If you want to support me in any way, consider either reaching out and being my friend. Or use a ref link or two from any of the dexes mentioned:

• GMX

• dYdX

• Level Finance

• Drift Protocol

• Deri Protocol

• Hyperliquid

• Mux

• Zeta

Introduction

I originally wasn’t gonna write this because it seemed like a lot of work for what would inevitably be limited payoff. As I spent more time lurking around some of the ecosystems I was interested in, I found tons of marginally-tweaked forks, new protocols leveraging existing solutions (see: like 50% of Arbitrum and how it’s all just GLP), as well as new interesting approaches that all required a certain depth of understanding that I felt I didn’t have. I also found that plenty of friends and mutuals were interested in my initial findings, which gave me enough motivation to kick this thing off.

My first exposure to finance was in trading stocks, and my first deep exposure was in learning options trading. I did quite well in this to start out (during a bull market, mind you) and learned a lot, but once I learned what a smart contract was my tradfi journey was over before it started. Despite that, I was drawn to defi from the start, and subsequently trading (despite actually being a horrendous trader). So - this writeup is a sort of return to my roots in a way, paying homage to what brought me here in the first place.

I find too often in crypto people have these gaps in their knowledge and instead of addressing them, they either ignore them or will make assumptions to try and fill in the gaps (A common shortcut to take in any field). Here’s my attempt at helping me and helping you be just a little bit better than that. I also understand how daunting it can be to understand this vertical (or any other) when just getting into crypto for the first time. So this will also be intended to take you from 0-100 (or maybe like 70?) on a fundamental level if you’ve never heard of this stuff before but are crypto-native.

If you’re a trader, a market participant looking for the next hidden gem, or a builder within defi - it’s crucial you understand the top players in this space and what differentiates them. Whether for your own usage, upside, or leveraging for a product you’re building - derivatives are one of the few verticals in crypto that have gained actual traction within our bubble.

What are they? Why so many?

A derivative refers to any type of financial contract (or trade) that is dependent on the value of some underlying asset. When trading against something like BTC, you have (generally) two options: derivatives (leveraged in nature) or spot (not leveraged in nature). The reason decentralized derivatives exchanges are so prominent in crypto boils down to two simple reasons in my eyes: (1) we like to gamble, and (2) blockchain allows for the creation of a trustless financial layer in which two parties can transact efficiently (these two points also represent the beautiful duality of our industry).

As a result of these two facts, we’ve seen rapid innovation in the derivatives space, leading us to where we are today with some damn good products. As long as you can design a place for folks to come and gamble that’s comparable/better than one of the projects listed in this article, you’ve got a market. I’ll only be focusing on options/perpetual futures from a high level for this series. Interest rate swaps are a fascinating category as well that is well on its way to being one of the strongest usecases in crypto post-Merge/Shanghai, but for the sake of focus we’ll be ignoring that category until it’s matured a bit more (I still own lots of $PENDLE though ngl).

Perpetual Futures (Perps)

A perpetual futures contract is an agreement to buy or sell an asset at some point in the future (onchain perps are often cash-settled, meaning the delivery of the underlying asset doesn’t actually take place but rather the cash position equivalent). Decentralized perp exchanges have become top dog in the decentralized derivatives space, thanks to a couple of notable protocols emerging along with clear product-market fit. Most perp dexes follow one of three approaches in their architecture - I’ll give a quick look at each one below to make things easier:

While we’re at it - let’s also define a couple of key terms that I found important:

Open Interest (OI) - The collective total of all open positions on a given day (often separated into long/short OI), denominated in USD.

Funding Rate - The periodic rate at which longs pay shorts or shorts pay longs. Used to correct long/short imbalances. Positive funding rate = longs pay shorts, Negative funding rate = shorts pay longs.

Orderbook (OB) - A collection of all the active buy/sell orders against a given asset. Requires very high throughput - often not feasible to do onchain.

Maker/Taker - A maker order is one that sits on the orderbook/adds liquidity. A taker order is one that fills an existing order. Taker fees are often higher than maker, since maker orders add liquidity while taker orders subtract. Not applicable to non-orderbook infrastructure.

