“It’s the future of finance”, “it’s a digital revolution”, “DeFi will eat TradFi”. These are all statements that get thrown around a lot in crypto circles. While this may be true, we are yet to see DeFi protocols get adoption outside of crypto-native circles. So for the time being, the sentiment is more hope than reality. However, the one thing that has gotten mainstream attention is trading.

Whenever you mention crypto to any non-native person, it is synonymous with trading and making those 100x returns. Trading exchanges have long been amongst the most profitable businesses in crypto, but with the recent trust issues sparked by the FTX collapse, the race to become the best decentralized alternative platform began. In this race to the top, there are two models that dominate proceedings, orderbook-based exchanges & LP-based exchanges.

One of the current market leaders for DEX perps is GMX, who really hit the ground running with their LP-based model. They pull in hundreds of thousands in daily fees and continue to grow their user base. While they may have surpassed orderbook-based exchanges such as dydx, the race is far from over.

If you’ve been keeping up with us at blocmates then you may have seen that we previously covered the Vertex protocol and their unique orderbook-based model here (What is Vertex Protocol? – A Complete Guide. - blocmates.). That’s because we believe when Vertex is live it will not only give GMX a run for its money, but also some CEXs. 

In this article, we’ll build upon our previous article and explain the benefits of Vertex’s scalable orderbook model vs an LP-based model.

How do they work?

Let’s start off with the LP-based model. Since GMX is not only a pioneer of this model but has a proven track record of success, let’s understand this model through them.

At the heart of the exchange is the GLP pool, this pool is essentially the market maker for the exchange and acts as the counterparty to all trades executed. It is a multi-asset pool holding all the accepted tokens by the platform. Prospective LPs can mint a GLP token by depositing the required assets into the pool and in return they earn fees as a reward. To get back their assets LPs can simply burn their GLP token.

This effectively makes GLP an index of crypto assets, but to appropriately weigh the index, the platform uses fees to incentivize or disincentivize LPs from depositing. If an asset in the pool is underweight then LPs are charged lower fees when depositing that asset to incentivize higher liquidity for that asset and vice versa.

Pic credits: @riley_gmi 

But as we know, all traders care about is execution. Cheap fees and low slippage are all that matter. GMX achieves this through Oracle pricing rather than using an AMM. The Oracle is of course made by Chainlink and it aggregates prices from top exchanges such as Binance to get a true price of the respective asset at all times without being prone to large wicks caused by poor execution.

In comparison to this, let's look at the Vertex model.

Vertex uses a hybrid model. It combines the powers of an AMM and an off-chain orderbook to provide a highly efficient trading environment. The basic premise is to have the lazy liquidity management and long tail asset support of AMMs combined with the speed and cost-effectiveness of an off-chain orderbook. The result is a cheap and efficient trading platform that can scale with open interest and at much higher multiples of TVL compared to the LP-based design.  

The liquidity on the Vertex AMM is used to support both spot and perp markets. The pooled liquidity sits alongside both markets and the Vertex API allows for automated traders, such as market makers, to plug into the DEX and execute bespoke strategies. 

With the sequencer orderbook layered on top of the AMM, traders receive the 10-30 millisecond trade execution with very low slippage.

Not only does this mean that trades are always executed at the best available price, but it also means that downtime won't be an issue. If the off-chain sequencer is undergoing maintenance or has a bug, then the on-chain AMM will be there to keep trading activity live.

Now let's understand the sequencer, which is one of the most important aspects of this protocol. It essentially matches orders off-chain with low latency, then settles them on Arbitrum. This gives you significantly improved scalability, latency, and MEV protection. But what really makes Vertex a game changer is their default state called “slo-mo mode”.

The AMM pools liquidity getting injected into the off-chain order book to offer much deeper liquidity, tighter spreads, and lower slippage with a super fast order matching engine.

Of course, there is much more to both of these models but for the sake of not making this article a PhD thesis, I will leave it at this brief overview.

So now let's move on to the main question, we are yet to see Vertex’s scalable order book model be used in practice, but theoretically, there are a lot of potential benefits that this product can give over its predecessors, so what are they?

Fees

As we know, fees are imperative to the functioning of any exchange because they are not only a source of revenue but also a source of incentivization. Therefore, a well-designed fee model can often make a world of difference for an exchange.

For starters, the fee model is the centrepiece of being successful at attracting volume. The whales, the HFT (high-frequency trading) firms, and regular trading desks will only use your platform if it's cheaper and more efficient to do so. Once you attract the volume and liquidity, a cycle of adoption gets kicked in as more traders come in bringing in more volume which brings in more liquidity which then further improves efficiency and attracts more volume and the cycle repeats.

