It’s been around two and a half years since we saw on-chain perps really take off after the launch of GMX, and since then, the sector has evolved massively.

Thus far, the sector has done roughly $380.4B in volume across all chains.

While this number is still small when compared to CEXs, it has steadily been increasing. All through 2023, with the fear around using CEXs thanks to our friendly neighborhood goblin Mr.SBF, DEX volume share, when compared to CEXs, has been only rising.

Now, I’m not going to sit here all day and bore you with the usual spiel about why DEX perps are important and how it’s risky to trust CEXs and have your funds there. You’ve heard it all a million times before and are well-informed by now.

What I will be doing is taking you through this innovative, exciting, and growing sector that is Perp DEXs. We will look at the different mechanisms and some of the top protocols in the sector, and finally, I’ll compare them to a very innovative product called IntentX. Which, by the way, if you don’t know about, then I highly recommend checking out our complete guide on them.

Alright, enough chit-chat, let’s get into it.

The Different Mechanisms

I know you all will love this section since we’re all in it for the tech, right? RIGHT?

Anyway, as with regular spot AMMs, perp DEXs also needed a way to bring perps on-chain in a way that’s well-suited to being functional and efficient on-chain. Blockchains can’t match the efficiency of a centralized orderbook, and therefore, new mechanisms need to be created.

There are four broad categories at the minute that all the popular perp DEX protocols can be split into:

• Orderbook
• Single Asset LP
• Multi-asset LP
• vAMMs

Orderbooks  

Orderbooks are most popularly used by centralized exchanges. The reason for this is that they are very quick, easy, and efficient, making them ideal for perp trading. An orderbook is essentially a list of buy and sell orders. It is used by almost every major exchange and often involves market makers who ensure that the different books for different trading pairs are liquid.

Orderbooks have an order-matching engine of sorts, which essentially ensures that the orders from parties on opposite sides of the trade are matched appropriately. However, to do this effectively, transactions are made very frequently. This, of course, is not feasible when it comes to blockchains.

Transaction times are slow. On days of high demand, they are very expensive, and there is also no guarantee that your transaction will be included in the next block. So, if an entity is going to be making multiple transactions a second, it makes no sense to do so on a blockchain.

Therefore, when the current leader in the on-chain orderbook sphere, dYdX, came to market, they did so in a partial way. Users' funds will remain on-chain for self-custody, and assets can be deposited into smart contracts. However, the actual order matching engine and orderbook will run off-chain. So, transactions are routed from the on-chain interface to the off-chain engine, where they are executed.

This is how most on-chain orderbooks work, but with dYdX V4, they are looking to go completely on-chain with their own cosmos app-chain where every aspect of the orderbook, including order matching, will be on-chain and maintained by various validators.

We’re yet to see how dYdX v4 works in full flow, but in general, fully on-chain orderbooks are simply not feasible. Having a major part of the exchange’s functionality run off-chain defeats the purpose of the blockchain because there is a major trust element involved in that off-chain element being safe and secure.

Multi-Asset Liquidity Pool

As people came to the conclusion that on-chain orderbooks aren’t the best way forward. We started to see a ton of innovation using pre-existing technology in the space. That tech is liquidity pools.

GMX (Formerly Gambit Financial) were the pioneers of this shift in on-chain perp trading. The idea was simple. We’ve already seen that liquidity pools work really well for spot markets, but using separate pools for each pair is not feasible and inefficient.

Therefore, the solution was to have one multi-asset liquidity pool, in this case, the GLP, into which anybody can deposit the accepted assets. This pool will act as a counterparty to all trades on the platform.

Users who deposit assets into this pool will mint a GLP token to represent their position in the pool. The pool then earns through swap fees and leverage trading, which are given back to GLP holders. For traders, when they want to long an asset, they are essentially renting out the upside of an asset from the pool, and when they want to short an asset, they are renting out the upside of a stablecoin from the pool.

This model took the market by storm. They have done a total of $175B in volume and earned $280M in fees across the two chains (Arbitrum and Avalanche) at the time of writing.

While the success of this model is undeniable, there are some major issues with it.  

The main issue is with open interest (OI). The first issue is with LP exposure during trending markets. Since traders rent out assets from the pool, there can be serious issues. For example, in a trending bull market, where all traders want to be long, the GLP pool could be drained of its assets. An example is when OI was 97% in September 2021, albeit GMX had very few users back then.

Conversely, in a bear market, where traders want to be short, the GLP pool will pay a lot in stablecoins. The concern here is a large outflow of stablecoins in combination with the assets in the pool going down in value. In November 2022, OI for shorts peaked at around 76%.

A related issue is a lack of OI balancing.

Funding rates are supposed to be a way to balance OI. For example, when funding rates are positive, the price of the perp is higher than spot, so people pay to long and get paid to short. This acts as an incentive to drive back the price to parity.

With the multi-asset pool setup, it is difficult to have an internal mechanism for balancing OI, and therefore, funding rates are always positive. So, regardless of whether you are long or short, you will be paying funding. Not ideal.

Add to this the issue of oracle manipulation, and you can see that this system is far from ideal.

Single Asset Liquidity Pool

Inspired by GMX’s liquidity pool model, competitors emerged. While some directly imitated GMX with minor tweaks, some other competitors decided to use the liquidity pool model as inspiration but went with a completely different approach.

Rather than having multiple assets in a pool, why not have a single asset pool that acts as collateral for all trades on the exchange? The Gains Network has championed this model with the gTrade liquidity engine.

There is a singular gDAI vault into which users deposit DAI to make trades. All leverage on the platform is synthetic and backed by the DAI vault, the GNS/DAI liquidity pool, and the GNS token. When traders make a profit, they are paid from the gDAI vault, and when they lose money, the amount of DAI in the vault increases.