Market Makers (MMs) - Actors who use exchanges to quote both sides of a given market in an effort to capture the bid-ask spread or generate revenue via delta-neutral strategies. Often contingent on orderbook models as well, MMs provide liquidity for exchanges that individual actors can’t and as such often make up a bulk of the volume (e.g. Wintermute, Jump)

Unfortunately, it’s not so simple that these frameworks and definitions are all-encompassing. The fun part is that people are still innovating and coming up with new ideas to this day, and will often take an approach that doesn’t properly fit into one category (see Drift). But for the six main protocols I’m looking at, this is sufficient. For each protocol, I’ll (aspirationally) include some key metrics, a brief overview of how it works, my pros and cons, and how I use (or would use) the dex. Coincidentally, the protocols I’ve selected are (mostly) the market leader on their main chain.

Charts

Before we start comparing some top protocols, let’s take a quick snapshot of the current market to see where things stand + some points to note for each. All the data I’ve got here comes from tokenterminal.

GMX

Overview

GMX was initially known as Gambit Financial on BSC, but rebranded to GMX and moved to Arbitrum in June of 2021, about 2 months after launching on BSC chain. The pioneer of the GLP model, GMX took the phrase “The house always wins” and applied it quite effectively. Normally, with everything in crypto being peer-to-peer, there is no house and thus things are a bit more complex than that. GMX created the GLP pool which serves as a counterparty for all traders. The GLP pool consists of all the assets available for trading on GMX (Currently ETH, WBTC, UNI, LINK, and stables USDC/USDT/DAI/FRAX). Users can come and deposit any of these assets into the pool by minting $GLP, which traders then use as liquidity with zero slippage and market impact. $GLP holders also earn 70% of protocol revenue, which consists of fees from depositing/withdrawing from GLP (dynamically weighted to incentivize maintaining a specific percentage of each asset), opening/closing trades, and an hourly borrow fee charged on open positions.

This model has subsequently created one of the strongest yield-bearing assets within all of crypto. GMX consistently earns some of the highest fees of any other protocol, resulting in an entire sub-ecosystem of projects building on top of GLP in order to take advantage of the real yield (to name a few: Jones DAO, Plutus DAO, GMD, Rage Trade, Abracadabra). This makes GMX comparatively quite a liquid option, due to all the different parties depositing into GLP. The zero slippage/market impact component isn’t a massive innovation - in fact the fee model makes GMX a pretty poor place to trade with size comparatively. What’s made it so effective was the GLP innovation creating such a strong sustainable yield source, which in turn made for one of the most liquid onchain perp dexes. The flywheel in place began turning as people started to realize this, and by now it firmly has the top spot in the Arbitrum ecosystem.

Catalysts/My thoughts

Now time to add some nuance - GLP has been performing quite poorly as of recently, crippling GMX’s biggest usecase. I don’t think this will last forever (especially with v2), but it is giving other dexes a chance to catch up in the perp markets. In addition - fees for GMX have always been high and don’t scale well with price, so options like Synthetix or dYdX are starting to become more attractive (or even some more niche dexes). Kwenta set up this fun little tool to add fuel to the fire, by directly showing the difference in execution between the two dexes.

That being said, you can never count out the GMX team. GMX v2 has long been in discussion (testnet actually just launched, try it out here), but some ideas as to how it’ll be implemented are still unclear. From what I can glean, some key updates in v2 will be: the introduction of synthetic assets, isolated LP collateral (as opposed to 1 large GLP pool), and in general a more sustainable system via tweaked fees and increased capital efficiency. Since GLP operates off of traders being net unprofitable, they’ve never had to consider what happens when they win (given they haven’t been operating at scale during a bull market). v2 attempts to address that issue before it becomes one.

There were also some really cool ideas about X4, a protocol controlled exchange over a year ago. The article details numerous upgrades to the swap function as its own exchange, as well as GMX improvements and additional potential approaches to the AMM model. From hanging out in Discord I’ve seen no updates on this initiative, which is a bit disappointing. However, the team certainly has not forgotten about it, as there are more than a couple hints towards X4 playing a role in v2 or post-v2. We’ll see an updated roadmap after v2 launches that should provide clarity on this.