The fee structure for an LP-based model for GMX is as follows. A 0.1% on the size of your position and a borrow fee which is paid to the counterparty (GLP pool). The borrow fee is:

In contrast, the Vertex fee model looks like this:

As you can see, the price makers are charged 0 fees in addition to being given VRTX rewards as an added incentive to keep them around. With the price takers, there is a standard fee model:

• 0.02% for spot stables
• 0.01% for futures/perps stables. 

With BTC, ETH, and other coins there is a 0.03% on spot and 0.02% on futures/perps.

The market makers and HFT firms are often making hundreds of trades per day, so getting charged a 0.1% relative to size works out to be a lot more expensive when compared to a flat 0.02% fee. In addition, traders will also receive VRTX rewards for a certain amount of time which will help them offset a portion of the cost they incur from the fees. With significantly cheaper fees than the LP-based model, the incentive for makers to trade on Vertex also provides retail traders with a better trading experience. Lower fees produce more maker activity, leading to more interest in market pairs for takers since spreads are tighter and liquidity is deeper. The zero-sum CLOB model is also market-driven, removing the upper ceiling on open interest, where traders can open multiple positions without incurring high fees as liquidity scales to multiples of TVL. 

A cheaper execution environment will then naturally continue to attract many more participants. So what are the spillover effects of this?

Open Interest

Naturally, the OI (open interest) levels of the platform will be large and consistently increase. For the platform to then be successful and compete with likes of Binance, it needs to have highly scalable liquidity for OI. If you don’t know what this means, then don’t worry, I got you.

Open interest refers to the sum total of all open positions for an individual perpetuals market. A high OI means there’s a lot of trading activity and unsettled positions on the platform, and a low OI means the opposite. Having highly scalable liquidity for OI simply means that the platform can handle multiple large orders without having a significant impact on market price. It means there’s a sufficient amount of large buyers and large sellers to ensure large trades can consistently be executed without major slippage or price impact.

Vertex achieves this through a combination of 2 factors. Their matching engine & the adoption flywheel which was mentioned earlier in this article.

The Vertex engine is designed to be extremely fast and efficient but is also designed to give you the absolute best possible execution price due to their hybrid model, so slippage won’t be a major issue. When you combine this with the fee model, things start to make a bit more sense. The cheap fees ensure more volume which attracts more liquidity which further tightens spreads and reduces slippage which then attracts more volume and the cycle repeats.

Furthermore, the LP model is fairly limited. The liquidity in an LP model is limited to the multiple of LP liquidity that has been provided whereas the Vertex model scales with the increase in OI making for a much more efficient system. 

So a platform designed around the centrepiece of tight spreads and low slippage will inadvertently have more scalable liquidity for OI, therefore, creating a user-friendly, hyper-efficient, and scalable exchange. 

Funding

Funding rates are something unique to crypto because we trade perps or perpetuals. There needs to be a way for the price of the perps contract to be in line with the underlying asset and this is done through funding which is paid between the buyer and the seller. If the perp contract is trading at a premium the funding rate will be positive and the buyer pays the seller a fee. If the perp contract is trading at a discount then the funding rate will be negative and the seller pays a fee to the buyer. 

If we look at the LP model of GMX which uses the GLP pool, we can see that they have fees in place for the rebalancing purposes of the pool. In order to ensure the pool is balanced and the utilization of assets is in accordance with their parameters, the fees are constantly adjusted to incentivize depositing more or less of a certain asset into the pool. Oftentimes, these fees can be very high and turn out to be rather extractive from traders. 

The Vertex model on the other hand doesn't depend on this sort of extractive rebalancing but rather is a zero-sum and direct market-driven funding rate. It is P2P like the CEXs and creates a much more attractive environment for traders who do not need to worry about exorbitant fee extraction cutting into profits or adding to losses on positions. 

Cross-margining

https://www.youtube.com/watch?v=g1JUnnYfV1Y

Another important aspect of exchanges is offering cross-margin (explained in the video above). For those who don’t know, cross-margining is a tool that allows you to share open positions’ liabilities across a single trading account. It often reduces the total amount of collateral required and is typically used for correlated financial instruments.

At Vertex Cross-margining exists by default. Users can deposit collateral into a trading account and that account automatically balances margin usage between positions on the back end, which effectively reduces the collateral a trader is required to put down. This creates a much more capital-efficient system as the cherry on top of a hyper-efficient and scalable hybrid AMM orderbook model.

Concluding Remarks 

All of us in the industry are well aware of the benefits that things like self-custody and transparency provide. It is one of the best things as a user because auditing the exchange you use becomes easy if you’re willing to put in the work. However, the industry has long been starved of an on-chain trading venue that is as efficient as its centralized counterparts. 

With the launch of Vertex just around the corner, that wait may finally be over. But I don’t expect you to take my word for it, go out there and try the exchange for yourself and then you decide if this is the one that can finally bring trading completely on-chain.