While this is a very innovative mechanism, it has its downsides. Highly profitable traders with large size could easily take a lot of DAI out of the vault, resulting in the whole protocol being at systemic risk due to under-collateralization. Even though there are limits set in place to prevent this, limits dissuade whales and other large capital participants from then using your platform.

Virtual AMMs (vAMMs)

vAMMs were the first real on-chain perp experience for traders. This model was spearheaded by the brilliant minds at perpetual protocols, and it simply iterates upon the traditional AMM model that we all know and love, x*y=k.

vAMMs like perpetual protocol essentially use this same formula, but the difference is that there is no real asset pool (k) stored in the vAMM itself. The real assets are in a separate smart contract that is used as collateral for the entire vAMM.

The AMM formula is only used as a price discovery mechanism.

Suppose the trading pair is ETH/USDC, and the price of ETH is $1k. The creator of the vAMM can put 100k vUSDC and 100 vETH in the virtual pool. A trader can come along with 100 USDC for a 10x lev long on ETH and deposit it with the protocol.

They will receive 1000 vUSDC to purchase 1vETH, and the same can be done on the short side. After the trade is closed, the profit/losses are settled from the collateral pool.

This model is simple and effective but faces problems regarding slippage. Another issue is that there is often an imbalance between longs and shorts, making it easy for traders to earn funding. The problem is that a lot of this payment, at least in the case of perpetual protocol, comes out of the insurance fund, which, over the long run, is not sustainable.

So, is there a solution?

Fear not, anon, you know we at blocmates always got your back.

As the industry has continued to grow and propel itself forward, new technology that can be applicable to all sorts of applications continues to emerge.

Recently, a piece of tech that was on the tip of everyone’s tongues was intents. For those of you who aren’t already familiar, I highly recommend reading this, but here’s the quick TL;DR.

An intent is simply an expression made by a user to perform a certain action that has a certain outcome. A user can say, “I want to farm XYZ pool on Avalanche.” Therefore, they have expressed their intent. The protocol, depending on what it is, can then automate the execution of this intent.

This is just one example. The concept of intents can be extrapolated and used in various applications. One of them being Perp DEXs.

The IntentX Advantage

IntentX has come to market with its unique intent-based architecture to take on the current market leaders that use liquidity pools, vAMMs, and orderbooks. While all of these are good individually, they are far from the answer.

Just like the on-chain spot markets had their 0-to-1 innovation with AMMs, the on-chain perp markets need its own 0-to-1 innovation. We believe IntentX is that solution.

The idea is simple. It is to iterate upon the pre-existing RFQ (request for quote) based trading system.

A traditional RFQ system works like this. A trader expresses their intent to long or short a specific contract at a certain price with a specified amount of size. This request is sent to market makers who come up with tailored quotes, which are presented to the trader. The trader then selects the best one, and the trade is executed.

Of course, for this to work for on-chain perps, it needs to be automated.

Therefore, IntentX created the AMFQ, or the automated market for quotations.

Rather than going through the arduous process of manually choosing from quotes, the AMFQ continuously streams quotations in advance to show traders the best available quotations at all times. This completely circumvents the inefficiencies of the traditional RFQ system, making it more amenable to on-chain derivatives trading.

There are two parties involved here. The trader and the solver. The solvers are the market makers on IntentX, the entity that creates and fulfills the quotations. These two parties are connected via the IntentX trading engine.

The beauty of this system is that the solvers are not required to commit any upfront capital. Only once a quote of theirs is selected are they required to fulfill the liquidity needs of that trade in a just-in-time liquidity fashion.

This system gives you the same efficiency as an orderbook without the constant need for order-matching and executing transactions. It can have an internal mechanism to balance OI so that the funding rates work properly and the protocol doesn’t end up paying exorbitant amounts in funding.

Even though it has been shown that the majority of traders lose money over time, the liquidity pool models still have the risk of major asset outflow at certain times, which can lead to under-collateralization. Since there are no liquidity pools, that same risk doesn’t exist with IntentX.

Additionally, its liquidity can continue to scale as the protocol grows. Since liquidity can be sourced from anywhere and provided in a just-in-time fashion, the protocol can have deep liquidity without needing to place limits on whales or other high-capital traders.

The cherry on top is trading pairs. Something that’s often overlooked with on-chain perp DEXs is the number of trading pairs on offer. Most venues offer between 5-40 trading pairs (40 is generous), which is very low for an industry like crypto. Not having enough trading pairs is a major deterrent for new users.

IntentX offers 250+ pairs. Almost anything you want, you can trade. Their intent-based architecture uniquely allows them to offer many pairs, which gives them a major advantage over their competitors.

Also, a bonus feature that was recently added was the introduction of TP and SL. A huge game changer in terms of UX.

Concluding thoughts

I know most of you think that Intents are just some new buzzword that’s been strummed up by some VCs to create new shiny coins so we all can get rich off them. But if you actually use IntentX, then you yourself will see that these guys are the real deal.

The perp DEX market is in dire need of finding a candidate that can meaningfully compete with the CEX experience, and the intent-based architecture from IntentX could be the solution. Of course, it’s still early days. We cannot say for sure if this is indeed the answer, but it sure as hell looks like it.

In case you need some extra incentive to give it a shot, then I think this will excite you.

They are going to be doing an airdrop very soon. All you have to do is make an account and start trading on it. You will collect xINTX, representing the airdrop you will eventually receive. If you would like to get started, then click here.

That’s all from me. Hopefully, this article helped pique your interest in intents and IntentX because I believe this is one you really won’t want to fade.