GMX is where I go if I want to earn reliable yield (GLP), or trade only BTC or ETH (This is solely because my size is not size).

dYdX

Overview

dYdX uses an orderbook model with a hybrid centralized/decentralized system. Users submit orders onchain which are matched to offchain orderbooks (often managed by market makers, large traders, and dYdX), preventing frontrunning and allowing for instant balance updates while retaining some degree of decentralization. The orderbook model is commonplace in tradfi/centralized exchanges due to the presence of large traders and MMs. Orderbook models also rely on the presence of sophisticated market makers as opposed to something like the GLP model where you are simply trading against a pool of assets. This is what leads to so much volume being generated regularly on dYdX, market makers can operate here more profitably than anywhere else. The intended goal from the start was to create a product that felt user-friendly for the average crypto power user as well as institutional market makers - and quite honestly they’ve done a damn good job of getting that done.

Catalysts/My thoughts

dYdX chain (or v4) was announced almost a year ago to date - the chain is currently in private testnet with a public testnet planned for July and mainnet planned for September this year. v4 will be on it’s own sovereign L1 built using the Cosmos SDK. This means they’ll have their own validator network as well as complete customizability for their chain infrastructure. With this approach, they intend to have validators run orderbooks offchain so that anyone can come in and provide infrastructure for the dex while being rewarded by the network. Decentralization while maintaining throughput was top of mind for this transition, and in theory it seems to pass the requisite tests for both concepts. Personally, I’m very excited for v4 as a trader and as a $dYdX holder. Some key features v4 will boast include:

• Gasless trading (fees will be more similar to cex-structured fees)

• Greater throughput

• More yield sources for $dYdX

• IOS + Android apps

Given that the protocol is already in a leading spot, the efficiencies that v4 boasts should only help solidify its spot there, assuming they can deliver. Not to mention the value proposition of $DYDX flies if it becomes its own L1 token rather than just a governance token for a dex. One key thing to mention however is that the private investor + team cliff for the token will be unlocked in December of this year, resulting in a pretty significant boost to the floating supply; between this and v4 launch coming at a similar time I’m expecting to see some serious volatility in the token.

dYdX is where I go if I want to trade specific assets with low fees & low bridging hassle from mainnet. I expect this usage to grow post-dYdX chain.

Synthetix/Kwenta

Overview(s)

Synthetix

Synthetix is a liquidity provisioning backend that enables the creation of derivative products. They don’t offer any frontends of their own, but rather enable protocols like Kwenta, Decentrex(perps), Lyra (options), and more to build frontends of their own. Synthetix operates using a debt pool against synthetic assets (synths). Users can stake SNX and earn exchange fees + emissions, as well as mint synths which reflect the price of an underlying asset via pricing oracles. A certain collateral ratio (C-ratio) needs to be met in order for users to claim rewards on their staked SNX, in order to maintain healthy debt within the system. If a users C-ratio dips below that, they have 12 hours to either bring it above the threshold by minting/burning, liquidate their position, or be liquidated after 12 hours, paying a penalty fee to SNX stakers in case of liquidation.

At it’s core, Synthetix is a similar approach to GMX (although SNX was first) except by using synths that act as debt to the SNX token, Synthetix is able to offer a wider variety of assets but less capital efficiency due to a certain ratio needing to be met (in GMX, 100% of the GLP pool could be utilized. Synthetix requires a certain C-ratio throughout the entire pool, so only ~85% can be used at any given moment). Both rely on pricing oracles of said assets though, making them more similar to each other than to something like dYdX which uses an orderbook. One key thing to note is that minting synths will require exposure to SNX, which not everyone is keen on. Options for hedging against your staked SNX are available though, but to me, this creates a bit of friction - making SNX LPs more segmented from SNX traders.

Kwenta

Kwenta is where almost all of Synthetix liquidity gets utilized (check out this great Synthetix Perps dashboard from @synthquest). A spot/perp dex built on Optimism using the Synthetix protocol, Kwenta accounts are margined using sUSD via their proprietary margin engine. The way this system works is a bit complicated, but essentially it boils down to having different actions be more discrete, so the user has more creative options when it comes to opening/closing trades, resulting in better execution and funding rates for users. The KWENTA token governs the protocol and earn emissions when staked. There’s not too much more to get into here now that we know how Synthetix works, but one important thing to know is that Kwenta is offering some solid trading rewards in the form of emissions as well as $OP incentives, so this could be a great place to trade if you’re looking for additional passive rewards as well as a solid variety of assets.

