated by the GMX team.

Regarding price slippage, since trading is conducted on pools based on oracles, there is actually no price slippage. This may lead to some special cases where traders can long or short low liquidity assets without any price impact. However, some protocols have introduced solutions to this issue, such as introducing synthetic slippage fees or requiring minimum price deviation to profit from closing positions.

Settlement is carried out for the assets provided by liquidity providers to the pool. Ideally, long and short positions should be balanced so that the pool has no real Delta exposure and liquidity providers only profit from fees. However, this is not always the case as liquidity in the pool is the counterparty of traders. In situations where the market is highly volatile, this approach works well because traders lose money in the long run, but in a one-way market (such as the Terra Luna crash), the situation becomes unstable.

The funding cost exists, but not for the purpose of balancing prices, but for balancing positions. This is a typical risk mitigation method implemented in the agreement.

Power Perpetuals

Again, it should be emphasized that due to the different implementation methods of each Power Perpetual, it is difficult to summarize their designs. In addition, some Power Perpetuals can be classified into other categories, such as Deri; or have not yet disclosed their models, such as Polynomial.

Index Price: The index price of Squeeth is ETH².

标记价格：Mark price is the current trading price of Squeeth.

Power Perpetuals rely on arbitrageurs to keep the index and mark prices close. But it's not just about the price.

Volatility: The market may overestimate/underestimate volatility. The funding rate may be too high/low.

Price: The difference between the index price and the marked price.

When buying on AMM, slippage may occur, which can be problematic because liquidity is usually not very high.

Similar to AMM, in order to make a profit, traders need to buy and sell simultaneously in the pool. This, combined with the above viewpoint, may sometimes result in traders not being able to obtain the best user experience.

Conclusion

In this article, we provide a comprehensive overview of the perpetual contract market, including liquidity provision and pricing of these contracts. In this report, we studied over 60 different protocols.

Just by looking at the number of perpetual contract protocols deployed in the past 4 years, it is clear that the perpetual contract market has achieved substantial product market fit, which may lead to market saturation to some extent. A simple example can illustrate this point: the amount of Bitcoin held increased from about 300,000 BTC (worth about $3 billion) in 2020 to about 450,000 BTC (worth about $13 billion) in 2023.

The order book is likely the most suitable type for trading, which is why dYdX is currently the largest trading volume perpetual decentralized exchange (perp-DEXs). The reason why most market makers prefer to use order books is that they give market makers (liquidity providers) flexibility and finer control over the prices at which traders buy and sell. However, the technological limitations of today's blockchain protocols have led to the use of other methods for liquidity guidance.

AMM and vAMM were very popular in the early stages of the DeFi cycle. Most of them eventually adopted the route of Uniswap v3 (centralized liquidity). AMM ensures that participants continue to receive liquidity as long as liquidity exists, but may absorb toxic order flow at a cost. InfinityPools and Predy use a new AMM method, where Uniswap LP is used as a source of liquidity.

Between the order book and AMM, there are two protocols - Drift and Vertex - that use a hybrid approach to provide liquidity and pricing. By combining these two models, they aim to balance the advantages of both, allowing users to have fine control over their trades while using AMM as a backup to ensure continuous liquidity.

On the other hand, oracle-based protocols rely on third-party infrastructure - oracles - to operate. Users can trade without worrying about the liquidity depth of a specific trading venue or the convenience of providing liquidity (just by adding liquidity to the pool), which has contributed to the success of these protocols.

In the future, we plan to conduct more in-depth research on the risk management strategies adopted by perpetual protocols, their cost structures, and their integration into the wider DeFi Lego ecosystem. By studying these aspects, our goal is to fully understand the evolving landscape of perpetual futures exchanges.