Catalysts/My thoughts

Synthetix and Kwenta have been in the spotlight recently due to a surge in trading volume. I attribute this to two main factors - (1) GMX v1’s issues we detailed above (high fees, GLP underperforming) and (2) $OP and $KWENTA incentives. Currently, weekly incentives for trading are close to $600k in USD terms - this is proportional to the fees you pay but in general, you are paid to trade on Kwenta (and Polynomial as well, but rewards are better on Kwenta). I think these rewards were enough to siphon activity from GMX and other dexes, but in the long run people will realize that fees/funding rates/execution on Kwenta isn’t all that bad, and even post $OP rewards they’ll stick around (after Aug. 30th).

However, Synthetix v3 just recently launched their alpha and should have the first version of perps liquidity going live any day now. I don’t think (I can’t find direct info on this) Kwenta will be integrating v3 immediately, but it will be happening soon I’d imagine. v3 makes the minting of synthetic assets more akin to a CDP protocol with support for a range of different defi protocols - really doubling-down on the “liquidity-as-a-service narrative” which I’m a big fan of. I wouldn’t be surprised if we see some new token design implementations around $SNX after v3 as well. Synthetix is much more than just a perp protocol, and as such it’s difficult to compare directly to the rest of the protocols I have here (especially from a token perspective). But, Kwenta as a perp dex has some of the most competitive fees, a solid asset offering, and strong incentives (until August), so I’m bullish even pre-v3, and will be reevaluating post $OP-incentives.

Kwenta is what I would use if I want the most direct trading incentives + as a general solution for a wide variety of assets.

Level

Overview

Level’s approach to liquidity provisioning (known as Risk Management for Liquidity Providers, or RMLP) involves splitting LPs into 3 different tranches (Senior, Mezzanine, and Junior) of increasing risk and subsequent expected yield. Imagine exactly what we saw with GMX v1, except with 3 different GLPs, each with varying degrees of risk/yield. When a user opens a trade, the liquidity is drawn proportionally from the 3 tranches (and fees are distributed proportionally) based on the volatility of the asset. You can get a decent idea of how these tranches work as opposed to a single LP below.

There’s some numbers I didn’t include here that can all be found in this section of their docs. I’d highly recommend taking a read if you’re keen on learning more, I think they did a great job explaining. From the graphics I’ve made above, a couple things should stick out to you:

• The mezzanine tranche is the most greatly incentivized via $LVL, if you see the risk distributions for different assets you can see that it (on average) requires the most liquidity across BTC-ETH-BNB, so it stands to reason that incentivization should be grater there.

• APRs are super high. They fluctuate weekly obviously, but in general 3-digit APY is not uncommon. You can also see that most of these rewards are real yield, while only about 10-20% is $LVL rewards

• The most important takeaway for your mental models should be: this is just GLP, but compartmentalized.

I want to fixate on that third point for a second - with GMX v1 we have one large liquidity pool with a basket of assets where certain assets have greater/less fees in order to maintain a certain weight. Here we have the same thing, but with different options for LPs so that liquidity can be segmented to fit demand a bit better. If higher volatility assets are in demand, junior tranches will earn greater yields, and LLPs will take advantage of this as well as provide liquidity where it’s most needed. It’s a simple yet very welcome improvement upon what was already an elegant solution to liquidity provisioning for perp dexes.

Despite being the newest perp dex on my list here, it’s not uncommon for Level to claim the third spot (behind dYdX* and GMX) in terms of 24h volume. There are even days when it rivals or tops GMX. Despite this, $LVL has the 2nd lowest FDV and by far the lowest circulating market cap. That being said, Level uses a dual token model with a governance ($LGO) and utility ($LVL) token, so comparisons can’t exactly be held equal (though, even if you sum both LGO and LVL FDV/circulating market cap, it’s still 4th and last, respectively). I don’t think it’s impossible that this volume consists of some wash trading - something worth digging deeper into there.

I won’t get too deep into the numbers here, but basically both tokens receive 10% of protocol revenue for staking (no lockup, in BTC/ETH/BNB/USDT), both are still quite inflationary currently, but the $LGO token can be redeemed for treasury assets at any given time, so it’s a bit more backed than $LVL. The token design system is a bit more complicated than usual, so I’d recommend checking out their economic flowchart here.

Catalysts/My thoughts

I don’t yet use Level to trade, but I think the token is worth a look given current metrics + upcoming Arbitrum launch.

Gains Network

Overview

gTrade uses a synthetic architecture similar to Synthetix. The key difference is that while Synthetix uses solely the $SNX token to mint assets, gTrade uses primarily $DAI which is supported by the $GNS token. All trades are opened in $DAI collateral via the gDAI vault, which uses the ERC-4626 standard which allows for greater functionality and composability - meaning you can trade your $gDAI and use it across defi despite the fact that it represents a staked position (this is where something like GND protocol) comes into play.

LPs stake $DAI in order to mint $gDAI (users can opt-in to discounts by locking for up to a year or minting when the collateralization ratio is closer to 100%, up to 5%), which increases in price against $DAI as it accumulates fees and PnL of traders. Withdrawing from the vault incurs a 3-9 day unlock period, depending on the collateralization ratio (c-ratio). Let’s follow two different trades (a win and a loss) to help us understand how exactly this works:

1. Trade is opened using $DAI as collateral which is sent to the storage contract.

2. Fees from opening trade are sent to gDAI vault.

*On Arbitrum, currently 5% of ALL trading losses are used to buyback and burn due to high c-ratio.

Catalysts/My thoughts

gTrade occupies a space in the onchain derivatives space that no other dex does currently. With (reasonably) liquid trading across a wide variety of synthetic assets and high leverage (50x on stocks, 150x on crypto, 1000x on forex, 250x on commodities), gTrade presents onchain trade opportunities not possible elsewhere. That being said, there are a few notable limitations the platform implements to maintain a certain level of risk management. Namely:

• Only 3 open trades per trading pair per wallet

• Max open interest per trading pair

• Max open collateral per asset class

• Max win on any trade capped at 900%

So the scalability is intentionally limited to protect LPs at this stage, but we can expect to see these limits rise across the board as the project progresses (OI cap increases are just around the corner actually). In addition, regulation is not far down the line for crypto, and a dex offering the trading of synthetic highly regulated assets with up to 4 digit leverage multipliers would probably be a point of interest for some three-lettered government organizations. This is the biggest risk facing Level - and especially with recent news I wouldn’t be surprised to see them start to delist assets*. But idk tho.

*Stocks and indices have been delisted since I wrote this lol

I would use gTrade if I ever felt compelled to trade commodities/forex/equities with balls-deep leverage onchain.

Drift

Overview

Of the 6 dexes I’ve talked about, Drift’s perp swap is probably the most complicated from a high level, but also one of the most novel. A single trade essentially routes through 3 potential sources with the trade being filled at the best prices first and the worst prices last to ensure best possible order execution for the user. Let’s break that process down by looking at each liquidity source for a market order:

1. Just-In-Time (JIT) Auctions

The first and most novel liquidity source that Drift offers is their JIT liquidity maker pool. Every market order triggers a 5 second Dutch Auction in which market makers can fill the order within a given price band, starting at the oracle price. This creates the opportunity for the user to get a better fill and for market makers to frontrun the orderbook + DAMM.

2. Decentralized Limit Orderbook/Keeper Network:

3. Drift AMM

The Drift AMM (DAMM) was inspired by another perp dex’s model - Perpetual Protocol’s Virtual AMM (vAMM). The original vAMM uses the same xy=k model we all know and love from the original AMM (UNI v1), but rather than actually maintaining a certain ratio (k) of asset x and y backing the liquidity pool, all collateral is held together backing the AMM. This offers a couple of notable improvements - namely being able to provide liquidity for traders without actual LPs since the traders are essentially providing liquidity for each other. Drift v2’s DAMM allows for external LPs to supplement liquidity (single-sided, no impermanent loss as compared to traditional AMMs), adjustable price multipliers, fee tranches, liquidity depth + adjustments, and dynamic spread that programmatically updates ahead of a trade. There’s a whole lot of cool shit going on here that is worth diving deep on (docs here), but without getting too into the weeds I’ll just say that the most current DAMM is a vAMM optimized to dynamically adjust to pricing imbalances and long-short imbalances in order to maintain a healthy market. The DAMM is the last possible liquidity source for a trade and thus has the least optimal execution price for a user.

Catalysts/My thoughts

Drift recently passed $600m in total volume - so it isn’t nearly as battle-tested as the other protocols we discussed (dYdX rarely does under $600m in any given week, but obviously comparing the two is unfair). The different liquidity sources it routes through could provide excellent execution even at scale, the JIT liquidity mechanism, as well as the orderbook, could provide good opportunity for market makers, and the DAMM could provide a decent yield source for users (this is all in theory). I’m rooting for a solid perp dex to emerge on Solana given how fast and cheap it is to use (when it’s not down), and Drift v2 is currently my favorite pick.

I don’t trade on Drift (and I won’t as long as briding options from Solana<>EVM suck so much), but I am certainly keeping an eye on it to see how it performs.

Honorable Mentions

So there’s obviously no way I can feasibly do a writeup with this level of detail for all the perp protocols in one report (I guess I can in theory, but I sure as shit won’t). Those seven alone took me about a month. However, there’s far too many talented builders and interesting products in the space to cap it at 7. 0xperp put together the most comprehensive list of onchain derivatives protocols here. I went scraping through the list as well as a couple of other resources and picked out some of my favorites - a short summary + link for each can be found below (ordered alphabetically). Also, keep in mind that a lot of these are not just perp dexes - I’m just tunnel visioning on their perp offering specifically.

Cap - Built on Arbitrum, off-chain orderbook managed by a keeper network that prices via oracles and executes. This seems to be a more common approach these days, you’ll see it pop up again after this. Cap isn’t all that interesting to me - but they are paying out $ARB in rewards right now. Not enough for me to consider it, but maybe enough for you to do so.

Derivio & Increment - I’m lumping these together because my thesis is the same for both. Both are on testnet still and are planning on launching on zkSync. I think the first native useful dapp (there are non currently) on zkSync could easily be a perp dex like it was for Arbitrum, and these appear to be the first two native ones. There’s too much idle liquidity on zkSync airdrop farming, so this market is itching for something like this. Derivio is currently my top pick given pair trading support (although I’ve gotten some weird vibes from the Discord mods, so not overly bullish).

Deri Protocol - These guys have their own algorithm (called Proactive Market Making) that fills the role of market makers by pricing contracts and using LPs to fill said liquidity. That being said, there are no fees here so LPs are incentivized with $DERI tokens, so TVL + token price since launch over two years ago. I’d be interested to see how the market-making model worked for them in terms of price execution, but otherwise, this one shouldn’t catch your eye.

Hubble - I know literally nothing about Avalanche, but this one still seems pretty cool. Hubblenet is an Avalanche subnet which has its own (multiple!) on-chain order books, gas in USDC, and much more. I find this one to be quite similar in theory to dYdX v4, which you already probably know I’m quite bullish on. They’re still on testnet and as such most of the info available is highly technical - all I’ll say is that it’s definitely one worth keeping an eye on.

Hyperliquid - Another fun one - recently launched on their own Cosmos-based L1 (though, you can’t tell because the UX is so smooth), Hyperliquid has an on-chain orderbook for each asset + a GLP-like vault to provide liquidity. There’s also a liquidator vault that rivals the HLP vault in yield and performance, along with user-created vaults. With this, we end up with some of the most liquid markets for tail assets. I’ve been hanging out a bit in the discord here and am pretty happy with what I’ve been seeing here, excited to see this one become battle-tested.

Injective - Not exactly a perp exchange but rather an L1 with built-in primitives to support the creation of one. Injective is IBC-enabled, provides an on-chain orderbook, and supports both EVM/IBC compatible tokens. Helix is the leading dex on Injective - boasting gasless trading and liquidity provisioning for users. I’m not as involved as I’d like to be in the Cosmos ecosystem, but this seems to be a decently big name over there.

Mango - Home of the infamous $100m exploit (or “highly profitable trading strategy,” depending on who you ask), Mango Markets is nothing more than a fun case study at this point in my opinion given the severity and popularity of the exploit. Offering and on-chain orderbook and socialized losses (lol), Mango is still functional somehow with a couple hundred thousand in volume daily and a banger of a Twitter account.

Mux - Mux is an interesting one - it’s been trending upwards in TVL all year, has sustainable revenue coming in, a simple but effective token, competitive fees + discounts, but very little traction socially. This is often the mark of an overlooked project, so what I’m looking for here is for the trend to continue enough for markets to eventually realize, or for some new novel features that catch the eye of the public. Mux operates similarly to GMX v1 but with multi-chain liquidity.

Perpetual Protocol - Pioneer of the Virtual AMM model. Built on Optimism, offers a decent pool of assets with reasonable liquidity as well as yield opportunities (+$OP rewards). Perpetual generally does the most volume outside of the top protocols I wrote about - and while the first iteration of the vAMM model wasn’t overly successful, it had huge implications throughout the perp space (as seen in nftperp, Pika, and Drift).

Pika - Pika recently launch a token, creating a bit of hype around the protocol. They utilize a vAMM model with liquidity concentrated around the current oracle price. Nothing here really jumps out at me besides a recent token launch + $OP incentives. Fees are quite high, and the vAMM model in this state hasn’t proven its validity over any of the other top approaches.

Polynomial - Polynomial was put on my radar when I saw they had flipped GMX in DAU for a day or two, but it didn’t sustain whatsoever and the volume was still suspiciously low for the number of traders (could have been wash). Anyways, Polynomial is built on Synthetix like Kwenta, has $OP rewards (actually quite high, about $3.5 per $1 fee), but also has higher fees. Given the $OP rewards this could be lucrative for the summer (especially if you have $sUSD debt) but past that I don’t see much.

Rage Trade - Rage Trade is probably my favorite on this honorable mentions list, due to their differentiation in approach yet simplicity in goals. They are aiming to build the most liquid $ETH (and only ETH) perp dex, backed by “recycled liquidity.” Recycled liquidity is composed of 80-20 LP positions across defi (currently supports Curve Tri-Crypto). The 80% in this vault is the LP position (from Curve, for example), while the 20% provides concentrated liquidity on Rage Trade. I’ve been using this LP for quite some time and have been quite happy with its performance. That being said, things are still super early (only 5x leverage, high slippage + fees, only 1 recycled LP pool).

RabbitX - Bet you didn’t expect to see a Starknet project on here, but here we are. Despite the fact that Starknet is barely usable right now, there are updates planned for the network that should drastically increase throughput. RabbitX has a similar onchain settlement + offchain orderbook model that we’ve seen a handful of times already, but this time on a ZK-rollup. The team seems to be quite strong here - and it’s a leading name on Starknet so it’s probably one of your best bets for exposure to that ecosystem.

Unidex - Every time I tweet about Unidex their supporters always come out in hordes to support. Love to see a small project having a community on that level. Unidex aggregates perp exchanges along with their own GMX fork type perp exchange. Still very early, but as far as I can see they’re the most developed perp aggregator we have currently.

Zeta - Formerly an options exchange, Zeta recently relaunched as a perp dex, using Serum infrastructure to provide onchain orderbooks. I don’t know that Serum is necessarily the answer for onchain orderbook infra, but I’m still glad to see projects trying to put this crucial piece of infrastructure onchain. Zeta is second to Drift in the derivatives category on Solana, and it was actually one of the first onchain derivatives platforms I ever used (back when they offered options).

What’s Next?

Who knows? Does Synthetix retain this volume post-$OP incentives? How do SNX v3, GMX v2, and dYdX v4 shake up the market? Does a newer player like Hyperliquid, Hubble, or Rage Trade capture a specific market? Do we finally see an efficient onchain orderbook model emerge (and from what ecosystems)? We could sit and speculate all day on individual outcomes, and to be honest it would probably be a pretty fun discussion. But we don’t have all day, this thing is already like 8k words and if you’ve made it this far you’re probably sick of me by now. So instead, I’ll give you some broad predictions of mine in the hopes that the 3 of you that have read this far will see this and debate some of them with me on Twitter.

If you made it this far, please don’t hesitate to contact me on Twitter (@cryptoaioli) or Telegram (@alib425). Anyone who actually reads this whole thing is a friend of mine <3. Thanks again, and be on the lookout for my options write-up.